Theme and thematic patterns in Spanish and English history texts

Tags: Theme, Eng, Contrastive Rhetoric, definition of Theme, the sentence, verbal processes, material processes, Halliday, material process, thematic considerations, similarities, textual analysis, prepositional phrase, English Corpus, participant types, Theme and Rheme, history textbooks
Content: THEME AND THEMATIC PATTERNS IN SPANISH AND ENGLISH HISTORY TEXTS VOL. I ANNE M. McCABE Doctor of Philosophy ASTON UNIVERSITY June 1999 This copy of the thesis has been supplied on condition that anyone who consults it is understood to recognise that its copyright rests with its author and that no quotation from the thesis and no information derived from it may be published without proper acknowledgement. 1
Aston University
THEME AND THEMATIC PATTERNS IN SPANISH AND ENGLISH HISTORY TEXTS
Anne M. McCabe Doctor of Philosophy
SUMMARY
The aim of this research project is to compare published history textbooks written for upper-secondary/tertiary study in the U.S. and Spain using Halliday's (1994) Theme/Rheme construct. The motivation for using the Theme/Rheme construct to analyze professional texts in the two languages is two-fold. First of all, while there exists a multitude of studies at the grammatical and phonological levels between the two languages, very little analysis has been carried out in comparison at the level of text, beyond that of comparing L1/L2 student writing. Secondly, thematic considerations allow the analyst to highlight areas of textual organization in a systematic way for purposes of comparison. The basic hypothesis tested here rests on the premise that similarity in the social function of the texts results in similar Theme choice and thematic patterning across languages, barring certain linguistic constraints.
The corpus for this study consists of 20 texts: 10 from various history textbooks published in the U.S. and 10 from various history textbooks published in Spain. The texts chosen represent a variety of authors, in order to control for author style or preference. Three overall areas of analysis were carried out, representing Halliday's (1994) three metafunctions: the ideational, the interpersonal and the textual.
The ideational analysis shows similarities across the two corpora in terms of participant roles and circumstances as Theme, with a slight difference in participants involved in material processes, which is shown to reflect a minor difference in the construal of the field of history in the two cultures. The textual analysis shows overall similarities with respect to text organization, and the interpersonal analysis shows overall similarities as regards the downplay of discrepant interpretations of historical events as well as a low frequency of interactive textual features, manifesting the informational focus of the texts. At the same time, differences in results amongst texts within each of the corpora demonstrates possible effects of subject matter, in many cases, and individual author style in others.
Overall, the results confirm that similarity in content, but above all in purpose and audience, result in texts which show similarities in textual features, setting aside certain grammatical constraints.
Key words: Genre Analysis, Contrastive Rhetoric, Rheme, History Discourse, Textbook Discourse
2
To my father, Jack McCabe, in whose patience and love I have learned to love learning. 3
ACKNOWLEDGEMENTS I would like to express my gratitude to the following people, who have provided so much help and support during my research: Thomas Bloor, my thesis supervisor, whose detailed, critical, and deeply insightful comments and discussion have helped me down new paths of discovery and knowledge. I am also indebted to him for his consistent faith in my research abilities, and his concern for my personal well-being, as well as that of my family. I am also grateful to Meriel Bloor for her perceptive discussion and comments, and for helping make my stays in Birmingham that much more pleasant. David Pritchard, for his vital technical support, without which the analyses carried out in this study would not have been possible. Isabel Alonso Belmonte, for our endless and countless discussions on Theme. Julie Arata, at Saint Louis University Madrid Campus, for helping to locate texts, and Fliss Tumkin for photocopying them to fit specifications. Rodney Stephens, for reading and commenting on a draft of this thesis. Julian, Ingrid, and Karo Edge, for their kind hospitality during summers in Birmingham. And to my husband, Luis, and daughters, Elisabeth and Diana, whose love, care, support and understanding have given me constant encouragement and inspiration, as have my parents, Jack and Shirley McCabe, for so many years and across great distances. 4
LIST OF CONTENTS
1. Introduction
10
1.1 Background ...........................................................................................................10
1.2 Genre.....................................................................................................................12
1.2.1 Theme and Genre............................................................................................12
1.2.2 Definition of Genre .........................................................................................13
1.3 Rhetoric.................................................................................................................16
1.4 Choice of Texts .....................................................................................................24
1.5 Analytical Tool and Methodology ..........................................................................29
1.6 Summary and Statement of Hypothesis ..................................................................32
1.7 Design of the Study ...............................................................................................33
2. Comparative Studies: Contrastive Rhetoric and Genre Comparison
35
2.1 Language Comparison ...........................................................................................35
2.1.1 Comparative Linguistics..................................................................................35
2.1.2 Comparative Typological Studies....................................................................35
2.1.3 Contrastive Analysis .......................................................................................35
2.1.4 SLA: Universal Grammar and Typological Studies..........................................37
2.1.5 Other Studies..................................................................................................38
2.2 Contrastive Rhetoric ..............................................................................................38
2.2.1 Contrastive Rhetoric Explained .......................................................................38
2.2.2 Studies in Contrastive Rhetoric: English and Spanish ......................................40
2.3 Comparative Studies Using Theme/Rheme Analysis ...............................................48
2.3.1 Comparing Genres ..........................................................................................48
2.3.2 Comparing Languages ....................................................................................49
2.4 Conclusion.............................................................................................................52
3. Theme/Rheme
54
3.1 Definitions of Theme..............................................................................................54
3.1.1 Theme as Topic ..............................................................................................54
3.1.2 Theme as Given or Known..............................................................................56
3.1.3 Theme/Rheme in Communicative Dynamism...................................................58
3.1.4 Theme as Initial Sentence Element ..................................................................61
3.2 The Model Used: Theme as Message Onset ...........................................................62
3.3 Rheme ...................................................................................................................67
3.4 Unit of Analysis .....................................................................................................69
3.4.1 Background on Choice of Unit........................................................................69
3.4.2 Choice of Unit ................................................................................................73
3.4.3 Rationale for Analyzing Independent Clauses as Single Units ..........................74
3.4.4 Rational for Analyzing Subordinate Clauses as Part of Their Related
Superordinate Clause.................................................................................................76
3.5 Theme Illustrated ...................................................................................................79
3.5.1 Theme in English ............................................................................................79
3.5.1.1 Simple Theme in Declarative Clauses ......... 80
3.5.1.2 Multiple Theme in Declarative Clauses....... 83
3.5.1.3 Theme in Interrogative and Imperative Clauses ........................................84
3.5.1.4 Summary of Theme types in English........... 85
3.5.2 Theme in Spanish............................................................................................85
3.5.2.1 Simple Theme in Declarative Clauses ......... 86
3.5.2.1.1 Process as Theme.................................................................................87
3.5.2.1.2 Se.........................................................................................................88
3.5.2.1.3 Hay......................................................................................................91
5
3.5.2.2 Multiple Theme in Declarative Clauses....... 92
3.5.2.3 Theme in Interrogative and Imperative Clauses ........................................93
3.5.2.4 Summary of Theme in Spanish ................... 93
3.6 Analytical Methodology.........................................................................................94
4. Transitivity And Theme/Rheme
96
4.1 The Experiential Function ......................................................................................96
4.2 Hypotheses for Transitivity Analysis ......................................................................99
4.3 Models ................................................................................................................102
4.3.1 Systemic Functional Grammar.......................................................................102
4.3.1.1 Halliday's Model...................................... 102
4.3.1.2 Other Systemic Models............................ 108
4.3.2 Fillmore's Case Grammar..............................................................................110
4.4 Halliday Revised: The Model Chosen for Analysis................................................113
4.4.1 Process Types...............................................................................................114
4.4.2 Participant Types ..........................................................................................121
4.4.3 Grammatical Metaphor .................................................................................123
4.4.4 Participant Types Revisited ...........................................................................126
4.4.5 Circumstances...............................................................................................128
4.5 Method of Analysis..............................................................................................132
4.6 Results.................................................................................................................133
4.6.1 Functions of the Themes in the Corpus..........................................................134
4.6.2 Participant Types in Theme position..............................................................135
4.6.3 Circumstance as Theme ................................................................................145
4.6.4 Process as Theme .........................................................................................151
4.6.5 Results from Theme Analysis: Participants, Processes and Circumstances......154
4.6.6 Participant Types in the Rheme .....................................................................155
Process Type ..............................................................................................................157
4.6.7 Circumstantials in Rheme..............................................................................158
4.7 Conclusions: Transitivity Analysis ........................................................................165
5. Thematic Patterning
168
5.1 Thematic Structure and information structure .....................................................168
5.2 Thematic Progression...........................................................................................170
5.3 Method of Analysis..............................................................................................176
5.3.1 Thematic Progression Chains ........................................................................176
5.3.2 Peripheral Themes ........................................................................................180
5.4 Expectations on Thematic Progression.................................................................189
5.5 Results of Thematic Progression ..........................................................................192
5.5.1 Results of Thematic Progression Patterns......................................................192
5.5.2 Results of Problematic Peripheral Themes....................................................204
5.6 Textual Themes ...................................................................................................207
5.6.1 Textual Theme Types....................................................................................207
5.6.2 Expectations on Textual Themes...................................................................208
5.6.3 Textual Theme Results..................................................................................213
5.7 Conclusions .........................................................................................................220
6. The Interpersonal and Theme/Rheme
222
6.1 Interpersonal Elements in Theme and Rheme .......................................................222
6.1.1 Interpersonal Function ..................................................................................222
6.1.2 Writer Intervention in History Textbooks......................................................222
6.1.3 Modality .......................................................................................................228
6.1.3.1 Modal Adjuncts ....................................... 230
6.1.3.2 Modality through Grammatical Metaphor 231
6
6.1.3.3 Finite Modal Operators............................ 236
6.1.3.4 Modality in Theme................................... 237
6.1.3.4.1 Modal Adjuncts in Theme ..................................................................237
6.1.3.4.2 Other Types of Modality in Theme Position........................................241
6.1.3.5 Modality in Rheme................................... 245
6.1.3.5.1 Modal Adjuncts in Rheme Position.....................................................245
6.1.3.5.2 Other Types of Modality in Rheme.....................................................248
6.1.4 Reader/Writer Interaction .............................................................................253
6.1.4.1 Reader Reference..................................... 253
6.1.4.2 Author Reference..................................... 255
6.1.4.3 Questions................................................. 259
6.1.4.4 Total Reader/Writer Interaction ............... 263
6.2 Conclusions .........................................................................................................266
7. Conclusions and Implications
268
7.1 Introduction.........................................................................................................268
7.2 The Theme/Rheme Construct: A Viable Tool.......................................................268
7.2.1 Theme in Spanish..........................................................................................268
7.2.2 Thematic Progression....................................................................................270
7.3 Summary of Results: A Cross-Cultural Picture of History Textbooks ...................271
7.4 Limitations and Theoretical Implications ..............................................................274
7.5 Practical Implications...........................................................................................282
7.6 Implications for Future Research..........................................................................284
8. References
286
VOLUME II
Appendix I: The Texts.................................................................................3 Appendix II: Theme Analysis......................................................................151 Appendix III: Rheme Analysis....................................................................228
LIST OF TABLES Table 1.1: Information on Texts Analyzed ........................................................................28 Table 1.2: Clause and Word Counts..................................................................................28 Table 3.1: Theme in English .............................................................................................85 Table 3.2: Theme in Spanish.............................................................................................93 Table 4.1: Process Types and Their Participants (Halliday, 1994)....................................102 Table 4.2: Fillmore's Cases.............................................................................................111 Table 4.3: Process Types ................................................................................................114 Table 4.4: Process and Participant Types ........................................................................128 Table 4.5: Types of Circumstantial Elements...................................................................129 Table 4.6: Themes in Terms of Transitivity Roles ...........................................................134 Table 4.7: Percentage of Actor as Theme per Total Themes: ..........................................137 Table 4.8: Circumstantials and Clauses as Theme............................................................145 Table 4.9: Ellipsis and Inversion of Verbs in the Spanish Data.........................................151 Table 4.10: Configurational Tendency of Participants: English........................................157 Table 4.11: Configurational Tendency of Participants: Spanish .......................................157 Table 4.12: Circumstantial Types in Rheme ....................................................................158 Table 5.1: Overall Thematic Progression Results ............................................................193 Table 5.2: Theme Types outside Progression Patterns.....................................................194 Table 5.3: Distribution of Constant Theme Progression Patterns (according to length)....195 7
Table 5.4: Distribution in English Corpus of Long Constant Theme Progression Patterns196 Table 5.5: Themes outside Progression Patterns in English: ............................................204 Table 5.6: Themes outside Progression Patterns in Spanish:............................................204 Table 5.7: Themes outside Progression Patterns in English (revised):..............................206 Table 5.8: Themes outside Progression Patterns in Spanish (revised): .............................206 Table 5.9: Textual Theme Results from other Studies .....................................................212 Table 5.10: Textual Themes Results ...............................................................................213 Table 5.11: Textual Themes Percentages with Author and Subject Matter ......................215 Table 5.12: Conjunctive Adjunct and Coordinator Results ..............................................216 Table 5.13: Coordinator Use with Non-problematic Clauses: ..........................................217 Table 5.14: Conjunctive Adjunct Use with Non-problematic Clauses: .............................218 Table 5.15: Orienting Themes with Problematic Clauses: English Corpus........................218 Table 5.16: Orienting Themes with Problematic Clauses: Spanish Corpus .......................219 Table 5.17: Conjunctive Adjuncts in the Rheme ..............................................................220 Table 6.1: Interpersonal Theme Results from other Studies.............................................227 Table 6.2: Modal Adjuncts in Theme ..............................................................................237 Table 6.3: Modal Theme Percentages with Subject and Author.......................................239 Table 6.4: Total Modality in Theme Position, English Corpus.........................................242 Table 6.5: Total Modality in Theme Position, Spanish Corpus ........................................242 Table 6.6: Differences in Percentages: Modal Adjuncts in Theme/Total Modality in Theme243 Table 6.7: Modal Adjuncts in Rheme ..............................................................................246 Table 6.8: Total Modal Adjunct (Themes and Rhemes) Percentages with Subject and Author ....................................................................................................................248 Table 6.9: Total Modality in Rheme, English Corpus ......................................................249 Table 6.10: Total Modality in Rheme, Spanish Corpus....................................................249 Table 6.11: Total Modality Percentages with Subject and Author ...................................250 Table 6.12: Reader References in Theme ........................................................................254 Table 6.13: Reader references in Rheme .........................................................................255 Table 6.14: Author reference in Theme:..........................................................................256 Table 6.15: Author reference in Rheme:..........................................................................257 Table 6.16: Frequency of Questions................................................................................263 Table 6.17: Total Interactive Features, English Corpus ...................................................264 Table 6.18: Total Interactive Features, Spanish Corpus...................................................264 Table 6.19: Total Interactive Features Percentages with Subject and Author...................265 LIST OF CHARTS Chart 4-1: Percentage of Process Types in English and Spanish ......................................133 Chart 4-2: Percentage of Participant Types per Total Number of Themes .......................136 Chart 4-3: Adjusted Percentages: Participant Type as Theme per Participant Themes .....136 Chart 4-4: Circumstantial Adjunct Types as Theme.........................................................146 Chart 4-5: Circumstantial Sub-Types in Percentages per Total Number of Themes..........146 Chart 4-6: Clause Types as Theme..................................................................................148 Chart 4-7: Clause Sub-types ...........................................................................................149 Chart 4-8: Combined Circumstantial Adjunct and Clause Sub-types................................149 Chart 4-9: Combined Circumstantial Adjunct and Clause Themes ...................................150 Chart 4-10: Most Frequent Theme Types .......................................................................155 Chart 4-11: Participant types in Rheme...........................................................................155 Chart 4-12: Circumstantial Adjuncts in Rheme................................................................159 Chart 4-13: Circumstantial Adjunct Sub-types in Rheme.................................................159 Chart 4-14: Circumstantial Clauses in Rheme..................................................................163 8
Chart 4-15: Combined Results: Circumstantial Clauses and Adjuncts in Rheme...............163 Chart 4-16: Clause Circumstantial Sub-types in Rheme...................................................164 Chart 4-17: Combined Results: Circumstantial Clauses and Adjunct Sub-types in Rheme 164 Chart 4-18: Frequent Circumstances and Participants......................................................166 LIST OF FIGURES Figure 4.1:Ergative Pairs ................................................................................................104 Figure 4.2: Ergative Pairs ...............................................................................................105 Figure 4.3: Transitive Interpretation ...............................................................................105 Figure 4.4: Ergative Interpretation..................................................................................105 Figure 4.5: Agent and Medium .......................................................................................106 Figure 4.6: Actor, Agent and Medium ............................................................................107 Figure 5.1: Mapping of Simple Linear Progression..........................................................171 Figure 5.2: Mapping of Constant Theme Progression......................................................171 Figure 5.3: Mapping of example 5-3: ..............................................................................172 Figure 5.4: Derived Theme Example (Dubois, 1987: 94).................................................173 Figure 5.5: Derived Theme Example (Danes, 1974: 120) ................................................173 Figure 5.6: Derived Theme Example (Weissberg, 1984: 490)..........................................174 Figure 5.7: Possible Derived Theme (Analyzed as Constant Theme): ..............................174 Figure 5.8: Example of Split Rheme Progression ............................................................175 Figure 5.9: Mapping of Split Rheme Progression ............................................................175 Figure 5.10: Split Theme Progression .............................................................................175 Figure 5.11: Mapping of Split Theme Progression ..........................................................175 Figure 5.12: Mapping of Example 5-6: ...........................................................................178 Figure 6.1: Modal Operators in the English Corpus.........................................................236 Figure 6.2: Modal Operators in the Spanish corpus.........................................................236 9
1. Introduction 1.1 Background The aim of this study is to compare/contrast history textbooks written in two languages using the notions of Theme and Rheme. Many comparative studies in English and Spanish have been carried out (see Chapter 2); however, the vast majority of these have been at the sentence level or below. While these studies have provided a wealth of information as to differences especially at the phonological and syntactic levels, comparison at the level of text merits more attention than it has received in the past, as the "basic unit of language in use is not a word or a sentence, but a `text'" (Halliday, 1970: 160). To clarify, in quoting Halliday here, I am understanding `text' as a technical term, in accordance with Brown and Yule (1983: 6), as "the verbal record of a communicative act". In this sense, a text may consist of only one word or one sentence, but it is its reflection of a communicative act which makes it the basic unit of language in use. The number of comparative language studies at the text level has increased considerably since the establishment of the field of contrastive rhetoric three decades ago. The purpose of contrastive rhetoric is to identify problems which writers have in composing in a second language, and to attempt to explain these problems with reference to the rhetorical strategies of the first language (Connor, 1996). Thus, the bulk of these studies has centered on comparison of student writing across cultures and languages. Grabe (1987: 116) points out that while "...such an approach offers many insights ... contrastive rhetoric must include comparisons of edited texts in different languages, where the writing is somehow comparable". Indeed, the last two decades have seen a great deal of interest on the part of ESP practitioners in examining edited and more professional texts through genre analysis, which looks at the defining linguistic and rhetorical features of text types and links them to the writer's communicative purpose. A great deal of this analysis has focused on the research article, in English and in other languages (Swales, 1981; Swales, 1990; Crookes, 1986; Taylor & Chen, 1991; Myers, 1989, 1994; Ventola & Mauranen, 1991; Rйgent, 1985; to mention only a few), mainly taken from articles written for the natural sciences (Holmes, 1997), although there has been some work on research articles in the social sciences, e.g. economics, linguistics and sociology (Dudley-Evans & Henderson, 1990; Whittaker, 1995; Mauranen, 1993b; Clyne, 1987; inter alia). 10
While little research has been done on history writing in general, some highly interesting work has been carried out in literacy research programs in Australia on history textbooks in English (cf. Veel and Coffin, 1996; Martin, 1991; Eggins, Wignell and Martin, 1993; Taylor 1983) based on genre theory and systemic-functional grammar. These have shed light on linguistic features of history textbooks, such as nominalization, thematic patterning, and grammatical metaphor, and have connected the way these are used with the context and situation of the textbooks. To the best of my knowledge, very few, if any, studies have analyzed textbooks across languages, and at the time of writing I have come across no studies which look at history textbooks across languages. For this reason, I have chosen history textbooks designed for instructional purposes at the upper secondary/tertiary educational level written in English and in Spanish. The Theme/Rheme construct is the analytical tool chosen to analyze the texts as "[i]n its relevance to the moment-by-moment focus of a discourse, theme is especially pertinent to relating written texts to the contexts in which they were produced" (Brandt, 1986: 98, in Vande Kopple, 1991: 341). Furthermore, although the use of the Theme/Rheme construct as an analytical tool has also increased in the last two decades, it has not been used frequently as a tool for analysis in contrastive rhetoric proper, yet it can provide a number of insights into how texts are constructed. Indeed, for Halliday (1985a: 53), "the textual function of the clause is that of constructing a message" and the Theme/Rheme structure is the "basic form of the organization of the clause as message" (ibid: 53). Central to the idea of text are the concepts of `organization' and `choice', as "the `textual' component in language is the set of options by means of which a speaker or writer is enabled to create texts" (ibid: 161). This brings to mind the concepts of `genre' and `rhetoric', as rhetoric "is concerned primarily with a creative process that includes all the choices a writer makes..."(Young, Becker, and Pike, 1970: xii) and genres "constrain the choices of discourse structures in complete texts" (Mauranen, 1993a). Therefore, this study sets out to explore the concept of Theme/Rheme and its relationship to genre and rhetoric across two languages, specifically within the context of history textbooks. The Theme/Rheme construct has been chosen to illuminate text choices which are assumed to be constrained by generic and/or rhetorical forces in the contexts and situations for which the texts have been produced. This places this study in the areas of genre analysis, contrastive rhetoric, and systemic-functional theory, specifically in the notion of Theme/Rheme. Thus, Theme, genre, and rhetoric are explained further, before the general hypothesis of this study is stated. 11
1.2 Genre 1.2.1 Theme and Genre Both the Theme/Rheme construct and the analysis of the thematic patternings of texts have been used by a number of linguists as a means of genre identification, i.e. to distinguish genres or establish similarities within the same genre. Vande Kopple (1991) suggests that, in this area, a profitable line of research might be to examine the relationships between textual, interpersonal, and ideational Themes, their positions and their progressions, as "investigating such intertwinings might help distinguish genres" (ibid: 331). Eiler (1986: 49) also picks up on this notion when she explains that analysis of thematic choice and thematic distribution "can reveal heuristic structures defining a genre". Francis (1989) points out that a number of writers have found that genres can be distinguished by the type of information contained in the Themes of the text, and that patterns of thematic progression seem to be genre-related. Fries and Francis (1992) include mention of linguists who report some correlation between thematic content and certain genres, as well as a correlation between thematic content and elements of generic structure. An example of this type of analysis is Taylor (1983), who compared clause Themes from a science textbook with those found in a history textbook. He found some differences: a preference for introducing propositions in history with a circumstantial setting and a noticeable trend in science towards setting up some condition as Theme. Nwogu (1990) and Nwogu and Bloor (1991) found differences across three different medical genres, not only in thematic progression patterns, but also in choice of syntactic and lexical items functioning as theme. Francis (1990) compared news reports, editorials, and letters from newspapers in terms of transitivity (the choice of participants and processes as Theme) and also as to lexical items chosen for the clause Themes) and found differences in both areas between the news reports on the one hand and the editorials and letters on the other. Basically, in the news reports there were higher numbers of material and verbal participants/processes, and fewer relational ones. Francis (ibid) attributes this clearly to genre and writer purpose, which, in the news reports, is to inform people about events. Ghadessy (1995) finds that Themes are organized differently with respect to grammatical and lexico-semantic features across genres. Fries and Francis (1992) report on work by other linguists (e.g. Berry, 1987) in which correlations between thematic content and certain genres were found. 12
According to Nwogu (1990: 21) differences in genre are "due to the pragmatic conditions of their production. That is, the influence of purpose, audience and medium of discourse and the assumptions of the writer about the reader". Therefore, if the pragmatic conditions of two texts are similar, that is if they were written with a similar purpose in mind, and for a similar audience, e.g. history textbooks written with the purpose of instruction, it seems likely that we could classify the two texts as belonging to the same genre. The question then remains as to whether two texts fitting the just mentioned conditions but written in two different languages can be considered members of the same genre. Before attempting an answer to that question, I shall turn to defining "genre". 1.2.2 Definition of Genre What constitutes a genre? According to Swales (1990: 46) "the principal criterial feature that turns a collection of communicative events into a genre is some set of shared communicative purposes". For Mauranen (1993a: 18) a genre "...is understood as a class of discourses which have the same primary social function". Christie (1985: 12) uses genre to refer to "any staged and culturally purposive activity leading to the creation of a text" and goes on to say that "we create texts representative of particular kinds of genres to serve different social purposes" (ibid: 12). Bamforth (1992: 93) calls the special communicative purpose by which a genre may be recognized its "privileged property", as purpose, rather than form, "seems to be the vital criterion". Bhatia (1993: 13) also underscores purpose: "Although there are a number of other factors, like content, form, intended audience, medium or channel, that influence the nature and construction of a genre, it is primarily characterised by the communicative purpose(s) that it is intended to fulfil". A genre, then, is defined not just as a text type; Martin (1992b: 503) states that "genres are social processes". It is clear from the above that the communicative purpose and social function of the events are key factors in the characterization and creation of genre. In sum, a genre represents an event or set of events which leads to the creation of a text based on a common social/functional communicative goal or purpose. An example of the above is a medical report written in an emergency room in Madrid. All those patients who enter the emergency room on foot go through a series of basic steps upon admittance to the emergency room: vital signs are checked, blood and urine samples are taken, and an interview is held with a doctor, followed by an examination. The obvious purpose of this activity is not just to diagnose the patient's ailment, but to determine if further emergency steps need to be taken. Once the results of the analyses are 13
all in, the doctor writes up a report for another doctor to read, either a specialist or the patient's general practitioner, for further study. Here there is a set of events, the analyses and their results, the interview and examination, which lead to the creation of texts, the reports, which have a common goal: to inform other doctors of the procedures, the diagnosis, and suggestions for further analysis; and all of this revolves around the overriding goal of easing the patient's suffering. In the above example, tapings of the interviews between the patients and the doctors would constitute a different genre from the reports, as their social purpose is different. This can be analyzed in terms of the context of situation, or the `contextual configuration' of field, mode, and tenor (Halliday, 1994: 390). For Halliday (ibid) field "refers to the nature of the social action: what it is the interactants are about". In the case of the interviews and the reports, the field is that of medicine, of hospital emergency procedures and further medical procedures, and specifically about the particular patient's health and symptoms. Mode "refers to the rhetorical channel and the function of the discourse" (ibid). In the above example, the interview is spoken, and its function is to find out the patient's symptoms, while the report is written, and its function is to indicate to another professional what procedures have been carried out, to diagnose, and to indicate further treatment. Tenor refers to the status of the participants and their role relationships. Here there is an asymmetrical role relationship in the first instance in terms of the social roles of doctor/patient, and a symmetrical relationship in the report of doctor/doctor. The interviews and the reports share some of the events that take place in their creation, yet their communicative purposes are different. Therefore, while they may share some content, or belong to the same field, the ultimate shape of the text, the features of the text, will be different, and we can expect some common features within each of the genres. The view taken here with respect to the register variables of field, mode and tenor accords with Martin (cf. 1992b) with genre realized through register, thus establishing genre as superordinate to register. Another example is that of history textbooks. In terms of field, the nature of the social action centers on historical events, on what happened in the past. The mode is written, with a pedagogical function; in other words they are written to be assimilated for study purposes. The tenor is asymmetrical: the writer is the one who is in possession of historical information which the reader presumably does not know, but needs to learn. The set of events leading to the creation of a history textbook include the historical events themselves, the existence of history courses in high schools and universities, the widespread 14
use of textbooks in courses in schools and universities, and processes of writing, editing and publishing. Thus, genres are social processes which bring together the variables of field, mode, and tenor in differing ways; those having a similar social function/goal will tend to do so in ways similar to each other, thus allowing a distinction to be made between, for example, an emergency room report genre and a history textbook genre. These similarities then feature in the texts produced as verbal records of genres, which are also referred to as genres. Thus, the definition of genre can be extended to include the texts, e.g. as Holmes (1997: 322) does: "A genre then can be briefly defined as a class of texts characterized by a specific communicative function...". Connor (1996: 11) also includes the product, or text, as part of her definition, as for her genre refers to "texts formed according to cultural and traditional expectations as required by specific purposes and tasks such as a Research Report in biology". In addition to register variables, genres also typically display similar overall structural organization, or schematic structure, e.g. methods-results-discussions sections of research articles, as well as content and style (Swales, 1990). At the same time, over time certain ways of writing texts become standardized, and thus, "[i]n more sophisticated versions, one recognizes a dialectic mutual constitution between text and context, rather than a determinative relationship from context or situation or activity to text." (Lemke, 1998). Devitt (1993: 576) also stresses this reciprocity when she states: "In practice, of course, genres already exist and hence already constrain responses to situations". At the same time, she emphasizes the notion that generic features develop as appropriate responses to situations, and then "the similarities among those appropriate responses become established as generic conventions" (ibid: 576). Indeed, texts are often produced in imitation of other texts. In other words, it is possible to produce an emergency room report whose goal is not to inform other doctors, but simply to illustrate what a typical emergency room report looks like. We can "hybridize genres and registers, parody them, ventriloquate them, subvert them, and generally play with them" (Lemke, 1998). For purposes of the discussion here, it is important to distinguish between defining a genre based on its social purposes and classifying a text as representative of a certain genre due to its textual features. As Mauranen (1993a:18) points out, genres "tend to display typical lexicogrammatical and textual features, which facilitate their identification, but such features do not constitute obligatory or definatory criteria for genres". For Francis (1990), however, an approach which first classifies a text as representative of a certain genre can be 15
limited in that any feature found has to be interpreted then as characteristic of that genre; she suggests that another approach would be to analyze texts which have not been precategorized and to classify them according to shared features. It would be interesting to then see if those texts which ended up in the same category could be defined as belonging to the same genre according to the definition of genre given above. At any rate, the approach adopted here is to work from texts which, given the similarity in their communicative purpose and the social function which they fulfill, are assumed to belong to the same genre. Then, the assumption is that, barring grammatical constraints, history textbooks will display similarities with respect to thematic content, i.e. textual, interpersonal and experiential elements, and thematic progression patterns across languages, specifically here across English language American history textbooks and Spanish language history textbooks published in Spain. 1.3 Rhetoric The assumption just stated would seem to fly in the face of the notion of contrastive rhetoric, which, according to Kaplan (1995: 21), is a hypothesis that claims that "the logic expressed through the organization of written text is culture-specific; that is, it posits that speakers of two different languages will organize the same reality in different ways". According to Kaplan, when faced with a cross-linguistic writing task, the writer must consider four questions: What can be discussed? What is evidence? How can that evidence most effectively be organized? To whom may a text be addressed? (Kaplan, 1995: 22). It is interesting to look at these four questions with regard to register variables and to the history textbook data around which this study centers. The first two questions refer, in the first instance, to whether it is allowable in one culture or another to discuss things such as sex and death, and in the second, to notions such as plagiarism. In some sense, these questions fit under the notion of field; historical events constitute a vast field, and different cultures will select different events as more or less important, and some events may be taboo in a given culture. In the Theme-based studies by Francis mentioned above, genre constraints were found with regard to thematic content, mainly in terms of lexico-semantic content of Themes, e.g. whether they referred to people, abstract entities, things, and so on, and in terms of the participant selected from the system of transitivity functioning as Theme, 16
e.g. whether it refers to an Actor1 in a material process, a Senser in a mental process, a Carrier in a relational process, and so on. While it is certain that the history textbooks across the two cultures will choose different specific content for Theme (e.g., and very simplistically, Felipe II in Spanish and George I in English), this is also the case for textbooks written within the same language, depending on the historical philosophy of the authors and what they wish to stress. However, a genre-based hypothesis would predict categorical similarities with regard to thematic content, i.e. Felipe II and George I are both people, and both can be, for example, Actors in material processes. Kaplan's third question above could be related to mode, to the organization of the text, and for Kaplan "is the crux of the matter" (Kaplan, 1995: 23). Kaplan (ibid: 23) cites Berman and Slobin (1994:12) in explaining that "[w]riters arrange evidence in terms of their ...abilities to convey just those analyses of the event that are most compatible with the linguistic means provided by their languages...". They go on to give an example which shows that, in a narrative, English and German utilize a number of locative particles combined with verbs of manner, while Spanish and Hebrew rely more on change-of-state and change-of-location verbs, thus providing a different style of encoding motion events. This example used by Kaplan seems to obscure the issue somewhat, as the focus here is more on what is encoded in the verb phrase - more detailed motion events as opposed to changes in the motion event - whereas the order of presentation of the event is similar in the examples given. Indeed, linguistic resources available will have an effect on the text organization, for example in terms of thematic progression patterns (e.g. Spanish allows for Subject-verb inversion and is a pro-drop language, thus allowing the verb to appear in thematic position; see Chapter 3), but contrastive rhetoric goes beyond simple lexicosyntactic differences (Connor, 1996: 15). Thus, it is not clear how the organization of text is affected by cultural perceptions and norms, beyond syntactic differences. Kaplan's fourth question above, regarding text addressees, relates directly to tenor. He gives examples of languages such as Chinese and Japanese, where elaborate politeness structures are syntactically marked. These types of structures do not appear in American and Spanish textbooks; indeed, as will be seen in Chapter 6, there is very little direct address to readers. Kaplan also refers to symmetrical differences with reference to the student writer writing in a setting where the distribution of power is unequal, yet often students are asked to write as if it were equal. In the case of the history textbooks looked at in this study, the 1 Labels indicating a linguistic function begin with a capital letter. 17
writer/reader relationship is asymmetrical, and that asymmetry would seem to be similar across the two corpora. From the above discussion, no clear conclusions can be reached as to what might be hypothesized based on the theory of contrastive rhetoric. However, it is clear that the field of contrastive rhetoric was conceived to address problems in L2 student writing, and its purpose is to attempt to explain L2 student writers' difficulty by reference to rhetorical strategies of their L1 (Connor, 1996; Houghton, 1980, but cf. Connor and Lauer, 1998: 138, where contrastive rhetoric is defined as "the comparison of the writing of students and accomplished writers from different linguistic and cultural backgrounds"). In fact, the studies in Spanish included in Connor's (1996) review of studies in contrastive rhetoric are all based on student writing (both L1 and L2 writing, mainly in response to a prompt). It is not at all the purpose of this study to question this aspect of contrastive rhetoric. At the same time, the definition of contrastive rhetoric provided by Kaplan is not based on student writing, but on writing. And Mauranen (1993a: 3) indicates that "[i]t is assumed that, despite a relative uniformity of academic papers imposed by requirements of the genre, there is significant intercultural variation in the rhetorical preferences of the writer". Is there an explanation, then, which would allow for the co-existence of a genrebased hypothesis, which would predict similarities in thematic organization and content across languages, and a contrastive rhetoric hypothesis, which posits that across cultures writers organize text in different ways? One possibility would be to make a distinction between genre and rhetoric as, by doing this, then "it can be argued that aspects of academic writing which tend to be universal are conditioned by genre, while the more variable aspects fall under the domain of rhetoric" (Mauranen, 1993b: 4). In order to draw this distinction, then, a difference between genre and rhetoric needs to be established. Thus, some discussion of how rhetoric is viewed is needed, in order to provide a background within which analysts of contrastive rhetoric are working. As we have already seen, according to Young, Becker, and Pike (1970: xii), rhetoric "...is concerned primarily with a creative process that includes all the choices a writer makes...". This description can be combined with Mauranen's idea of variability: while certain choices made by writers may be constrained by genre, within those constraints there is a certain amount of creative variability which falls under the domain of rhetoric, where perhaps there is more of a cultural aspect. Indeed, Kaplan has been criticized for his narrow view of rhetoric (Connor, 1996: 31), a view which is based mainly on organization and on the classification of discourse into description, narration, argumentation, and exposition. 18
Purves (1988: 9) defines rhetoric as "the choice of linguistic and structural aspects of discourse - chosen to produce an effect on an audience. Rhetoric, therefore, is a matter of choice with respect to the uses of language as opposed to those uses that are determined by lexical and grammatical strictures [sic]". Here we do see the notion of rhetoric wanting to stretch beyond lexico-syntactic limits. An interesting aspect of Purves definition is the inclusion of producing an effect on the audience. Indeed, the criticism which Connor (1996) levels at Kaplan is that of leaving out a major component of classical rhetoric, which is persuasion. This was replaced in Kaplan's model by argumentation, with a focus on the appeal to reason and logic, while left out were appeals to credibility and to the emotions. For Mauranen (1993a: 8), appealing to credibility is a key aspect of rhetoric, as "[t]hrough rhetorical choices, a writer aims at increasing the credibility of certain propositions in the reader' s mind". Andrews (1992: 7) approaches a definition of rhetoric from a different viewpoint, from that of literary studies. He laments the reduction of rhetoric to a different element of classical rhetoric in this field, to that of diction, as the focus of rhetoric seemed to be reduced to stylistics, since "it was seen to operate only at lexical and syntactic levels and via metaphor and metonymy", ignoring, thus, invention, arrangement, memory and delivery. He would like to see especially invention, arrangement and delivery brought again under the umbrella of rhetoric. He feels that rhetoric offers a general picture of the relationship between speaker/writer and audience, subject matter and text or utterance, which "can array the possible routes a speaker/writer can take to achieve his or her end" (ibid: 13), and that "we are now in a position to play much more inventively with arrangement than the classical treatises suggest" (ibid: 13). Thus, rhetoric is concerned with the choices that a writer (in this case) can make in order to achieve an intended effect. For example, Andrews (ibid: 13) suggests that "[i]t can service propriety, but also generate humour through impropriety". In this sense, rhetoric goes beyond only that which is persuasive. Or looked at another way, "all communication is persuasive" according to Winterowd, for whom rhetoric means "the use of language" (Winterowd, 1968: vii, in Andrews, 1992: 4). This more general use of the term "rhetoric" is supported by Liu (1995), a professor of rhetoric, who shows that the field of rhetoric currently goes beyond the limits originally set in classical Greece in his treatise on the subject. He argues against Derrida's (1990: 15, in Liu, 1995: 9) contention that rhetoric should stay with the traditional "limits of verbality, formality, figures of speech", and against Ricoeur's (1989: 140-141, in Liu, 1995: 12) admonition that it should stay in its "original sites" or "generative seat", where it must limit 19
itself to "the right to command the public use of the word in the typical situation described by the political, legal, and the festive assemblies". Liu (1995) cites Cicero (1954) in support for his view of modern rhetoric, which is that it should be concerned with all of the speech acts, as, if the view of rhetoric is limited to the tri-part division, On what kind of oratory are we to consider ourselves to be employed, when we complain, console, pacify, excite, terrify, encourage, instruct, explain obscurities, narrate, plead for mercy, thank, congratulate, reproach, abuse, describe, command, retract express our desires and opinions, to mention no other of many possibilities" (Cicero, 1954: 3.4.3, in Liu, 1995:14). Within systemic linguistics, White (1998) talks of the "rhetorical potential of texts", which he equates with "how texts are organised not only to persuade explicitly but also to influence, and ultimately to naturalise attitudes, beliefs, and assumptions by more indirect, more implicit means". From this discussion, some difference between genre and rhetoric emerges with regards to text. While genre is concerned with discerning similar patterns in text types, due to their similar contextual situation, purpose and audience, rhetoric seems more centered on the individual production of the text and the effect a particular text will create. If we look at the examples given above in the section on genre (Section 1.2.2), doctors writing emergency room medical reports may wish to add a few "!!!" or use underlining in key places to convey to the reader a sense of importance with regard to some aspect of the patient's diagnosis. In history textbooks, authors may use pictures, charts, graphs, tables, and timelines, or insert text boxes with reproductions of primary textual documents in order to take the reader beyond the lines of the textbook. This is to state the difference between genre and rhetoric superficially, since the two are intertwined in the production of a text. It is also to blur the notion of text as explained above, as a verbal representation of a communicative act, because some of the conventions just mentioned are visual, or iconic, representations. With regards to the lines of text itself, that is the written words on the page produced by a writer in a given social situation, rhetorical effect can be achieved as well. If we take a genre such as a research article, authors may conform to generic conventions by including Introduction, Methods, Results and Discussion sections, but may, through choices of lexis and syntax (e.g. inclusion of overt interpersonal markers, passive voice, etc.), show a greater or lesser alignment with other authors in order to achieve an effect. This is also to simplify the issue, as it is not the point to say that genre choices have to do with overall schematic organization and rhetorical ones with lexico-syntactic choices. The point is that generic choices in this view would be those which conform to certain expectations, while 20
rhetorical choices may or may not, depending on the hoped for effect. Valero-Garcйs (1996: 281) arrives at a similar conclusion when she says: In other words, some features of scientific discourse are provided by the genre of the text, others by the culture they belong to and also by the writer's own style. The first features help us to distinguish an economics text from a novel, and the second ones (rhetorical features) add some cultural characteristics to the text. However, this view of rhetoric cannot ignore the writer's own style, which Valero-Garcйs also admits, which would mean that there could be as many individual differences between texts as there could be cultural differences. As to cultural differences, Valero-Garcйs gives the example of English writers placing topic sentences at the beginning of an article, while studies have shown that writers in other cultures (cf. Clyne, 1987, for German and Mauranen, 1993b for Finnish) have a tendency to include them later. Moreno (1997) notes varying results in studies in contrastive rhetoric, some which show differences, some which show similarities (see Chapter 2). She feels that those studies which show similar results "might be taken to suggest that there might be some factor, more powerful than the writing culture of the different language groups, that is able to unify the nature of discourse patterns across languages". Her paper explores whether the research article genre might not be one such factor, basing this conjecture on Widdowson's (1979: 61) claim that within scientific exposition there is a universal rhetoric "with some tolerance for individual stylistic variation" but "which imposes a conformity on members of the scientific community no matter what language they use". This use of the word "rhetorical" here is in tension with contrastive rhetoric, it would seem, unless we can, for purposes here of establishing a distinction between rhetoric and genre, tie the notion of "individual stylistic variation" with rhetoric, and perhaps leave the conformity under the domain of genre. However, Moreno raises a problematic issue: that of which aspects of the text are more subject to genre constraints and which to cultural peculiarities. The literature is not at all clear on this issue, and, in fact, it is usually not addressed at all. Moreno raises this at the end of her paper, and indicates that more studies of genres and rhetorical variables are needed. The problem is that if similarities are found, they are usually attributed to genre constraints, and if differences are found, they are attributed to rhetorical variables. What of the Theme/Rheme construct in regards to genre/rhetoric? While the studies mentioned earlier link Theme choices and thematic progression with genre constraints, Hawes and Thomas (1996: 159) link thematic choices with rhetoric in their paper on rhetorical uses of Theme in newspaper editorials. They argue that "thematization can be 21
used to control hortatory rhetoric" by "[v]arying the complexity of thematic structure" which "can produce minimalistic, informal text or an academic formal style". They also argue that the number of discourse participant Themes correlates with outspokenness or impartiality, and that marked Themes are an indication of hortatory content (cf. a similar statement made by Francis and Kramer-Dahl, 1991). For Mauranen (1993a: 36), Themes organise sentences so that the content of each sentence can be seen as a relevant unit in the evolving content of the text as a whole. Rhemes normally carry the focal content in a sentence. The choice of placing constituents in thematic or rhematic position is therefore rhetorically significant. The rhetorical character of thematic choices lies in the organising power of the theme-rheme configurations. Lotfipour-Saedi and Rezai-Tajani (1996: 243) feel that different modes of thematization "rather than being genre specific are discoursally motivated" as they find in their data of literary, journalistic and advertising texts that "there is no one-to-one correspondence between a specific mode of thematization and a single genre" (ibid). Rather, authors may use the same strategy type for different purposes in different text types, e.g. a strategy may be used to attract readers' attention in an advertising text, and the same strategy may be used in a literary text to lengthen text processing time in order to add to the text's imaginative nature. However, while they did find heterogeneity in these types of texts, in scientific texts (from a range of 5 areas of science) they found greater homogeneity in terms of thematic selection and distribution, which they attribute to genre constraints. They attribute the homogeneity in scientific texts to it being an "`information' oriented genre where what is said assumes more importance than how something is said"; the heterogeneity of literary texts is attributable to their being a "more `imagination' oriented genre, one in which the way something is said assumes higher importance than `what should be said'" (ibid: 241). This difference is of interest here, as rhetoric has been linked more with persuasive texts, such as advertising texts, and with more imaginative texts, such as literary texts (and, indeed, contrastive rhetoric has its roots in composition studies, which are in turn rooted in English, or literature, departments in the U.S.). There are two points which recur with regard to the notion of rhetoric. The first is that of intended effect of the speaker/writer on the audience. In the literature, genre seems to take in all of the variables of field, tenor and mode, while rhetoric seems to be restricted mainly to this notion of effect, linked more, then, with tenor, but to some extent with mode. This is tied in with a second point. It will be seen in Chapter 2 that the speakerwriter/hearer-reader relationship seems to be an area where differences are found in contrastive rhetoric, and notably in the Spanish/English studies reported on there. Thus, for 22
purposes of this study, genres are social process having a common social/functional communicative goal or purpose, while rhetoric is a discourse process by which writers (in this case) make textual choices based on the effect they would like to achieve and on the relationship they have with their readers. Mauranen (1993a) also arrives at a similar conclusion, as for her, the limits of genre are set by the purpose and convention forms of the social activity and the limits of rhetoric by the desire for effect: the likely success is regulated by the expectations of readers. Given that the texts chosen for this study are textbooks, the writers presumably do not have a great necessity of persuading their readers to agree with them, although they do want to sound objective and factual so that the readers will believe them. Also, the reader/writer relationship is asymmetrical: the writer is in possession of knowledge which the reader presumably does not have. Thus, it is hypothesized here that genre constraints will override rhetorical differences. This also accords with Mauranen's (1993a: 8) view: Through rhetorical choices, a writer aims at increasing the credibility of certain propositions in the reader' s mind. This view of genre and rhetoric then postulates rhetoric as an aspect of genre. Genre constrains rhetorical choices, and in this sense logically precedes rhetoric. One final point with regard to this tangle of genre/rhetoric needs to be addressed here. It was established in section 1.2.2 above that texts such as research reports in biology, emergency room reports, and history textbooks are posited as generic. Francis (1990) uses the term "genre" at a greater level of generality when she classifies editorials as belonging to the genre of analytical exposition and letters of complaint as belonging to the genre of hortatory exposition. Likewise, Lotfipour-Saedi and Rezai-Tajani (1996) include narrative, descriptive and argumentative as different discourse genres. However, in composition teaching in the American tradition, these are normally referred to as rhetorical patterns. In this study, they are taken as textual devices or patterns which are used within genres to help writers achieve their purposes. For example, in history textbooks, at times the writers use narration to provide, in a sense, pictures of the past for readers, and at other times they use analytical exposition to provide explanations, usually causal, of events. To sum up, from past studies based on genre, it was hypothesized that the two corpora would show similarities with respect to thematic organization and content. However, this hypothesis would seem to clash with the notion of contrastive rhetoric. Given the lack of a precise definition of what rhetoric is in the literature, it was posited for purposes of this paper that rhetoric is concerned mainly with producing an effect on the reader and with the writer/reader relationship. It is assumed that text types which are more 23
"imagination-oriented", where how something is said may have more importance than what is being said, will feel the force of rhetorical differences more than "information-oriented" texts. Thus, the conclusion has been reached that the constraints of the genre of history textbooks would produce similar rhetorical choices as regards the Theme/Rheme construct. 1.4 Choice of Texts The question still remains as to whether or not we can consider history textbooks from two different cultures as representative of the same genre. I have chosen a definition of genre which is based on the social/functional goals or purposes inherent in an event or set of events, leading to the creation of text. Theme is a central element in text, and it is interesting to note that Berry (1996) relates the study of Theme with the question of how a speaker\writer's main concerns in producing a text are determined. She suggests that the simplest way of determining the main concerns is by asking those involved. It has not been possible in this study to interview the writers concerned. At any rate, in the case of the texts examined here, it would not be sufficient to interview the writers, as all of these texts have gone through a process of publication, which would necessarily bring in the editor(s) and publisher(s) and, thus, an additional set or sets of concerns, which may parallel those of the author(s) in many respects, but may not in others. Therefore, here the establishing of concerns is based on the context and on the purposes inherent in the event for which the texts have been produced. The event leading to the creation of the texts used in this study is similar in both cultures, as their communicative purpose, that of imparting knowledge of historical events, and their social function, that of instruction, are similar. However, this still begs the question of whether instruction itself is similar in the two cultures, and more specifically, whether the instruction of history is similar. Indeed, there are differences, which has had an effect on the selection of texts for the corpus. The original intent of this study was to look at textbooks chosen for two history courses taught at an American university. Both of these courses are offered in both Spanish and English, and the students can choose the course in either language to fulfill the general history requirement set by the university. The objectives and content of the courses are very similar across the two languages. A number of the texts included in the corpus were in use during the initial phase of this study. This led to a discrepancy in Publication dates from the start: the Spanish texts used (Spa 1, Spa 2, Spa 3, and Spa 8) were all published much earlier than the English texts (Eng 1, Eng 2, and Eng 3). Another problem arose soon after data collection began: The professors of the section taught in Spanish were not happy with 24
the materials used, due in part to the date of publication, but due in greater part to the content of the texts, which the instructors felt did not fit well with the proposed content. This stems mainly from the different way historical events are divided in the two countries. In the U.S., the tradition has been to focus on "Western Civilization", as a means of providing American students with a greater sense of the shaping of their own culture. While the notion of Western civilization as a "reality" is currently under heavy attack in the U.S. and abroad, it still underlies the history courses taught as a general requirement at American universities, and provides the backbone for the texts produced for the courses, as illustrated by the following quote, taken from the introduction of one of the texts used in this study: The view that this [Western civilization] is one civilization with America tied more closely to ancient Greece than Greece is to Egypt or Spain to Islam can be easily challenged in every respect save the conscious tradition that has shaped our culture... The Western Experience, then, honors that tradition...(Chambers, et. al, 1991, xii) Indeed, there are a number of textbooks written in this tradition, and which are used for the type of course in question here, that of a general requirement of university students in the U.S. Since the initial phase of this study, the textbook in use has been changed, and chapters from that text have been included (Eng 6 and Eng 9). Eng 4, Eng 5, and Eng 7 are all taken from textbooks which have been considered for the course in English, and which are kept in the library as further reference for the students on the course. Finally, and in order to offset the earlier publication date of many of the Spanish texts, two older texts in the same tradition have been included (Eng 8 and Eng 10). However, history is not carved up in the same way in Spain, where the focus is rather on countries or continents, or on world history, than on a notion of Western experience. Therefore, the teachers of the history course at Saint Louis University in Madrid found it difficult to choose an instructional text, and often resorted to assigning readings from different books. The solution in place now has been for the teachers to write their own materials for the course. I had originally planned to use these materials as part of the corpus; however, since they have not been published, there are problems with them in terms of, for example, errata and layout, which render them not very comparable to texts which have undergone the process of editing and publishing. Therefore, in addition to the texts in use at the initial phase of this study (Spa 1, Spa 2, Spa 3, and Spa 8), also included are supplementary texts which are given as assigned reading in the courses (Spa 4, Spa 6, and Spa 7). At any rate, given the problem of publication dates and teacher written materials, I 25
decided to look beyond the context of these American university courses, in order to find more data in Spanish. My first choice was to look for texts used at Spanish universities to teach history. However, the type of course for which the texts in English were written is a general history course required of all university students at most, if not all, American universities. This type of general requirement does not exist at Spanish universities. The texts used to teach in the university History degree program in Spain are highly specialized in one specific area of history. Then, history professors themselves suggested that history textbooks used during the last year of secondary school might be more comparable to the type of text used in the English sections of the course. In the number of different secondary schools consulted, the same textbook was in use, and is included in the corpus (Spa 5, Spa 10). However, at the same time a balance of topics was needed in the corpus. The curriculum for the above mentioned secondary level course is modern history, and deals with the 19th century onwards. As two chapters from the prehistoric period were included in the English corpus, a chapter from a Spanish university textbook for a course in prehistory was included (Spa 9). This leads back to the original question of whether or not we can classify the texts chosen as belonging to the same genre. If the texts are looked at in terms of content, audience and purpose, I feel that they can be classified as such. The content in both instances is historical events; the audience is 16-20 year old students, and the purpose is instructional. This still leaves open one final question, which is whether or not the purposes of instruction are the same across the two countries. However, purposes of instruction can vary widely among academic institutions, publishing companies, and individuals, as well as between countries. For this reason, it will be assumed here that the texts chosen belong to the same genre. In addition to content, audience, and purpose, the texts chosen are similar in another way: they all belong to the "grand narrative" tradition of history textbook writing. These texts give emphasis to: "...cause-and-effect links between events; notions of continuing, unending development over time; unitary, en masse explanatory principles (i.e. recurrent themes or motifs in history) and the segmentation of chronological time into historical periods" (Veel and Coffin, 1996: 196-197). Many of the texts include some primary sources, either at the end of sections or chapters, or as a separate part of the textbook itself, and/or they include a bibliography containing further references, including original sources. The Chambers text acknowledges another type of history, termed "social history", which 26
focuses on "more than chronology, more than kings and battles" (Chambers, et.al., xiii). This social history seeks to focus on life in all of its aspects, especially on the experiences of ordinary people, in order to compensate for the focus on the rich, powerful, and educated, as in traditional history. They, then, include short essays on daily life in the chapters. However, the focus on the whole across the texts is on interpretation of what have traditionally been considered "major" historical events as they have occurred chronologically. The importance of interpretation is echoed in several of the texts: ...La Historia cobra un valor trascendental cuando los hechos de los hombres son interpretados y se les busca el sentido que en conjunto tienen, investigando el porquй del suceder de los acontecimientos. (Ballesteros and Alborg, 1973: 7) The tradition of the introductory course in European history (and our cultural tradition as well) is recognized by keeping the book's chapters essentially chronological in sequence, sometimes using groups of chapters to cover a particular period of Western history. At the same time each chapter is presented as an interpretative essay, introducing a set of historical problems important to the understanding of the period treated. (Chambers, et.al, 1991, xiv) The table below lists the texts and provides information on authors, subject of the chapter, number of clauses, and number of words, with the text reference to be used for examples in the rest of this paper in the first column. (Full biographical information is given in the references section.) 27
Table 1.1: Information on Texts Analyzed
Text ref. Year Author
Subject
Eng 1
1991 Chambers, et.al
Introduction to the study of history
Eng 2
1991 Chambers, et.al
Prehistory
Eng 3
1991 Chambers, et.al
14th & 15th centuries
Eng 4
1992 Greer & Lewis
Preface
Eng 5
1992 Greer & Lewis
Prehistory
Eng 6
1996 McKay, et.al
13th & 14th centuries
Eng 7
1997 Perry
Ancient Greece
Eng 8
1981 Roy
Ancient Rome
Eng 9
1995 McKay, et.al
19th century population
Eng 10 1976 Brinton, et.al
18th century
Spa 1
1973 Ballesteros & Alborg Introduction to the study of history
Spa 2
1986 Fernбndez
19th century population
Spa 3
1973 Ballesteros & Alborg 12th cent France & 12th & 13th cent Spain
Spa 4
1991 Mangas
Ancient Greece
Spa 5
1988 Palacio, et.al
14th & 15th centuries
Spa 6
1986 Santos
Ancient Rome
Spa 7
1983 Domнnguez
18th century
Spa 8
1986 Fernбndez
Preface
Spa 9
1989 Cerdб, et.al
Prehistory
Spa 10 1988 Palacio, et.al
18th century
Clauses Words 135 2880 133 2318 135 2218 34 746 165 3202 208 3216 181 3003 170 3330 135 2323 200 3625 133 3117 149 2538 165 3259 153 3385 122 3373 143 2761 164 3455 28 719 103 3350 103 2660
It is interesting to note from the Table 1.1 one difference between the two corpora,
especially in light of the discussion to follow in Chapter 2. This is with reference to the
number of clauses and the number of words. The difference can be seen more clearly in
Table 1.2:
Table 1.2: Clause and Word Counts
Text No. of Clauses No. of Words Average no. of
Text
Words per Clause
No. of Clauses No. of Words
Average no. of Words per Clause
Eng 1
135
2880
21.33 Spa 1
133
3117
23.44
Eng 2
133
2318
17.43 Spa 2
149
2538
17.03
Eng 3
135
2218
16.43 Spa 3
165
3259
19.75
Eng 4
34
746
21.94 Spa 4
153
3385
22.12
Eng 5
165
3202
19.41 Spa 5
122
3373
27.65
Eng 6
208
3216
15.46 Spa 6
143
2761
19.31
Eng 7
181
3003
16.59 Spa 7
164
3455
21.07
Eng 8
170
3330
19.59 Spa 8
28
719
25.68
Eng 9
135
2323
17.21 Spa 9
103
3350
32.52
Eng 10
200
3625
18.13 Spa 10
103
2660
25.83
Total
1496
26861
17.96 Total
1263 28617
22.66
As will be seen in Chapter 2, a number of contrastive analysis studies of Spanish and English also find this difference. What I have not been able to discover from other studies is
28
an explanation for the longer clauses, i.e. whether in Spanish there are more grammatical words, or greater adjective use, or whatever an explanation might be. It does not enter the scope of this study either to pursue this difference. 1.5 Analytical Tool and Methodology Enkvist (1984: 45) states: It is by now a truism that all contrastive linguistics must be based on some theory and model of linguistic description. Without such a foundation we simply cannot compare and contrast languages in a systematic way. Indeed, it is hard to imagine how one could go about comparing two vast areas, such as two different languages, without some basis in mind as to how to go about doing so. However, the use of "theory" or "model" is not without problem, especially for researchers in this day in age, when, among other concerns, the structure of theories and the relationship between theory and evidence, e.g. the question of whether theories are generated from, or determined by, evidence, have come under increasing scrutiny. Phillips (1987: 12) explains the current debate on theory as owing to the conception that whatever evidence is available, a variety of theories can exist that are compatible with it; furthermore, as new evidence accumulates, there are a variety of ways in which every one of these competing theories could be adjusted in order to take account of the new material - no specific change in any theory is necessitated by new evidence, all that new evidence necessitates is that some accommodation be made somewhere. As will be seen later in Section 3, the Theme/Rheme construct fits into this description perhaps too comfortably, although it is not a theory as such. It is embedded in systemic theory, which is "a theory of meaning as choice, by which a language, or any other semiotic system, is interpreted as networks of interlocking options" (Halliday, 1994: xiv). Also of central importance in Halliday's theory of language use is the goal of explaining the relationship between the options displayed in a text to the contexts of situation and culture (Malinowski, 1923), in other words "how the linguistic features of a text relate systematically to the features of its environment" (Halliday, 1985a: xv-xvi). Thus, Theme is one of the textual choices the language producer makes at the clause level, as the options in thematization are realized (in English) by "the sequence of elements in the clause" (Halliday, 1967b). Phillips (1987) points to further problems with regard to cross-cultural studies, based on emic and etic distinctions. Traces of this can be seen in the literature on Theme/Rheme studies; for example, Hakulinen (1989) finds a strictly positional delineation 29
of Theme problematic, and argues that "the notion of theme ...seems...still to suffer from an Anglo-Saxon bias" (56). Indeed, Halliday's systemic theory is a theory based on English. He does suggest that in Japanese, Theme is not realized through position, but through the postposed particle wa (Halliday, 1994: 37). However, Spanish has no such particle, and throws up many of the grammatical constructions which Hakulinen describes for Finnish, such as verb-initial constructions due to inversion, impersonal expressions and so on, as being problematic for a Theme/Rheme analysis. There are several positions from which to defend the use of the Theme/Rheme construct as an analytical tool for a cross-language study of extended written texts. First of all, in spite of the existing confusion as to terminology and concepts in Theme/Rheme analysis, some serious research has been carried out in which there is a large degree of consensus among researchers, and which has offered interesting insights into the notion of text, insights which again have been corroborated by work done by different researchers working on unrelated projects, in different languages and fields, and in different areas of the world in the same way it is used in this study (cf. Fries, 1983, 1992, 1994, 1995; Francis, 1989, 1990; Francis and Kramer-Dahl, Taylor, 1983; Dubois, 1987; Vande Kopple, 1991; Bдcklund, 1992; Ventola, 1995; Martin, 1995; Cummings, 1995; Ventola and Mauranen, 1991; Mauranen, 1993a, b and c; Whittaker, 1995; Ghadessy, 1995; Alonso Belmonte, 1997; inter alia). This would not be the case if individual researchers were constantly accommodating the theory, or in this case, the Theme/Rheme construct, to fit new data. Secondly, this study is based on a relatively large corpus for the type of analysis it entails. Some problems in application of the construct arise in both languages, in English as well as in Spanish, while at the same time the texts in both languages fit the categories provided by the construct rather well, despite differences in word order between English and Spanish. Mauranen (1993a: 5) uses the notion of Theme/Rheme in her contrastive study while being perfectly aware of problems in ethnocentrism in contrastive text research. Cummings (1995) points out that, while the notion of Theme is presented as universal, its realizational statements are presented as descriptive of modern English. He applies a Hallidayan thematic analysis to old English, as he feels it is useful to test the model to see how far its utility extends beyond English. This aim is also kept in mind during this study. A further consideration of Theme is related to genre analysis. It has been seen, above, that Theme has provided interesting insights in establishing similarities within and differences between genres. This study means to add to that literature in an in-depth way. Nwogu (1990) opts for an eclectic approach in his genre analysis of medical texts and their 30
popularized accounts through his study of schematic, thematic and cohesive elements in the texts analyzed. He feels that this is necessary to avoid partial descriptions of genres, which could lead to confusion with regard to genre differentiation and classification, and states that a meaningful characterization of genre "necessitates the isolation of patterns of discourse structure which in their totality reflect the communicative potential of a genre" (ibid: 355). Thus, the question here is whether the focus on Theme allows for a reflection of the communicative potential of the history textbooks analyzed in this study. Bell (1991: 123) explains that in order to comprehend what it means to understand a text, "i.e. what the text `means' and how the reader gains access to it", what is necessary is a "multilevel approach which treats the text as the product of at least three types of choice which express different kinds of meaning, reflected in the content, purpose and organization of the text". These he relates to the ideational2 metafunction, in its expression of cognitive meaning, the interpersonal, in its expression of the speech functional meaning, i.e. its expression of the speaker's relationship to others, and the textual, in its expression of discoursal meaning, which includes its function of ordering clauses cohesively. In the present study, the Themes and Rhemes of the clauses (for the basic unit of analysis, see section 3.4) in the texts are analyzed for their ideational, textual and interpersonal content. Thus it is an analysis of the clause in its complete textual configuration for three major types of meaning. The Themes and the Rhemes are analyzed in terms of their ideational content (Chapter 4), their role in the unfolding of the text (Chapter 5), and their interpersonal significance (Chapter 6). Thematic structure "gives the clause its character as a message" (Halliday, 1994: 379), and "it is only because we can select the desired form of the message that we can also use language effectively, both to represent an experience (experiential) and to interact (interpersonal) with those around us" (Gosden, 1996: 13). Thus, a focus on Theme/Rheme does seem to allow for a composite picture of the texts in terms of their communicative potential in the context and situation for which they were written. This is not to dismiss the contribution of other textual features to the characterization of genre (Gosden, 1996). A study which looked at how textbooks are divided into sections would provide a different view of the texts; at the same time, a cursory glance at the corpus shows great variety in division of chapters and sections, and, at the time of writing this, I am aware of no study which looks into conventions of overall layout 2 Halliday (1994) explains that the ideational metafunction is comprised by two functions: the experiential and the logical. The experiential is concerned with "meaning as organization of experience" (ibid: 179), while the logical is concerned with logical relationships between ideas. The main concern of this study within the ideational metafunction is the experiential. 31
in textbooks to allow for further comparison. Thus, given the groundwork laid by others in connecting Theme with genre, and given the potential for analysis of the three linguistic metafunctions in the clauses, it seems that the Theme/Rheme construct is an optimal lens with which to scrutinize the texts. Furthermore, it is not in the purpose of this study to examine all of the characteristics necessary to distinguish a history textbook genre from other genres, but to examine significant textual choices in relationship to their contextual configuration. The overall methodology of this study is quantitative, in that things are counted. However, it is not a statistical analysis; the quantitative data is used as a basis for interpretation, to link the textual elements with the context and situation of the text. It is based on Hallidayan grammar, generally on the notions of Theme/Rheme, and within Theme/Rheme, on notions of transitivity, modality, and textuality. This calls for interpretation on the part of the analyst, as textual elements are divided into categories, e.g. as being a certain type of participant in a certain type of process, and so on. While for the bulk of the data, this can be a fairly straightforward procedure, it may be thought of as quite subjective at times, given that there is no one-to-one form/function correspondence in language. Throughout the analyses chapters, these "gray areas" of analysis are discussed. 1.6 Summary and Statement of Hypothesis It has so far been stated that there is a place for a study such as this one, which analyzes Theme and Rheme in American and Spanish history textbooks. It can contribute to systemic-functional linguistics, specifically Theme-based genre studies, and to genre analysis, and contrastive rhetoric. It has been assumed, based on the perceived similarities in the communicative function and purpose of the history textbooks, that they belong to the same genre, that of history textbooks. It has been seen that texts belonging to the same genre present a similar contextual configuration; in other words, they will show similarities in linguistic choices related to the field, mode and tenor of the text. Furthermore, it has been stated that an analysis of the ideational, interpersonal, and textual choices in the Themes and the Rhemes of the clauses will provide a composite representation of their communicative potential, which can then be interpreted with respect to the context of situation for which they were written. It has also been assumed that genre subsumes rhetoric in terms of the linguistic choices the writer makes, especially in the case of more information-oriented genres. A number of questions which this study proposes to address can be raised based on the discussion so far in this chapter: 32
1. Given that Halliday's discussion of Theme is linked specifically to English, can a Hallidayan notion of Theme be applied to Spanish? Does it allow for a comparison of equal elements across the two languages? These questions are procedural, in that they deal with the application of the construct to the data. The rest of the questions are related to the results of the application of the analytical tool. Obviously, however, the first set of questions is an important one to keep in mind, as the validity of the results rests on the validity of the tool used for comparison. 2. Are there similarities between the Spanish and American history textbook corpora with regard to Theme and Rheme choices and thematic patterning which can be attributed to generic considerations, i.e. to the contextual configuration of the field, mode, and tenor of the texts? 3. If there are differences, to what can these differences be attributed? This question recognizes that there will be differences which can be attributed to differences in the grammatical nature of the two languages. For example, Spanish allows for verb-initial clauses due to its VSO and pro-drop nature. Thus, the question here refers to differences which cannot be explained by looking at the grammar of the language, but which, like question number 1, lead the analyst to the context and situation of the texts, and which thus tie the texts to the audience, purpose and content. 4. If there are more similarities than differences, can it be said that the texts belong to two different genres? Or can a cross-cultural history textbook genre be posited and, thus, differences be explained by a difference in rhetoric? Is it simply impossible to talk about generic constraints across two cultures, where the field may be construed differently, and the relationship between writer and reader may be constructed differently? While the above questions guide the study, the basic overall hypothesis for this study, based on a theory of genre which posits that texts are products of social processes, and that texts having similar audiences, purposes and content will display similarities in linguistic choices, is, that, barring grammatical constraints, history textbooks will display similarities with respect to thematic and rhematic content, i.e. choices from the textual, interpersonal and ideational metafunctions, and thematic progression patterns across languages, specifically here across English language American history textbooks and Spanish language history textbooks published in Spain. 1.7 Design of the Study Chapter 2 provides a background for this study by rooting it in comparative linguistics generally, and more specifically in studies in contrastive rhetoric and in genre comparison using the Theme/Rheme construct. Chapter 3 gives a comprehensive review of the notions of Theme and Rheme, and provides the definition of Theme followed in this study, as well as an explanation of Theme/Rheme in Spanish and English. Chapter 4 centers on the relationship between Theme/Rheme and the ideational metafunction, and looks at the 33
transitivity roles represented in the Themes and the Rhemes of the clauses, and provides comparative results from the two corpora. Chapter 5 also provides results, in this case centered on the textual metafunction: specifically on thematic progression chains and textual Themes. Chapter 6 is related to the interpersonal metafunction, and explores the use of interpersonal Themes, overt addresses to the reader, and the use of questions across the two corpora, again with quantitative results. Chapter 7 provides overall conclusions for the study. 34
2. Comparative Studies: Contrastive Rhetoric and Genre Comparison The purpose of this chapter is to provide the background in which this study is rooted. The focus is on studies in contrastive rhetoric of English and Spanish, as well as on genre studies carried out using the Theme/Rheme construct. However, before turning to contrastive rhetoric, a brief review of different areas of language comparison is given. 2.1 Language Comparison 2.1.1 Comparative Linguistics Comparative linguistics, the aim of which is to establish genetic affiliations among languages, was a major current of linguistic study in the nineteenth century, and indeed came to be considered a respectable academic subject by the end of the same (Hartmann, 1980; Mathesius, 1975). Linguists in this field have carried out detailed comparisons of identical or similar features in phonology, grammar, and lexicology among different groups of languages, looking at historical changes and at the possibility of determining characteristics of an earlier common stage, with the aim of reconstructing a proto-language. 2.1.2 Comparative Typological Studies Studies in language typologies also had their start in the early 19th century. The purpose of these studies is to classify languages in a more synchronic fashion, according to the presence or absence of certain features. In this century, language typology is most frequently associated with Greenberg (1963), who proposed that there are three common word order types (SVO, VSO, SOV) in the world languages, and that these carried with them a number of "implicational universals" (e.g. SOV languages have postpositions rather than prepositions). English is classified as an SVO language, as is Spanish (Teskey, 1976); however, the latter frequently allows VSO, as it has more flexibility in word-order. 2.1.3 Contrastive Analysis Also in the twentieth century, the `diachronic' approach of comparative linguists came under criticism by those who felt that a `synchronic' approach was needed. Structuralists especially held the view that, prior to establishing similarities in isolated features across languages, what was needed was a detailed description of the patterns of the whole system of the language in itself. American structural linguists provided the model of 35
language used initially in the field of Contrastive Analysis (CA). In spite of the stress by structuralists on the lack of effective comparison among languages, as it is not possible to set up a system of classification that would fit all languages (Bloomfield, 1933, in Ellis, 1985: 25), the need starting in the 1940's to teach languages in the most efficient way possible caused this to be ignored in the large number of contrastive studies carried out. These studies were motivated by the consideration that, by taking into account the similarities and differences between the learners' native language and the language to be learned, teachers would be able to predict areas where the students would experience problems or difficulty and act accordingly, providing a more effective pedagogy. C.C. Fries, perhaps one of the most important advocates of contrastive studies during that time, states that "it is not enough to have teaching materials based upon a descriptive analysis of the language to be learned" (Fries, C.C., 1942a: 279); also necessary is a descriptive analysis of the native language, and the two must then be "systematically compared" (ibid). An example of this is his course in English for native speakers of Spanish (Fries, C.C., 1942b). Indeed, Krzeszowski (1990) states that throughout history, the emphasis has been to contrast languages for pedagogical purposes. Well known contrastive analytical studies of English and Spanish include Stockwell & Bowen (1965) & Stockwell, Bowen & Martin (1965). Their analysis was in many ways more sophisticated than other CAs in that they distinguish between structural and functional/semantic correspondence (Larsen-Freeman & Long, 1991). Also, Stockwell, Bowen and Martin did not go along with other analysts' belief of the greater the difference, the greater the difficulty, especially in new or missing categories of language structures. This position was born out in later studies of English speakers learning French (Buteau, 1970). In fact, this is one criticism leveled at CA: "At best CA can only predict some (italics original) of the difficulties a learner will have, and it does so in an unhappily vague way at that." (Oller, 1971, in Taylor, 1979: 8). This led Taylor to claim that "[c]ontrastive analysis itself is only barely alive and definitely not well under that name in applied linguistics today." (Taylor, 1979: 11). However, CA has a more modern advocate in Carl James, who feels that CA is alive and well, in spite of belief to the contrary on the part of many (James, 1994). Also, in 1980, just a year after Taylor made his claim, James, while noting the insecurity plaguing CA at the time, also noted that it was still "highly vigorous, to judge from the large number of practitioners it enlists" (James, 1980: 166). One such practitioner from the 1980's is Whitley 36
(Whitley, 1986), who, acknowledging the limits of CA in pedagogy, advocates a CA approach in his book on Spanish and English. He goes beyond some of the earlier studies in that he includes some mention of pragmatic/functional aspects of language, such as the Theme/Rheme distinction and its manifestation in Spanish and English. James (1980) also extends the discussion of CA from the traditional focus on phonology, morphology and syntax, or microlinguistic CA, to macrolinguistic CA, including contrastive textual analysis and conversation analysis. He points out that at the time of writing, macrolinguistic CA "is a relatively new field of enquiry, awaiting exploration" (James, 1980: 140). Before moving on to this level of contrastive studies, a brief mention is made of other studies at the micro level. 2.1.4 SLA: Universal Grammar and Typological Studies Research in the field of Second Language Acquisition (SLA) based on the comparison/contrast of languages through the study of linguistic universals proliferate. As mentioned above, not all differences between languages point to difficulty for learners; what may cause difficulty or not is if the property in question is a universal property or a language specific one (Ellis, 1985). In the field of SLA, two different contrastive study approaches have been used: Universal Grammar, in the Chomskyan tradition, and typological studies. Gregg (1989: 30) states that "a generative theory of grammar is a necessary component of a theory of second language acquisition". Liceras (1989) points out that the parameterized model proposed by Chomsky (1981, 1982) has provided the theoretical framework for a number of studies in SLA (Flynn, 1987; Phinny, 1987; White, 1988; see also Flynn, 1989; White, 1989;). These studies differ from the spate of contrastive studies of the 1960's in that hypotheses are formed as to difficulties learners will face due to differences in the setting (or resetting) of parameters, and then the hypotheses are tested by examining learners' interlanguage. At the same time, interlanguage studies also serve to confirm or adjust the theoretical model proposed by Universal Grammar (de Miguel, 1988) Typological studies in SLA have been carried out in a similar fashion; examples of typological studies in SLA include those related to topic-prominent vs. subject-prominent languages (Jin, 1994; Fuller & Gundel, 1987; Huebner, 1983; Rutherford, 1983). The former are characterized by invariant, overt marking of topic, in which semantic roles (such as Subject) play no role in controlling word order, which is pragmatically constrained; in the latter, word order is syntactically constrained, the Subject playing a major role. Languages 37
have been described as being more or less topic-prominent: English is categorized as a nontopic prominent language and Spanish is intermediate in terms of topic-prominence (Fuller & Gundel, 1987). It is hypothesized in typological studies that learners will have less difficulty with areas of the L2 that are typologically similar to their L1. 2.1.5 Other Studies Many contrastive studies have been carried out not with the purpose of pedagogy or SLA in mind, but with the intent of discovering the motivation behind why languages are the way they are. For example, Barcelona Sanchez (1986, 1990) attempts to provide a pragmatic motivation for phenomena such as raising to subject or raising to object, for which he finds a transformational explanation inadequate. He does this by citing examples from both English and Spanish. Thompson's (1978) paper is a descriptive typological study based on the way languages differ in how they utilize word order; English uses word order to indicate grammatical relationships, and languages like Spanish, use it to indicate pragmatic relationships, such as Theme-Rheme distinctions. Also, based on Yagua language, Payne (1990) discusses throughout her book verb initial languages, and concludes that we can broadly distinguish among (a)languages in which word order is primarily used for syntactic functions such as signalling grammatical relations, (b)languages in which order is primarily used to signal the discoursepragmatic status of information (e.g. identifiability, focus of contrast, etc.) and (c) languages in which order displays a good mixture of syntactic and discourse-pragmatic functions. (25) and goes on to point out that languages could be placed on a continuum from almost fully syntactic to almost fully pragmatic. 2.2 Contrastive Rhetoric 2.2.1 Contrastive Rhetoric Explained As manifested by the above discussion, the majority of contrastive studies have been done at the clause level or below: phonetics and syntax making up the major focus of study. Kaplan (1966) pointed out that this focus was also the trend in language teaching, but he also held that students who master syntactic structures of a second language often cannot write what would be considered an adequate composition in that language: "The foreignstudent paper is out of focus because the foreign student is employing a rhetoric and a 38
sequence of thought which violate the expectations of the native reader" (Kaplan, 1966: 4). In this seminal article, Kaplan then explains how an English expository paragraph reflects the cultural thought patterns which readers of English expect: it is dominantly linear in its development....usually begins with a topic statement, and then, by a series of subdivision of that topic statement, each supported by example and illustrations, proceeds to develop that central idea and relate that idea to all the other ideas in the whole essay, and to employ that idea in its proper relationship with the other ideas to prove something, or perhaps to argue something. (Kaplan, 1966: 4-5) He then illustrates this with an example. He compares this typical development with several hundred student compositions written in English by speakers of various languages, as well as some professional translations. He concludes that the students employ text patterns which differ from English in their development. He does this by examining, for example, parallel constructions, which he finds typical of Semitic languages, and which make use of coordinated structures (while English prefers a greater degree of subordination over coordination). This may result in the tendency for Arabic speakers to over-use coordinating elements in their writing. Oriental writing, Kaplan notes, is often quite indirect, writing around the subject, not directly on it. He arrives at this conclusion by looking at how sentences are related to the topic and to each other in the sample compositions. He characterizes Romance language development of texts as allowing for greater digression than English, looking at how the ideas included are related to the overall topic of the text. Purves (1988) explains further the reasoning behind contrastive rhetoric by comparing it to interdisciplinary differences: If such differences [e.g. reference placement and use, use of first person and the passive, and use of interpretation or inference] divide academic disciplines and even permeate these disciplines to the extent that the conventions obtain regardless of the language in which the article is written, it would seem reasonable to expect that similar kinds of differences separate ethnolinguistic or geographically distant cultures and societies. (ibid: 14) Also according to Purves, there are basically two methodologies used in contrastive rhetoric studies. One is to isolate certain linguistic features of texts, while the other is to look at the criteria used by different raters of written texts. The purpose behind these studies is quite similar to the purpose behind CA: contrastive rhetoric is presented as a tool to help students who are in the process of learning a second language to better understand its paragraph patterns. Given the focus in this thesis on Theme as a feature of text, included 39
here is a sampling of the types of studies which have been carried out in contrastive rhetoric in Spanish and English in terms of linguistic features in order to give an idea of the range of text analysis tools used. 2.2.2 Studies in Contrastive Rhetoric: English and Spanish Grabe (1987) proposes in his research to define text comparability in two or more languages by presenting a methodology for defining expository prose. He does this by comparing text types within one language: English. The variables used in the study are divided into two groups: syntactic (e.g. prepositions, nominalizations, pronouns, etc.) and cohesion (reference, repetition, etc.). To discover distinctions between text types within expository prose, two types of analyses were carried out on the texts: factor analysis and cluster analysis. The results showed that certain surface linguistic features were indeed more prevalent for different text types. He suggests that this type of multivariate/multidimensional approach could be carried out in order to discover similar text types across languages, especially in edited texts of different text types. While not following Grabe's recommendation of analyzing and comparing edited texts across languages, a number of researchers have applied multidimensional/multivariate analyses to student writing in Spanish. For example, Lux (1993) compares essays written in response to a prompt across four different groups: Anglo-Americans writing in English, Anglo-Americans writing in Spanish, Latin-Americans writing in English, and Latin Americans writing in Spanish. In his initial discussion, Lux distinguishes between linguistic features, which form the code, and include such items as word endings and grammatical constructions, and rhetorical features, which include organizational patterns and argumentative appeals. A central question in his paper is "whether syntactic and cohesive features in two languages can be interpreted in such a way as to serve as reliable parameters for measurement of a universal discourse style" (ibid: 6). Lux chose 19 of Grabe's (1987) 33 text features and added 3 variables of his own, and thus analyzed his corpus for a total of 22 features (i.e. prepositions, nominalizations, 1st and 2nd person pronouns, 3rd person pronouns, past tense, adjective clauses, noun clauses, locatives, repetition, hedges, present tense, words per sentence, infinitives, adverbial subordinators, passive voice, questions, conjunctive adverbs, coordinators, and wh-clauses). He then determined English/Spanish correspondences with regards to these features, having as a goal defining the text variables (which originally had been defined in English) in Spanish which would "postulate a level at 40
which a text feature could be said to function as passive voice, pronoun reference, [emphasis original] etc., while at the same time allowing for a number of surface forms in the two languages" (ibid: 14). Lux hoped that this would "ignore the purely linguistic differences between the texts but reveal any rhetorical difference" (ibid: 14-15), and often required defining the text feature at a more general level than in a monolingual study. He then applied a factor analysis statistical procedure in order to find co-variance among the text features. The results of this showed that the ways in which variables were grouped was similar to factor scales of previous studies, even in spite of the difference in text types (bilingual/monolingual, professional/non-professional, L1/L2). In fact, four identifiable factors resulted: elaborated/reduced, informational/interactional, abstract/conversational, narrative/non-narrative. Also, Lux carried out ANOVA measures, which suggest that some of the variance shown by the individual variables and factors was linguistic, some cultural, and some based on L1/L2 proficiency differences among writers: L1 writers used more grammatically elaborated prose, although elaboration was also a characteristic of Latin American writers regardless of the language of the task; L2 writers were more `interactional'; and Anglo-Americans tended more towards conversational language This last feature is important here as the text features to which the abstract, formal tone were attributed (prepositions and nominalizations) seem to be more prevalent in Spanish. Lux's conclusions based on this study deal mainly with the practical application of this type of analysis to a different language, not to a comparison between the languages. But he does suggest that there exist a good number of structures in both languages which seem to serve similar functions (e.g. first and second person pronouns and passive voice). Reid (1990) also follows a large-scale study format, as she finds the possibility of multidimensional statistical approaches to provide less contradictory results than discourse analysis studies using laborious handcounts of linguistic features. The main purpose of her study was to demonstrate the effect of topic type (prompt) on written response. The variables Reid uses in her study are divided into three categories: fluency (number of words), syntactic (average sentence length, short sentences, complex sentences, and passive-voice verbs), and lexical (average word length, content words, and pronouns). The four language backgrounds represented in the study are: Arabic, Chinese, Spanish, and English. While the major interest of the study is in the differences between topic tasks, and not amongst cultures, Reid did find one syntactic variable for which there was a significant 41
difference: passive-voice verbs: the Arabic and Chinese writers used more passive-voice verbs for one topic type than for another, which was not the case for the other language groups. Reid also reports on a previous study (1988) of the same corpora, in which she examined the co-occurrence of language features, where she found that native speakers of English used few coordinate conjunctions, a high percentage of passive voice verbs, and a high percentage of prepositions, while Arabic and Spanish writers used significantly more coordinate conjunctions, and significantly fewer prepositions (which is in contrast to Lux's study) and passive-voice verbs. Montaсo-Harmon (1991) looks into problems students may have in writing, which she contributes to conflicting discourse patterns, or "the organization and development of text via the logical arrangement of ideas" (ibid: 417). She analyzed and compared compositions written by 14-15 year old students from four linguistic groups: 1)Mexican students in Mexico writing in Spanish; 2)ESL students (native speakers of Mexican Spanish) in the U.S. writing in English; 3)Mexican-American/Chicano students (dominant speakers of English) writing in English; and 4)Anglo-American students writing in English. A statistical analysis of the following discourse features was then carried out: sentence length; types of sentences (fragments, run-ons, simple, compound, complex, compoundcomplex); lexical cohesion; syntactic cohesion (reference, substitution, ellipsis, conjunction); and coherence (topic sentence, enumerative, additive, summative, resultative, explicative, illustrative, contrastive, transitional words and phrases). Montaсo-Harmon found that Mexican-Spanish students wrote longer sentences, and employed more run-ons than their Anglo-American counterparts, who employed more simple sentences consisting of SVO and often used because subordinate clauses. As to lexical cohesion, the Mexican students relied heavily on synonyms given within commas after a given word, often adding quite a few and becoming more and more "fancy" or formal. Anglo-American writers depended more on lexical chains via repetition of same word or on collocations. While there did not seem to be much difference in the use of syntactical cohesion, there were some differences in the patterning used in the texts. The rhetorical pattern used most often by Anglo-American students was enumeration (through use of connectors such as "first", "second", "then", and "finally", either to denote chronological or spatial order, or emphatic order). In fact, according to Montaсo-Harmon, all the Anglo-Americans used it at some point, while it was not used in any of the Mexican compositions. The latter tended to organize their texts 42
through additive or explicative relationships, and they also had more deviations, or more complete breaks from one idea to the next. These seemed to be conscious deviations, as they often used transitional words or phrases to return to the previous idea (e.g. Pero me he salido del tema. Volvamos a lo que habнa dicho antes..). Anglo-Americans did not have as many deviations, and these tended to be unconscious, i.e. the writers did not refer to them in a meta-textual way. Montaсo-Harmon also mentions three other studies which contrast rhetorical patterns: Santiago (1971), Strei (1971), and Santana-Seda (1975). These show "striking" differences in the organization of written discourse in texts written in Puerto Rican Spanish and English; compositions written in Puerto Rican Spanish had significantly higher proportions of coordinate structures, nonsequential sentences, additive constructions, and one-and two sentence paragraphs. Pak (1995) summarizes a number of Spanish and English contrastive studies and finds differences with respect to sentence length, number of sentences per paragraph, length of introductions, digression, and style. This corroborates Kaplan's (see above) depiction of Romance language with respect to digression, and Montaсo-Harmon's (see above) finding with regard to sentence length. Cook (1988), on the other hand, in a study focusing on paragraph structure and overall organization of content, found no significant differences in texts written in both English and Spanish by Spanish-speaking graduate students taking ESL. In a study on cohesion and coherence in native English speaker writing in comparison with Japanese and Spanish college-level ESL students, Connor (1984) found that general cohesion density was not perceived as problematic in the non-native vs. native speaker writing. However, she did find that the ESL writers were lacking in the variety of cohesive devices which the native speakers had at their disposal. McCabe-Hidalgo and Gomis (1994) carried out a small scale study in order to determine what is valued by Spanish teachers in student (secondary age level) writing, and compared it with what Anglophone teachers valued. This was done by choosing two texts from a sample of student writing in Spanish; one contains long sentences with heavy adjectival and adverbial modification, great variety of lexis, and cultural references which the reader is expected to know. The other is simpler in lexis, and includes same-word repetition, has some errors of spelling and punctuation, and uses enumeration to guide the reader, and a transition to the conclusion. They asked a number of Spanish, British and 43
American language teachers to choose which text they preferred and to comment on why. The Spanish consistently chose the first text when asked which text they found more valuable. They wrote comments such as "good expression" and "original", while with the second, they expressed that the writer had "serious problems with punctuation and a lack of vocabulary". Others dismissed it as "poorly written" or "messy". With the American and British teachers, the comments were almost reversed. While some found the first text interesting in a "literary way", many dismissed it as "wordy" and pretentious. Most felt that the second text was much better organized and easier to follow. McCabe-Hidalgo and Gomis show that in Spain, writing instruction in schools focuses on syntax and lexis, and that Spanish society values language for its own sake: words are used to decorate and to fill space, and are rarely considered superfluous. There may be some influence here from Arab cultures, where "the beauty of form is valued more highly than content" (Ostler, 1987). Also along the lines of student writing, but in an experimental, rather than a textlinguistic, approach, Connor & McCagg (1983) carry out an analysis of text recall through paraphrase. The text analysis applied to both original text and to written paraphrases by both native and nonnative English speakers is based on Meyer (1975): texts are broken down into clauses, and each clause assigned a rhetorical function based on its role in conveying the overall meaning of the text. The paraphrase task was carried out by eleven native speakers of English, eleven native-Japanese speakers, and eleven native-Spanish speakers. The task involved the students in reading a text in English, after which they were asked to write down as much as they could remember. The authors conclude that no indication of the transfer of culture-specific rhetorical patterns as observed by Kaplan occurs in text reconstruction; this may be explained by the seeming constraint on non-native speakers by the original text structure. A difference they did find involved the greater amount of recall of subordinate ideas by English speakers who included more detail to support ideas. Moving back to text-linguistic based studies, but this time of professional, rather than student, texts, studies of academic articles across English and Spanish include ValeroGarcйs (1996), who analyzes metatext in economics texts written in English by native and Spanish speakers. She looks specifically at connectors (e.g. however), reviews, which are summaries or repeats of previous text (e.g. so far we have assumed that, as previously mentioned), previews, which anticipate a later stage of text (e.g. we show below that), and 44
action markers, which are indicators of discourse acts (e.g. the explanation is...., to illustrate this). She considers 4 articles, 2 written by Anglo-American writers in English, and 2 by Spanish writers in English. She finds that the latter use less metatext than the Anglo-American writers, and that there is a major difference in paragraph length. She also finds differences in the ways writers manifest their presence, and the presence of other authors, in the texts. She finds the Spanish writers' texts to be more writer- than readerbased: "texts in Spanish are less designed to be easy reading" and "the reader, not the author, is primarily responsible for effective communication" (Valero-Garcйs, 1996: 289). Much of this seems to support the differences found in studies on student writing mentioned above. Another study by Valero-Garcйs (1997) compares verbal syntagms in economics texts written in Spanish and English. However, the results are presented in a very inconclusive manner: she finds similarities, such as a predominance of active over passive forms, a high percentage of present tense, and almost no presence of continuous form, which she attributes to the genre of the texts (which she confusingly refers to in the conclusion as "scientific texts"). She attributes other similarities to coincidence (e.g. use of present perfect), as their "rhetorical functions" are different, yet their frequency is similar across the two languages. She also finds differences, which she does not attribute to any cause, such as a greater tendency in Spanish towards the avoidance of non-personal verb forms, especially -ndo (-ing) forms, as well as a preference in this language for the reflexive passive and avoidance of the imperative. As the form-function relationships in this study are not clearly established, it is difficult to get a feel for what the genre constraints are, and how they might operate on verbal syntagms. The results of Valero-Garcйs' (1996) study on metatext in economics articles contrasts with that of Moreno (1997), a study of metatextual signals of cause-effect in research articles written in English and Spanish also in the field of economics. Moreno found more areas of similarities than of differences with regard to frequency of explicit signals, and to strategies used to express them, as well as to the distribution of these strategies. Moreno puts forth as a possible explanation for this similarity the fact that Spanish business education is influenced by English language materials and procedures. Lуpez (1982) also questions the notion of contrastive rhetoric in genre specific writing. She found that the rhetorical structure of research report articles in science and 45
technology, in terms of type and amount of information included, and the way that it is organized, was similar across the two languages. She feels that her findings "tend to deny Kaplan's theory of a cultural-bound rhetoric" (ibid: 80), a least for a genre as highlyspecialized as the one examined in her paper. She, like Moreno above, attributes the similarities to the fact that a lot of this information filters through to Spanish-speaking countries through the English language, as well as to "the fact that the activities in the ST field are framed by the rigorous principles of scientific enquiry" which "has to be reflected in a restricted way of expressing and presenting their concepts, notions and rhetorical elements"; thus, "a set of universal conventions for communicating ST are shared by the different cultures" (ibid: 82). She does conjecture, however, that Kaplan's theory may prove true in other more standard, less formalized texts, such as those produced for storytelling and journalism. Thus, for Lуpez, the international ST community sets expectations for texts that override individual groups' cultural expectations. Chartrand and Rising (1995) compare versions of the same company memo in Spanish and English using an analysis based on Rhetorical Structure Theory (Mann, Matthiessen and Thompson, 1992). This analysis showed a similar rhetorical structure across the two versions, except in two points. One was that the English version used more relations of the justification type. When they asked the writer of the Spanish memo about this point, she said that there was no need to explain actions, only to inform. A second point of difference has to do with organization of one of the sections in the text. Yet, overall they conclude that there is a high degree of similarity in rhetorical structure. Thus, they decided to look at syntactic and lexical features with the relational propositions of the text spans. Here, they also found few differences. The English version used first person singular pronouns to address a second person, while the Spanish version used first person plural or impersonal verb forms to carry out the same function. Also, the people mentioned in the memo are used as agents of processes in the English clauses; in the Spanish, they appear as compliments of nominalizations or recipients of actions. For Chartrand and Rising (1995: 35), these differences, along with those of rhetorical structure, show a difference in the reader-writer relationship which stem from differences in the corporate culture. In the Spanish memo, the writer writes from a superior position, and "represents an entity rather than him/herself"; thus, the Spanish memo is "company-centered". This would account for the lack of justification for decisions, and the use of impersonal verb forms, etc. The English 46
memo, they argue, is "employee-centered". This is evidenced by references to employees by first name, first and second person singular pronouns, and the more frequent use of justification moves. In sum, there are few text-based studies, especially of professionally published texts, from which generalizations can be made. Furthermore, some of the studies, especially those on student writing, take as their base Mexican Spanish or Puerto Rican Spanish. If different cultures employ a different rhetoric, it could be argued that Spain, Puerto Rico and Mexico would each produce a rhetoric specific to their culture. However, from the above there do seem to be some characteristics which emerge with respect to differences in Spanish and Anglophone writing. One difference which is repeated in a few of the studies is that longer sentences and more varied paragraph length are both features of Spanish texts, a notion which seems to be supported by the data in this study as seen in Table 1.2 above. Another notion that seems to be repeated is that Spanish writing is perceived as more formal by the analysts. Indeed, these two differences together reflect a difference in the relationship between writers and their readers. The above studies, taken together, seem to hint at a difference between the two languages and the cultures they represent along the lines of Hinds' (1987) typology of reader versus writer responsibility. In writer-responsible prose, the concern of writers is to make themselves understood by using clear, well-organized prose; to illustrate reader-responsible prose, Hinds draws on the work of Suzuki (1975) in putting forth the claim that "Japanese authors do not like to give clarifications or full explanations of their views. They like to give dark hints and to leave them behind nuances" (Hinds, 1987: 145). Hinds cites Suzuki's view of Japanese readers, who "anticipate with pleasure the opportunities that such writing offers them to savor this kind of `mystification' of language" (Suzuki, 1975: 31, in Hinds. 1987). Hinds finds this difference reflected in textual features of unity and coherence, e.g. transition statements are more prominent in English, while in Japanese they are more subtle or absent. Perhaps, then, Spanish might be thought of as somewhere along the cline between English and Japanese with respect to reader/writer-responsible prose. However, this tendency is only tentatively suggested, and it may be counteracted by more universal constraints in certain genres, or by the communicative purposes of particular texts or text types. 47
2.3 Comparative Studies Using Theme/Rheme Analysis 2.3.1 Comparing Genres As mentioned in the previous chapter, a number of studies have been carried out which find correlations between thematic content/thematic progression and genres. Already mentioned there were Taylor (1983), Nwogu (1989), Nwogu & Bloor (1991), Francis (1990), Ghadessy (1995), Lotfipour-Saedi and Rezai-Tajani (1996), and Berry (1987). Other studies include Whittaker (1995), who compares economics and linguistics journal articles in English; she finds similarities across the two in terms of interpersonal and textual Themes, which she attributes to similarity in audience and purpose, that of presenting research to peers. She finds differences, however, in ideational Themes, specifically in those relating to material processes, "that is, differences stemming from field" (ibid: 117). While not exactly a genre study, Francis and Kramer-Dahl (1991) compare a less standard, more popularized narrative in the field of neuropsychology, written by Oliver Sacks, with a case report published in a professional journal. They show that the Theme choices made by Sacks show involvement and concern with the patient, while the professional journal report chooses Themes which are seemingly more remote and objective. Also, Sacks uses more marked Themes (e.g. predicated themes, thematic equatives, fronted Complements), which Francis and Kramer-Dahl point out are more typical of persuasive argumentation. In the professional journal report, however, the vast majority of the Themes coincide with Subject. Francis and Kramer-Dahl (ibid: 354) argue that this is not meant to convey the idea that in this report the author "does not intend to persuade, but rather that he feels constrained to allow the "facts" to speak for themselves. Hence he relies on "standard" information-ordering resources of the language to signal importance and relevance". This study also shows the correspondence between Theme choice and purpose: for Francis and Kramer-Dahl (ibid: 340), Sack's purpose with his texts is not only to popularize; in fact, his "novel texts...are at the same time motivated by epistemological necessity, and represent his attempt to communicate his unusual vision of neuroscientific practice". For Sacks (1985: x) neuropsychology needs to retain its scientific rigor, while it must also "restore the human subject at its centre". Hawes and Thomas (1996) compare editorial articles from The Times, an "ancient and respectable organ of the establishment" (ibid: 159) with The Sun, "know for its 48
sensationalist scandalmongering" (ibid). Included in their study is an analysis of the discourse participants encoded in the Themes of the clauses, marked Themes, and thematic progression. They found that The Sun used many more visible discourse participants (e.g. "The Sun", "We") in Theme position than The Times; a sense of impartiality is conveyed by the latter, and a sense of candidness and outspokenness by the former. They also found differences in pronoun use; they relate this to Bernstein's (1966) restricted code, which employs more pronouns, and elaborated code, which employs more content words. The higher pronoun count in The Sun is closer to a restricted code, which they take as a reflection of its status as a working class paper; on the other hand, the higher lexical content in The Times is attributable to a middle class mode which also has academic overtones. Furthermore, they found more marked Themes in The Sun, and, like Francis and KramerDahl, attribute this to the greater amount of hortatory rhetoric in this newspaper. They also find differences with respect to thematic progression: while both newspapers are similar with respect to simple linear progression, The Sun makes more use of constant topic progression patterns than does The Times. This they find revealing "because this is the type favoured by demagogues, who will hammer their audiences over and over with the same word or phrase" (Hawes and Thomas: 165). The Times uses more derived Theme (where the Themes are cohyponyms, related to an implicit or explicit overall Theme; see section 5.2 below). For Hawes and Thomas (165), derived Theme is "the indirect style of the academic who does not like to be too obvious"; it also assumes of the reader a great deal of background knowledge. 2.3.2 Comparing Languages In addition to the studies mentioned above, in which thematic development or progression is used to highlight genre differences/similarities, the notions of Theme/Rheme have formed the basis of many cross-language comparative studies. In fact, Enkvist (1984) states that "all studies of Theme and Rheme originally started from linguistic comparison and contrast" (54), and cites his own examples from Swedish, which allows object-initial patterns not allowed in English. It needs to be pointed out here that Enkvist conflates the terms "theme" and "topic" (this tendency is discussed in Section 3.1.1. below). (Typological studies relating to topic/comment are discussed in section 2.1.2 above.) Danes (1970) discusses contrastive analysis of Czech and English using Functional Sentence Perspective (Theme/Rheme) and comes to some interesting conclusions regarding 49
differences in the function of word order and intonation related to these concepts; in essence, Czech, with a relatively free word order, has a more uniform intonation pattern, while English, with a relatively fixed (according to the rules of grammar) word order, allows for greater variety in the location of the center of intonation in the clause. A number of studies which use the Theme/Rheme construct are centered on translation and analyses of the effectiveness of translated texts. Ventola (1995) compares source and target texts in German and English and illustrates the problem that ensues when translators do not try to render the Theme/Rheme configuration of the original text: often the flow of the writer's original argument is lost. She thus argues for a more textual approach to translation which would involve overt attention to thematic progression patterns in translation. Vasconcellos' (1985; 1992) main purpose in her study is to show that the Hallidayan notion of Theme is valid across languages, and she uses translation as a tool to prove this. She looks at 32 texts written originally in Portuguese and their translations into English. While Portuguese is a language which allows for VSO constructions (46.2% of the clauses illustrated this construction in her corpus), she shows that translators found ways of getting around this in order to maintain the original thematic flavor of the texts as well as the information structure: in fact, the rate of consistency for maintenance of focus of new information was 87.6%. Munday (1997) looks at Theme choices and thematic progression in a text written by Gabriel Garcнa Mбrquez in Spanish and its translation by Edith Grossman into English. He finds that, when adjusting the analysis for systematic grammatical differences between the two languages, e.g. the prodrop nature of Spanish and the initial "no" in negative constructions in Spanish, the translator retains the thematic structure of the original. Where differences do occur, Munday discusses the dilemma the translator faces: often, in attempting to retain the Theme/Rheme flavor of the original, the translator must "foreignize" the target language text by not conforming to the thematic requirements of the latter. At times this may be necessary, but at other times, especially with the Adjunct shifts he finds in the translation, Munday, like Ventola, laments the alteration of some of the stylistic structures, which can affect the point of view of the narrative. Also of interest here are descriptive corpus-based studies. Aziz (1988) is a study of thematic organization in a narrative paragraph from a novel written in Arabic, the results of which he compares to other studies of thematic organization in English and in French. Aziz 50
bases his delineation of Theme and Rheme on Prague theory, where the Theme is that which contributes less communicative dynamism to the sentence (see section 3.1.3 below). Aziz's results seem to suggest that the structure of Arabic paragraphs is more similar to French, in that both these languages favor a constant Theme organization over a simple linear organization (see section 5.2 below for explanation of these types of organization) and that Arabic allows more freedom than both English and French, as it favors more than one pattern. He bases his comparison on the work done by Newsham (1977, in James, 1980) on paragraphs selected at random from English and French university level textbooks from different disciplines. Newsham found differences across the two languages in terms of preference of thematic patterning: English preferred either the simple linear pattern or paragraphs in which the Themes of subsequent sentences all related to the Rheme of the first sentence (reminiscent of Valero-Garcйs' mention of greater topic sentence inclusion, see section 1.3 above), while French preferred the constant topic pattern or a pattern not used at all in English, that of Rheme-Rheme. This latter finding tentatively suggests that French allows for multi-topic paragraphs, while English does not. Ventola & Mauranen (1991) is a preliminary report of a study on Finnish researchers' academic articles from different fields, written by non-native speakers of English, and their revisions by native speakers. Further data in their study include texts written by the same authors in Finnish. Approaches used in this study are based on genre analysis, contrastive rhetoric, and a semiotic approach to genre, which assumes "that the global generic structures constrain the combinational variable realizations of register, which in turn determine discourse and other linguistic structures in texts" (p. 462). They point out that generic organization in terms of global structures, which plays an important role in text formation, specifically in the way such structures are realized through register and discourse choices, has not played as important a role in early descriptive work on academic writing, which focused on the lexicogrammatical and orthographical strata. The discourse systems chosen for their investigation are connector use, thematic development, and reference, the latter two being of particular interest to this study. They discovered that the Finnish writers writing in English had problems with the use of connectors, which they attribute to the less frequent use of connectors by Finnish writers as compared to native speaker writers. There were differences from native speaker writers with respect to thematic patterns: the Finnish writer texts in English showed less thematic pattern variation than native speaker texts. 51
Also, Finnish writers employed fewer textual Themes, and provided less lexical cohesion between Themes. It is difficult to know whether the perceived differences stem from cultural differences (i.e. L1 transfer) or from the Finnish writers having to write in an L2. This study can be compared with Mauranen (1993), who compares paragraphs taken from academic journal articles in medical science and economics. She analyzes texts written in English by native speakers, texts written in English by native Finnish speakers, and texts written in Finnish by native speakers. Her analysis of Theme and thematic patterns points to no substantial rhetorical differences between Finnish and native-English writers. The main conclusion is that the Finns writing in English were not able to manipulate themes as skillfully as in Finnish, so they are at a rhetorical disadvantage there. While not at the level of genre (see section 1.3, above), Alonso Belmonte (1997) looks at strategies of thematization in Spanish texts of instructional, narrative, descriptive, expository, and argumentative patterns. These are taken from different genres, e.g. the instructional text patterns are taken from instruction booklets and from recipes, the expository text patterns from textbooks and encyclopedias, and so on. She compares her results to those of Lavid Lуpez (1994a, 1994b), who carried out a very similar study in English. Alonso Belmonte concludes that the same thematization strategies are used in Spanish and English text patterns (e.g. time references or verbal process as Theme in instructional texts, time references or participant as Theme in narrative, and so on), while there are differences in the linguistic realization of these strategies, due to the grammatical idiosyncrasies of each of the two languages. She explains that the similarities are the result of the similarity in communicative purpose of the text types. 2.4 Conclusion The above discussion shows that, first of all, the number of studies at the text level between Spanish and English is low. Secondly, it shows that there are differences between Spanish and English at the morpho-syntactic level, which will have an effect on Theme choices, as will be seen throughout this study. These differences include that Spanish allows for VSO as it has a more flexible word order; thus word order in Spanish may be more pragmatically controlled, while in English it is more grammatically controlled. At the level of contrastive rhetoric, results are not conclusive, as different researchers reach often conflicting conclusions. However, there does seem to be some tendency for student writers 52
in Spanish to be more elaborate and to write longer sentences; this could be confirmed in published writing by the longer clauses in the Spanish history textbooks analyzed in this study (see Table 1.2 above). Given that there is some evidence that writers in English provide more metatextual markers for their readers, e.g. markers of enumeration, (although there is some evidence that there is no such difference in, for example, markers of causeeffect), it has very tentatively been suggested that perhaps writing in English is slightly more writer responsible than is writing in Spanish. Student writing in Spanish also is perceived as being more formal in tone; this finds some support in professional texts, i.e. company memos, where the impersonal is used in Spanish in contexts where personal forms are used in English. This point will be touched on again in Chapter 6. Genre studies using the Theme/Rheme construct show that Theme content and thematic progression patterns correlate across texts belonging to the same genre, and that they differ in texts which belong to different genres. However, differences are found across texts purportedly belonging to the same genre. These differences can be explained by differences in how the writer construes the field (e.g. Oliver Sacks' construal of neuropsychology) and by differences in the effect the writer wishes to make on the reader. As these differences are related to field, mode, and tenor, it could effectively be argued, perhaps, that the texts belong to different genres, or that they are different sub-genres of the same genre. With regard to Theme studies across languages, while there is some evidence that differences exist across languages with respect to thematic patterning, it is not clear that these results (e.g. Newsham, 1977) would be replicated in studies comparing same genres. Alonso Belmonte (1997) provides evidence that similar communicative purpose correlates with similar Theme choice in texts written in Spanish and English. 53
3. Theme/Rheme 3.1 Definitions of Theme Theme has been defined in a variety of different ways, and for this reason, some analysts find the dismissal of the Theme/Rheme distinction easily defensible. For example, Stephen Levinson (1983: x) scathingly attributes the exclusion of the concept from his book to the following: "Terminological profusion and confusion, and underlying conceptual vagueness, plague the relevant literature to the point where little may be salvageable." An answer to this will be attempted further on. It is interesting to juxtapose this view of Theme with the fact that it is an analytical tool which has been increasingly gaining in popularity in the last few years, as has been seen in the previous chapter. Mauranen (1993a: 104) attributes this popularity to "its interesting position at the interface of grammar and discourse". The definition of Theme used in this study places Theme at the point where the grammar of the clause meets the surrounding text and also relates to the thought in the speaker's mind. However, before the position is further expounded, a discussion of the different definitions and uses of Theme/Rheme as an analytical tool is warranted before settling on the most optimal definition for the purposes of comparing extended written text of the same genre across two languages. Basically, the different uses can be divided into four categories, based on different definitions of the Theme itself: Theme as topic, Theme as having the least degree of communicative dynamism, Theme as given or known, and Theme as message onset. Each of these will now be discussed in turn. 3.1.1 Theme as Topic Vande Kopple (1986: 74) points out that Mathesius was one of the first linguists to describe the Theme/Rheme distinction "and his words probably still capture what most theorists mean by theme". According to Mathesius, Theme is the segment "that is being spoken about in the sentence" (in Danes, 1974: p. 106). Mathesius (1975) elaborates further by stating: "...an overwhelming majority of all sentences contain two basic elements: a statement and an element about which the statement is made." (81): the element about which the statement, or Rheme, is made is the Theme. Halliday seems to provide a double definition of Theme when he explains "...the Theme is the starting-point for the message; it is what the clause is going to be about" (Halliday, 1985a 39), or "The theme is what is being 54
talked about, the point of departure for the clause as a message" (Halliday, 1967b: 212). I
would agree with Downing (1991) in her criticism of this double definition of Theme, as
point of departure and topic may coincide, but this is not necessarily the case. To illustrate,
the following example has been taken from the history textbook corpus (Themes are
italicized in all examples in this study):
3-1. Eng 7
(159) In 594 B.C., Solon [c. 640-559 B.C.), a traveler and poet with a reputation for wisdom, was elected chief executive. (160) He maintained that the wealthy landowners, through their greed, had disrupted ... (161) Solon initiated a rational approach to the problems of society ... (162) he sought practical remedies for these ills;
Here, in clauses 160, 161, 162, the Theme and the topic coincide. This passage is about
Solon, and he is Theme in those three clauses. However, in clause 159 a time adverbial is
thematized: In 594 B.C. However, this passage is not "about" 594 B.C.
Researchers, then, often prefer the use of the term topic to Theme when referring to
"aboutness". For Dik (1980), the topic presents the entity about which the predication
predicates something, while the Theme is something which is outside the predication
proper, coming before the clause set off by breaks in intonation. Also, Witte (1983: 338, fn
29), defines the topic of a sentence as "what the sentence is about". As topic can be
considered a "rather elusive category" (Downing, 1991: 121), pinning it can be problematic.
For Connor and Farmer (1990), topic can be explained as "simply the main idea or topic of
the sentence" and can occur in many places in a sentence - beginning, middle, or end" (ibid:
128). They do, however, point out that this type of topic identification is fairly intuitive
although they cite research which indicates high interrater reliabilities on sentence topic
identification. Yet, there are many cases where disagreement could occur, as in the
following bit of text,
Without care from some other human being or beings, be it a mother, grandmother, sister, nurse, or human group, a child is very unlikely to survive. (Witte, 1983: 319)
Witte takes the Subject of the main clause - a child - to be the topic of the sentence.
However, given that the writer begins with a prepositional phrase which could also have
been placed after the main clause, one could just as easily argue that lack of care is topical
in this particular sentence, not a child.
While the term topic is the most commonly used term for aboutness, whether it be at
the clause level or at a higher discoursal level (Schlobinski and Schutze-Coburn, 1992),
55
Theme has been used, by, for example, Jones (1977), to refer to the "main idea" or "central thread" of various levels of discourse. For the non-linguist, synonyms for theme include "subject" and "topic", while for some linguists, the terms "theme" and "topic" are used interchangeably. At the same time, many linguists use the terminology of topic/comment and Theme/Rheme interchangeably. Indeed, Levinson (1983: x) does so when he explains the omission of "the topic/comment (or theme/rheme) [emphasis original] distinction" from his book. It is clear from his further explanation that he is referring to aboutness, and given the above-mentioned elusiveness, intuition and problems equated with topic identification, perhaps it is not surprising that Levinson has little regard for it. There are further terminological problems with "topic". Witte and Cherry (1986) assign four functions to topic: 1) to a greater or lesser degree, to express old or given information, 2) to announce what the sentence or t-unit is about, 3) to provide local coherence between individual sentences or t-units, 4) to guide the reader in constructing gists and identifying discourse topics. Halliday (1994) brings up a similar problem with topic-comment terminology: "The label `Topic' usually refers to only one particular type of Theme...; and it tends to be used as a cover term for two concepts which are functionally distinct, one being that of Theme and the other being that of Given" (p. 38). In this section, then, we have seen that the notion of Theme as topic brings about problems due to the elusiveness of the term topic in discourse, for which reason Theme as topic will not form the basis for analysis in the present study. We have further seen that topic is often equated with given or known. In the next section, I will discuss the notion of Theme as given, an equation which often comes about via the term topic. 3.1.2 Theme as Given or Known The previous section on Theme as topic opened with Mathesius and his definition of Theme as topic. Danes (1970) in his explanation of the principle of utterance and text organization termed Functional Sentence Perspective (henceforth FSP), which he attributes to Mathesius, expounds further: Analyzing the organization of the utterance, we state as a rule its bipartite structure. The two parts (more precisely, the partial communicative functions assigned to them) may be defined from two different points of view: (1)Taking for granted that in the act of communication an utterance appears to be, in essence, an enunciation ...about something..., we shall call the parts THEME (something that 56
one is talking about, TOPIC) and RHEME (what one says about it, COMMENT)... 2. Following the other line, linking up the utterance with the context and/or situation, we recognize that, as a rule one part contains old, already know [sic] or given elements, functioning thus as a `starting point' of the utterance.... (Danes, 1970: 134). However, he later clarifies that Mathesius maintained from the onset of his FSP studies that "...theme need not be a known piece of information" (Danes, 1989: 25), and argues for a differentiation of Theme as topic on the one hand, and known information on the other. Yet many analysts do not make this distinction. In many studies based on the Theme/Rheme construct, these terms are often conflated with those of given/new, including in comparative cross-language studies. Whitley (1986) seems to rely on a definition of Theme as given to preface his brief discussion of Theme/Rheme analysis on spoken Spanish. He cites the Real Academia's explanation of the use of unmarked SVO order in which "el sujeto representa de ordinario el termino conocido, la continuidad del discurso." (Real Academia Espaсola 1979: 395, in Whitley, 1986: 237) He then explains that English more commonly uses a phonological procedure to show a changed thematic structure, (e.g. "DIANA called") and that, whereas Spanish can also use the latter, a change in word order is preferred (e.g. "Llamу Diana"). This may provide support for the hypothesis that Spanish is more pragmatically controlled while English is more grammatically controlled (see section 2.1.4 above). Babby (1980, in Vande Kopple 1991) also identifies Theme as conveying old or given information. Gutiйrrez Ordoсez (1997) does as well. In his view, everything occurs in an informative declaration as if the speaker were answering a question; the speaker makes a mental composition of what the hearer knows and does not know and orders the proposition in the same way as a question. And, in fact, he ultimately rejects the terms Theme and Rheme, because of their "terminological inflation": "Dada la actual inflaciуn terminolуgica de los signos tema/rema, optamos por las denominaciones transparentes de soporte (informaciуn conocida)/aporte (informaciуn nueva)." (Guttierez Ordoсez, 1997: 21), thereby moving to a given/new function of the concept of Theme. While it is the case that many studies are more explicit as to the terms used for the concepts being analyzed and specifically use the terms `given' or `known' instead of Theme (likewise using `new' in place of Rheme), there are problems in analysis for these concepts. Vande Kopple (1991) underscores the nature of the problem in his discussion of Prince's 57
(1979, 1981) work on given and new information. Prince's analysis (based on a taxonomy of given/new on a scale which goes from unused and brand-new entities on one end to situationally or textually evoked on the other) provides "one of the clearer and more insightful analyses of these phenomena available" (Vande Kopple 1991: 315). Yet it is not without problems. The main problem is that the line between unused entities and inferable entities (in the middle of the scale) can often become blurred (a problem which Prince herself acknowledges). What is unused information for one reader might be inferable information for another. The key to the problem of given (or Theme as given) here is that the perspective is that of the reader (listener). In order to establish what is given in the clause, it is difficult to establish the beliefs and presuppositions that the speaker has about the hearer's knowledge about the world, the context and the co-text. Halliday (1967b, 1994) states unequivocally that, while the concepts of Theme and given are related, they are not the same thing; "The Theme is what I, the speaker, choose to take as my point of departure. The Given is what you, the listener, already know about or have accessible to you. Theme + Rheme is speaker-oriented, while Given + New is listener-oriented." (1994: 299) Both are speakerselected, but while there seems to be a general correlation between Rhematic status and New information, and most Themes are presented as Given, there is no perfect correlation (Halliday, 1994; Fries, 1994). Therefore, in this study Theme will not be defined as given or known. But before pinning down a definition of Theme for my purposes here, another configuration of Theme/Rheme will be discussed; that is Theme/Rheme in Communicative Dynamism. 3.1.3 Theme/Rheme in Communicative Dynamism Jan Firbas has been more explicit about and consistent with the definitions of terms such as Theme/Rheme, and has employed his analysis with more precision than other analysts associated with FSP (Vande Kopple, 1986). In response to criticisms of the conflation of Theme/Rheme with given/new, Firbas proposed a tripartite configuration of the utterance - Theme, transition, and Rheme -depending on the degree of Communicative Dynamism (hereafter CD) the linguistic element contributes "toward the development of a given purpose" (Firbas, 1986). Within written language, the distribution of degrees of CD is affected by the interplay of three factors. The first factor is that of linear modification, a term Firbas borrows from Bolinger (1952), and which means that, provided there are no 58
interfering factors, communicative importance gradually increases as it moves towards the end of sentence. The second and third factors, the context and the semantic structure respectively, are interfering factors in that they can work counter to linear modification. The factor of context means that an element expressing known information, which is defined by Firbas as information retrievable from the immediate context - the immediately relevant preceding verbal context and/or the immediately relevant situational context -, carries a lower degree of CD than an element conveying a piece of irretrievable, new, unknown information. Therefore, context is hierarchically superior to linear modification and to semantic structure. The factor of semantic structure refers to some types of semantic content and some types of semantic relations, which, if they are context-independent, can work counter to linear modification. For example, some subjects are context independent, especially in the case of verbs which denote appearance or existence on the scene, e.g. A boy came into the room. Distribution of CD over sentence elements, then, is determined by an interplay of the above factors. Therefore, development of communication cannot be considered as "a merely linear phenomenon" (Firbas, 1986: 46). The dynamics of communication belong to one of two processes. One of these is the foundation-laying process: all context-dependent elements are foundation-laying in that they provide a foundation upon which the remaining elements complete the information and fulfill the communicative purpose. The second is the core-constituting process, in which elements completing the information constitute the core of the information. The Theme is formed by the foundation-laying elements, while the coreconstituting elements form the non-Theme, consisting of the transition (or those elements performing a linking function) and the Rheme. While CD is more highly detailed than other analyses involving the Theme/Rheme construct, it is not without problems. The first has to do with practicality: as Martin (1992a) points out, communicative dynamism is a matter of degree, there is no need to draw a line between Theme and Rheme. However, although there is some arbitrariness involved in deciding how far the Theme of an utterance extends, "it generally proves more practical to draw a categorical line between Theme and Rheme" (ibid: 151). Secondly, there is still some confusion as to how Theme is defined in CD: as Firbas states: "The theme expresses what the sentence is about and constitutes the point of departure in the development of the communication" (54) 59
A third problem I have found with CD is in the interpretation of one of the clauses from the foreword of one of the textbooks used for the Spanish corpus (Spa 2): Aparece este volumen dedicado al siglo XIX...; <>3 "este volumen" is highly context-dependent, so the question is why does it appear after the verb? (In this case in Spanish, the Subject can either come before or after the verb). While CD takes verbs which denote appearance/existence on the scene as providing less CD in the case of a context-independent Subject (e.g. A boy appeared, which in Spanish would have to be rendered through SV inversion: Aparece un niсo), there is no interference from the context factor in the Spanish example above, so according to the linear modification factor, there is no reason why "este volumen" should not be placed first in the clause. Teskey (1976) argues that Spanish allows for VSO word order due to CD, yet the examples he gives are not taken from real language in use and thus he makes assumptions for the possible context-dependence/independence of the examples he gives. Also, he states that Spanish is basically an SVO language, and that, unless criteria such as semantic weight and contextual dependence are brought into play, "this basic grammatical order will tend to be followed" (Teskey, 1976: 13; emphasis mine). He does not offer an explanation for examples such as the one just given above. Hatcher (1956) suggests that verbs of existence will tend to take a VSO order (see section 4.6.4 below). However, that is not always the case, as in: 3-2. Spa 2(17) La primer ediciуn de su Ensayo sobre la poblaciуn aparece en 1798; <> In this case, there is only one element of the thematic nominal group functioning as Subject which is context-dependent, the "su", or "his", referring to Malthus, who is mentioned in the Rheme of the previous clause. Therefore, it is not as highly a context-dependent item as is "este volumen" in the example given above, yet here the writer chooses not to invert the Subject and the verb. Unfortunately, at the time of this study, I have not been able to find a corpus-based study which would explain these differences. I have asked several Spanish informants, teachers of Spanish at our university, and they attribute the choice to style. A number of grammars in Spanish (El Esbozo de una Nueva Gramбtica de la Lengua Espaсola, 1986; Fernбndez Rodrнguez, 1986, inter alia) mention stylistic motives or 3 Triangular brackets <<>> are used to indicate a literal translation of the text. 60
reasons of expressive force as the motives for Subject-verb inversion or verb-Object inversion. Yet this does not seem to get us very far in explaining the Theme choice in an example like this one, or in any clause, for that matter. As Quirk, et al (1972: 23) point out, "`style'... is a term which is used with several different meanings". As these informants did not seem to be using the term in any technical sense, I asked them to clarify what they meant by style. One summed it up by saying simply "que suena bien", or "it sounds good". Thus, it is not clear what might be overriding context-dependency here. Hawes and Thomas (1997) point out other problems with CD, including the difficulty in deciding which element of the clause is lowest or highest in CD, and the limitation of Firbas' work to the isolated sentence. For these reasons, CD will not provide the basis for analysis of Theme in this study. The next section looks at an equation of Theme with initial position in the clause, which will then lead way to a definition of Theme which will be followed in the analyses carried out in the present study. 3.1.4 Theme as Initial Sentence Element There is some confusion in the interpretation of Theme as the initial sentence element. The cause of the confusion stems from a consideration of position as a way of defining Theme. Halliday provides a positional explanation for Theme: "...the element selected by the speaker as theme is assigned first position in the sequence" (1976: 179). However, he later points to a definition of Theme when he discusses "...the meaning of theme as point of departure for the clause as message" (ibid: 180, emphasis added). Halliday (1988 in Fries and Francis, 1992) calls attention to the misrepresentation of his intentions behind the statement "The Theme of an English clause is the element that is put in first position", is here meaning ` is represented by' n ot ` represents' : In other words, a clause that was intended to say how the Theme in English is to be recognized was taken as a statement of how it is to be defined - one of the most fundamental confusions in linguistics (Halliday, 1988: 33 in Fries and Francis, 1992: 45) This distinction is clear in Halliday (1976), where he says that thematization "is realized by the sequence of elements in the clause" (ibid: 175), and later states, to repeat one of the quotes above, that "the element selected by the speaker as theme is assigned first position in the sequence"(ibid: 179). Fries and Francis (1992) stress that Theme is identified by its 61
function, not its form, and Fries (1994) reiterates that initial position of Theme in English is a realizational statement for English, not a definition of Theme. At any rate, there are some analysts who base their assignation of Theme solely on position. For example, Barcelona Sanchez (1990: 98) uses Theme in a positional sense in his comparison of pragmatic motivation in raising to Subject in Spanish and English, and categorically states "Desde mi punto de vista, el "tema" es el primer elemento integrado o integrable sintбctica, prosуdica, y semбnticamente, de una oraciуn". 3.2 The Model Used: Theme as Message Onset In the above section, the notion of Theme is strictly formal: Theme is equated with first position and no mention of function or meaning is made. In this section, both the meaning aspects and the formal aspects of Theme are considered, as they are both crucial for an understanding of the Theme/Rheme construct and its function. Further specification of the meaning of Theme according to Halliday involves reference to `point of departure': "The Theme is the element which serves as the point of departure of the message" (Halliday, 1994: 37). While this conception of Theme may seem strictly positional, Fries (1983) makes the point that "there are good and sufficient internal grammatical reasons to say that the beginning is special for some reason" (118) and goes on to argue that "initial position in the sentence, or sentence level Theme, means `point of departure of the sentence as message'" (ibid: 119, emphasis added). Martin (1992a) provides evidence which indicates that "point of departure does indeed mean something more than coming first" (151). For Halliday, Theme "provides the environment for the remainder of the message, the Rheme." (Halliday, 1994: 67). Davies (1997) echoes this when he explains that the Theme initiates "the semantic journey" of the clause, and he adds that if a different starting point is chosen for the journey, a different journey results (Davies, personal communication). This idea is not a new one, as Bentham (1839: 268, in Bowers 1988: 92), while not using the label Theme, talks about the initial element of a sentence in a similar way: "If a word expressive of another idea come before it [i.e. the `the principal object'] the mind is in the first instance put upon a wrong scent; and a sort of correction and partial change of conception must have place, before the idea meant to be conveyed is apprehended". For Bloor and Bloor (1995) the Theme in English is "the idea represented 62
by the constituent at the starting of the clause" (72, emphasis added), while for Vasconcellos (1992), the function of Theme is "to signal the relationship between the thought in the speaker's mind and its expression in discourse" (147). She bases her conception of Theme on work done by Travnicek (1962), who believed the nature of Theme to be universal and invariant, expressing the view that all utterances have Theme, and that they occur without fail in initial position (1962, in Vasconcellos, 1992: 166, emphasis added). Travnicek's conclusion was that the principle of the Theme flows from the relationship between the speaker's object of thought and its expression in text. The key notion here is the relationship between the thought in the speaker/writer's mind and its expression in discourse. According to Travnicek, and later Vasconcellos, Theme realizes the connection between the thought and the continuation of the discourse. Speakers/writers choose one element or another for thematic position based on the thought they wish to express and based on how they wish to connect that thought to the surrounding text. The Theme, then, has a least a double function: to express the thought of the speaker and to indicate the relationship of that thought to the co-text. Does bringing in of the speaker's thought, of cognition, go too far beyond Halliday's conception of Theme? For cognitive linguists, the psychological dimension of language is necessary in models of language and language use in order to "specify in what different ways natural language users can play on [the complex instrument defined by Functional Grammar]" (Dik, 1997b: 58). Butler also feels that "a theory of language as communication must explore the important relationships between language and cognitive structures and processes" (Butler, 1998: 71). Givуn is also committed to "taking cognition and neurology seriously", and states: "We will continue to assume here that language and its notional/functional and structural organization is intimately bound up with and motivated by the structure of human cognition, perception and neuro-psychology" (Givуn, 1995:16). Within Systemic Functional Grammar, Halliday himself has never been particularly concerned with the cognitive side of language, preferring to concentrate on sociological factors. Nevertheless, the two elements are by no means incompatible, and attempts have been made by Fawcett (1980) to integrate them into a single, socio-psychological framework which has been applied to the description of English. All the same, in a descriptive text analysis such as this one, it is necessary to work from the evidence at hand: the texts themselves. From a cognitive/psychological point of 63
view, in a Theme/Rheme study, it might be of interest to engage in a protocol analysis of the Thematic choices made by a writer. Conversely, it might also be of interest to carry out a similar type of study from the reader's perspective, analyzing the cognitive processes taking place during reading with respect to textual organization (cf. van Dijk & Kintsch, 1983; Vande Kopple, 1986). Neither of these types of studies are used here; the study centers on the linguistic outcome of cognitive processes, not on the cognitive processes themselves. Through analysis of the text itself and its Theme/Rheme structure "...we can gain an insight into its texture and understand how the writer made clear to us the nature of his underlying concerns" (Halliday, 1994: 67). Halliday describes the Theme of the clause as always containing an experiential element: either a participant, a circumstance or a process (Halliday, 1994: 52). These are the elements of the system of transitivity, which is "the set of options relating to cognitive content, the linguistic representation of extralinguistic experience, whether of the phenomena of the external world or of feelings, thoughts and perceptions" (Halliday, 1967b: 199). The experiential component is: the linguistic expression of the speaker's experience of the external world, including the inner world of his own consciousness - his emotions, perceptions and so on. This component of the grammar provides a conceptual framework for the encoding of experience in terms of processes, objects, persons, qualities, states, abstractions and relations; it is sometimes referred to as `cognitive', and experiential meaning as `cognitive meaning', although as suggested earlier all components presuppose a cognitive level of organization...The term `experiential' makes it clear that the underlying function is seen not as the expression of `reality' or `the outer world' but as the expression of patterns of experience; the content given to an utterance by this portion of the language system derives from the shared experience of those participating in the speech situation. (Halliday, 1968: 209, emphasis added). For Lуpez Garcнa (1994: 7) "[t]oda expresiуn lingьнstica es una forma peculiar de verbalizar el mundo". Enkvist (1984) also brings in the notion of expression of experience in his discussion of word-order variation, which is key in a discussion of Theme-Rheme choice. For word-order choice, he posits three possible principles, which he terms "textual" as "they involve parameters outside or beyond the sentence" (Enkvist, 1984: 56). The first of these is information structure - old information first, new later. The second is what he terms "textual iconicism", where a clause may mirror the pattern of a preceding clause exactly or chiastically (e.g. he came in triumph and in defeat departs). The final principle relates to the previous quote by Halliday, and Enkvist terms it "experiential iconicism", which means "isomorphy between the text and our experience of the world...symbols of the 64
text are ordered in the same way as their referents in the world of things or the world of events" (ibid: 56). He underscores three major types: temporal, spatial, and social, and states that "...corpus studies show the importance of experiential iconicism in texts that have to order experience for the benefit of the receptor. Such texts are instructions, for instance guidebooks and cookery-books" (ibid: 57). For Halliday, moving from thematic Given to a rhematic New allows a "movement in time" which "construes iconically the flow of information" (Halliday, 1993b: 92). Bowers (1988) finds evidence for the benefits of this kind of inconic arrangement of clauses in a treatise on legislative language from the last century by Jeremy Bentham (Bentham, 1839). Bentham provides a general rule for sentence word order: "Rule: Whatever be the principal object which your sentence is designed to bring into view, bring forward as early as you conveniently can the word employed in the expression of it - if you can make the sentence begin with the same word, so much the better" (ibid: 268). Bowers (1988) points out that this iconistic "bringing into view" or "bringing forward" allows for ease of interpretation for the reader. However, this is only one of the pressures put on Theme choice. For Enkvist, the outcome in terms of the form a sentence may take, including its thematic arrangement, is the result of a process of several different forces: "What is optimal in textual terms depends on the speaker/writer's intentions and motives, on the text type and on the text strategy" (ibid, 58). Bowers (1988: 90) also sees this: "Motivation of thematic selection is not at one level alone, but embraces factors of semantic role, vocative address, ease of processing by the reader, and representation of real time in the order of elements in a sentence". Thus, the reality, or better stated, the perceived reality, of an event or events is one of the forces which is brought to bear on Theme choice. Martin (1995) discusses the range of information available for selection as topical Theme in a particular text as being the characters in the narrative and the entire contents of the room. This notion of the range of items available for Theme as limited is also shared by Fries: "the perceived simplicity or complexity of the development of the ideas in a text will correlate with the degree to which the experiential content of the Themes of the text may be seen to be derived from a limited set of semantic fields" (Fries, 1995: 324). Therefore, the use of "cognitive" (or "idea" or "thought in the speaker's mind") related to the notion of Theme is used here to refer to the expression of the speaker's perception of reality. Thus, the Theme of the clause, which formally is the initial element of 65
the clause, functionally combines the expression of the speaker's perception of reality and the concerns of the speaker to communicate that perception of reality to the listener. It is, thus, both cognitive, in the sense that it refers to the world of experience, and communicative, in the sense that it has a discoursal role. This combined function of Theme goes some way in explaining some of the problems involved with pinning down the function of Theme. At times, the speaker/writer chooses as the point of departure a Theme which relates to the surrounding text and reflects concerns of the overall text. At other times, the speaker/writer chooses a Theme which reflects the concerns of the immediate clause. As Berry (1996: 18) expresses it: [discourse Theme] is something that a speaker or writer has in relation to a text or large section of a text, a priority set of types of meaning that reflects his/her underlying concerns for the duration of the text or large section of text, and that [clause Theme] is something that a speaker or writer has in relation to a particular clause, a (set of) meaning(s) that reflects his/her priority for that particular clause. As explained above, the choice of Theme manifests a number of different pressures (not least grammatical constraints): discoursal pressures include the wish to relate the point of departure of one clause with something that has come before, and experiential pressures, which include the desire to represent iconically the pattern of experience through the clause. Added to this is the notion that "the priority concerns, discoursal or clausal, of a speaker or writer need not be ideational" (Berry, 1996: 19). To support this statement she refers to Brown and Yule (1983: 141-3) and their analysis of a passage of spoken English, in which the interactional aspect, through the subject pronouns I and you are thematized. The main concerns in this spoken passage are interpersonal. This conception of Theme separates it from notions of givenness and aboutness. While, as stated above, the view of Theme taken here is separate from that of given, it is the case that often Theme and given are conflated. If we take the function of Theme as connecting the speaker/writer's perception of reality with the surrounding discourse, it is often the case that the item chosen for thematic position will be something already given in the previous discourse: "other things being equal, the speaker will choose as `that from which to proceed in his discourse' something that is `known or at least obvious' to the listener" (Halliday, 1985b: 99). However, that is not always the case, "other things are not always equal" (ibid: 99), which means that, at times, speaker/writers will choose to introduce a new concern into the discourse and will do so by encoding that item as Theme. 66
Still, Fries (1993: 464) states that "[w]riters tend to construct their sentences so that New information falls at the end of the clause". And this brings the discussion to the Rheme of the clause and its function. 3.3 Rheme As seen earlier, for Halliday (1994: 37), while the Theme serves as the point of departure of the message, "the remainder of the message, the part in which the Theme is developed, is called...the Rheme". In the previous section, it was explained that Theme is both a clausal and a discoursal notion, and writers are dealing, more or less consciously, with both of these considerations in shaping their texts. Some Themes are chosen in order to provide for the flow of discourse, and look outwards from the clause to larger stretches of text, perhaps to one other clause or to many others, thereby taking a greater part in the global coherence of a text. An example of this type of Theme is the following: 3-3. Eng 1 (16) The circularity helps to keep the study of history alive,... In this text, the circular nature of the study of history has been the concern of clauses 8 through 15, and the Theme The circularity in 16 summarizes that part of the text, and the Rheme goes on to introduce a new idea with which a subsequent stretch of text is concerned: that of new research in history. However, other Themes are chosen for more local reasons at the level of the clause, reasons which can include, but are not limited to, grammatical constraints. An example of this type is: 3-4. Eng 2 (10) Expansion caused these early societies to develop still more complicated forms of political and social control. The idea of expansion is new in the discourse, yet it would be difficult for the writer to express the idea of expansion being the cause of the development in the early societies of more complicated forms of political and social control through any other grammatical structure. One possibility would be to express "these early societies" as a circumstantial Adjunct of location, thereby maintaining thematic progression. However, the result would be something like this: In these early societies, expansion caused still more complicated forms of political and social control to develop. 67
However, in this way, "early societies" loses its implied role of having a part in bringing about "still more complicated forms of political and social control". Therefore, this rewording loses an important point made in the original. Another possibility for ordering the constituents of this clause is through passivization, resulting in: The development of still more complicated forms of political and social control by these early societies was caused by expansion. Nonetheless, besides sounding somewhat convoluted, the author's purpose in this section of the text is to focus on different forms of political and social control. For Fries (1994), one of the expectations of the Rheme is that its content should correlate with the goals of the text as a whole, while we should expect the choice of Thematic content usually to reflect local concerns, as it orients the reader to the message of the clause. This idea is reiterated by Mauranen (1993b), who underscores the rhetorical significance of thematic and rhematic choices, the latter normally used to carry the focal content. The original wording allows the author, then, to place in the Rheme the more global concern of the text and the focus of this particular sentence. It is important to clarify here that the systems of information structure (Given-New) and thematic structure (Theme-Rheme) are separate structures, although they are related. We have already seen that other things being equal, Theme will conflate with Given. In correspondence, other things being equal, speakers will "locate the focus, the climax of the New, somewhere within the Rheme" (Halliday, 1994: 299). In example 3-4, we have just seen that "expansion" is new to the discourse. At the same time, "still more complicated forms of political and social control", is also new, and quite clearly encapsulates the global concerns of the text. This is made evident by the subsequent clauses, in which these forms of control are developed. Thus, this clause introduces a global concern of the text in Rheme final position, and then this is taken up in the following stretch of text. Fries (1993; 1994; 1995b; 1995c) coins the term "N-Rheme" for this final constituent of the clause, which is of interest "because it is the location of the unmarked placement of New information". For Martin (1992b: 452) "Themes angle in on a given field, reflecting a text's genre; News elaborate the field, developing it in experiential terms". Tied to notions of Theme and New is the concept of prominence. Halliday (1994: 336) speaks of two points of prominence in the clause through Theme and focus, which "gives a kind of diminuendo-crescendo movement to the typical clause of English: the 68
downward movement from initial, thematic prominence being caught up in the upward movement towards final, informational prominence". Hence, writers can organize their clauses in order to bring one element into thematic prominence while leaving another element for final prominence. Thus, here the Theme, other things being equal, is seen to reflect the concerns of writers in tying their perception of reality, of the field, to a starting off post for their readers, while the Rheme is seen as elaborating further on some aspect of the field; hence "a much greater range of meanings will be realised in New than Theme" (Martin, 1992b: 487). For this reason, the major focus of analysis in this study is Theme, as its realization is easier to illustrate, although in the analyses, Rheme content will also be examined to some extent. But before looking at the realization of Theme in English and Spanish, the unit for analysis is defined. 3.4 Unit of Analysis 3.4.1 Background on Choice of Unit One of the first considerations in an analysis of this type is the unit or segment of text for which Themes will be specified. Research in Theme/Rheme structures is by no means uniform in this respect. Some researchers take the clause as their unit of analysis; however, even within this group there are some differences. For example, Francis (1989, 1990) analyzes the Theme/Rheme structure of all clauses, including hypotactically related clauses such as projected clauses, as well as minor clauses (those without a finite verb). Ghadessy (1995) analyzes all clauses for Themes, except rankshifted clauses and minor clauses. Cloran (1995) also takes the clause as the unit of analysis and, basing her reasoning on Hasan (1991), states that a clause must have two characteristics in order to realize a message: it must be ranking and it must be non-projecting. Thus she analyzes all clauses except embedded and projecting clauses for Theme/Rheme structure. There are other researchers who take not the clause but the orthographic sentence as the unit for which Theme is analyzed. Whittaker (1995: 107) does so, with the justification that "[w]riters organize content into grammatical units and give orthographic signals of their segments. The orthographic sentence is a particularly important unit". She also quotes Halliday as saying that the sentence is "a significant border post, which is why writing systems are sensitive to it and mark it off" (Halliday 1985a: xxi, in Whittaker, 1995: 107). A 69
further justification has to do with the purpose of her study, which deals with text readability for non-native speakers: "researchers in reading consider the sentence to be the unit used by readers in processing written text." (ibid: 107). Lowe (1987) also works from sentences; his main interest is sentence initial elements which initiate a span (a unit larger than the sentence, for which the sentence initial element, or point of departure, serves as a framework) in the discourse. Nwogu (1990) and Nwogu and Bloor (1991) identify Themes in utterances, which Nwogu (1989) describes as a functional unit whose structure is dependent on the act of communication, which is determined by "the general character and regularities of linear materialization and linear perception" as well as by "extralinguistic content of the message, by the context and situation and by the attitude of the speaker towards the message and towards the addressee" (Danes, 1964, in Nwogu, 1990: 214). Nwogu goes on to say: "That structure is the division of the utterance into two portions: the theme (or topic), conveying the known (given) elements, and the rheme (or comment), conveying the unknown (not given) elements of an utterance" (Nwogu, 1990: 214). However, and in spite of this elaborate definition, a cursory look at the examples given by the researchers in order to home in on precisely what is the unit under analysis seems to point to the sentence. Aziz (1988) studies thematic progression in sentences in Arabic, and Lotfipour-Saedi and Rezai-Tajani (1996) also take as the unit of analysis the sentence in their study of the contribution of thematic elements to global coherence of the discourse. Other researchers who use the sentence as the basis for their analyses are Thomas and Hawes (1997), in their study of Theme in academic and media discourse, and Gosden (1996), in his study of Theme in research articles. While some researchers, then, take the clause or the sentence as a unit of analysis, Fries (1995) professes that, in his work on Theme, he has "found it useful to deal with a unit slightly larger than the clause, but smaller than the sentence" (49). He terms this the "independent conjoinable clause complex", also referred to as a "T-unit" by many analysts. This consists of an independent clause plus all of its hypotactically related clauses. In analysis based on the independent conjoinable clause complex, paratactically related clauses are each analyzed for an ideational Theme, while in the case of hypotaxis, only one Theme is analyzed for the whole of the clause complex. If the/a dependent clause comes first, this is taken to be the Theme of the entire clause complex. If the independent clause comes first, then the first ideational element of that clause is understood to be the Theme, with the rest 70
of that clause and any other subsequent dependent clauses forming the Rheme. Fries justifies this analysis by pointing out that the order of independent clauses cannot be reversed, as exemplified by "?And Bill left, John came" (Fries, 1983). This means that there is no Theme at the level of the clause complex, so each of the paratactically related clauses is treated separately. However, in the case of hypotaxis, the order of clauses can be changed around, and important in Theme is the notion of choice: "thematic structure means what it does because of selections from the range of lexico-grammatical options available for first position" (Fries and Francis, 1992: 47). Undeniably, the subordinate clause has its own thematic structure, as do rank-shifted clauses, and, indeed, smaller units, such as nominal groups. However, Berry (1995) also justifies the focus on independent clauses by citing Halliday: "the main contribution [to the method of development of a text] comes from the thematic structure of independent clauses" (Halliday, 1985a: 62, in Berry, 1995: 63). This point is echoed by Fries and Francis (1992) in relationship to the researcher's purpose and the linguistic unit to be analyzed. They further justify here the notion of the independent conjoinable clause complex as a choice linguistic unit: With regard to purpose, it can be suggested tentatively that if one chooses to examine only the `main' clauses within a clause complex (i.e. paratactic sequences and primary clauses in hypotactic sequences, with beta clauses forming part of the Rheme), then it becomes easier to discern the method of development and thematic progression of a text. In other words, if one wants to look at the larger units, and at whole texts, it may be as well to ignore the Themes of hypotactically related (subordinate) clauses. One important reason for this is that the structure of beta clauses, including their thematic structure, tends to be constrained by the alpha clauses. For example, the range of thematic choice is more limited in relative clauses, in dependent non-finite infinitive, -ing and -en clauses and so on. Such constraints upon Theme, and especially upon the cooccurrence of Subject and Theme in beta clauses, may oblige the writer to break out of a thematic pattern. Or, more interestingly, it may be that the writer is consciously using a particular method of development, and sees this as being predominantly a function of his or her main clauses (47). Here, Fries and Francis defend the analysis of the independent conjoinable clause complex on the basis of purpose, which in this case is to look at the method of development of text. The notion of purpose helps to explain why Fries in other studies uses not the clause complex but the clause as his unit for analysis (as does Francis, as mentioned above). For example, Fries (1992) analyzes all clauses, including minor ones. This is not surprising in light of the relationship between purpose and choice of unit, to which can be added the text type under consideration. In this study, Fries analyzes written advertisements to see where 71
information directly related to the goals of the advertisement is placed, whether in the Theme or in what he terms the N-Rheme, and to look at the differences in information placed in one position or the other. Also, the corpus contains many sentence fragments, or as he calls them "minor sentences", which calls for a different analysis. It is interesting to note that he finds similarities in the function of the information contained in postposed `because' clauses and `because' clauses which are `minor sentences', or enclosed between periods with no alpha clause. However, there are differences between their corresponding alpha clauses: in the case of the postposed `because clauses', the alpha clause contains minimal new information, while in the case of the minor sentence `because' clause, the preceding orthographic sentence, which operates in effect as the alpha clause, contains information which is relevant to the purpose of the advertisement. As this type of information normally appears in the N-Rheme of the clause, then the advertising writer would be motivated to encode the `because' clause as a separate sentence. In Fries (1994), he again reiterates the usefulness of treating Thematic structures in independent conjoinable clause complexes. Yet in this study, in which he looks at the difference in function between information appearing in Theme and N-Rheme positions in a fund-raising request letter, he also analyzes all clauses, including non-finite ones. Fries does point out in one case that major conclusions cannot be drawn from analyzing the structural Theme when in a dependent clause when analyzed separately. However, when this same clause is analyzed as part of its clause complex, the whole of the clause complex fits the pattern established by the other clauses in the analysis, the pattern in this case consisting of temporal adverbials. Indeed, during his analysis, he moves back and forth between discussing individual clauses and segments (here referring to each punctuated sentence). This underscores the importance of purpose and corpus type in unit selection. A further example of justification of the analysis of the clause (not the independent conjoinable clause complex) is Cloran (1995). As mentioned above, Cloran takes as the unit of analysis all non-rankshifted and non-projecting clauses. She demonstrates in this study the semantic functions of Subject and Finite in the mood structure, showing that they "define and delimit segments of text" (361).Given that Subject and Finite are clause level constituents, not clause complex level constituents, the choice of clause as the unit of analysis is logical here. 72
Other research which is carried out along the lines of analysis spelled out by Fries in terms of analyzing independent clauses with all of their dependents as one unit, and considering a preposed dependent clause as Theme, include Taylor (1983) (with the difference that he analyzes non-restrictive relatives and reported clauses separately for Theme/Rheme structure), Dubois (1987), Vande Kopple (1991), Bдcklund (1992), Ventola (1995), Martin (1995), Cummings (1995), Ventola and Mauranen (1991), and Berry (1995). Bдcklund (1992) is an interesting case, as she analyzes spoken language, a text type that would not seem to lend itself as readily as formal written text to clause complex analysis. However, once again the researcher's purpose is key in determining the unit of analysis; while she does not explicitly state the justification of the unit of analysis chosen (and, indeed, many researchers do not do so), one feature she analyzes is thematic progression, and, as seen above, the independent conjoinable clause complex seems to be the optimal unit for this type of analysis. 3.4.2 Choice of Unit The choice of unit for this study is the independent conjoinable clause complex, (for ease of reference hereafter referred to in most cases as `clause complex' in the case of an independent clause with dependencies, and `clause' in the case of an independent clause on its own). One of the reasons for this choice is the corpus chosen. The length of the corpus dictates a consistent unit of analysis for all of the functions: ideational, interpersonal, and textual. As explained above, the optimal unit of analysis for Thematic progression is the independent conjoinable clause complex; therefore, this is the unit analyzed consistently throughout this study. Still, the above does not justify the use of only one type of clause complex, that involving hypotaxis, as the unit of analysis, while the clause complex involving parataxis is divided into its separate clauses for analysis. This practice is normally defended on the grounds that the independent clause can stand on its own, while a dependent clause cannot, but the point is that writers often choose not to have the independent clause stand on its own and link it with another. Further justification, then, is warranted, and can be provided on structural, formal, and functional grounds. 73
3.4.3 Rationale for Analyzing Independent Clauses as Single Units In section 3.3, we have seen that one important function, but not the only one, of Theme is discoursal in nature. Related to this is the notion of scope, which provides some of the reasoning behind analyzing paratactically related clauses as separate, while hypotactically related clauses are analyzed as one unit. One scope related problem raised by Taylor (1983) is how to determine the Theme of a clause complex in cases such as the following: After the concert we drove round to Alec's but Jim stayed back with Penny. (Taylor, 1983: 206) One possible analysis Taylor offers is to take after the concert as the Theme of the clause complex, much in the same way that it is suggested that a subordinate clause in initial position be taken as a clause complex level Theme, and then we and Jim as Themes of the individual clauses. However, there are problems with this; one which Taylor points out is that this would then restrict Theme to noun phrases. Another is that it means analyzing a text at different levels - one at the level of the clause complex, the other at the level of the clause. A final problem, also pointed out by Dubois (1987) is that the notion of scope occurs at levels higher than the clause complex, i.e. there are elements of scope which range over a series of clauses or sentences. This means that taking the sentence as a unit of analysis in order to avoid problems related to scope in determining Theme in clauses such as the one above does not really avoid the problem at all. Also, there are instances when an element of scope is introduced in the first pair of a coordinated set of clauses which does not extend to the second; for example, in the following set, the scope of "For a long time" does not extend to the second clause. 3-5. Eng 1 (68) For a long time historians emphasized politics as the activity that summarized a whole society and determined its course, a view reinforced by emphasis on the development of the national state; (69) and that familiar perspective with its convenient chronological framework is generally maintained in this account. In fact, following a coordinating conjunction, a writer can insert a new item of scope, as seen with the following: 3-6. Eng 3 (35) a century later it had shrunk to something between one-third and one-half its earlier size, (36) and only after 1470 did it again begin to increase. 74
The independent clause is the basic unit in which interpersonal meanings, through the mood structure, and experiential meanings, through the transitivity structure, are expressed (Downing and Locke, 1992). Writers have choices with independent clauses: they can choose to write them as simple sentences, existing on their own, or they can string them together using coordinating conjunctions. It is interesting to note that writers often use coordinating conjunctions at the beginning of sentences; for example, in the history textbooks corpus analyzed in this study, there are a number of the following types of constructions: 3-7. Spa 5 (114) Pero los resultados no fueron satisfactorios 3-8. Eng 10 (145) But the parlements were to return, more arrogant than ever, under his successor. Structurally, then, independent clauses are complete, independent units and are analyzed as such. A major problem with this reasoning related to coordinated independent clauses is ellipsis. In both languages, there are a large number of the following types of coordinated sets: 3-9. Spa 2 (80) En ocasiones guerra y epidemia se alнan (81) y suman sus vнctimas; 3-10. Eng 6 (75) In June 1348 two ships entered the Bristol Channel (76) and introduced it into England. In English, the tendency has been to analyze the second clause as only having a Rheme. However, in Spanish, subject pronouns are not always included before the verb, and, in fact, are marked if included where the antecedent is clear, as the morphology of the verb indicates the person and number of the elided Subject. As this is the case, the independent clause in Spanish is considered complete without an expressed Subject outside the verb, and can make up a complete sentence. Along the same lines of reasoning, it can be argued that the Subject is encoded in the verb in English. Granted, English verbs are not marked for person or number. Yet, in spite of the verb markings, in Spanish the reader still has to provide the referent: without a previous clause, 81 above would not be interpretable. Therefore, Spanish verb initial constructions without overt Subjects are anaphoric; English verb initial clauses have anaphoric ellipsis. Thus, in order to compare like with like to the greatest possible extent, subjectless verb initial independent clauses are analyzed separately. 75
This leads to a further problem, however, which raises the question of how to analyze constructions such as the following: 3-11. Spa 1 (52) Al menos como tal naciу y se desarrollу en sus orнgenes. 3-12. Eng 1 (118) The timelines in each chapter have been expanded and improved because students requested it. 3-13. Spa 3 (7) Todos los procesos histуricos que en los diversos continentes se han ido fraguando hallan trabazуn entre sн, (8) culminan y maduran entre el comienzo y el final del 1200. These are cases where the line between independent clause with Subject ellipsis and group becomes blurred. These have been analyzed as one unit, mainly because of the limiting scope of the Adjunct in the first case (example 3-11) and the dependent clause in the second (example 3-12). The third set (example 3-13) is more complicated. There is actually a series of three coordinated verbs: "hallan trabazуn entre sн, culminan y maduran". However, the scope of "entre si" does not extend to the last two verbs, which are then analyzed as a separate independent clause together, because of the scope of the final Adjunct. Another example is the following set 3-14. Eng 2 (45) only recently, in the long period of human history, have people been able to produce food (46) and thus enjoy the security of a stable means of living. where not only has the Subject been elided, but also the finite verbal operators. I have chosen to analyze them as separate here because of the discourse Adjunct included in the second one. These problematic cases are limited to the above examples. While their analysis is not wholly satisfactory in that they are only further along the line in ellipsis than the subjectless clauses mentioned above, in a linguistic analysis such as this one there are bound to be borderline cases such as these. 3.4.4 Rational for Analyzing Subordinate Clauses as Part of Their Related Superordinate Clause While independent clauses are analyzed separately, dependent, or subordinate clauses, are analyzed as part and parcel of the independent clause to which they are attached. One reason for this is that this type of analysis is optimal for a textual analysis of thematic progression. Another reason is their similarity, in the case of what have traditionally been known as adverbial clauses (projected clauses are considered later in this section), to circumstantial Adjuncts, both formally and functionally. Formally, they can be placed in several different positions in relationship to the main clause, just as a 76
circumstantial Adjunct can. For example, in the following clause complex, the subordinate clause (underlined for ease of reference) is placed after the Subject: 3-15. Eng 10 (111) Its warships, though admirably designed, were inadequate in number. It could, however, be placed in clause initial or clause final position as well, much as a circumstantial Adjunct, or a conjunctive Adjunct such as "however", for that matter. This shows that formally, and as one unit, dependent clauses are similar to these other items which also can be moved around the clause for thematic or other purposes. Also, an initial element may occur, such as a textual Adjunct or a coordinating conjunction, which provides a scope for an independent clause, yet there may be an intervening dependent clause, such as: 3-16. Eng 3 (32) But although they give us no reliable figures for total population, they still enable us to discern with considerable confidence how it was changing. There is a further justification for the analysis of preceding dependent clauses in a clause complex as Theme, in this case not in the textual function, but in the ideational function. From a functional viewpoint in terms of transitivity, the clause is divided into three parts: processes, circumstances and participants. The subordinate adverbial clauses which appear in the history textbook corpus have the same semantic function as the circumstantial Adjuncts in that both of these constituents of the clause complex are used to express meanings of cause, contingency, manner, location and so on. An example is: 3-17. Eng 3 (76) In spite of the virulence of pneumonic plague, it is hard to believe that medical factors alone can explain the awesome mortalities. This could also be expressed as: Although the pneumonic plague was highly virulent, it is hard to believe that... In both cases, in the first through a prepositional phrase and in the second through a dependent clause, a semantic expression of concession precedes the main clause and provides the context in which it is to be interpreted. This similarity of function between circumstantials and dependent clauses is, then, a further reason for the analysis of the independent conjoinable clause complex. Vasconcellos (1992: 156) also considers the subordinate clause in initial position as having thematic status as a "stage-setting adjunct". There are, inevitably, problems in this analysis. As Halliday points out (1994: 231), the line between parataxis and hypotaxis in the case of some adverbial clauses is not very 77
sharp. There are some cases in the corpus of this type, involving the use of "for" in English and "pues" in Spanish, e.g.: 3-18. Eng 1 (91) New interests and new techniques have in fact greatly expanded the range of useful historical sources, (92)for important information can be gleaned from anything that records the pulse of human life: coins; pots and lamps; fiscal records... 3-19. Spa 9 (44) Estos contactos son de enorme trascendencia por su segura cronologнa y por su repercusiуn cultural, (45)pues representan el preludio de los influjos coloniales fenicios, .... "Pues" and "for" function similarly in that they both express relationships of cause and could feasibly be substituted by "because". However, in the cases of a clause with an initial "for" or "pues", the clause cannot be placed in front of the other clause in the complex. In one instance, this feature of "for" is underscored in that the textbook writer encodes it as the first constituent of a separate sentence: 3-20. Eng 5 (85) As a descriptive term, however, it fails to suggest the radically different character of human affairs after 7000 B.C. (86)For during the next two or three thousand years, substantial portions of the human species shifted from migratory hunter-gatherer patterns of existence to forms of settled agrarian life. Halliday (1994) suggests a working rule: in the case where the dependent clause can be switched around, then it is a case of hypotaxis, with the implication that if this is not the case, then it is a case of parataxis. This is the rule of thumb followed in this study. The whole notion of clause is not that straightforward. For Halliday, there is an "area of overlap between prepositional phrases and non-finite clauses" (Halliday, 1994: 212). He gives the example of "with no-one in charge" (ibid: 241) as a type of non-finite dependent clause. This is a one end of a scale which moves from the prepositional phrase to the independent clause, with the dependent non-finite and dependent finite clauses forming the middle steps. As mentioned above, some analysts choose to draw the line for the unit of analysis at different stages in this scale, some analyzing all independent clauses, some all finite clauses, and some all clauses including non-finite, (although I am unaware of any analysis which also includes the prepositional phrase). There is another main type of hypotactic clause relationship in addition to those which function as adverbials: projected clauses. These are similar to adverbials in that they are dependent on a primary clause, yet they are different in that their order cannot be reversed, or is very, very highly marked if done so. Also, they do not function as 78
circumstantials; they do not express relationships of manner, contingency, and so on. In spite of this difference, the same functional argument applies for analyzing hypotactic complexes involving projection as one unit. In the following example: 3-21. Eng 3 (83) Others argue that acute, widespread malnutrition had severely debilitated the population and lowered resistance to all kinds of infections. There is a verbal process "argue", with a Sayer "others" and with the whole of the projected clause "that...infections" functioning as the verbiage. Again, undeniably each clause has its own Theme: Others in the projecting clause and acute, widespread malnutrition in the projected clause. However, for the sake of the transitivity analysis, and for the sake of thematic progression, this type of hypotactic clause complex is analyzed as one unit. With regard to thematic progression, if we look at the previous clause, 3-22. Eng 3 (82) Some scholars consider that the weather of the age -it seems to have been unusually cool and humid- somehow favored the disease. there is an obvious mini-method of development at work here in thematizing Some scholars in one clause, and Others in the subsequent clause. Also, the global concern of this portion of the text is the plague and what may have caused its high virulence. The projected clauses encode this concern, and therefore exhibit more of a rhematic flavor (see Fries, above). In sum, dependent clauses fit into the transitivity structure of the independent clause, while one independent clause can not fit into the transitivity structure of another independent clause. For this reason, and for the sake of thematic progression, the unit of analysis chosen for this study is the independent conjoinable clause complex. 3.5 Theme Illustrated 3.5.1 Theme in English Halliday provides a very comprehensive account of Theme in English in his Introduction to Functional Grammar (1994). This section is a summary of that account. However, before looking at Halliday's account of Theme, it is important to note that other researchers working within systemic functional grammar differ from Halliday in determining the Theme of the clause. As will be seen below, for Halliday, the Theme of the clause in English is the ideational element which appears structurally first, i.e. in the left-most position of the clause. Thus a clause initial circumstantial element, such as temporal "In 1500" in example 3-23 below, is considered the Theme of the clause: 79
3-23. Eng 3 (13) In 1500 Europeans doubtless remained fewer in number ... For other researchers (cf. Berry, 1995; Gosden, 1996, Thomas and Hawes, 1997; Davies, 1989), the Theme of the clause extends up to and includes the Subject of the clause. Thus, in the just mentioned example (3-23), the Theme of the clause for these researchers would be In 1500 Europeans. Thomas and Hawes (1997: 35) base their decision to include the grammatical Subject on evidence from Chafe (1976) which suggests that "an item in Subject position serves as a more effective prompt for what a passage was about than another item which is not the Subject of the sentence...This suggests that the Grammatical Subject is closely associated with what the message is about". Gosden (1996) also rests his support on aboutness in his analysis of a clause from his corpus: However, recently, Fraas reported... (73). He argues: "if recently is seen to exhaust thematic potential as the obligatory ideational component,...it may be a point of departure as the clause-initial element, but it is very hard to see it as a contribution to the definition of what the clause is going to be `about'" (78). Given that in the present study Theme is taken to be the point of departure and that `aboutness' has been rejected (see sections 3.1.1 and 3.2 above), here Theme is taken as formally described by Halliday. 3.5.1.1 Simple Theme in Declarative Clauses For Halliday, the Theme of the declarative clause consists of one structural element which is represented by a nominal group, a prepositional phrase, or an adverbial group. e.g.: 3-24. Eng 1 (1) Everyone uses history. 3-25. Eng 6 (63) With ships continually at sea, their rats too were constantly on the move, 3-26. Eng 3 (131) By the middle of the fifteenth century, agricultural prices tended to stabilize, These representations may be complex in that they bring more than one group or phrase together, e.g. through coordination (underlining is used here to illustrate the different groups involved in the Theme): 3-27. Eng 8 (15) Warehouses and shops expanded to deal with the rising commerce of the empire. or apposition: 3-28. Eng 1 (32) The biographies of great men and women, dramatic accounts of important events, colorful tales of earlier times can be fascinating in themselves; In the case of a nominal group as Theme, this may coincide with the Subject of the clause, as in examples 3-24, 3-27, and 3-28 above. It may also coincide (although rarely in the corpus examined) with a nominal group functioning as Complement, e.g.: 3-29. Eng 10 (93) Of juvenile delinquency we do not hear much. 80
In addition to nominal groups, prepositional phrases and adverbial groups, dependent clauses can also be thematic. While Halliday analyzes all clauses for Theme, as explained above, the unit of analysis in this study is the independent clause with its dependencies. Thus, when the dependent clause comes first, it is analyzed as the Theme, much the same as a prepositional phrase, e.g.: 3-30. Eng 10 (32) When Anne, Mary's sister and the last Stuart monarch, died in 1714, Parliament had already arranged for the succession of the House of Hanover. Halliday (1967b: 221) provides further support for this type of analysis when he says: Probably only independent clauses...exhibit the option of theme in its full interpretation; ...while the theme system does operate in dependent clauses, the interpretation of theme in this environment requires the recognition of it as secondary to the underlying theme of such a clause, its relation of dependence to another clause. Another Theme type in the declarative clause is what Halliday terms the "thematic equative". It is a simple Theme in that a clause has been nominalized and it thus functions as a single constituent in the clause. An example of this is: 3-31. Eng 7 (114)What made Greek political life different from that of earlier Near Eastern civilizations,... was the Greeks' gradual realization that community problems ... Here the Theme is the italicized portion of the clause. It is termed a thematic equative as it construes the Theme/Rheme relation as one of equality. Other Theme types include "there" and "empty it" constructions. Now, according to Halliday, the Theme of the clause must contain one, and only one, experiential element as a Theme; that is, some element from the system of transitivity of the clause, either the process, a participant in that process, or any circumstantial feature of the process, is always present in the Theme. However, empty "it" in cases of extraposition and existential "there" are not participants in the transitivity of the clause, and criticism has been leveled at Halliday for labeling something like "there" as Theme (c.f. Huddleston, 1991). Martin (1995) contends that existential there does realize ideational meaning, in that it helps distinguish existential from relational clauses. Matthiessen and Martin (1991: 45) explain that existential there as Theme "realises the feature `existential'; it sets up as the point of departure that an Existent will be presented. The new information comes with the Rheme as the Existent...". This accords with Martin's (1992a: 164-165) view: ...existential clauses...are ideally designed for introducing participants as unmarked news at the end of the clause...and reinforcing their introduction by taking their existence as point of departure (i.e. Theme) ... there is 81
anticipatory; it signals that something is coming - namely the new participant at the end of the clause. Thus, what is thematic here is the existence of some new participant in the discourse, and this thematization is encapsulated in "there". An example of this is: 3-32. Eng 7 (49) there was a spectacular rise in the use of metals,... "Empty it" constructions are of two types: cases of predicated Theme and cases of extraposition4. Predicated Theme is exemplified by the following: 3-33. Eng 5 (61) it was this ability, no doubt, that permitted H.s. sapiens to survive, ... and extraposition by: 3-34. Eng 2 (91) it must have been clear that a new supply of food would have to be found. Superficially, these two types may seem similar. However, as Halliday points out, they are different. In the second instance, the clause can be rephrased as: That a new supply of food would have to be found must have been clear. However, this rephrasing is not possible in example 3-33: їThat permitted H.s. sapiens to survive was this ability. In fact, in this instance, the rephrasing would need to take place through a thematic equative: What permitted H.s. sapiens to survive was this ability. Another difference can be seen through another type of rephrasing. Halliday points out that the function of theme Predication is contrastive, to show that the predicated item, and no other, is the focus. Thus, in example 3-33, it was "this ability" and no other ability that made possible that H.s,. sapiens survived. Predicated Themes have an agnate form where the Theme of the agnate form is the contrasted item, e.g. for example 3-33: This ability permitted H.s. sapiens to survive. There is no such agnate form in the case of extraposition, i.e. we cannot say: їClear is a new supply of food would have to be found. In the case of predicated Theme, as the name implies, Halliday takes the whole element it was this ability as Theme. In the case of extraposition, Halliday takes only the initial pronoun as Theme. This analysis is followed here. This contrasts with other researchers 4Halliday uses the term `postposition' for this and for noun phrases which follow a clause, e.g. "they don't make sense, these instructions (Halliday, 1994: 60, underlining mine). In order to distinguish this latter type with the "it must have been clear" type exemplified here, I use the term "extraposition", following Quirk, et al (1972). 82
(e.g. Martin 1995; Whittaker, 1995). To illustrate, Martin (1995: 244) uses the following
example from a text:
Theme [it would be] irresponsible
Rheme to ignore...
thus analyzing "irresponsible" as Theme. This differs from Halliday's analysis, in that the
latter would analyze the empty "it" as Theme of the clause. I argue later (Chapter 6, section
6.1.3.2) that this type of construction allows the speaker to move the modalized adjective
(in this case "irresponsible) away from the position of thematic prominence. Thus, following
Halliday, in cases of extraposition, only "it" is counted as Theme here.
However, cases of extraposition and existential there are counted separately; thus,
the empty "it" of the extraposed clauses in the corpus and there of existential clauses do not
appear in the total count of Participant as Theme, as, obviously, they are not participants in
the clause. They are counted as a separate categories of extraposition and there.
All of the above are examples of what Halliday terms "simple Themes", as they form
a single constituent in the structure of the clause. Thus, as explained above in section 3.4.3,
in the case of the coordinate clause with Subject ellipsis, the process is considered as
Theme, e.g.:
3-35.Eng 6 (76) and introduced it into England.
Again, this is a case of simple Theme (the rationale for not considering the elided element as
Theme is explained in section 3.5.2, when discussing process as Theme in Spanish).
3.5.1.2 Multiple Theme in Declarative Clauses
However, there are other elements which can be thematic, yet which do not use up all of the thematic potential of the clause. These other elements can be divided into two groups: textual, which includes continuative, structural and conjunctive elements (these are explained more fully in section 5.6.1), and interpersonal, which is any combination of vocative, modal or mood-marking elements (these are explained more fully in section 6.1.3.1). Textual and interpersonal Themes are exemplified below in that order: 3-36. Eng 7 (110) Thus, the religious-mythical tradition never died in Greece 3-37. Eng 2 (80) Probably no single cause can provide the answer, Conjunctive and modal elements are outside of this experiential structure of the clause as they have no status as participant, process or circumstance. And, according to Halliday (1994: 53), until a participant, process or circumstance appears, "the clause still lacks an
83
anchorage in the realm of experience". Yet they are thematic given that the speaker/writer can choose to place them in different places in the clause, e.g.: 3-38. Eng 6 (39)The amount of food yielded, however, did not match the level of population growth. This is in contrast with the coordinating conjunctions and, but and so on. These obligatorily thematic, as there is no choice as to their positioning in the clause. Thus, clauses can consist of multiple Themes, i.e. any combination of textual, modal and experiential Themes, when textual and/or modal Adjuncts appear prior to the experiential Theme. When they appear in a posterior position to the experiential Theme, as in example 3-38, they are considered as part of the Rheme of the clause. Examples 3-36 and 3-37 are typical examples of multiple Themes: "Thus, the religious-mythical tradition" exemplifies a textual Theme followed by the experiential Theme of the clause, while "Probably no single cause" exemplifies a modal Theme followed by the experiential Theme. 3.5.1.3 Theme in Interrogative and Imperative Clauses The above examples and discussion of Theme all pertain to the realm of the declarative clause. In interrogative clauses of the "yes/no" type, the finite verb, which is the element that represents the polarity, is Theme. It is considered a modal Theme in its moodmarking function, and thus does not use up the thematic potential of the clause. The experiential element which follows the finite verbal operator is also considered Theme. Thus, in the following example, "Did formal beliefs, economic activities, established institutions and social classes" is all considered the Theme of the clause: 3-39. Eng 1 (42) Did formal beliefs, economic activities, established institutions, and social classes sustain each other? In the case of interrogatives with Wh- elements, the latter are considered Theme as they function in the transitivity of the clause, either on their own when functioning as the Subject of the clause, as in: 3-40. Eng 6 (9) What were the social and psychological effects of repeated attacks of plague and disease? or as part of a nominal group when functioning as Complement, such as: 3-41. Eng 6 (8)What economic difficulties did Europe experience? Finally, in the imperative mood, the verb is considered as Theme if it comes first, as in: 3-42. Eng 6 (200) Imagine an entire society in the grip of the belief that it was at the mercy ... 84
3.5.1.4 Summary of Theme types in English
The Theme types in English are summarized in Table 3.1 below:
Table 3.1: Theme in English
Mood
Textual Theme
Declarative Structural and Conjunctive Elements
Interrogative Structural and Conjunctive Elements Imperative Structural and Conjunctive Elements
Modal Theme Modal Elements Mood-marking element (finite verbal operator)
Experiential Theme Participant Circumstance Process Clause There Other: Extraposition, Predicated Theme, Thematic Equative Participant Wh-element Wh-element + participant Process
While the use of conjunctive and modal elements (besides the finite verbal operator) are not that common in the interrogative and imperative moods, their appearance is certainly not impossible, as evidenced by: 3-43. Eng 2 (94)But, again, why was this region the cradle of Near Eastern agriculture? This leads to a final point here, which has to do with the notion of markedness. Halliday considers that the unmarked option for thematic choice in the declarative clause is a nominal group functioning as the Subject of the clause. An adverbial group or prepositional phrase functioning as circumstance in Theme position is more marked than a Subject, and the most marked is Complement as Theme. It has been pointed out to me (Paul Thibault and Leo Hickey, personal communication) that markedness can only be determined in context. That is to say, there may be contexts in which Subject as Theme is the marked option. At the time of this study, I am not aware of any text type which exhibits a different pattern of markedness to that suggested by the above grammarians. Bloor and Bloor (1995: 82) use markedness as concerning "the language as a whole rather than in any particular use of the language in context", which would explain Halliday's statement on unmarked choices and Theme. At any rate, as is evidenced by Table 4.6 in section 4.6.1 below, this is indeed the pattern of markedness in the history textbook data in English.
3.5.2 Theme in Spanish
Much has been written about Spanish and its use of word order, which is thought of as being freer than in English, mainly because of its allowance for VSO order. This is
85
generally thought of as being pragmatically motivated in order to express relationships "such as theme-rheme distinctions" (Hinds, 1987: 142). Indeed, many of the studies on Theme/Rheme in Spanish come to this same conclusion: that word order in Spanish is pragmatically motivated (Thompson, 1978; Teskey, 1976; Silva-Corvalбn, 1984; Ocampo, 1990; de Miguel 1988, inter alia). However, these studies conflate Theme with given or known, or use Firbas's notion of Communicative Dynamism (see section 3.1.3) in their discussions. For Jimйnez Juliб (1984) the Theme is "el elemento que abre la claъsula, presentando el бmbito sobre el que se va a predicar algo, el punto de partida...". This accords with the definition of Theme given above, and also ties Theme/Rheme formally to the distribution of elements in the clause. There have been studies carried out in Spanish which follow this delineation of Theme (Alonso Belmonte, 1997; Munday, 1997), and Portuguese, a language which is grammatically similar to Spanish (Vasconcellos 1985, 1992). Thus, in this study, as with English, formally, the Theme is realized in Spanish by first position in the clause. 3.5.2.1 Simple Theme in Declarative Clauses The above means that there are similarities with Theme in English: a simple Theme consists of some experiential element, either a participant, a process or a circumstance. These can be realized by nominal groups, verbal groups, adverbial groups or prepositional phrases. Some examples follow: 3-44. Spa 1 (86) La sнntesis busca desentraсar la marcha de la Historia, 3-45. Spa 3 (6) Podrнamos calificarlo en sнntesis con la palabra "madurez". 3-46. Spa 4 (56) A finales de este milenio se dio en Creta un gran desarrollo del urbanismo As has already been explained, when a dependent clause precedes its independent counterpart, it is considered thematic for the whole of the clause complex, as in: 3-47. Spa 4 (2) Aunque los mismos griegos tenнan vagos recuerdos de un pasado anterior nunca se esforzaron por estudiarlo cientнficamente; Spanish, like English also has predicated Themes and thematic equatives. An example of a predicated Theme in Spanish is: 3-48. Spa 6 (130)Es entonces cuando entra en crisis la esclavitud << Is then when enters into crisis slavery>> It is at that time that slavery enters into a crisis. While there are no examples of thematic equatives in Spanish in the corpus, they do exist, e.g.: Lo que hay que mirar en un hombre es lo que hay dentro de su cabeza. 86
<> What we must look for in a man is what is inside is head. 3.5.2.1.1 Process as Theme A major area of interest in a Theme analysis in Spanish is that of process as Theme. Spanish allows for verb-initial constructions in a variety of contexts. One is when a verb of a subsequent clause refers to the Subject of a previous clause. In this situation, there is an obligatory pro-drop movement that takes place: it is considered very marked if the pronoun is included, and the speaker/writer would have to have a special purpose in including the pronoun. In his study on Spanish-English translation, Munday (1997) analyzes the elided subject pronoun as the Theme of the clause. This is because if he analyzed the verb as Theme, there would be different thematic progression profiles for the source and target text he analyzes due to a syntactic difference between the two languages, and not due to a difference in intent with regard to thematic progression on the part of the translator. He admits that this may be to some extent an imposition of an English framework on Spanish, yet he finds it necessary for his purposes. What I am suggesting here may be thought of as the opposite: imposition of the possibility of verb-initial structures on English (as suggested in section 3.4.3 above). Thus, in these cases, the verb, and not its elided pronoun, is taken as Theme. Suсer (1982) concludes that the verb serves as a way of presenting an object "...it is like the tray on which the delicacy is presented" (1982:126). Travnicek (1962, in Vasconcellos, 1985: 238) also shows the verb as Theme in his example from Czech "Trefila kisa na kamen" (literally - Struck the scythe against the stone), where the verb `trefila' is the Theme. According to Travnicek, what is being made thematic here is "the conflict of wills established through the verb" (ibid). Also, there are other cases of verb-initial constructions in Spanish besides pro-drop. The others are inversion and impersonal constructions. Inversion, as the name implies, consists of a verb-initial construction followed by the Subject. Thus, the Subject appears in the Rheme of the clause, and the verb in the Theme. Motivation for these constructions is taken up in sections 4.6.2, 4.6.4 and 5.5.2 below. Impersonal verbs in Spanish do not allow for the inclusion of an explicit Subject (Alarcos Llorach, 1994). These include verbs which refer to meteorological phenomena such as Llueve (`It is raining), Ha nevado (`It has snowed'). Ser and estar, also in their meteorological or temporal reference, do not allow for an explicit Subject, e.g. Son las tres 87
(`It is three o'clock'). This is also the case with the verb hacer and temporal references, e.g. Hace tres meses que no le veo (`It's been 3 months since I've seen him'). While none of these types of impersonal verbs appear in the Spanish corpus, there is another type which does. It is similar in form and purpose to extraposed clauses in English. Examples include: 3-49. Spa 4 (32) Conviene tener presente que casi todo lo que conocemos sobre la Creta antigua se lo debemos a la Arqueologнa. 3-50. Spa 7 (126) Es indudable que algъn progreso existiу; The first clause, which could be glossed as "It is convenient to keep in mind that almost all that we know...", and the second, which could be rendered as "It is without doubt that some progress took place" have the same configuration as the extraposed clauses in example 3-34 above. Indeed, like their English counterparts, they have a parallel construction in which extraposition does not take place, such as: 3-51. Spa 4 (142) El que algunas islas de las Cнcladas y tal vez el Дtica estuvieran sometidas al rey Creta parece probable; Here, the noun clause, which could be glossed as "That some of the Cyclades islands and perhaps Atica were under the authority of the Cretan king " functions as the Subject of the sentence. In Spanish it would be possible to say "Parece probable que algunas de las islas...", and in English to say "It seems probable that some of the Cyclades islands...". While in English an "empty" it is used in place of the extraposed Subject, in Spanish no such pronoun may be used, thus the process itself is thematic. In order, however, to provide for comparison across the two languages, these clauses have been counted as "extraposition" in the Theme counts, rather than including their numbers under the "Process" category for the Spanish data. 3.5.2.1.2 Se 3.5.2.1.2.1 Extraposition There is a slightly different but related construction to this just explained type of extraposition in Spanish. It also involves extraposition, but takes a proclitic pronoun "se". An example is: 3-52. Spa 4 (57) y se cree que aparecieron los primeros Estados de Grecia. This construction is called impersonal (Alarcos Llorach, 1994) or the impersonal reflexive voice (Valdйs, Dvorak & Pagбn Hannum, 1984). It is used in cases where there is no agent, or when the agent is not an essential part of the message (Valdйs, Dvorak & Pagбn Hannum, 1984).The extraposed element is the Complement of the verb, and, as will be seen 88
shortly, object pronouns in initial position are unmarked in Spanish. In English, we can use a passivized construction with the empty it referring to the Complement, which then becomes the passivized Subject of the clause. Thus, the gloss for example 3-52 is: and it is believed that the first Greek states appeared. Spanish does not allow for this type of construction, where the verb "ser" is combined in a passive construction with a participle of a mental state verb. Thus, a more literal translation of example 3-52 would be: and one believes that the first Greek states appeared. However, that gives a false impression that the Spanish impersonal reflexive focuses on agency, as the implication here is that someone believes something. Indeed, it is used in order to avoid expressions such as "one", "people", or the potentially ambiguous "we" in terms of inclusivity. At any rate, in this type of extraposition, through the impersonal reflexive, while se is considered thematic, once again, these constructions are counted as cases of "extraposition". There is another related type of impersonal expression which takes proclitic "se" which is also impersonal reflexive, but does not involve extraposition of a noun clause, but of a non-finite clause or a prepositional phrase. An example of this from the corpus is: 3-53. Spa 9 (56) Por tanto no se puede hablar propiamente de una unidad cultural atlбntica,... In English, in order to retain the lack of presence of a Sayer in this verbal process, this could be rendered through the passive: Therefore, a cultural Atlantic unity cannot be spoken of. However, in order to retain the thematic configuration, with the "new" information `cultural Atlantic unity' left for Rheme position, we would more likely use either a neutral Sayer, such as we: Therefore, we cannot speak of a cultural Atlantic unity. or a non-finite extraposed clause: Therefore, it is not possible to speak of a cultural Atlantic unity. In all of these cases, the clause is classified in the Theme counts under the category of extraposition. Once again, the reason for this is to provide for clearer comparison of Themes across the two languages. But there is another motivation. These impersonal reflexive constructions have counterparts with enclitic, rather than proclitic, se, such as: 3-54. Spa 1 (42)y no puede decirse concretamente que sea Historia en realidad, ... 89
The function of the two types are identical. Alarcos Llorach (1994) states that in present day Spanish, enclitic pronouns are used much more infrequently than proclitics (except in the case of the imperative). Counting these two different positionings as extraposition, rather than as "se" as Theme in the first case, and finite as Theme in the second, allows for a comparison of the use of extraposition across the two languages. One interest here is the interpersonal force of these constructions (see section 6.1.3.2 below). 3.5.2.1.2.2 Other Functions of Se However, the use of the impersonal reflexive is not limited to these cases which correspond to extraposed clauses in English. Another use also involves keeping the verb in the singular third person case, and the verb is then followed by a prepositional phrase (Alarcos Llorach, 1984). This confers to se in these cases the function of referring to the indirect object. Examples (with glosses) include: 3-55. Spa 6 (38) y no se presta atenciуn a los factores econуmicos y sociales. <> and no attention is given to social and economic factors 3-56. Spa 7 (138) se dio mбs importancia a la ganaderнa... <> more importance was given to livestock breeding... Se in these cases is a reflection of the indirect object. It refers to "a los factores econуmicos y sociales" in example 3-55 and to "a la ganaderнa" in example 3-56. Thus, there is no way to gloss this function in English. What is of importance to note in these cases that the number of the noun in the ensuing prepositional clause does not affect the verb, which is always singular. The purpose of this discussion is to distinguish this use of se from another use, which is termed passive reflexive. In contrast with the impersonal reflexive, the passive reflexive implies the function of Subject for the nominal phrase to which it refers (Alarcos Llorach, 1984), e.g.: 3-57. Spa 2 (50) se produce un impulso demogrбfico de compensaciуn, << it is produced a demographic boost of compensation>> a demographic boost of compensation is produced, The function as Subject of the passive reflexive se is clear when the nominal phrase is in the plural, as in: 3-58. Spa 4 (46) y se quebraron los cuadros de la organizaciуn comunal primitiva <> and the profiles of primitive communal organization were broken. 90
Example 3-58, where the verb is in the third person plural form, can be compared with example 3-55, where the verb is in the third person singular form, and thus the difference between the impersonal reflexive and the passive reflexive can be seen. There is a further use of se, which is simply reflexive in nature, and could be glossed as the reflexive pronouns in English. In these cases, the agent of the verb is the same as the Complement. An example of this is Juan se lavo (John washed himself). There is one instance in the corpus where it is not clear whether the usage is reflexive or the impersonal reflexive. This is a clause which comes after a discussion of Alfonso X and Richard of Cornwall's attempts at becoming crowned emperor of Germany. According to the text, the latter went ahead: 3-59. Spa 3 (53)y se coronу en Aquisgrбn, <> OR <> This could thus be translated either as "and crowned himself in Aquisgrбn" or "and was crowned in Aquisgrбn", although the implication from the text seems to be the former. At any rate, in all of these uses of se, se is considered Theme. Se in these cases is not categorized as a participant, as it is not a participant in its own right, but a reflection of a participant which is left for the Rheme, thus giving it a function similar to empty "it". It also is anticipatory, not unlike existential there according to Martin as commented above, except that it is not a marker of existentiality, as it can appear in any kind of process. Se allows information to be left for Rheme position, while at the same time it indicates that agency is either non-important or non-attributable (except in the case of the reflexive, and the only example is the clause in example 3-59 which is ambiguous). Thus, a separate category of Theme simply called se has been set up to take this into account in the analysis. 3.5.2.1.3 Hay There is a further case in Spanish which is considered separately: hay. Hay is from the auxiliary verb haber, yet the hay form is autonomous. It is impersonal in function, and can only be used in the third person singular. It cannot be used with a pronoun. It can be glossed as "there is" or "there are" in English, e.g.: 3-60. Spa 1 (55) Hay muchas clasificaciones de las fuentes, mбs o menos afortunados, There are many classifications of sources, more or less felicitous. It can also be used in other tenses, e.g.: 3-61. Spa 7 (151) No habнa suficiente estнmulo porque la mano de obra era abundante y barata. 3-62. Spa 7 (87) Hubo una verdadera revoluciуn agraria paralela a la industrial 91
While hay, no habнa and hubo function as the process of the clause, in order to provide for comparison with the English data and the use of existential there, they have been considered separately here in a category simply labeled Hay. Teskey (1976: 50) provides further justification for such an analysis when he says that hay "seems to occupy a subject-like position in the grammatical structure of the sentence, and the indefinite noun functions as a grammatical object". 3.5.2.2 Multiple Theme in Declarative Clauses As with English, Spanish can also have multiple Themes. That is, in addition to the experiential Theme, there may be preceding textual and/or modal Themes. and their function and placement is similar to that of English, e.g.: 3-63. Spa 1 (80) Sin embargo, el concepto "bosque" es en sн imprescindible 3-64. Spa 1 (46)Quizб esta moderna tendencia sea la mejor para hacer historia. There is a further point with regard to the interpersonal function and Theme as to the formation of the negative in Spanish. The negative is formed by inclusion of the pre-verbal morpheme no, e.g.: 3-65. Spa 1 (91) No es asн, sin embargo, << no is like this, however, It is not like this, however, In his initial analysis, Munday (1998) includes "no" as interpersonal Theme. As he is comparing thematic choices across a text and its translated version, he later factors "no" out of his counts. However, in English, Halliday (1994: 88) holds that "polarity does not figure as a separate constituent". While it is the case that the negative in English is realized through a distinct morpheme nґt or not, "this is an element in the structure of the verbal group, not in the structure of the clause" (ibid). Thus, in Spanish also, here the morpheme no is taken as part of the verbal group. Thus, no es is considered verbal process as ideational Theme in example 3-65. An analogous case in English is the nominal group with "no" as determiner, e.g.: 3-66. Eng 6 (182) No estimates of population losses have ever been attempted for Russia ... Here no clearly forms part of the nominal group and it would not be considered separately as an interpersonal Theme. The same is the case in Spanish where no forms part of the verbal group and is considered as Theme along with the group, but not as a separate interpersonal Theme. 92
3.5.2.3 Theme in Interrogative and Imperative Clauses In Spanish, as in English, there are two types of interrogative clauses: those questioning the polarity of something (yes/no), and those questioning the identity of something (wh-). As with English, the latter can be Theme either as part of a nominal group when functioning as Subject: 3-67. Spa 2 (4) їQuй interйs ofrece para el historiador este aumento espectacular del nъmero de los hombres? or as Theme on their own when functioning as Complement: 3-68. Spa 2 (63) їCuales son estos frenos? Polarity questions are indicated solely by orthographical conventions in written Spanish. Word order is no different for interrogative and declarative clauses. While in the spoken language, the difference is signalled through intonation, in the written language it is signalled by a clause initial orthographic symbol ї and a final ?, e.g.: 3-69. Spa 4 (139) їLas Cнcladas y Grecia formaron parte del reino de Cnosos? A temptation here is to take the orthographic ї as Theme (much as the finite verbal operator in English) along with the subsequent ideational element. At any rate, it does not really affect comparison across the two languages as in both cases the ideational element is considered thematic. As the finite verbal operator is attached as a morpheme to the verb, the process is considered as Theme. In the case of the imperative, as with English, the process is considered Theme, e.g.: 3-70. Spa 10 (67) recuйrdense, en el caso de Espaсa, las Reales Sociedades Econуmicas de Amigos del Paнs.
3.5.2.4 Summary of Theme in Spanish
Table 3.2 below provides a summary for Theme types in Spanish:
Table 3.2: Theme in Spanish
Mood
Textual Theme
Modal Theme Experiential Theme
Declarative Structural and Conjunctive Elements Modal Elements Participant
Circumstance
Process
Clause
Se
Hay
Other: Extraposition, , Predicated
Theme, Thematic Equative
Interrogative Structural and Conjunctive Elements
Participant Process
Wh-element
Wh-element + participant
Imperative Structural and Conjunctive Elements
Process
93
This section will also leave off with a comment on markedness, as did the previous section on Theme in English. Teskey (1976) notes that Spanish is basically a SVO language, meaning that, as with English, the Theme will most often be conflated with the Subject of the clause. This would be the most unmarked order. However, there are exceptions to this. One is the use of object pronouns: they must be placed before the verb in the declarative and the interrogative, as we have seen already in the case of se, and which can be seen by other pronouns, e.g.: 3-71. Spa 3 (67) Despuйs de derrotar a las tropas castellanas, llegaron rбpidamente ante Sevilla (68) y la sitiaron. After defeating the Castilian troops, they arrived quickly at Seville <> This is the unmarked case, as the grammar of the language requires that the object pronoun be placed before the verb. Another case of unmarkedness is inversion in the case of existential verbs (Hatcher, 1956; see section 4.6.4 below). The notion of markedness will be returned to in section 4.6.1 below, along with a comparison of the results of the Theme types in the English and Spanish corpora. 3.6 Analytical Methodology As explained in Chapter 3, all independent clauses in the corpus were analyzed first of all into Theme/Rheme. This was done manually, and Microsoft Excel was used to record information for each of the clauses. This record can be found in Appendix II. It includes an initial column with the code name of each text (Eng 1, Spa 1, etc.), followed by the subject of the chapter, and the clause number. After this there is a column for conjunction types, one for modal Themes, one for finite as Theme, and another for textual Themes. Then, each ideational Theme was given a code, based on the previous discussion in this chapter, according to its Theme type: - Participant (all Subjects) - Object (Complements in initial position) - Circumstance (prepositional and adverbial phrases, non-finite clauses) - Clause (dependent clause in initial position) - Verb - Wh- Fronted 94
- Extraposition - Predicated Theme - There - Hay - Se Also, clause initial circumstances types were analyzed and code types were entered for those as well as for thematic dependent clauses. In addition to this thematic analysis, a rhematic analysis was carried out as well. This information is also included in the Excel analysis, and consists of: participant roles (see section 4.4.4), circumstantial types, textual and modal Adjuncts, and finite modal operators. This analysis can be found in Appendix III. In each of the analysis chapters below (Chapters 4, 5, and 6), further specification of how results were tabulated is made clear. 95
4. Transitivity And Theme/Rheme 4.1 The Experiential Function The purpose of history textbooks is to portray `reality' to readers in order for them to come to a greater understanding of historical events, processes, changes, and so on. According to Halliday (1994), the experiential function concerns the clause in "its guise as a way of representing patterns of experience" (106). Thus, this chapter focuses on Theme/Rheme and its relationship to the experiential function. The experiential function realizes patterns of experience through the system of transitivity. Transitivity is "the set of options relating to cognitive content, the linguistic representation of extralinguistic experience, whether of the phenomena of the external world or of feelings, thoughts and perceptions" (Halliday, 1967b: 199). Here we are in the realm of field, particularly, in the realm of the field of history. History is a vast field, and can be viewed in many different ways. We have already seen in section 1.4 that history is carved up in different ways in the U.S. and in Spain, the former having a notion of Western Civilization, a notion which takes in Europe from Greece westwards and America, and the latter focusing the field on different continents or countries. Thus, an initial approximation might assume that differences may occur in the ways of representing historical experience due to differences in culture in its broader sense: However hard we struggle to avoid the prejudices associated with colour, creed, class or gender, we cannot avoid looking at the past from a particular point of view. Cultural relativism obviously applies as much to historical writing itself as to its so-called objects. Our minds do not reflect reality directly. We perceive the world only through a network of conventions, schemata and stereotypes, a network which varies from one culture to another. (Burke, 1991: 6) While there may be differences across cultures, there are other factors which may serve to mitigate them. First of all, that there is a common tradition of history writing in western societies is reflected in writings on the philosophy of history (e.g. Berkhofer, 1989, Tilly, 1985). Tilly writes: "European [emphasis added] social history' s central activity, as I see it, concerns reconstructing ordinary people' s experience of large structural changes" and involves "the search for links between small-scale experience and large-scale processes" (Tilly, 1985: 13). The existence of this common tradition in Spain can be seen in studies carried out by educational psychologists in the area of learning and understanding history 96
(Blanco and Rosa, 1997; Carretero et.al., 1997). Secondly, even within one tradition, there are disputes as to what should constitute the writing up of history, as evidenced by the structural/narrative debate (Burke, 1991) or that of the individualist vs. the holist (Gellner, 1973). So, on the one hand, there may be more similarities across cultures than Burke's statement above may imply, while at the same time there may be more differences within the same culture, depending on the historian's philosophy of history. Nevertheless, there is an underlying current of similarity, even between the different schools of historiography. While historians may differ in what they choose to portray as well as in their interpretations, i.e. explaining things in terms of individual character/intention or in terms of the conditions of society, geography, demographics, etc., the common ground is indeed that of explanation: of attributing cause, whether to individuals or to conditions, and of analysis of historic evidence. This leads to another common ground for all historians: they are working from a limited body of evidence: all of the evidence historians may gather together represents only "a minute portion of full living past reality" (Berkhofer, 1989: 185), yet it is the task of the historian to capture the past "in its full complexity or plenitude" (ibid: 185), a task which Charles Tilly, the historical sociologist, argues is futile, as "it is not humanly possible to construct a coherent analysis of the history of all social relationships: the object of study is too complex diverse, and big" (Tilly, 1985: 12). Still, this has been the main goal of "normal" history, as reflected in the following quote by Henri Marrou: This brings us to the essential point: explanation in history is the discovery, the comprehension, the analysis of a thousand ties, which, in a possibly inextricable fashion, unite the many faces of human reality one to another. These ties bind each phenomenon to neighboring phenomena, each state to previous ones, immediate or remote (and in like manner to their results). (Marrou, 1966: 192, in Berkhofer, 1989: 186). These words by Marrou show two important and interrelated features of history writing: interpretation on the one hand, and showing relationships between past events on the other. "What happened" is described and explained in terms of its context, in terms of "when it happened and what happened around it at the same time or over time" (Berkhofer, 1989: 186). This contextualizing approach entails explaining why things occurred as they did by the revelation of the specific relationships they bore to other events occurring in the circumambiant historical space ...[T]he Contextualist insists 97
that "what happened" in the field can be accounted for by the specification of the functional interrelationships existing among the agents and agencies occupying the field at a given time (White, 1973: 17-18, in Berkhofer, 1989: 186-187). The historian is thus involved in unifying events into a process, to explain events in light of their relationship to each other and to the process as a whole. This unified flow of events is organized into a narrative, into a unified exposition of events. As explained in Chapter 1, all of the texts chosen for this study belong to this "grand narrative" portrayal of history. It is felt that an analysis of the transitivity of the clause, of the participants, processes and circumstances, can reflect this tendency of history writing across the two cultures. An analysis of transitivity can illuminate the participants involved, and whether they are individuals or whether they are events, conditions, processes and so on. Also, it can show whether the authors are involved more in description as a means of explanation, or whether they use more cause/effect links, either through the process used and the connection the process affords between the participants in the clause, or through the use of cause circumstantials. What is of interest here is the relationship between the textual function and the experiential function, as already mentioned above. Therefore, the analysis of the texts that is explained in this chapter is the coming together of the two functions in the Theme of the clause, specifically by identifying the element of the system of transitivity which appears as Theme, while also identifying which process type that element proceeds from. Of special interest for the purposes of this study are the elements of the system of transitivity which writers of the history texts under analysis choose to place in Theme position. They may choose to thematize one of the participants in the clause, or some circumstantial element, or the process itself. In Spanish, especially, the process may be placed in thematic position. This serves to emphasize the idea that writers have choices as to which part of the experiential or ideational component of the language they choose to place in Theme position. Not only is analysis for Theme in terms of transitivity interesting from the point of view of determining which aspect of `reality' writers thematize, but it also may serve the purpose of showing differences/similarities between/within genres. Francis (1990) analyzed news reports, on the one hand, and news editorials and letters, on the other, to determine whether there exist differences in Theme choices in terms of transitivity in order to 98
determine whether there are typical choices in a given genre. The process types she found in her data fell into the following categories: material, mental, verbal, relational, and existential. Her results show a basic distinction between News reports, on the one hand, and both types of exposition on the other. Typical themes in the News reports were participants involved in material and verbal processes, with fewer relational participants/processes. Francis attributes this clearly to genre and the writer's purpose, which is to inform people about events; Also the high proportion of Sayers and verbal processes in the Themes is because a good deal of news is about what people say about the events. On the other hand, editorials and letters use fewer material and very few verbal processes, and there is a much higher percent of relational processes, reflecting more nominalization. Francis (1990: 53) states: nominalisation is a synoptic interpretation of reality: it freezes the processes and makes them static so that they can be talked about and evaluated. In other words they are no longer about what is happening, but what is being internalised and `factualised` by society as to the status of what has already happened: the relationships between events rather than the events themselves [emphasis original]. hence, more metalanguage and more self-reference is reflected in the choice of relational processes. Martin (1993) also finds differences in text types and their processes. He points out that the grammar of explanation depends more on material clauses, while the grammar of reports in science, which focus on classification and description, depends more on relational clauses. Even within types there are differences of delicacy. In reports, definitions are often used, and these rely on the grammar of identifying relational clauses, while classification requires a different type of relational clause. Martin also links writing in the humanities with a high use of abstraction, the effect of which is to "foreground relational clauses at the expense of material ones and to at the same time foreground nominal groups at the expense of clause complexes" (Martin, 1993: 219), and which often involves the expression of processes not through the verb but through nominal groups, thus again emphasizing the importance of nominalization in interpretive types of writing, such as history textbooks. 4.2 Hypotheses for Transitivity Analysis Based on the above discussion, several hypotheses can be advanced with regard to results of the transitivity analysis: 99
1. First of all, in keeping with work on Theme and genre previously mentioned in this study, there will be a tendency to thematize a circumstance of time or place over other types of circumstances, for example, condition. This is in keeping with the Taylor's (1983) study, who found that this was the case in the history textbooks he looked at, while science favored setting up a condition as Theme (see section 1.2.1), and in keeping with the general tendency in narrative to thematize circumstantials of time/place. 2. It is predicted that there will be a fairly even spread of material and relational participant types, as the texts involve both narrative of events (material) and description (relational), as well as interpretation and analysis. Interpretation and analysis, which in history are linked to cause/effect, involve both participants of material processes in this analysis in the form of Factors (see section 4.4.4 below), and participants of relational processes, usually in the form of Carriers. Related to this expectation is another: that the most common choices for Theme as participant will be the salient participant types in these categories, namely Actor, Factor and Carrier. 3. It is expected that existential processes will be in evidence, although in much smaller proportion than relational or material processes. For Halliday, it is not one of the main process types; however, in history textbook writing, there is some emphasis on the existence of things: on phenomena coming into existence and on the fluctuating state of the existence of phenomena in terms of growth and change. 4. Mental and verbal processes are expected in small proportion, as there is inclusion of what other historians have said and thought, and of what historical characters have said and thought. Francis (1990) found in her news reports that a full 22% of the clauses were verbal processes, and it has been said by Geoffrey C. Ward that "Journalism is merely history' s first draft". Does history writing rely as heavily on what people say/have said as journalism seems to do from Francis' study? Given the remoteness of the source material, this is doubtful. Thus it is predicted that Existents, Sensers and Sayers will be thematized to a limited extent, and that the frequency will be much lower than in report writing. 100
5. The thematization of Behavers in behavioral processes is expected to occur very infrequently, as history textbooks concentrate more on deeds and events, rather than on human physiological and psychological behavior. 6. Given that this study is based on the notion of genre, predicting that in texts written for similar communicative purpose there are similarities across cultures and languages, setting aside grammatical constraints, it is expected that the findings will be similar across the two languages. However, and related to this point, it is expected that there will be a higher number of participants as Theme in English, given that Spanish has the option of placing the process itself as Theme in a greater number of contexts than does English. Thus, there is also the expectation that Spanish will thematize the process itself more than English. While processes themselves are not the focus of this study, the participants cannot be determined without first identifying the process identified by the verb of the clause; it is the type of process being carried out which determines the functions distributed in the participant roles. Thus, notions of Theme/Rheme will be left aside in order to first explain systems of verbal processes. This is necessary to delineate clearly the model chosen for analysis in this chapter and to explain necessary adjustments to the model. In the textual and interpersonal analyses which follow in subsequent chapters, Halliday's models are used, and it would be consistent to use here his model of processes in the experiential function. However, for several reasons, which are explained below, Halliday's process model has been modified here in order to best reflect the process types found in the data. Before elucidating the final model chosen here, Halliday's model is discussed in section 6.2 below, along with two others. Then, space is given to explaining the difficulties inherent in assigning process labels to data from the textbooks. The assignment of process labels is necessary to thus determine the participant functions in the Themes and the Rhemes of the clauses. Then, we will return to Theme/Rheme in showing the participants, circumstances and processes appearing in the Themes and the Rhemes of the corpus. 101
4.3 Models
4.3.1 Systemic Functional Grammar
4.3.1.1 Halliday's Model
Transitivity is the system by which the grammar of the clause expresses the experiential aspect of meaning. According to Halliday (1994), reality is made up of processes, which can be divided into several types. The clause represents reality through these processes, which he initially, and informally, divides into "happening, doing, sensing, meaning, and being and becoming" (Halliday, 1994: 106). The process types are organized under, and construed by, the system of transitivity. The system of transitivity involves the process itself, the participants in the process, and the circumstances related to the process. Each of these processes has certain participant roles which are in some way related or involved in the different processes. These are as set out in Table 4.1:
Table 4.1: Process Types and Their Participants (Halliday, 1994)
Process type
Meaning
Participants
Material: action event
`doing' `doing' `happening'
Actor, Goal Range, Beneficiary (Recipient & Client)
Behavioral
`behaving'
Behaver
Mental perception affection cognition Verbal
`sensing' `seeing' `feeling' `thinking' `saying'
Senser, Phenomenon, Sayer, Target, Verbiage, Receiver
Relational attribution identification
`being' `attributing' `identifying'
Token, Value Carrier, Attribute Identified, Identifier
Existential
`existing'
Existent
There are several problems in applying Halliday's process types as they stand to the data from the history textbooks. One major problem is grammatical metaphor, which is not a problem inherent in the model or in the labels of the process types, but is due to the nature of the way reality is expressed through language. This problem will be discussed in section 4.4.3 below. The problems with applying the model in this study have more to do with the fact that Halliday's participant types do not allow for finer distinctions which recur in the data to be made. For example, applying the label `Actor', which Halliday defines as `the one
102
that does the deed' (1994: 109) in material clauses, seems to fit in clauses such as the following:
4-1. Eng 5 (5) Human beings roamed the earth for several million years... 4-2. Spa 6 (141) Los propietarios abandonan las ciudades. 4-3. Eng 8 (147) He governed Rome sternly and efficiently for nine years.
However, there are a number of clauses of the following type:
4-4. Eng 2 4-5. Spa 2 4-6. Spa 3
(109) The resulting expansion of families increased the importance - and the burdens... (71) Un aсo de escasez de cereales, ...podнa provocar todavнa un cataclismo. (14) ...la consolidaciуn de poderes monбrquicos estables,..., da como resultado un incremento notabilнsimo del comercio,...
While it certainly is possible to say that the resulting expansion of families, a year of scarcity of cereals, and the consolidation of stable monarchical powers all ' did the deeds' expressed in the clauses, an important distinction is being missed if we consider these Actors in the same way we would the human participants in the previous examples. It is not simply a distinction of animate vs. inanimate `causers' of something; in history textbooks, there is a great deal of expression of analysis, of attempting to explain why things happened, in short, of establishing cause-effect relationships. In many cases, it is the historian looking back on historical events who establishes the cause/effect links in retrospect: "the historian' s efforts to understand and explain the more recent past lead him to attribute to the more ancient past a role or significance which it did not possess until more recent events had occurred" (von Wright, 1971: 155, in Hallden, 1997). It is not the same to look back in time and state a material process in which an entity can be said to have done something at a given moment in time as it is to look back in time and state a material process based on an interpretation of what happened after an event or a number of events. In the case of human beings roaming the earth, there are participants in a material process who are doing something. But in the instance of, for example, the resulting expansion of families, it is more difficult to say that the resulting expansion of families is doing something in the same way. It does do something, but only in a retrospective view, in that it brings about an increase of the burdens on females. And this type of event, usually one which has taken place over a period of time, such as the expansion of families and the consolidation of power, is often involved in bringing about, leading to, allowing, something else to occur. As this is a feature of the texts, it seems important that this difference in type of Actor be reflected in the participant labeling.
103
Francis (1990) found that fewer relational and more material processes meant less nominalization in her data; that does not seem to be the case in the history textbook data. Eggins, Wignell and Martin (1993: 77) found that history textbooks use a great deal of nominalization, which allows them to "remove people, turn actions into things, and turn sequence into setting". Therefore, while, as will be seen later in the results, there are similar amounts of material and relational processes reflected in the total data, the inclusion of a great deal of material processes does not mean that there is not a lot of nominalization. And these nominalized entities do appear as participants in the clauses found in the data. The above account is based on a system of transitivity, in which processes may have one participant, in which case, the clause would be considered intransitive, or the process may extend beyond this participant to some other entity (e.g. Actor Goal in a material process, as in "John kicked the ball"), in which case the clause would be considered transitive. However, Halliday offers an alternative interpretation in which "the variable is not one of extension, but of causation" (1994: 163); this is ergativity. In a system of ergativity, there is one main, and necessary, participant, which he terms "Medium", as it is "the entity through the medium of which the process comes into existence" (ibid). One attractive aspect of this interpretation is it allows for an explanation of the following types of clauses, or ergative pairs: Figure 4.1:Ergative Pairs The boat sailed. Mary sailed the boat. The boy woke. The mother woke the boy. In these cases, Halliday analyzes the boat and the boy in each of the sentences in which they appear as the Medium, while "Mary" and "The mother" in the second sets both function as Agents, or external causers of the process, which still comes into being through the Medium. This account of processes seems attractive in the cases mentioned above of the resulting expansion of families and so on as it hinges on the idea of causation. An ergative explanation is also attractive in clauses of the following type: 4-7. Eng 8 (15) Warehouses and shops expanded to deal with the rising commerce of the empire. 4-8. Eng 10 (146) ...power passed to one of the few statesmen of prerevolutionary France... 4-9. Eng 2 (115) and the pot, the raft, and the wheel combined to provide the means to transport grain and other goods. 104
These clauses can form ergative pairs, i.e. we can say:
Figure 4.2: Ergative Pairs Warehouses and shops expanded power passed the pot, the raft, and the wheel combined
Warehouses and shops were expanded power was passed the pot, the raft, and the wheel were combined
In a system of transitivity, the second member of each pair, in the right-hand column, would be the more congruent one, in that there it is clear that the participants "warehouses and shops" and "power" and so on receive the action; in that interpretation, their function is that of Goal, and the Actor is omitted. However, and still in the system of transitivity, for the first member of each pair, in the left-hand column, the analysis would have to change, and "warehouses and shops" and "power" would now be Actors. This analysis can be seen in Figure 4.3 below:
Figure 4.3: Transitive Interpretation
Warehouses and shops expanded
Actor
process: material
Warehouses and shops were expanded
Goal
process: material
However, in Halliday's system of ergativity, in both cases the sentence Subjects are analyzed as Medium, which allows for more consistency of analysis:
Figure 4.4: Ergative Interpretation
Warehouses and shops expanded
Medium
process: material
Warehouses and shops were expanded
Medium
process: material
Thus, this analysis would allow the same interpretation for the examples above as it would for the following:
105
4-10. Eng 10 (122) The number of warships was increased. Hence, the number of warships, warehouses and shops, and power are all caused to do what is indicated in the verb process, and they are all Mediums in agentless constructions. Halliday, however, does provide for another participant type in this analysis: precisely that of Agent, which is some participant functioning as external cause. The following example illustrates the roles of Agent and Medium:
Figure 4.5: Agent and Medium
The infection causes the lymphatic glands to swell,
Agent
Medium
In Spanish, the type of agentless construction where the Subject of the verb is the medium of an ergative construction in which it can be presumed that an external agent exists, but is not expressed, also exists in the corpus. However, Spanish has a different way with dealing with this, as was seen in section 3.5.2.1.2, the passive reflexive with se. An example from the corpus illustrates this: 4-11. Spa 10 (69) El pensamiento ilustrado se extendiу a toda el бrea euro-atlбntica...
Here, we see that `illustrated thought' extended, or was extended, to all of the Euro-atlantic area. Thus, "se" is used in an ambiguous way; it is not clear if there is an external agent or not. In spite of this way of dealing with this type of construction, there is at least one instance in the corpus where it is not used; instead, an active construction is used:
4-12. Spa 7 (109) El pan mejorу de calidad.
which could form the ergative pair: "El pan mejorу/El pan se mejorу". To say that "Bread improved in quality" means to say "Bread was improved in quality", as bread probably cannot improve on its own. In the ergative analysis, then, "El pan" would be Medium in both members, while in a transitive analysis, it would be Actor in the first case, implying it did its own improving, and Goal in the second, which is more congruent. In sum, Halliday's system of ergativity is attractive for these types of clauses, and would allow the analyst to be consistent with the types of clauses that fit into ergative pairs and also to make distinctions in the data that through the system of transitivity might remain obscure; for example, the distinction between Medium and Agent, or outside causer, is an interesting one for purposes of a corpus of history textbooks, especially given that in a
106
transitivity analysis the two might be considered as having the same function, that of Actor, e.g.5
Figure 4.6: Actor, Agent and Medium
Things
change
Actor
transitive process: material
People Actor
change
things
transitive process: material Goal
Things Medium
change ergative process: material
People Agent
change
things.
ergative process: material Medium
The system of ergativity allows for more consistency in the analysis in this sense. Often a number of sociological, political, economical and geographical factors together affect some process or entity in some way, and it would be odd to express these factors as Actors, as doing deeds. The ergativity analysis would allow them to be expressed as Mediums. However, in spite of the attractiveness of basing the analysis for participant types on Halliday's ergativity system, it has a major drawback for the purposes of elucidating the different participants in the history textbook corpus. As outlined by Halliday (1994), the transitive function is divided into several different process types (see Table 4.1 above), while the ergative function is not. Therefore, in the ergative system, whether the process represented is material, mental, relational, and so on, the participant types remain the same. What might be functioning as Medium in an ergative gloss on a clause could be glossed in the transitive system as either Actor, Behaver, Senser, Carrier, etc., depending on the process type taking place. This lack of refinement would mean that many important distinctions would be lost, distinctions such as when a historical character is functioning as an Actor carrying out some material action or when the same character is being described, e.g. as a Carrier in a relational process. Indeed, of great interest to this study is a comparison of the amount of description with the amount of narration of events, of material
5 This example has been taken from Jeanette Winterson's Written on the Body. It is interesting to note that this phenomenon is noted by a non-linguist as a feature of the language; for Winterson, it is a mis-use of language in that: "Things don't change People change things. There are victims of change but not victims of things" (56-57). 107
happenings. The difference can be elucidated through a more detailed participant analysis than the ergative system allows. Given that Halliday's systems of transitivity and ergativity both have drawbacks as optimal tools for analyzing the history textbook corpus, another system, Fillmore' s case system, was considered. Before looking at that, however, two other models within the systemic school are considered briefly. 4.3.1.2 Other Systemic Models Berry (1975) and Fawcett (1980: 1987) base their discussions of the system of transitivity on the work done by Halliday. However, their models differ somewhat from Halliday's, and are interesting from the point of view of the problems raised here. Their models will not be discussed in full, however. Only those points which are of interest in terms of useful distinctions to be made in the history textbook corpus will be considered. Berry' s (1975) system is more ' delicate' than Halliday' s in that vshideedsimaterial processes into two types: action process and event processes. Action processes are those which usually have an animate participant in the role of Actor, while event processes are typically performed by an inanimate being. She gives the following as examples of event processes: A stream flows through that part of the valley. The car backfired noisily.(Berry, 1975: 151) A further difference between event and action processes is that event processes are not brought about intentionally, while action processes can be either intentional or unintentional. For unintentional action processes, Berry uses the term `supervention', and describes this as "a process which just happens" (Berry, 1975: 151). This discussion of events, actions and intentionality is obviously of interest to this study. In a clause such as the following: 4-13. Eng 3 (54) A plague broke out at Messina we have "a process which just happens", and this kind of declaration of the occurrence of some event is a feature of the history textbook data. Halliday's system does not allow this distinction to be made. Berry also offers a category of process type based on causation, which, as already mentioned, is a point of major concern in the textbook data. This category is a subdivision 108
of unrestricted processes, which are processes which can have either one or two participants (as distinct from restricted processes, which have a relatively fixed number). As Berry points out "[w]hen unrestricted processes occur with two participants there is usually an implication of `causation' in their meaning" (Berry, 1975: 157), i.e. the Actor can be thought of as causing the other participant to undergo the process, as in "John opened the door" meaning "John caused the door to open". Unrestricted processes, then, can be divided into `causative' and `non-causative', an example of the latter being "The door opened". This is a useful distinction as the Actor in a causative is not really the Actor of the process, but Actor of the causing, i.e. John does not open himself but causes the opening of the door. This type of analysis is attractive for clauses such as example 6.4 above (repeated here for ease of reference): 4-14. Eng 2 (109) The resulting expansion of families increased the importance - and the burdens - of the females. Here, the resulting expansion of families causes the increase of the importance and the burdens of the females, which is really the entity that undergoes the process of increasing. There is, however, a major drawback to applying Berry' s system to the data in a study of this type, which is based on Theme. In applying her analysis, the processes would be labeled accordingly - whether they were action/event, restrictive/unrestrictive, causative/non-causative. However, the participants would not show the difference; for example, with John opened the door and The door opened, both ' John' in the first sentence and ' door' in the second would be labeled as Actor. Therefore, the distinction between these types of participants would be lost. Given that this study analyzes participants in the Themes and the Rhemes of the clauses, and the analysis of the verbal processes of the clauses is only a step along the way to elucidating the participants, Berry's model is not optimal here. Fawcett' s (1980) system, on the other handi,s based more on the participant roles, rather than on the processes themselves. He distinguishes between two basic types of action which can be found "in varying proportions throughout the languages of the world" (Fawcett, 1980: 140): one `affected-centred' and the other `agent-centred'. These correspond roughly with the terms `ergative' and `transitive', respectively: however, Fawcett' s terms "have the advantage of reflecting the fact that it is the ' centrality' of a particular inherent role in each process type that determines its nature" (Fawcett, 1980: 140). In an affected-centred process, the central role belongs to the affected entity, while 109
the agent is an optional extra. Thus he introduces into the network the role of 'affected', which is that participant affected by the process of the clause. This is useful in analyzing clauses which were included as problematic in the discussion of Halliday's system above, and which are repeated here: 4-15. Eng 8 (15) Warehouses and shops expanded to deal with the rising commerce of the empire. 4-16. Eng 10 (146) ...power passed to one of the few statesmen of prerevolutionary France... 4-17. Eng 2 (115) and the pot, the raft, and the wheel combined to provide the means to transport grain and other goods. Each of the italicized items in the above examples would be classified in Fawcett's system as Affected. In Halliday's system of transitivity, the italicized items would be classified as Actors, which clearly does not represent the state of affairs expressed. Fawcett's system allows the state of affairs to be expressed more aptly for purposes of this analysis in the participant roles. Fawcett's account of transitivity, then, includes more categories of participant roles than Halliday's does. The motivation for this is that he is concerned with designing a set of realizational rules for English through a systemic network in order to develop a generative model. The purpose of this study, on the other hand, is descriptive, not generative. Therefore, the complexity of Fawcett's systems may serve to confuse, rather than to clarify, features of the texts in the corpus. Thus his model has not been chosen as a basis for analysis here. 4.3.2 Fillmore's Case Grammar Another model which could be used to analyze the data is Fillmore's (1968; 1971) case grammar. Halliday's focus in his account of transitivity is on the capturing of experience in the clause through the experiential function, "its guise as a way of representing patterns of experience" (Halliday, 1994, 106); in his account, the clause is taken as a "mode of reflection" of social interaction and events, and of our inner experience and states of being. Fillmore's account, on the other hand, comes out of a more mentalist approach, and seeks to establish a universal system of case relationships in the deep structure of sentences; he presumes these notions of case to be innate. Fillmore's case notions can be thought of as similar to Halliday's since they are given as a reflection of human experience, as they embody concepts "which identify certain types of judgments human beings are capable of making about the events that are going on around them, 110
judgments about such matters as who did it, who it happened to, and what got changed" (Fillmore, 1968, 24). This is not to imply that Fillmore and Halliday have the same motivation, but to imply that there is some similarity in the categories of case grammar and those of Halliday's account of transitivity. Indeed, Halliday (1968) refers to Fillmore's work in several instances in his account of transitivity. Fillmore does not posit different types of verb processes, as does Halliday; instead, he focuses on the particular case relationship(s) existing between the verb and the noun phrase(s) associated with the verb. The basic cases that he lists in his 1968 paper are:
Table 4.2: Fillmore's Cases
Agentive
the case of the typically animate perceived instigator of the action identified by the
verb.
Instrumental the case of the inanimate force or object causally involved in the action or state
identified by the verb.
Dative
the case of the animate being affected by the state or action identified by the verb.
Factitive
the case of the object or being resulting from the action or state identified by the verb,
or understood as a part of the meaning of the verb.
Locative
the case which identifies the location or spatial orientation of the state or action
identified by the verb.
Objective
the semantically most neutral case, the case of anything representable by a noun
whose role in the action or state identified by the verb is identified by the semantic
interpretation of the verb itself; conceivably the concept should be limited to things
which are affected by the action or state identified by the verb.
Later he changes Dative to Experiencer and Factitive to Result, and adds the following
cases (Fillmore, 1971):
Counter-agent Source Goal
the force or resistance against which the action is carried out. the place from which something moves. the place to which something moves.
Using Fillmore's cases to analyze the history textbook data has its attractions. First of all, he posits them as universal and includes data from other languages in his explanations. This makes it especially appropriate for a cross-linguistic study such as this one. Secondly, it was explained above that for purposes of a study based on participant/processes/circumstances as Theme, a focus which distinguishes different types of participant roles is optimal. Fillmore' s model allows for this type of focus, and useful distinctions can be made . For example, the category of Instrumental allows for a distinction between the "doer" of a deed (an Actor in Halliday's terms, and an Agentive in Fillmore's) and some sort of artifact or thing mediating between an often unexpressed "doer" and the
111
entity which receives the result. In a Hallidayan analysis, "values" in the following example would be analyzed as Actor: 4-18. Eng 1 (9) We are shaped by values,... (Eng 1: 9) The active counterpart of this passivized construction is Values shape us. This active/passive pair can be compared to Fillmore's exemplificatory use of "the key" as in: The door was opened by the key. The key opened the door. In the history textbook corpus, the distinction between an instrument used to carry out a process on some other entity and the entity actually instigating that process is an important one. In the sentence "We are shaped by values", if values is taken as Actor , as it would be in Halliday' s system, the notion that values are actually in some sense an instrument used by societies to help regulate the behavior of individuals and groups is glossed over: it is as if the instrument values exists on its own, independently of the very people they shape. Just as a key does not open a door of its own volition, values shape people only through society; therefore, they are instrumental, not agentive. Fillmore's category of Instrumental allows this distinction to come more to the forefront. Other useful categories from the point of view of the history textbook data are Objective and Factitive. Objective resembles Fawcett' s category of Affected above, thus the same advantage applies. This category illuminates the type of role played in a causative process where the emphasis is on the participant affected, and where agency is often difficult to place, or is placed on a series of events or historical phenomena. If we combine this with the previous category, Instrumental, a neat analysis results for clauses such as the following examples from section 4.3.1.1 above, repeated here (for ease of reference): 4-19. Eng 2 (109) The resulting expansion of families increased the importance - and the burdens of the females 4-20. Spa 2 (71) Un aсo de escasez de cereales, ...podнa provocar todavнa un cataclismo. Rather than analyze the underlined items as Goal, to which they would correspond in a Hallidayan analysis, they can be analyzed in Fillmore' s classification as Objective, which has the advantage of showing that they are an effect of the verbal process. To be more precise, the underlined item in the second example in the set, number 4-20, would be analyzed in Fillmore' s terms as a Factitive, asun cataclismo comes into being because of the action represented by the verb and set into motion by a year of scarcity of cereals. This is similar to 112
Halliday's Resultant (Halliday, 1970: 149) as in Sir Christopher Wren built this gazebo, with gazebo being the Resultant. In later writings on transitivity (e.g. Halliday, 1994), however, Halliday seems to have discarded this category. Nonetheless, this type of category, Factitive or Resultant, is useful for the history textbook data. While Fillmore's case of Factitive aptly captures the affected or resultative nature of the underlined elements, there is a problem in applying his categories to the above examples, namely in the italicized elements in examples 4-19 and 4-20. These do not quite fit the category of Instrumental, as they are not instruments being used by some other entity to bring about the desired result. They are more agentive than instrumental in nature. However, to label them Agentive does not quite fit, as they do not concord with the description given by Fillmore for Agentive. In sum, using Fillmore's cases does not resolve a problem mentioned above, which is the desire to reflect through the analysis the type of agent, or `causer' which is inanimate, and mainly, collective; indeed, "Notoriously the grammatical Subject of sentences written or uttered by social scientists is often not a man, or enumerated or characterized men, but groups, institutions, `cultures', etc. The proper study of mankind is human groups and institutions" (Gellner, 1973: 1). Fillmore himself points out the inability of the case system to cover this type of participant, or case, type; he describes the Agentive case as "the typically animate [emphasis added] perceived instigator of the action identified by the verb". Fillmore qualifies this by explaining that the "typically" is there because of such things as "robot" or, more interestingly for purposes of this study, "nation" as a "human institution" noun, for which, he points out "I know of no way of dealing with these matters at the moment, I shall just assume for all agents that they are ` animate'." (Fillmore, 1968: 24 fn31). A further problem, and the major one, with Fillmore's model is that it is based on "action" verbs, on processes which are dynamic, and not stative. This would not allow the analysis to show differences in use of descriptions of events, individuals, places, and so on, or of recounting actions, or of reporting what others have said about the past. Therefore, Fillmore's model has not been chosen as the basis for analysis in this section. 4.4 Halliday Revised: The Model Chosen for Analysis From the above discussion, it is clear that the model chosen needs to be able to illuminate certain phenomena in the data. It needs to distinguish between different types of 113
processes, i.e. descriptive states, mental states, actions, verbal projections, and so on. It needs to distinguish between different types of participants: especially those participants which are more affected by the process than instigators of it, and those which are phenomenal causers. Halliday's system is the most complete in that it distinguishes amongst the different type of processes. However, the participant types do not allow for the illumination desired for the data entirely. In order, then, to best analyze the data, the basic model chosen is that of Halliday, 1985a (1994), with some additions. Those additions are culled from the above discussion on Berry, Fawcett, and Fillmore, and on previous work done by Halliday himself and are included in order to illuminate some of the features specific to a corpus of history textbooks. These additions in terms of participants, processes and circumstances are discussed below, and the full system for each of the categories is given in its corresponding section.
4.4.1 Process Types
As mentioned above, the systemic models are optimal for analysis of the data in
terms of process types, as they distinguish among the three major types of processes:
material, mental, and relational, with the additional more minor categories of behavioral,
existential and verbal. Table 4.3 below reproduces the process types laid out in Table 4.1
above, and offers a brief explanation of each. The participant types will be discussed in
Section 4.4.2.
Table 4.3: Process Types
Process
Explanation
Material active processes, actions: someone/thing is actively doing something, and/or some-
one/thing is affected by the action expressed in the verb.
Behavioral active processes: typically a conscious being carrying out physiological-psychological
behavior.
Verbal
active processes with an associated result of verbal expression.
Mental
active processes, but having no associated overt external action, can be voluntary or
involuntary; processes of sensing .
Relational stative process; something (Carrier) is being described as having some attribute, or
something is being related to something else (identification: Identified-Identifier)
Existential stative process; something (existent) is posited as existing, coming into existence,
occurring or happening.
Initially, I had been planning on including two additional categories: Inchoative and Occurrence. Inchoative refers to anything which is coming into existence, as in examples such as the following:
114
4-21. Eng 1 (10) the way we look at the past and what we see there -the questions we ask and the problems we pose- arise out of current concerns. 4-22. Eng 2 (114) A trading economy thus began to emerge. 4-23. Spa 2 (10) Incluso en la polнtica aparecerбn fenуmenos de masas... Occurrence processes include the following: 4-24. Eng 2 (69) About 10,000 B.C. occurred the most important single event... 4-25. Eng 2 (92)The Neolithic Revolution first occurred among the hills... Halliday does include both of these types under Existence processes. In addition, they were not that frequent in the two corpora. Therefore, rather than consider them separately as originally thought, they have been subsumed here under Existential. In his discussion of each of the categories, Halliday points out the fuzzy nature of their boundaries. In fact, in analyzing the data, it was often difficult to decide exactly which type of process was represented by the verb. This is due to these "soft" boundaries, in combination with the metaphorical nature of the texts, in terms of both grammatical metaphor (see section 4.4.3 below) and to more traditional types of metaphor, such as personification. While it is tempting to simply state the obvious, that history textbooks rely heavily on metaphor for their explanations and give some examples to illustrate, that might obscure many of the similarities/differences in the types of reality expressed in the textbooks across the two cultures. Therefore, while deciding how far to analyze metaphor is always a controversial issue and definitely not an exact science, some participants have been categorized as representing a participant with which they would not normally be associated. One example of this is the use of verbs normally associated with verbal processes with non-congruent participants for that type of process, where the participant projecting the proposition would not be considered a congruent Sayer, especially with verbs like imply, indicate, show, demonstrate, signify, suggest. Halliday (1994: 142) says that these verbs have the dual property of functioning either as "saying" (a verbal process) or as "being (a sign of)" (a relational process). In other words, they are on the borderline between verbal and relational. He goes on to say that if the Subject is a conscious being, or if the clause it is contained in is projecting, then it is verbal. The instances of interest here are those involved in projecting clauses. They can be paired with similar uses which do not involve projection. Examples from the history textbook corpus follow; the first clause complex in each set shows projection (in parentheses): 4-26. 1. Eng 2 (48) The evidence from some nonindustrial societies today may suggest (that women provided most of the food supply through gathering,...) 115
2. Eng 3 (132) and this [that agricultural prices tended to stabilize] suggests a more dependable production. 4-27. 1. Eng 2 (68) scars on the walls seem to show (that people threw spears at the animals after painting them, as if thus to symbolize killing them). 2. Eng 3 (33) Almost every region of Europe from which we possess such records shows an appalling decline of population between approximately 1300 and 1450. 4-28. 1. Eng 5 (26) Excavations of fossils (remains of organisms) at a number of locations indicate (that H.s.sapiens goes back only about thirty thousand years). 2. Spa 9 (50) Esto [finding similar metals in different locations] indica una doble relaciуn de causa a efecto en el desarrollo de la navegaciуn, ... According to Halliday's direction, the first clause in each pair would be classed as a verbal process, given that they involve projection, while the second would be categorized as relational, given that they involve neither a conscious sayer nor projection. This is clearly a fuzzy border area, as the verbs are being used in similar ways in the examples. However, this study follows Halliday's direction due to an interpersonal consideration. Later, in Chapter 6 (section 6.1.3.2) it will be seen that a means which authors use to present a proposition as open to interpretation, or as a claim which is not their own, is through projection. Thus, the first type of each of the examples is considered to be a verbal process, while the second is considered to be relational. There are times when verbs are used in a more congruent manner in relationship to process type. At the same time, it may be the case that while the verb is used in a more congruent way, the participant(s) chosen may seem incongruent; for example: 4-29. Spa 1 (99) Grecia y Roma perfilan la Historia como un quehacer sistemбtico... 4-30. Eng 2 (16) Israel permanently conquered people' s imagintaions. It is difficult to imagine countries carrying out these processes, i.e. Greece and Rome do not congruently portray history, nor is Israel capable of congruently going out and conquering people's imaginations. One possibility here is to analyze the participants as "location", a relational participant category which Fawcett (1980) includes in his model, and thus as participants in a relational process. However, I have preferred to look at them as verbal and material processes respectively. The reason for this is that the first case is obviously an example of metonymy, where the country is used to refer to Greek and Roman writers, or historians. And in the second case, it is not the location which has conquered people's imagination, but the culture, in which case the Subject (here conflated with Theme) is a Factor in a material process. (for explanation of the participant types, see section 4.4.4) 116
Another difficulty with verbal processes is tied up with speech act theory, where saying is doing, and doing involves saying. This problem illustrates the blending of one type of process into another. For example, in the following clause 4-31. Spa 3 (47)...y podнa alegar sus derechos... the participant involved, in this case Alfonso X, can allege his rights to the throne as a direct descendent of the royal family. Presumably, the way to do this involves a verbal process: he would allege his rights by saying so. In fact, "podнa" conveys a somewhat ambiguous meaning. It can mean either "he could allege his rights" (but perhaps did or perhaps did not) or "he was able to allege his rights", meaning he did. I have chosen the latter analysis here, as further reading shows that he did claim his rights to the throne, and presumably did so verbally. To complicate the picture even more, at times mental, verbal, and material are blended in one process, e.g.: 4-32. Spa 3 (58) ...el Papa Gregorio X se opuso resueltamente a la elecciуn del castellano,. Opposing something can be a description of a mental state, similar to liking or disliking something. However, if we know he opposed it, he must have done so in some outward way; he must have expressed his refusal in some way. Therefore, in this case, again a verbal analysis was chosen, with "a la elecciуn del castellano" as Target. A similar analysis was given to the following: 4-33. Eng 8 (152) His successor Caligula [reigned 37-41) demanded deification at the age of twenty-five 4-34. Eng 9 (129) These settlers demanded a halt to Asian migration. In these examples, "deification" and "a halt to Asian migration" function as Targets in verbal processes. It is interesting to note that in discussing verbs such as insult, Halliday comments that "this type of clause is closer to the Actor + Goal structure of a material process" (Halliday, 1994: 141). Indeed, in the following example: 4-35. Eng 7 (63) In depriving "the swift and excellent" Achilles of his rightful war prize (the captive young woman Briseis), King Agamemnon has insulted Achilles' honor the analysis is one of a material process, as it is clear from the initial circumstantial Adjunct of manner that the way in which Achilles' honor has been insulted has been through a material process. 117
Another example of a blending, in this case of mental and material processes, is: 4-36. Spa 3 (77) y decidiу dar una compensaciуn a los infantes Alfonso's deciding to compensate the princes in return for not disputing the throne implies that he then did so. However, it is clear that the writers have decided here to express this historical happening as a mental process on the part of Alfonso rather than as a direct material process of giving; therefore, this is analyzed as mental. Another problematic area involves adjectives which have the same form as past participles. Quirk, et al (1972) point out the lack of a clear-cut distinction between the adjective and the past participle, and Alarcos Llorach (1994) says the same for Spanish. Quirk, et al (1972) add that the function is often indeterminate when no clear indicator is present (e.g. a by agentive phrase). They suggest that the "participle interpretation focuses on the process, while the adjective interpretation focuses on the state resulting from the process" (ibid: 244). This is not very comforting here, as then it is up to the text analyst to interpret the function of the participle/adjective. The following set provides an interesting case in this sense: 4-37. Eng 7 (129) They were trained in the arts of war 4-38. Eng 7 (134) Spartan soldiers were better trained and disciplined The first denotes an action and has been analyzed as a material process. This interpretation is arrived at also by the surrounding context. If we look at the clauses preceding example 437: Eng 7 (127) The Spartans learned only one craft, soldiering, Eng 7 (128) and were inculcated with only one conception of excellence: dying in battle for their city. we see the focus on learning/teaching, thus leading to an interpretation of clause 129 in example 4-37 as a passivized construction with `trained' as past participle, and thus encoding a material process. However, example 4-38 denotes a state, evidenced by the modification of `trained and disciplined' by `better'; also, the further context here serves to provide support for this interpretation. The clause immediately following clause 134 in example 4-38 is: Eng 7 (135) and were more physically fit than other Greeks. Thus, clauses 134 in example 4-38 and clause 135 are both relational clauses. The focus in these two clauses is on attributes of the Spartans, while the focus in clauses 127-129 is on the material process of teaching/learning. 118
However, there are cases where the interpretation could more easily go either way, and are evidence of the vexing nature of the slipperiness of participles and adjectives. An examples from the Spanish corpus is: 4-39. Spa 4 (80) No todos los problemas estбn resueltos. This can be analyzed as a passive construction, with an implied agent (someone has not solved the problems) and therefore material.. However, the more congruent form of the passive construction with no agent in Spanish is: No se han resuelto todos los problemas The construction in example 4-39 seems to focus more on the state of the problems, rather than on the process of solving as does the rewording. Therefore, in this example the adjectival interpretation has been opted for, thus classifying the clause as a relational process. An example from the English corpus is: 4-40. Eng 7 (99) for its citizens were intimately involved in the political and cultural life of the community. Here again the verbal phrase "were intimately involved" seems to reflect more an attribute of the citizens rather than an action, which could perhaps be better expressed as " its citizens involved themselves intimately". As the quote by Quirk, et al (1972) above reflects, these cases are contentious. Other examples which are contentious in that I have opted for an analysis which reflects the "reality" behind the process, rather than the congruent meaning of the verb, include the verb "suffer", which in many senses seems to indicate a mental state. However, the examples in which it is used often have a country as participant e.g.: 4-41. Eng 10 (102) France suffered particularly from the rigidity of its colonial system, the inferiority of its navy, and the very mediocre abilities of most of its statesmen. 4-42. Eng 10 (179) Spain suffered comparatively little damage from the great war over the succession to her throne that was fought in the early 1700s. 4-43. Eng 6 (21) Almost all of northern Europe suffered a terrible famine in the years 13151317. 4-44. Eng 6 (47)Without woolen cloth, the businesses of Flemish, Hanseatic, and Italian merchants suffered. 4-45. Spa 3 (10) Asia sufre al igual que Europa la invasiуn mongola. 4-46. Eng 6 (175) Densely populated Italian cities endured incredible losses. 4-47. Eng 7 (151) In the next century, aristocratic regimes experienced a social crisis. These have been analyzed as material processes in which one participant is affected by an action initiated by another as Affected/Material/Factor (see section 4.4.4 below for 119
participant types). This is similar in a sense to Senser/Mental/Phenomenon in that one participant is affected and perhaps altered in some way by another. However, the clauses here do not fit Halliday's criteria very well for mental processes. First of all, the first participants are not human. They could be said to be endowed with human consciousness; however, the process seems to refer more to a change undergone (or not undergone) by a place, rather than a mental sensing on the part of a conscious being. Also, the second participant cannot be a fact in these clauses. It is different to say: Almost all of northern Europe suffered from the fact that there was a terrible famine in the years 1315-1317. than it is to say: Almost all of northern Europe suffered a terrible famine in the years 1315-1317. The first suggests something about northern Europe, that people suffered from a famine (thereby also endowing Northern Europe with a consciousness) while the second suggests that a terrible famine did something to Europe, affected it somehow, without it needing consciousness. This is a subtle difference, but the second clause type goes further in not fitting the criteria of mental processes due to the fact that, in the present tense, these clauses do work in the continuous quite well; we can say "Almost all of northern Europe is suffering from a terrible famine". This serves to reinforce the notion that the two clauses above are indeed different, as it is possible to say: Almost all of northern Europe was suffering a terrible famine in the years 1315-1317. but it sounds odd to say: Almost all of northern Europe was suffering from the fact that there was a terrible famine in the years 1315-1317. Therefore, these have been classed as material processes, although the participants are not Actor and Goal, but Factor and Affected (see section 4.4.4 for participant types). Then there are the following examples, which are expressed as relational processes; however, they are essentially material in nature: 4-48. Eng 6 (63) With ships continually at sea, their rats too were constantly on the move, 4-49. Eng 6 (101) A determined rat had little trouble entering such a house. The first, 4-48, is obviously expressing an action, not a state, so this one has been analyzed as a material process. The second, however, is not so clear cut. It refers in a sense to a quality of a determined rat, the quality of having no problem or difficulty in entering such a 120
house. It does not reflect the process of the rat entering a house in the same way as the first one reflects the process of rats moving. Therefore, while the first has been analyzed as material, the second has been analyzed as relational. Other clauses are difficult to analyze simply because of their highly metaphorical nature. For example, in the following clause: 4-50. Spa 7 (14) ...nos hallamos ante un crecimiento del 70... there is a highly metaphorical representation, "we find ourselves before a growth of..." of a process which could perhaps more congruently be represented as an existential process: "There was a growth of 70 percent". However, the authors have presented it more as a mental process, with the growth as Phenomenonon, and "we" as Sensers, which is how is has been analyzed here. This is similar to: 4-51. Eng 7 (44) The Dark Age saw the migration of Greek tribes from the barren mountainous regions of Greece to more fertile plains, and from the mainland to Aegean islands and the coast of Asia Minor. What seems to be happening behind the grammar of this clause is an expression of an existential process: that a migration of Greek tribes took place during the Dark Ages. However, in this case, the authors have expressed it as a mental process, and this is how it has been analyzed here. The above are an illustration of the problems involved in analyzing the clauses for process types. These problems are inextricably related to problems in analysis of participant types, and to notions of grammatical metaphor, which is discussed in section 4.4.3 below. There is the difficult question of how far to go in analyzing the semantics of the clause in terms of congruence and "reality". For example, should the analyst rephrase the above examples as existential clauses in doing the analysis? The short answer here is "no", and a longer answer, which explains the "no" is given in section 4.4.3 on grammatical metaphor, following an initial discussion of participant types. 4.4.2 Participant Types Before looking at the participant categories, a discussion of other problems in applying the analysis to the data, especially in relation to the discussion in section 4.3.1.1 above on transitivity and ergativity, is necessary in order to provide some background as to decisions made in modifying Halliday's participant type scheme. An illustration of the problems is Halliday's major participant type involved in a material process, Actor, which is 121
an entity that "does" something. As mentioned above, Francis (1990) found fewer material processes as compared to relational processes in some of her data. She attributes this difference to a difference in nominalization: fewer material processes meant less nominalization, as the material processes in her corpus were used to inform people of events. History textbooks also inform people of events, and these events can fall into two rough categories. The first are actions carried out by individuals or specified groups of people, and can be transitive or intransitive: Intransitive: 4-52. Spa 2 (8) Estos hombres se desplazan; 4-53. Eng 3 (3) The Christian West fought to halt the expansion of the Muslim Turks. 4-54. Eng 6 (192) They retaliated with such measures as the English Statute of Laborers 4-55. Eng 8 (147) He governed from Rome sternly and efficiently for nine years; Transitive: 4-56. Eng 8 (164) He murdered his wife and mother and Claudius' son, 4-57. Spa 4 (62) Y, efectivamente, A. Evans encontrу documentos escritos en Creta. 4-58. Spa 2 (133) Algunos estados pusieron trabas a la emigraciуn de sus hombres; These processes + participants fit nicely into a Hallidayan transitivity analysis of material processes. It is clear that a specific entity is `doing' something; in the first set, the `doing' is confined to the Actor, and in the second set, the `doing' is directed at, or extends to some other entity (Halliday, 1994: 109). This is the type of event we can expect to find in a narrative such as one finds in a newspaper account of a newsworthy story or in a recounting of a past event. As Francis (1990) found in her data, this type of event reporting does not feature heavy nominalization. There is another type of event reporting in history textbooks, however. With this type, the process typically extends to some other entity beyond the entity involved in `doing' the process, and is therefore usually transitive. The relationship expressed in the process is causal. Finally, this type can, and often does, present a great deal of nominalization: This type of reporting is exemplified below: 4-59. Spa 2 (31) la emigraciуn de los europeos a otros continentes hizo surgir "nuevas Europas". 4-60. Eng 3 (9) The partial failure both of the medieval economy and government and of the established systems of thought and value facilitated change The noun groups in Subject position in the above examples encapsulate whole events in themselves. Here, relationships are encoded in nominal groups, and could more congruently be expressed as clauses, for example "Europeans emigrated to other continents", and "The medieval economy and government partially failed". This brings the discussion to the notion 122
of grammatical metaphor, which must be considered before resuming the discussion on participant types. 4.4.3 Grammatical Metaphor Grammatical metaphor is "a substitution of one grammatical class, or one grammatical structure, by another" (Halliday, 1993a: 79) It occurs when "a semantic configuration that would be represented congruently (non-metaphorically) by one type of clause is represented metaphorically by another" (Halliday, 1994: 57). Martin (1991) explains congruence as the natural relation between semantic and grammatical categories: "people, places, and things are realized nominally; actions are realized verbally; logical relations of time and consequence are realized conjunctively, and so on" (328). In the examples cited at the end of the previous section, nominal groups encode semantic configurations which would more congruently be expressed by full clauses, as the nominalization encodes both participants (e.g. Europeans) and processes (e.g. emigrated). . For Halliday, nominalizing "is the single most powerful resource for creating grammatical metaphor" (1994: 352). Martin (1991) also discusses this phenomenon in relationship to its appearance in history textbooks, in which events are often realized nominally instead of verbally. He comments on the grammatical metaphor used in history textbooks with regard to cause/effect relationships, where these are expressed not through cause/effect conjunction, but through the encoding of the cause/effect relation in the verbal process. He gives the example: The enlargement of Australia's steel-making capacity, and of chemicals, rubber, metal goods and motor vehicles all owed something to the demands of war. (Martin, 1991: 314) In the history textbook corpus used in this study, this same phenomenon appears quite often, as the last two examples in the previous section demonstrate. The reasons for this appearance are most probably several and interwoven. This kind of abstract reasoning is a necessary tool for historians to explain the consequential flow of historical without sounding "childish and naive" (Martin, 1991:332). Martin (1991) questions the use of nominalization as to whether the underlying reason is one of function or one of status, and Martin (1993) provides thematic motivation for its use: The significance of grammatical metaphor is that it is the grammar's most powerful resource for packaging meanings - for grouping them together into Theme and New. Grammatical metaphor re-textures the clause, allowing it to participate in its context in ways appropriate to the organization of texts as text. (Martin, 1993: 242) 123
Halliday (1993b) reiterates this in saying that, through these verbal representations, there is a movement from a thematic Given to a rhematic New; indeed, "this movement in time construes iconically the flow of information" (Halliday, 1993b: 92); we have seen in section 3.2 that Enkvist (1984: 56) also speaks of "experiential iconicism", in which the clause construes experience in a pattern similar to our experience of the world. The above mentioned history discourse is similar to the type of scientific discourse which Halliday (1993a) describes when he explains the reasons why nominalization became prevalent, first in scientific writing and then spreading to other genres. In this scientific discourse, reports on experiments called for a step by step encoding, in which there was "a constant movement from `this is what we have established so far' to `this is what follows from it next'" (ibid: 81). This type of encoding is done most effectively through the single clause, in which each of the steps is transformed into a nominal group, joined by a verb expressing the way in which the second follows from the first. (ibid: 81). Thus, if example 4-60 in the previous section were to be expressed more congruently, the writer would have to express the nominalizations as clauses and explicitly express the causal relation as a conjunction. The result would be something like the following:
The medieval economy and government and the established systems of thought both failed partially. Therefore, there was change. This loses both the sophistication and the effectiveness of the original wording, as well as the thematic configuration, the iconic construal of the Theme/Rheme pattern of the single clause. However, there are other thematic motivations for nominalization besides iconic construal. Through nominalization, processes are reworded as nouns, and thus function in the transitivity of the clause as participants. This transformation makes it difficult to analyze a clause in terms of its transitivity; for example, an analysis of one of the examples from the history textbooks could be carried out as follows:
The timelines produced for this edition seek to overcome these limitations: (Eng 1: 121)
Actor
material process goal
However, a more congruent analysis might be the following:
We produced
timelines for this edition in order to overcome these limitations
Actor material process goal
Circumstantial Adjunct: Cause: Purpose
124
This more congruent analysis moves a process out of the participant function and into its more congruent role as an expression of the process in a way which also more aptly portrays reality. In the first instance, "timelines" have been personified as actively attempting to do something, where in the second, it is the authors who "do" something, and then the reason for doing so is given. This more congruent analysis, then, also shows the causal relation which is missed in the first analysis. However, changing the wording in order to give a more congruent interpretation of the reality expressed in the textbooks means, again, losing sight of issues of thematic configuration. One consideration here is the personal pronoun "we", thematized in the more congruent reworded version. The authors may wish to downplay their active role in producing the book (see Chapter 6), and thus thematize a feature of the book instead. The question becomes then: "How far should one pursue the analysis of ideational metaphors?" (Halliday, 1994: 353). Halliday's answer to that is: "There can be no universally valid answer to this question; it depends on what one is trying to achieve" (ibid). One of the aims of this study is to determine which participants, processes or circumstances the authors of the corpus place in Theme position, and what is the role of these participants processes and circumstances in the on-going, unfolding reality as portrayed in the different texts. Therefore, any unpacking which alters the order of clause constituents to the extent of the example given above is clearly not in the interest of this study in terms of thematic choice and thematic progression. However, authors of history texts seek to explain and to make causal connections between different events, in addition to reporting actions, on the one hand, and describing people, places, events, and so forth on the other. In order to compromise between the desire to show this feature of the texts without moving too far away from the author's encoding, new categories of participant functions have been created solely for the purpose of this study. The suggestion is not that Halliday's system is somehow lacking in terms of semantic categories; grammatical metaphor will always prove elusive to a classificatory system of semantic analysis. However, in order to best portray goings-on as they are expressed in history textbooks, with prevalent use of abstract language, nominalization, and causal relations, some additions have been made to Halliday's analysis in terms of participants. 125
4.4.4 Participant Types Revisited Halliday's categories (see Table 4.1 in section 4.3.1.1 above) of participant types has been retained here with some additions. These additions are included in an attempt to capture the essence of the type of participant mentioned in the previous section. Examples of the type of entity with which we are dealing are included again here (types are underlined): 4-61. Spa 2 (63) Factores culturales y psicolуgicos inciden en este proceso. 4-62. Eng 1 (9) We are shaped by values,... 4-63. Eng 1 (39) What customs and institutions sustained those beliefs? As explained above, there are many examples of this type of processes, which have some features in common. First of all, the processes are all related to some extent to cause. Also, the Actors, which are underlined, are inanimate; furthermore, most are abstract. It would be possible to classify these more incongruent forms according to Halliday' s process and participant types, and, again, simply point out that there is a high instance of grammatical metaphor at work in the texts. However, given the nature of the subject matter, in which authors often express cause/effect relationships, implicitly or explicitly, it seems to be of greater benefit to the analyst to reflect this in the analysis by expanding on Halliday' s participant roles in this case. So, for the type of process in which an inanimate, usually abstract, entity brings about or leads to some other event, the label of material process will still be used. However, rather than code the `doer' of the action as `Actor' , I have added a category of participant type which I shall call `Factor' . This is based in part on Fillmore' s (1968; see section 4.3.2 above) Instrumental for something which seems very similar: "the case of the inanimate force or object causally involved in the action or state identified by the verb". He gives the example of "The key opened the door", where the key is instrumental. However, a key is a concrete instrument, while a Factor is some inanimate, usually abstract, entity which acts in some way upon another entity. This acting upon some other entity is usually seen in retrospect under the analytical eye of the history textbook writer. Also, there are other entities involved in material processes in the history textbook data which differ in important ways from Halliday's participants, and which are worth distinguishing. I have termed these `Result' , based on Halliday's (1970)Resultant and Fillmore's (1968) Factitive, later changed to Result (Fillmore, 1971), and ` Affected ' , based on Fawcett (1980). The participant `Result' is one which comes into existence through the 126
process carried out by an Actor or a Factor; the following is an illustration, with the Factor in italics and the Result underlined: 4-64. Eng 6 (206) Lack of confidence in the leaders of society, lack of hope for the future, defeatism, and malaise wreaked enormous anguish... `Affected' refers to a participant which has in some way been changed or altered by an Actor or a Factor. It often includes another participant which is a state brought about by the action of the verb, and which is termed Resulting Attribute (Downing and Locke, 1992: 95). In the following example, the Factor is in italics, the Affected is underlined, and the Resulting Attribute is included within brackets: 4-65. Eng 2 (9)Permanence, complexity, and sophisticated social organization transformed the agricultural village (into something we can recognize as urban, the form of habitation crucial for the development of civilization). These additions to material process participant types are the only changes from Halliday's classification. Other participant types which had been under consideration for addition are a set which Fawcett (1980) includes as relational process participant types: Located/Location. (Fillmore also includes a similar type, Locative, in his model). Examples where the identification of these participant types might prove useful include the following (Locateds are underlined and Locations are in brackets): 4-66. Eng 9 (11) for between 1815 and 1932 more than sixty million people left (Europe). 4-67. Eng 9 (28) As Figure 26.3 also shows, people left (Britain and Ireland)... However, rather than treat the underlined participants as Located, I have preferred to treat them as Actors in a material process, as the emphasis is not so much on their state, or on where they are located, as on the process of their leaving. This means that the participant in brackets is then treated as Goal, rather than as Location, which may be a loss to the analysis in terms of an account of total treatment of location via participants and circumstances. However, locations act as different types of participants throughout the clauses, and to attempt to code them as a separate participant outside of a classification of processes (e.g. to code all examples of location as Location) means to lose sight of other roles they play, and to force other participants into the role of Located. Therefore, these participants are analyzed as taking on a participant role according to the process indicated in the verb. The complete set of processes and participant types appears in Table 4.4 below, with the above-mentioned additions. 127
Table 4.4: Process and Participant Types
Process type
Meaning
Participants
Material: action
`doing'
Actor
Factor
Affected
Goal
Range
Beneficiary
Result
Resulting Attribute
Behavioral
`behaving'
Behaver
Mental perception `sensing' `seeing' Senser
affection cognition `feeling' `thinking' Phenomenon
Range
Verbal
`saying'
Sayer
Target
Verbiage
Receiver
Relational
`being'
attribution
`attributing'
Carrier
Attribute
identification
`identifying'
Identified Identifier
Existential
`existing'
Existent
4.4.5 Circumstances
Circumstances comprise the third major transitive function of the clause.
Circumstantial elements serve to give additional information related to the process of the
clause and its participants in relation to location in space or time, condition, manner, cause,
contingency, and so on. According to Halliday (1994), they are typically expressed through
prepositional or adverbial phrases; to these I have added dependent clauses. The rationale
behind considering dependent clauses as part of their superordinate clause is explained
above in section 3.4.4. Halliday (1994: 156) provides an inkling of support for this analysis
in discussing the semantic relations of circumstances of contingency, which he points out
are often realized clausally. When circumstantial elements, whether they be adverbial or
prepositional phrases or dependent clauses, appear first in the clause, they are analyzed as
the ideational Theme for that clause (see section 3.5.1.1). The Rhemes of the clauses were
also analyzed for Circumstantial elements, but here only those phrases functioning as
Circumstantials which could have been brought out into Theme position are considered. In
other words, the elements which are analyzed in Rheme position are those whose order in
the clause could be changed; an example is:
128
4-68. Eng 1 (75) For some it was primarily the history of the working class, often narrowed to be the history of labor movements.
The underlined element in the example above can also be considered a reduced relative
clause, in that it could be expressed as "which is often narrowed to be the history of labor
movements". However, as it is expressed without the `which is', it could be placed at the
beginning of the clause, in Theme position. Therefore, it is counted in the analysis as a
separate element of the clause, as a Circumstantial element, while full relative clauses are
not considered.
Having decided on a Hallidayan approach to the analysis of processes and
participants, it follows logically that the analysis of third component of the semantics of the
clause, the circumstances related to the clause, also be guided by this same approach, with
one minor addition: Result, which is explained after Table 4.5. The Circumstantial types
can be seen below in Table 4.5. Most are exemplified through Adjuncts. However, Result
and Condition are exemplified through non-finite and finite dependent clause, respectively,
as this is their most common form of expression.
Table 4.5: Types of Circumstantial Elements
Type
Subtype
Example
Extent
Spatial
Throughout Europe
Temporal Durante muchнsimos siglos
Location
Spatial
In parts of the Middle East
Temporal in the 1950s
Manner
Means
In this way
Quality
Both individually and collectively
Comparison Unlike the Spartans,
Cause
Reason
As a result of this "Agricultural Revolution",
Purpose
For purposes of hunting and protection,
Behalf
Para Irlanda...
Result
...thereby rooting ourselves more deeply...
Contingency
Condition If the model we are constructing is valid,
Concession In spite of the virulence of pneumonic plague,
Accompaniment
Except in size,
Matter
By the word "civilization"
Angle
Para los historiadores,
An addition to Halliday's categorization of circumstances falls into the general category, Cause, under which Halliday (1994: 151) includes three subcategories: Purpose, Behalf, and Reason. A further subcategory in the history textbook data is that of Result. Circumstantials of Result represent a result which comes about because of the action expressed in the clause. Result is different from Purpose in that it is more of an accidental result of the 129
action, the action was not carried out with that express purpose in mind. Result is typically expressed in gerundial phrases, or by prepositional phrases of the `with [the result of]' type .Examples of Result from the corpus are (Results are underlined): 4-69. Eng 3 (134) Agriculture was now considerably diversified, benefiting the soil,... 4-70. Spa 3 (137) ...tuvo tambiйn que levantar el sitio..., perdiendo asн la ъltima oportunidad 4-71. Spa 5 (50) ...el rйgimen territorial...es sustituido por el vallado de las fincas...con la consiguiente eliminaciуn de los mencionados servicios... 4-72. Spa 9 (77) Su cronologнa puede establecerse del 1000 al 900 a.C., ...equivaliendo al Bronce Final II en Francia. 4-73. Spa 6 (4)...se encuentra exenta de racionalismo, siendo, por tanto, muy distinta a la moderna. It is interesting to note in the last example that the expression of Result is underscored by the inclusion of the term por tanto (therefore). Indeed, in the gerundial phrase types, the word `thereby' could be inserted in the case of Results. There are several constructions which do not fit under Halliday's classification as described in his work (1994). However, semantically they express meanings similar to the construction types he gives, and, therefore, they have been classified accordingly. Constructions formed with the present and past participles are, perhaps, the most slippery, as they represent a range of meanings, perhaps because they are whole clauses, and thus they contain a verbal process. Examples from the corpus follow (the item in question is underlined), followed by a possible gloss, which is then accompanied by the classification given in the analysis: 4-74. Spa 4 (123) Dejando a un lado las caracterнsticas de su arquitectura, se observa que.. This gerundial phrase can be glossed as "If we leave aside the characteristics of its architecture," , so it has been analyzed as Contingency/Condition. 4-75. Eng 5 (7) Viewed in relation to more ancient life forms..., humans are latecomers. At first glance, it is tempting to analyze this as a circumstantial element of Manner, specifically, Quality, as if the question "How are humans viewed?" had been asked, and the answer were "In relation to more ancient life forms". But then the process expressed in the main clause would have to be different; it would have to read "By viewing humans in relation to more ancient life forms, we see that they are latecomers". However, the main clause encodes a relational process which is contingent on the process expressed in the gerundial phrase; thus it could be rephrased as "If we view humans in relation to more 130
ancient life forms, they are latecomers". Therefore, the relationship of example 4-75 has been established as Contingency, specifically, Condition. There are some gerundial phrases which can be glossed as expressing the meaning "with"; therefore, they are analyzed as Accompaniment. Examples follow from the Rhemes of the clauses, with a gloss in parentheses: 4-76. Spa 9 (90) El mбs importante parece situarse hacia el centro de Portugal,...siendo caracterнsticas del mismo las hachas... (with axes being characteristic of the same) 4-77. Spa 7 (79) Mayor vitalidad mostraban las ciudades perifйricas, sobresaliendo Barcelona,... (with Barcelona standing out) Another meaning expressed often in textbooks and other academic texts is metatextual in nature. It is rarely Thematized, and it usually refers to another section of the text or to a chart or graph. Examples are: 4-78. Spa 1 (68) ...como vemos a continuaciуn. 4-79. Eng 10 (131) ...as we have seen. In addition to their having a metatextual nature, they can be compared to the following: 4-80. Spa 1 (83) ...como creen muchos,... 4-81. Eng 5 (70) ...as shown in prehistoric cave paintings. The above two sets have been analyzed as Angle. Perhaps this not quite in the same vein as Halliday's intention, as he relates Angle to the Sayer in a verbal process. However, they do attempt to give weight or authority to the proposition with which they are connected, much the same as "according to" as in the example in Table 4.5 under Angle "According to the chroniclers..." . In the case of the first set (examples 4-78 and 4-79), the weight is given through the text itself, while in the second, what many believe (example 4-80) and prehistoric cave paintings (example 4-81) provide the weight. There are some constructions which fall outside of the analysis in the experiential function, as they do not represent participants, processes or circumstances. These are also difficult to pin down; Quirk, et al (1972: 636) call these either "verbless adverbial clauses" or "attribution appositives", depending on their function in the clause. They provide the following examples to distinguish between the two: Ron Pall, a blatant liar, was expelled from the group. Ron Pall, a blatant liar, used to be in my class at school. (Quirk, et al, 1972: 637) The first in the set is a verbless adverbial clause as it can be interpreted as causal. Verbless adverbial clauses can be interpreted usually as either causal or concessive, e.g. "The heir to 131
a fortune, his friend did not need to pass the examination. The second sentence in the set above offers no motivation for cause or concession, and is thus an attribution appositive. There are five examples of this type of construction in initial position in the history textbook corpus: 4-82. Spa 10 (74)Filуsofo, poeta, crнtico literario y autor teatral, sus teorнas estйticas... ejercieron amplia influencia en el siglo XVIII. 4-83. Eng 10 (77)A heavy drinker and a devotee of bawdy stories, he established the English politician's tradition of the long country weekend in order to indulge his passion for hunting. 4-84. Eng 1 (100) A history of social processes that tend to move at a glacial pace, it often lacks sharp chronology. 4-85. Eng 1 (101) A history of structures more than events, it tends to be organized around problems rather than the biographies of individuals or the story of a nation. 4-86. Spa 2 (146) Proceso poco acusado hasta 1845, comenzу a adquirir volumen a partir de las crisis econуmicas y polнticas de los aсos 1846 a 1848, None of them seem to provide a direct causal or concessive link with their accompanying clauses. Thus, they have been analyzed as attribution appositives, and have therefore been left out of the transitivity analysis. 4.5 Method of Analysis The overall method of analysis for this study is explained in section 3.6. It was explained there that all of the clauses were analyzed for Theme type in terms of process, participant, circumstance, and so on. For this section, two overall types of analyses were carried out, one for the Theme and one for the Rheme. For the Theme, the following analyses were realized: - First, the type of process expressed in the clause was determined. If the verb itself was in initial position, it was labeled as Theme. - Participant roles were assigned to each of the Themes which function as a participant. - Circumstantial functions were assigned to thematic prepositional and adverbial phrases, as well as non-finite clauses. - Circumstantial functions were assigned to thematic dependent clauses. And for the Rheme: - Participant roles were assigned to Subjects and Complements appearing in the Rhemes. 132
- Circumstantial functions were assigned to prepositional and adverbial phrases and non-finite clauses. - Circumstantial functions were assigned to rhematic dependent clauses. In each of these two analyses, all instances of each of the analyzed types were counted and collated for each of the texts. Given that the number of clauses for each text is different, percentages were then calculated by dividing total number of instances of each of the types by the total number of clauses for each of the texts. 4.6 Results Before looking at participants and circumstances in Theme, the overall results for process types is given a cursory glance. Chart 4-1 below displays the results of the analysis of the clauses for process type, along with percentages that each type represents in terms of the overall number of clauses for each text. Chart 4-1: Percentage of Process Types in English and Spanish
45% 40% 35% 30% 25% 20% 15% 10% 5% 0%
English Spanish
Material Relational Mental Verbal Existential Behaviora l
From this chart, it is clear that there is a difference between Spanish and English in terms of process type, which will, of course, have an effect on the type of participant appearing as Theme. The English data employs almost 10% more material processes than is reflected in
133
the analysis of the Spanish data. The Spanish data makes up this difference with a slightly higher count of relational processes, one and a half times as many both existential and verbal processes, with mental and behavioral processes showing greater consistency across the corpora. In section 4.6.6, the implications of this for participant type in Theme position is discussed. But first Theme choice in terms of grammatical features is considered.
4.6.1 Functions of the Themes in the Corpus
The first consideration here is the second area of hypothesis: that of type of
participant as Theme. It was hypothesized in section 4.2 that Spanish would use a higher
number of processes as Theme than would English. Table 4.6 shows the Theme types in
terms of transitivity roles and in terms of those types which function outside of the system
of transitivity:
Table 4.6: Themes in Terms of Transitivity Roles
English
Spanish
Theme Type
#
%
#%
Participant (Subject) as Theme
964 64.44% 569 45.05%
Circ. Adjunct as Theme
323 21.59% 290 22.96%
Process as Theme
92 6.15% 231 18.29%
Se
-
- 51 4.04%
Clause as Theme
61 4.08% 40 3.17%
There/Hay
12 0.80% 8 0.63%
Wh
21 1.40% 3 0.24%
Extraposition
10 0.67% 32 2.53%
Participant (Complement) as Theme
5 0.33% 30 2.38%
Predicated/Thematic Equative
5 0.33% 7 0.55%
Fronted
3 0.20% 2 0.16%
Halliday (1994: 43) points out that in the declarative clause, the most typical arrangement of the clause is that Theme conflates with Subject. Thus, the Subject is the unmarked option for Theme choice. In the history textbook data, the above table shows that this is indeed the case in both corpora. Also, Halliday (1994: 44) posits the circumstantial Adjunct as the most usual form of marked Theme (marked Theme being any choice besides the Subject), and, again, this is borne out in the data. We have seen that in Spanish the process as Theme is unmarked in many cases, e.g. existential verbs (Hatcher, 1956; see section 4.6.4 below) and what De Miguel (1988) calls ergative verbs (see section 4.6.4 below). Thus, the results for the history textbook data fit in with general notions of markedness. Within the categories, it is interesting to note the great similarity across the two languages in terms of Circumstantial Adjuncts and Clauses as Theme. And, while the two 134
languages are different in terms of Verb as Theme, as was predicted, it is also of interest to note that by adding the categories of Subject, Verb, Complement, and Se together, the total for English is 70.92%, and for Spanish, 69.76%. The point here is that there is a very similar mapping across the two languages in terms of type of Theme if the results are adjusted to take into account the major grammatical differences which affect Theme choice: that in Spanish object pronouns must be placed before the verb, that Spanish employs the pronouns Se, which has no counterpart in English, and that in English all verbs must have a literal subject, except in the context of coordinated clauses, non-finite clauses, and imperatives. It is interesting to compare this similar mapping with other studies of different genres. If we focus on the similarity in circumstantial Adjuncts as Theme across the two corpora and compare it to other studies, we see, for example, that Ghadessy (1995) found that 12.5% of the clauses have a marked ideational Theme, which includes circumstances, in written reports of football matches. Francis (1989) found 9%, 11% and 12% in news reports, editorials and letters respectively. Both of these researchers, then, looked at texts in newspapers, and their findings show a lower number of circumstances as Theme than the present study. Whittaker (1995) found 30.34% of Circumstantial Themes in five linguistic articles and 23.10% in five economics articles, the latter being similar to the present study, perhaps showing an affinity between economics texts and history texts due to a similarity in studying trends and movements, and the effects of these on other entities and phenomena. Obviously, more studies are needed to make a statement on whether there is a generic influence here. At present, and with the small amount of data available at this point, it is mere speculation that there is a generic influence, and that it may be in force over the two languages. 4.6.2 Participant Types in Theme position A genre-based hypothesis predicts that the two languages will employ similar process/participant types in Theme position. We have already seen that there are more material processes in the English corpus than there are in the Spanish, which means that logically there will be more participants of material processes in Theme position in English than in Spanish (Actors, Factors, etc.). We have also seen that there are more Subjects chosen in English than in Spanish, due to grammatical constraints. This means that logically there will be a greater overall number of participant Themes in English than in Spanish, where the process itself is much more likely to appear as Theme than in English. This latter 135
point will obviously have an effect on a comparison of the two languages. In order to take this point into account, it is necessary to compare the overall results for participant types in Theme position in two different ways: the percentage of the different participant types as Theme with respect to the total number of all Themes, on the one hand, and, on the other, the percentage of the participant types as Theme in relationship to the total number of those grammatical Themes which can represent a participant (Subjects, Objects, predicated Themes, Thematic equatives). These results appear in Chart 4-2 and Chart 4-3 respectively. Chart 4-2: Percentage of Participant Types per Total Number of Themes
25% 20% 15% 10% 5% 0%
English Spanish
Actor Factor Goal Affected Result Beneficiary Carrier Identified Identifier Attribute Senser Phenomenon Sayer Verbiage Receiver Target Existent Behaver Range
Chart 4-3: Adjusted Percentages: Participant Type as Theme per Participant Themes
40% 35% 30% 25% 20% 15% 10% 5% 0%
English Spanish
Actor Factor Goal Affected Result Beneficiary Carrier Identified Identifier Attribute Senser Phenomenon Sayer Verbiage Receiver Target Existent Behaver Range
The first point to be made here is that the configurations of the two languages within each of the graphs are strikingly similar, except for a few major differences. The first is the number of Actors appearing as Theme: English employs twice as many in both the overall 136
and in the adjusted results. The other major difference is the number of Carriers; while in the
overall results it appears that the English data makes use of more Carriers as Theme, in the
adjusted results the number of Carriers is higher in the Spanish data. Similarly, in the
category of Identified, while this participant type is used in much the same number in the
two corpora when compared with the overall Theme results, in the adjusted results the
Spanish data shows a higher number. This can be explained by the grammatical difference:
in both corpora, the adjusted results are higher, which is not surprising given that the
participant Themes are being compared within a smaller category than in the overall results.
Nor is it surprising that this difference is more notable in Spanish, given that the number of
overall participants is lower, as already explained.
The major difference between the English and Spanish data in terms of participant as
Theme is clearly within the category of Actor. As already mentioned, the English data
shows over twice as many Actors as the Spanish, both in terms of percentage of Actor
Themes with respect to the total number of Themes, and also with respect to number of
Actor Themes per number of grammatical Themes which allow for a participant as Theme.
It was hypothesized that the participant types Actor, Factor and Carrier would be the most
frequent. This is the case with the English data, Carrier being most frequent, followed by
Actor, then Factor; however, in the Spanish data, Actor does not enter into the three most
frequent, which are Carrier, followed by Identified and Factor, which appear with equal
frequency. In sum, even taking into account the grammatical difference between Spanish
and English, and the lower number of material processes, the Spanish data employs a
noticeably lower frequency of Actors as Theme in the data. In order to analyze this
difference, a look at the number of Actors per each individual chapter or section is
warranted. These results appear in Table 4.7:
Table 4.7: Percentage of Actor as Theme per Total Themes:
Text Instances % (per Tot) Text
Instances % (per Tot)
Eng 1 Eng 2
15
11.11% Spa 1
11
8.27% Spa 2
2
1.50%
7
4.70%
Eng 3
9
6.67% Spa 3
17
10.30%
Eng 4
5
14.71% Spa 4
3
1.96%
Eng 5 Eng 6 Eng 7
4
2.42% Spa 5
14
6.73% Spa 6
17
9.39% Spa 7
3
2.46%
10
6.99%
6
3.66%
Eng 8
25
14.71% Spa 8
0
0.00%
Eng 9 Eng 10
31
22.96% Spa 9
41
20.50% Spa 10
0
0.00%
3
2.91%
Total
172
11.50% Total
51
4.04%
137
The greater appearance of Actor as Theme is fairly generalized in the English texts. The average percentage is 4.04 of the total Themes in Spanish, and only one of the texts is actually below that number in the English texts. The reverse is also the case: actually none of the texts in the Spanish corpus is higher in percentage of Actors as Theme than the average of the English texts. Therefore, there is a definite difference across the two corpora in this sense. In part this has to do with the type of Actor as Theme. For example, as can be seen further on in Chapter 6, section 6.1.4.2, Eng 4, which is a preface, has a high number of author reference as Theme: the authors portray themselves as Actors in material processes with positive outcomes. It is speculated there that perhaps this is the case because they may view the preface as a selling space. Spa 8, on the other hand, is also a preface, yet the author does not portray himself at all as an Actor. In the only two references to himself, the author includes himself (in the third person - "el autor") as a Senser in a mental process, expressing the hope that the readers will find the book useful. In comparing the two introductory chapters, Eng 1 and Spa 1, we can again see the difference. Eng 1 uses 11.11% of its Theme slots as Actors, while Spa 1 only 1.50%. Eng 1 uses the all inclusive "we" frequently, in explaining what people do with the study of history. It also refers frequently to the general category of historians. Spa 1, on the other hand, barely uses the participant Type Actor, referring more often to Factors in material processes and also using relational processes. That is because, rather than explaining what people use history for, Spa 1 analyzes the term history and explains its many different meanings. Eng 1 presents a more dynamic view in showing the uses of history, while Spa 1 presents a more static view in its analysis of history. The chapters on prehistory, Eng 2, Eng 5 and Spa 9 are all low in their use of Actor as Theme. This is not surprising, given the subject - there are very few specific Actors to refer to, except people in general. Eng 2 speculates for example on what men and women's roles might have been during that time, and these collectives then do function as Actors in material processes. Eng 5 refers on occasion to human beings and their actions. But Spa 9 employs no Actors as Themes. Spa 9 is a highly specialized text. It is heavily based on descriptions of tools and the implications that the evidence of the tools has for the development of history. In fact, over 70% of the clauses in that chapter are relational, showing the emphasis on description. 138
The rest of the chapters in the study relate the history of events from the 12th century onwards. Here again, with one exception, the differences between the two corpora hold. Eng 7 and Spa 4 are both concerned with Ancient Greece, and reserve 9.39% and 1.96% respectively of their Theme slots for Actors. Eng 7 refers to the citizens in general on a few occasions, and to Homer, to Achilles, and above all to Solon, who is given a Theme chain which stretches over 8 clauses, with some gaps between references. Spa 4 refers on two occasions to A. Evans and his archeological activity (activity which took place not during the time of Ancient Greece, but during the latter part of the 19th century and the beginning of the 20th century), and on one occasion to a political body which carried out a process at the time of the ancient Greeks. Here we can see a slight trend in English to narrate a series of events with a person (Homer, Achilles, Solon) as Actor. This trend is repeated in the two texts on Ancient Rome, Eng 8 and Spa 6, with 14.71% and 6.99% of the Themes as Actor respectively. Spa 6 uses Actor to refer to historians doing things, while Eng 8 uses that slot to refer to characters of the time; a large number of clauses refer to Augustus, and he is often thematized as Actor. The texts on the 18th century, Eng 10 with 20.50% of its Themes as Actor, Spa 7, with 3.66%, and Spa 10, with 2.91%, continue the trend. Spa 7 personifies countries, where 3 out of 6 of the Actors in material processes are locations, while Spa 10 employs as Actors philosophers and thinkers of the time to show their active participation in fomenting and changing political and philosophical thought of the time. Eng 10 uses as Actors kings, judges, and politicians, and then follows with much more narration of their actions and deeds. An example of this is: 4-87. Eng 10 (148) Fleury did not remedy the chronic and deep-seated injustice and inefficiency of French fiscal methods, (149) but he did stabilize the coinage, (150) and he put the farming of taxes on a more businesslike basis by restricting tax-farmers to the comparatively modest profit of 7 1/2 percent. The texts on 19th century population growth, on the other hand, Eng 9 with 22.96% of Actors, and Spa 2 with 4.70%, use their actor slots in similar ways. The Actors in almost all cases represent immigrant groups and the ensuing process in the clause depicts their actions, usually in terms of movement from one place to another. However, Spa 2 only does so at the end of the section. There are a total of 148 clauses analyzed, and the Actors as Theme appear in Clauses 8, 120, 126, 130, 133, 134, and 136. The rest of the chapter discusses the causes for emigration, setting the scene for the movement described at the end 139
of the chapter, in which immigrants appear as Actors. Eng 9, on the other hand, includes examples of immigrant movement throughout. It does analyze the phenomenon, as evidenced by: 4-88. Eng 9 (23) Before looking at the people who migrated, let us consider three facts. However, in looking at the three facts considered: 4-89. Eng 9 (24) First, the number of men and women who left Europe increased rapidly before World War One. (27) Second, different countries had very different patterns of movement. (38) Third, although the United States absorbed the largest number of European migrants, less than half of all migrants went to the United States. an interesting pattern emerges. The English text looks at the "facts" of the movement, and from these extrapolates trends, causes, etc. But the focus is on the movement; therefore, it is logical that immigrants involved in that movement will be thematized as Actors. The Spanish text, on the other hand, describes the advent of the field of demographics, and moves to a description of famines, epidemics, and wars and analyzes the role they played in the great migration of the 19th century. It moves from the causes to the effects, while the English text moves from the effects to the causes. This may correlate in part to a different philosophy of history. Halldйn (1997) identifies two types of explanation in history writing: structural and causal. Structural explanations explain events by describing their function in a system, while causal explanations explain the event through other events. If the texts are looked at through the Actors used as Theme, the Spanish text seems to take a system of demographics and explain the related events through their function in that system. The English text seems to explain the movements by then looking at other larger events. Halldйn goes on to state that "these two forms of explanation pertain to different historical methods. Causal explanations can be related to a narrative form of history, whereas structural explanations are more concerned with the method of colligation" (Halldйn, 1997: 205). Carretero, et al (1997: 245-246) draw support from von Wright (1971) to underscore the notion that "causal explanations in history are intentional and personalized, given that individual agents play a prominent role in the genesis of historical events". Leinhardt (1997: 224) explains the nature of events in history, and shows their relationship to narrative: Events are the paradigmatic, short, narrative episodes about wars, treaties, and people that are characteristic of history. Events include migrations, revolutions, 140
changes of people and offices; the connections are causal. Events have actors, purposes, motives, consequences and are narrative in flavor. Colligatory expressions, on the other hand, are higher order concepts, such as "the Industrial Revolution", which bring together a series of events. Thus, the actions of the individual are the focus of a causal or narrative conception of history, while impersonal structures are the focus of a structural conception of history. In an attempt to provide further explanation for this slight perceived difference, I had the opportunity of interviewing Dr. Francisco Garcнa Serrano, who holds an undergraduate degree in history from the Universidad Complutense of Madrid, and a Ph.D. from the University of California, Berkeley. He has taught in both the American and Spanish educational systems, and is thus quite familiar with a wide range of history textbooks. When I asked him what he felt was the major difference between American and Spanish history textbooks, his reply was to the effect that American textbooks provide a social history, and Spanish textbooks a political history. When I asked him what that meant, he said that political history involved events, while social history involved people. If we put these two explanations together, it would seem that social history correlates with causal explanation through narrative, while political history correlates with structural explanation. However, the distinction does not work this neatly. According to The Cambridge Dictionary of Philosophy (1995: 584), [a]cademic history in recent times has seen a strong turn away from traditional political history toward social, cultural and economic analyses of the human past. Narrative is associated with the supposedly outmoded focus on the doings of kings, popes and generals. This effectively associates doing political history with narrative. The question is whether it disassociates social history from narrative. Hexter (1993: 263.) suggests that this is the case: So the innovative sort of twentieth-century historical work [social history] marginalized writers of narrative history. It also marginalized human persons, or effectively disintegrated them, made into historical non-entities the actual players in those manifold stories of the past, that had been most of what most historians wrote about during the previous hundred years (ibid). There also seems to be support for associating structural explanation with social history. According to Stearns (1993: 247), social historians maintain that "history, in dealing with politics, or women, or eating habits, best captures the past by focusing on patterns and on processes of change in patterns, rather than simply on a series of discrete events". 141
However, the Cambridge Dictionary of Philosophy points out that the marginalization of narrative has "lost some of its steam and narrative history has made something of a comeback among historians" (ibid), and that what may be perceived has non-narrative history retains narrative features. Evidence for this comes from Eng 1, the introductory chapter to the study of history in the English corpus: The "facts" that follow include the events and figures that all interpretations must take into account -the famous names and dates that are an important part of our historical lore - but the "facts" also include basic information about social organization, about modes of production and of exchange, and about cultural activity that is crucial to understanding the way of life of a particular society. All this information - much of it narrative after all, some of it good stories and accounts of dramatic events - is used to explore the problems on which a particular chapter focuses. (xiv) Indeed, the authors of Eng 1 explain that in their text they use a combination of more traditional and more modern approaches in their text. And the type of history included in Eng 9, from which example 4-89 above is taken, where not a member of the elite is in focus, but immigrants, indeed shows that a large number of the Actor Themes do refer to people, not individuals, but groups. And there is a great deal of narration of their movements. There is other evidence to support the connection between social history and narrative. Briggs (1993: 595) explains that "[when] G.M. Trevelyan published his bestseller Social History in the midst of war in 1942 he was more interested in describing scenes, in narrating, in recapturing moods and in inspiring readers, than in theorizing about or explaining why society changed". In order to discover whether the texts have more of a social or political historical focus through a Theme analysis would involve categorizing the Themes more delicately, e.g. whether the Actors are individuals, groups, institutions, structures, and so on, an analysis which goes beyond the scope of this study. However, we have seen some evidence from the history textbook corpus that could indicate that American textbooks rely more on causal explanation, which is related to events and narrative, while Spanish textbooks rely perhaps more on colligatory, or structural explanations. However, this is not to say that the latter ignores narrative. Halldйn explains that historians (in general) favor a structural method of history: "With one foot in the humanities and the other in the social sciences, historians try to combine the actions of the individual with the effects of impersonal structures in models for the explanation of historical events"(Halldйn, 1997: 205). However, she goes on to say that students (in her study, upper secondary students in Sweden), "seem 142
to have both feet in the humanities" (ibid), and they "cling to a narrative method" (ibid: 207). Thus students often personalize structures, e.g. the Liberal party, and then turn them into agents who perform actions (ibid: 205). It may be very tentatively speculated then, that the English texts come from an American tradition in which, perhaps, the philosophy behind the writing of history textbooks is more rooted in the narrative camp, perhaps in order to provide a more readily available framework for students, while the Spanish texts come from a tradition more rooted in the structuralist camp, and present history to students more from the point of view of the professional historian. Of course, more evidence from other texts, and interviews with history teachers and textbook writers, is needed before this speculation can begin to have any general validity. Also, this may only be a slight tendency, as there are differences between texts of the same language. Indeed, it is worth noting here that not all of the texts conform to the trend seen above in the Ancient Greek and Roman and the 18th and 19th century texts. The texts on the Middle Ages, Spa 3, with 10.30% of its clauses having an Actor as Theme (the highest number in the Spanish corpus), Eng 6 with 6.73%, and Eng 3 with 6.67%, show the opposite trend of all the other texts. Here, although the difference is not very great, there are more Actors in the Spanish texts than in the English. The type of Theme chosen also reverses the trend mentioned above. Here, both the English and Spanish texts use collective nouns as Actors ("The Christian West", "The papacy", "los notables"), but the Spanish text uses to a much greater extent specific individuals, especially kings, such as Sancho IV and Alfonso X, and narrates their deeds and actions. This shows the tentativeness of the speculation that history textbook writing in the two different countries may be based on different philosophies of history, and shows that more texts would need to be analyzed in order to come to firmer conclusions. With regard to other participant types as Theme, Carrier as Theme is by far the most common Theme type (after Circumstance as Theme) in both corpora. This is not surprising in the Spanish corpus, given that relational processes are the most common, as we have seen in Chart 4-1 above. In fact, in the Spanish corpus, there are over 11% more relational processes than material (40.06% and 28.42%, respectively), while in the English corpus, the count for both is almost identical (37.23% for relational clauses, and 37.97% for material). Yet, when it comes to type of participant as Theme, 20.99% of the Themes in the English 143
corpus are Carriers, and only 11.50% are Actors. This is due in great measure to the preference in English in the data of beginning a material process clause with a circumstantial Adjunct more so than with any other type of clause. Roughly 42% of the circumstantial Adjuncts appearing as Theme do so in material clauses in the English data. This happens to a much lesser extent (approximately 27%) with relational clauses in the English data. At any rate, across both corpora, Carriers are the most popular choice for participant as Theme. These Themes seldom refer to individuals and their attributes, although there are a few examples: 4-90. Spa 3 (47) Alfonso X, como hijo de Beatriz de Suabia, era directo descendiente de la familia imperial 4-91. Eng 8 (106) His way of life was as simple as his house. However, most Carriers depict a more collective entity, such as a city, state or country, people in general, or some type of process or phenomena (the weather, growth, a theory, a revolution, an era, etc.). These are presented as things which have attributes, and their status as Carrier is thematized frequently. Added to this is the category of Identified. In much the same way as Carrier, rarely is it used to thematize an individual. It also tends to refer to a phenomenon, factor, cause, way of thinking, entity, process and so on, and in the Rheme some identification is given to that, e.g.: 4-92. Eng 7 (107) The great innovation introduced by the Greeks into politics and social theory was the principle that law did not derive from gods or divine kings, but from the human community. 4-93. Spa 7 (30) Un factor muy importante, quizбs el mбs importante, fue la disminuciуn de los grandes contagios epidйmicos. Factor as Theme is also common in the two corpora, and its appearance in terms of percentage of total Themes is similar across the two languages, 6.62% for the English corpus, and 5.86% for the Spanish. These are similar in substance to the majority of Carriers and Identifieds in that they depict a political thought, a reform, a revolution, population growth, and many other types of phenomena or processes. By adding these three categories together, they make up a third of the total number of Themes for each of the corpora (32% for English and 29% for Spanish). Thus it is clear that the focus of the Themes chosen in terms of participant types is not on individuals and their attributes and actions, but on processes and phenomena, with their descriptions and actions forming the Rhemes of the clauses. In terms of participants as Theme, this shows emphasis on description and analysis in the corpora, which is expected in this genre. 144
The other participant types, belonging to verbal, mental, and behavioral processes, as well as those participant types which are not usually chosen as Subject, do not warrant much of a mention here. There are no major differences between the corpora in their frequency, which is low; this does show a difference between the history texts and the journalistic texts analyzed by Francis (1990), in which she found frequent recourse to verbal processes. However, in the case of existential participants, a grammatical difference plays an interesting role. While in Spanish there are more existential processes than there are in English, there are more Existents as Theme in English than in Spanish. This is due to the use in Spanish of inversion in clauses which assert the existence of something (Hatcher, 1956; see section 4.6.4 below). In the Spanish corpus, material and relational clauses use about equal amounts of inversion and ellipsis in the case where the process is Theme, in other words, in verb-initial constructions. However, with the case of existential processes, ellipsis is rarely used, while inversion is highly prevalent. In sum, the analysis of participant type as Theme shows that there are similarities across the two corpora. These similarities point to a prevalence of descriptive and analytical clauses, thus showing more interpretation than narration of episodes or events. However, there is a tendency in the English corpus to use more Actors as Theme, thereby indicating perhaps a slight tendency towards more causal explanation than seems to be present in the Spanish corpus.
4.6.3 Circumstance as Theme
Circumstances, including circumstantial Adjuncts as Theme and clause as Theme, show similar frequency across both corpora. The results, given in Table 4.7 above, are repeated below to show this similarity:
Table 4.8: Circumstantials and Clauses as Theme
Theme Type
English
Spanish
Circ. Adjunct as Theme
323 21.59% 290 22.96%
Clause as Theme
61 4.08% 40 3.17%
Even the configuration of type of Circumstantial Adjunct as Theme is similar across the two languages, as can be seen in Chart 4-4.
145
Chart 4-4: Circumstantial Adjunct Types as Theme 14% 12% 10% 8% 6% 4% 2% 0%
En glis h Spanis h
Acc Angle Cau Cont Ex Loc Man Matt Other
The most common circumstantial as Theme is Location, which includes both spatial and temporal location, the latter being more common in both, followed by Manner, then Extent, which also includes both spatial and temporal types, and then Cause. Chart 4-4, then, confirms the genre-based hypothesis in that the configuration of circumstantial types is similar across the two languages. If we look at the circumstantial sub-types (see Table 4.5 above), it is clear that, while the configuration for most of the sub-types is very similar across the two languages, there is a slight difference between English and Spanish in terms of spatial and temporal expressions, as can be seen in Chart 4-5:
Chart 4-5: Circumstantial Sub-Types in Percentages per Total Number of Themes 12%
10%
8%
6%
English
Spanish
4%
2%
0% Mean Com Qua Beh Res Pur Rea Conc Cond Spa Temp
Spanish shows a slightly higher tendency than English to thematize a spatial Location or Extent (6.31% and 4.21%, respectively), while English has a higher percentage of temporal Locations and Extents (10.70% and 8.92% respectively). These differences are not very 146
high, and in looking at the individual texts, it can be seen that the majority of the 20 texts in the total corpus show a higher use of temporal sub-types than spatial sub-types. The only exception to this in the English corpus is Eng 4, with 11.76% for spatial circumstantials and 2.94% for temporal. In the Spanish corpus, however, there are two texts which show this pattern, one which shows a slight difference: Spa 1 (spatial: 5.26%; temporal: 3.76%) and one which shows a very high difference: Spa 10 (spatial: 15.53%; temporal: 2.91%). This difference is difficult to attribute to any one factor, given that Eng 4 is a textbook preface, as is Spa 8, yet Spa 8 does not show the greater spatial circumstantial configuration as does Eng 4. Eng 4 thematizes on several occasions spatial references to the book (e.g. "throughout the book"), while Spa 8 does not do this. Spa 1 is an introduction to the study of history, and thematizes spatial locations in terms of documents and history in general ("En la historia..."), yet Eng 1, also an introduction, does not do this. Spa 10 concerns itself with the 18th century, and uses spatial type circumstantials to refer to several different types of things: to refer to ways of political and philosophical thought, for example: 4-94. Spa 10 (16) Frente a esta concepciуn deнsta del mundo, (17) En la crнtica polнtico-social y en la elaboraciуn de un nuevo sistema de gobierno (40) En la concepciуn de Rousseau (48) En el campo de la economнa (86) En sus filas (59) En el capнtulo tercero They are also used to refer to people's writings (and once to the book itself): 4-95. Spa 10 (14)En su Ensayo (19)En la misma (28)En sus Discursos sobre los fundamentos y el origen de la desigualdad entre los hombres (43)En los escritos de Rousseau as well as to countries: 4-96. Spa 10 (70)En muchos paнses,. (87)En Italia, (90)En Espaсa, (94)Tampoco en Rusia Eng 10, which is also concerned with the 18th century, only thematizes spatial circumstantials on 6 occasions, 3 referring to countries and 3 to political entities. The difference may be attributable to author preference. Although Spa 5, written by the same authors as Spa 10, shows a higher temporal than spatial rate of circumstantial Adjuncts, it is not with the same rate of difference that the other texts in the corpus show. 147
Indeed, Spa 5 shows a much higher thematization of spatial circumstantials than the norm for the whole corpus (which is approximately 5%), with a total of 9.84% of the clauses having a spatial sub-type as Theme (while 12.30% of the clauses have a temporal sub-type as Theme). The topic of Spa 5 is the 14th and 15th centuries, and it thematizes circumstantials which refer to countries in 8 clauses, while Eng 3, written about the same topic, only does so on 3 occasions. It can only be tentatively suggested with the small amount of data available that the authors of Spa 5 and Spa 10 have a slight preference for taking as a point of departure a spatial element. It will be remembered that in this study, dependent clauses are considered Theme when they precede their superordinate clause (see section 3.4.4) and that they are analyzed according to circumstantial functions (see section 4.4.5). Clause as Theme shows a different semantic configuration than does Circumstantial Adjunct as Theme, as can be seen by comparing Chart 4-4 above with Chart 4-6 below: Chart 4-6: Clause Types as Theme
2%
2%
1%
Eng lis h
Spanis h 1%
0%
Acc Angle Cau Cont Ex Loc Man Matt Other
In both of the corpora, clauses are thematized for purposes of showing contingency more than for any other purpose; as we have seen (section 4.4.5), Halliday points out that contingency tends to be expressed clausally. The English corpus shows a greater preference for thematizing clauses to show cause and location than does the Spanish, while the Spanish shows a slightly higher preference for thematizing clauses of manner. The counts are so low for each of the chapters, however, that it is difficult to ascertain if there is any overall pattern in this difference. With regard to Clause sub-types as Theme, again, as we have just seen, the percentage of overall Clause types as Theme is very low across both languages,. Therefore, 148
here also, when this is broken down into Clause sub-types, the percentages are so low that it is difficult to say whether they are at all meaningful. At any rate, as can be seen in Chart 4-7, in the English corpus there is a higher percentage of clauses which express concession, reason, quality and notions of temporality, while the Spanish corpus shows a higher tendency with respect to clauses of comparison, condition and means:
Chart 4-7: Clause Sub-types 1%
1%
1%
1%
English
Spanish
0%
0%
0% Mean Com Qua Beh Res Pur Rea Conc Cond Spa Temp
With reference to the inclusion of Clause as Theme, it was explained in section 4.4.5 that one reason for their inclusion as Themes of clause complexes was their similarity in function and expression to Circumstantial Adjuncts. The analysis provides further support for this claim. Chart 4-8 below shows the results obtained by combining the percentage of Circumstantial Adjunct sub-types as Theme with Clause sub-types as Theme:
Chart 4-8: Combined Circumstantial Adjunct and Clause Sub-types
12% 10% 8% 6% 4% 2% 0%
En g lis h S p a n is h
Beh Com Conc Cond Mean Pur Qua Rea Spa Temp
149
The configuration is almost identical to the configuration presented in Chart 4-5. This would seem to indicate that writers use clauses and circumstantial Adjuncts to express the similar meanings manifested in the sub-types. The exceptions to this are expressions of concession and condition. Here again, there are similarities across the two languages, although in this case it may be due to syntax, rather than to genre. In both languages, there are ways of expressing concession through a circumstantial Adjunct, by using "in spite of" in English, and "pese" in Spanish, English prefers "although" and Spanish "aunque" to express this meaning, and both of these words require a clause in their syntactical configuration. With condition, the most common way of expression in both languages is through a word which also requires a clause: "if" in English and "si" in Spanish. In terms of a cross-cultural genre-based hypothesis, it is interesting to note the results for the circumstantial Adjuncts and clause types (rather than sub-types) combined. The semantic configuration in this case is remarkably similar, as can be seen in Chart 4-9.
Chart 4-9: Combined Circumstantial Adjunct and Clause Themes
14% 12% 10% 8% 6% 4% 2% 0%
En g lis h S p a n is h
Acc Angle Cau Cont Ex Loc Man Matt Other
This would seem to confirm a genre-based hypothesis related to the thematization of Circumstantial Themes of location in space and time as frequent in history textbooks. However, at the time of writing this, I am aware of no other genre studies which look at Circumstantial Themes in this way in order for a comparison of results to confirm that this is a finding which sets apart history textbooks from other genres. It may say something about textbooks in general, or about history writing in general. More comparative studies of different types are needed for this finding to be of any significance. One sub-type that is not thematized in either language is that of Result, the category that was added in this study to Halliday's model of circumstantial Adjunct types. This is a category which only appears in the Rhemes of the clauses. The next sections look at 150
participants and circumstances in the Rhemes of the clauses to ascertain whether the similar semantic configuration, except in the case of the participant category Actor, holds in the Rhemes of the clauses as well. But before moving to Rheme, Process as Theme is considered.
4.6.4 Process as Theme
In addition to participants as Theme, it is also possible to thematize the process
itself, as has been discussed previously in this study (section 3.5.2.1.1). Obviously, the
major portion of a discussion of process, or verb, as Theme belongs to the Spanish part of
the corpus, as a full 22.65% of the total Themes in Spanish are verbs, while only 6.08% of
the clauses in English begin with a verb. Of interest is whether the verb is Theme because of
ellipsis, in which case the motivation is either that the verb represents an impersonal process
or that the Subject of the verb is clear from the previous context, or because of inversion, in
which case the Subject has been left for Rheme position.
With regard to Spanish, 66.43% of the total verbs in Theme position present ellipsis
of the Subject, while 33.57% present Subject inversion. As can be seen in Table 4.9, in the
case of existential verbs inversion is much more likely to occur than is ellipsis, while the
opposite is true for the other processes. Those existential verbs which do present ellipsis in
Spanish are those which resemble more material processes, e.g. those of growth, maturing,
and so on. In Table 4.9, the total counts for each of the process types are given; the
percentages represent the percent of the total number of instances of that process type (e.g.
for Existential, 39) divided by the ellipsis and inversion instances:
Table 4.9: Ellipsis and Inversion of Verbs in the Spanish Data
Ellipsis
Inversion
Process Type # % of total # % of total
process type
process type
Existential
8
20.51% 31
79.49%
Behavioral
2
66.67% 1
33.33%
Material
56
83.58% 11
16.42%
Mental
22
78.57% 6
21.43%
Verbal
16
84.21% 3
15.79%
Relational
56
74.67% 18
25.33%
Several reasons for Subject/verb inversion taking place have been advanced in the literature. Jimenez (1995) suggests that Subjects which appears to the right of the verb are interpreted as having a contrastive focus. Ocampo (1990) also uses focus to explain word order in 151
Spanish. This could explain examples of inversion such as in clause 82 in the following
example :
4-97Spa 3
(81)El rey, abandonado por casi todos, buscу la ayuda del rey de Marruecos, mientras su hijo se atraнa la del de Granada. (82)Estaba ya el rey Sabio reducido a Sevilla, que le permaneciу siempre leal, cuando muriу.
Here, there are 3 kings involved: the first king, thematized in clause 81, is Alfonso X, also
known as "el Sabio"; the second king is the king of Morocco, and the third the king of
Granada, both mentioned in the Rheme of 81. The Subject of estaba in clause 82 is Alfonso,
and needs to be separated out from the 3 kings in the reader's framework at this point. This
may explain the Subject/verb inversion which takes place in this clause. It is interesting to
note further that the Subject splits up the verbal group in clause 82, leaving the lexical
information contained in "reducido", which is "newer" than "el rey Sabio" for after the
Subject.
Vasconcellos (1992) brings in another consideration, related to the point just made.
She cites Suсer (1982: 126) as concluding that the verb "is like the tray on which the
delicacy is presented", thus "It carries the existential assertion, to a greater or lesser degree,
that the "object" exists in the universe" (Vasconcellos, 1992: 153). Hatcher (1956) points
out that "inversion of subject in existential sentences may be taken for granted as
unproblematical", and gives English as an example, where the use of "there" allows for the
inclusion of the object whose existence is predicated to occur later in the sentence. This
allows for new information to be placed in the Rheme of the clause. This also ties in with
Firbas' notion of Communicative Dynamism, explained in section 3.1.3; inversion in
examples such as the following allow "new" Subjects to be placed later in the clause, in the
Rheme where they are more "expected":
Spa 3
(96)Firmуse entonces en Francia el tratado de Lyon, por el cual se pactaba la guerra contra el monarca aragonйs, Alfonso III.
In this example, the signing is made thematic through the initial placement of the verb, and the treaty of Lyon, which is new in the discourse, is left for the Rheme. There are other ways in this clause of leaving the treaty until the Rheme, e.g. Entonces en Francia se firmу el tratado de Lyon, por el cual se pactaba la guerra contra el monarca aragonйs, Alfonso III. thus thematizing time. However, this has the disadvantage of placing both the time (`entonces') and place (`en Francia') reference in a focal point of the clause, giving them a
152
more prominent position in the discourse at this point than perhaps the author feels that they warrant. De Miguel (1988) talks about a class of verbs in Spanish which she calls "ergative". What she means by this are those verbs which are "anticausative", that is, which refer to events which take place in the absence of an agent, e.g. abrirse, romperse, oxidarse. She also includes in ergative verbs those verbs which semantically express a process instead of an action, a process which usually has as a consequence a change of state, such as llegar, pasar, faltar, amanecer, or morir. The Subjects of these ergative verbs are semantically not agents, as they experience the effect of the verb rather than cause it. In Halliday's terms (see section 4.3.1.1) they are Mediums, and, indeed, de Miguel's anticausatives seem to coincide with Halliday's examples of verbs with Medium as Subject. According to de Miguel, then, these Subjects are generated in the structural position of the object, and those are postverbal, and, indeed, she finds that "ergative" verbs in Spanish show a preference for Subject/verb inversion. De Miguel gives other reasons for inversion: Subject as newer information, focus, end-weight, and style. The latter of these, style, which as has already been mentioned is a slippery term (see section 3.1.3), seems to tie in with Enkvist's (1984) notion of "textual iconocism", mentioned above in section 3.2, where a clause may mirror the pattern of a preceding clause exactly or chiastically. Indeed, this may be the motivation behind inversion in Clause 46 in example 4-98: 4-98. Spa 10 (45) Para йl, la razуn no lo es todo, (46) cuenta tambiйn el sentimiento. At the same time, el sentimiento is the new element in the discourse, and it is focused on in contrast to la razуn. A number of studies on Spanish (Ocampo, 1990; de Miguel, 1988; Teskey, 1976; Garcнa Ruiz, 1985, inter alia) attribute inversion in great part to pragmatic motivations, mainly to issues of Given/New, or of Focus. Other motivations put forward include stylistic and expressive force (see section 3.1.3). Indeed, there are cases of inversion where the motivation cannot be attributed to notions of focus or newsworthiness. One example is sentence 15 in example 4-99: 4-99 Spa 8 (14) Los tiempos modernos y la complejidad creciente de la administraciуn estatal han multiplicado hasta dimensiones inabarcables los fondos escritos que guardan el secreto de nuestro pasado inmediato. (15) Pero no debe aplastarnos el peso de estos fondos ingentes. 153
El peso de estos fondos ingentes is not new, as it is just what has been described in the Rheme of the previous sentence. Thus, the motivation for inversion here does to respond to pragmatic considerations (this clause is considered again in section 5.5.2) When I asked informants (teachers of Spanish) to explain why they thought the writers had chosen inversion in clauses such as the following, where the Subject (underlined) is a given in the discourse: 4-100 Spa 3 (115) Fue Sancho IV buen polнtico y protector de las ciencias y de las letras; 4-101 Spa 3 (134) De este modo quedaba el sultбn interesado en la lucha contra Castilla, they all assumed the motivation to be stylistic. It is interesting to note that in these cases, while the Subject is not new, it is not left for the end of the Rheme, which is also an option, but this would place it in the most newsworthy position in the clause. At the same time in each case it does provide more information than the verb, which is a copula in all three cases, so a sort of communicative dynamism in Firbas' terms may be in effect here. From the above discussion, for the most part it seems, then, that the motivation for inversion accords with Theme/Rheme focus and information patterns. 4.6.5 Results from Theme Analysis: Participants, Processes and Circumstances Chart 4-10 reflects the most common Theme types in both the English and Spanish corpora. In both, the most common point of departure is some kind of circumstance, followed by a participant in a relational process. Then, the Spanish corpus shows a wide difference from the English corpus, in that the process itself is the third most common point of departure, which has been explained as being due to the grammatical nature of the language. The third most common Theme for the English corpus is that of Actor in a material process; the motivation for the prevalence of this type of Theme in the corpus has been explained through a possible difference in focus in ways of explaining history in the two cultures: there may be a slight preference for causal explanations in American history textbooks and for structural ones in Spanish history textbooks. In sum, besides the major grammatical difference, and a minor difference in focus, the two languages show a great deal of similarity in point of departure of the clauses. 154
Chart 4-10: Most Frequent Theme Types 30% 25% 20% 15% 10% 5% 0%
English Spanish
Circ Carrier/Identified Verb Actor Factor Senser Existent Sayer
4.6.6 Participant Types in the Rheme
It will be remembered from Chart 4-1 above that there are differences between Spanish and English in terms of process type, and that this difference was reflected in the participant types which appeared as Theme in the clauses. Naturally, it is expected that this difference will also be reflected in the Rhemes of the clauses. Indeed this is the case, as can be seen in Chart 4-11:
Chart 4-11: Participant types in Rheme
35% 30% 25% 20% 15% 10% 5% 0%
Englis h Spanish
Actor Factor Goal Affected Result Recipient Res. Att. Carrier Identified Identifier Attribute Senser Phenomenon Sayer Verbiage Receiver Target Existent Behaver Range
155
Once again, there is a very similar configuration between the two corpora. One difference is the higher number in English of participant types which pertain to material processes: also here there are more Actors in the English data than in the Spanish, as well as more of the participant types which tend to be rhematic: Affected and Goal. A possible explanation for the greater number of material clauses overall, and thus material participants, has already been given in section 4.6.2 above, and there is no need to provide a different explanation for the same phenomenon in the Rheme of the clauses. However, it is interesting to note that by combining the Theme and Rheme results for the category of Actor, we see that in the English data there are far more Actors than in the Spanish data. The total for Actors in the English data is 19.25%, and in Spanish 8%. Thus, this Rheme finding does lend additional support to the tentative claim that the two cultures may bring to bear different ways of explaining to the writing up of history textbooks, thus meaning that there is a slightly different way of portraying history as field. What is of interest to the analyst here is not only the comparison/contrast across languages, but the comparison/contrast of participant types appearing in the Themes of the clauses, and those appearing in the Rhemes. For Halliday (1968: 214-215): ...thematic prominence tends to be assigned to the more `central' among the clause elements, the participants which occupy the active roles in transitivity; and this, together with the opposite tendency in information focus, which favours the more `peripheral' elements, especially circumstances, defines in general terms a preferred clause type for transitivity and theme. This is one in which the initiating, often anaphoric, element in the message is the element most closely associated, with the process; and the culminating, information-carrying element is that which is most remote. Thus, participants such as Actor, Carrier, Senser, Existent, and so on are more likely to appear in Theme position, while participants such as Goal, Attribute and Phenomenon are more likely to appear in Rheme. This does not mean to say that, for example, Actors always appear in Theme. It means that, other things being equal, Actors are likely to appear in Theme position. This is because the unmarked option for Theme choice is the Subject (see section 4.6.1). However, it will be remembered that for existential verbs in Spanish, the unmarked option is for inversion to occur, thus placing the Existent in the Rheme. Thus, the configurational tendencies of participant appearance in Theme or Rheme are slightly different between English and Spanish. These tendencies, i.e. the greater likeliness of a participant's appearance in Theme or Rheme, appear in Table 4.10 for English and Table 4.11 for Spanish: 156
Table 4.10: Configurational Tendency of Participants: English
Process Type Theme
Rheme
Material
Actor, Factor
Goal, Affected, Resulting Attribute, Beneficiary
Relational
Carrier, Identified
Attribute, Identifier
Mental
Senser
Phenomenon
Verbal
Sayer
Verbiage, Receiver, Target
Existential
Existent
Behavioral
Behaver
Table 4.11: Configurational Tendency of Participants: Spanish
Process Type Theme
Rheme
Material
Actor, Factor
Goal, Affected, Resulting Attribute, Beneficiary
Relational
Carrier, Identified
Attribute, Identifier
Mental
Senser
Phenomenon
Verbal
Sayer
Verbiage, Receiver, Target
Existential
Existent
Behavioral
Behaver
The discrepancy in Chart 4-11 between Spanish and English in terms of number of Existents
can be explained by the above configurational tendency: in Spanish, Existents are more
likely to appear in the Rheme. This, at the same time, explains the difference between
number of Existents appearing in the Themes of the Spanish clauses (2.53%) and in the
Rhemes (8.37%). English also has ways of placing the Existent in the Rheme of the clause,
although the more rigid Subject/verb order does not allow this to occur in as many contexts
as in Spanish. In English, one way of placing the Existent in the Rheme is through the use of
the word "there" as Subject. This is done in 13 out of the 118 existential clauses in English.
In other words, it is far more common in the English corpus to find an existential clauses of
the following type:
4-102 Eng 10 (53) for party discipline of the modern kind did not yet exist;
than there is of the "there" type.
Another way of leaving the Existent for the Rheme of the clause is by fronting it with a Circumstantial, as is done in 39 of the 118 existential clauses. This leads to a discussion of one of the most important ways of placing participants in Rhemes across both languages, that of fronting circumstantials. We have seen the thematization of
157
circumstantials in section 4.6.3 above. We now look to the placement of circumstantials in the Rhemes of the corpus, and discuss possible motivation for choosing one or the other position for their placement.
4.6.7 Circumstantials in Rheme
Table 4.12 below shows a very similar picture for the Rhemes of the clauses in terms
of circumstantial meanings across the two corpora. The percentage in each case shows the
percent of total number of clauses which employ circumstantials, through Adjuncts and
through dependent clauses.
Table 4.12: Circumstantial Types in Rheme
Type
English
Spanish
Adjunct
42.45%
43.25%
Sub. Clause
4.14%
4.81%
The distribution, however, is not over all of the clauses of the corpus, as some clauses employ two, or even in a few cases, three circumstantial expressions. At any rate, the percentages show a similar amount of circumstantial expressions in the Rhemes of the clauses of the two corpora. These percentages show almost double the amount of circumstantial Adjuncts in the Rhemes of the clauses as compared to the Themes (see Table 4.8 above). Halliday (1968: 213) explains the motivation for this: In the transitivity structure of the clause, there is a tendency for circumstances, where present, to follow participants; adjuncts, in other words, apart from those directly associated with the `closed' verbal categories of tense and polarity, and discourse and modal adjuncts (which have no function in transitivity), normally occur towards the end of the clause. If there is an expression of time or manner or other circumstance that element will tend to occur in a position in which, in the unmarked case, it will have, alone or with other elements, the value `new' (the next most favoured position for such items being clause initial, where they have prominence of the other kind). While circumstantial Adjuncts are almost double in the Rhemes of the clauses, the percentage of subordinate clauses appearing in Theme and Rheme position are very similar. (As a reminder, in the English corpus, 4.08% of the clauses had a thematic circumstantial Adjunct, and for Spanish, 3.17%). This means that it is more marked in the case of circumstantial Adjuncts to place them in Theme position than in Rheme position, but this is not the case with subordinate clauses, as they are more evenly distributed across Themes and Rhemes, at least in this corpus.
158
Before looking at the differences in placement of circumstantial expressions in the Theme or the Rheme of the clauses, it is worth looking at the distribution in the Rhemes of the clauses in terms of meanings expressed. This can been seen in Chart 4-12:
Chart 4-12: Circumstantial Adjuncts in Rheme
20% 18% 16% 14% 12% 10% 8% 6% 4% 2% 0%
Englis h Spanish
Acc Angle Cau Cont Ex Loc Man Matt Other
Again, here the configuration across the two corpora is similar, with the exception that Spanish employs more Adjuncts of manner than does English (19.20% and 15.44%, respectively), while English employs more Adjuncts of extent (4.01% and 2.13%, respectively). These differences are reflected in the circumstantial sub-types in Chart 4-13:
Chart 4-13: Circumstantial Adjunct Sub-types in Rheme
16% 14% 12% 10% 8% 6% 4% 2% 0%
English Spanish
Mean Com Qua Beh Res Pur Rea Conc Cond Spa Temp
Spanish uses a higher number of circumstantial Adjuncts of quality than does English, whereas English employs a higher account of temporal Adjuncts. While in the overall circumstantial types English and Spanish showed a similar configuration of Cause 159
circumstantial types, here it can be seen that they are divided differently amongst the subtypes. English uses more purpose Adjuncts, while Spanish employs more reason Adjunct types. At any rate, the differences are not very considerable. It is worth pointing out here that there are greater differences within each corpus as to the frequency of these Adjuncts than there are across the corpora. The higher number of Adjuncts of quality may indicate a slightly higher tendency in the Spanish corpora to characterize some variable of the process. There are differences in frequency between Adjunct in the Theme of the clauses and Adjuncts in the Rheme of the clauses. For example, Adjuncts of location are of high frequency in both Themes and Rhemes. However, in Chart 4-4 we saw that location is by far the most widely thematized circumstantial Adjunct in both languages, while in the Rheme it is surpassed by circumstantial Adjuncts of manner in Spanish and its frequency is almost equaled by manner Adjuncts in English. This underscores the popularity of location as Theme in the corpus. What are the differences between placing an Adjunct of location in the Theme or in the Rheme of the clause? One reason seems to be textual: Adjuncts can either make or break a thematic chain. For example, the following set of clauses all have as their point of departure a circumstantial Adjunct of location: 4-103.Eng 1 (11) Today we are likely to ask what industrialization did to the environment; (12) in the 1950s scholars more often asked about the conditions that led to economic... (13) a generation earlier the central question was how workers had gained ... If any of the circumstantial Adjuncts here were left for the Rheme, the thematic chain of moving back in time to look at what concerned historians would be broken. At other times, if the circumstantial Adjunct were moved out of the Rheme into the Theme, it would break up a thematic chain. For example: 4-104.Spa 1 (101) Jenofonto, en su Anбbasis o Retirada de los Diez Mil, hace el relato completo de un conjunto de hechos relacionados нntimamente; (102) Tucнdides, en su Historia de la Guerra del Pelaponeso, aнsla un secuencia de acontecimientos y los historнa. Here the chain consists of names of authors, and their works are included in the Rhemes of the clauses as the location in which they carry out the verbiage of the clause, which is left in both cases for the most newsworthy part of the clause, the end of the Rheme (Fries, 1994). The authors could have opted for thematizing the circumstantial Adjuncts in both cases, yet this section of the text discusses Greek and Roman authors and how they combine literary texts with historical accounts. Therefore, the authors are thematized and form a chain. This 160
also takes in the notion of thematic continuity, which "is a central consideration in determining which element gets the sentence initial position" (Lowe, 1987: 9) At other times, the circumstantial location is the newsworthy information in the clause, and is thus left for the Rheme, e.g.: 4-105. Eng 5 (47) A later species, designated Homo erectus, appeared about three hundred thousand years ago; Several of the clauses surrounding this one thematize circumstantials of time and space, yet this clause leaves the temporal notion until the end, as it is the important information, and hence that which advances the overall goals of the text (Fries, 1994). However, Fries' notion that the N-Rheme of the clause has the purposes of advancing the overall goals of the text, while the thematic content has more local concerns, appears to be challenged by Thompson (1985) through her analysis of initial and final purpose clauses. For Thompson (1985: 55), while these two clause types have an identical appearance grammatically, their functions are quite different: the initial purpose clause functions to state a `problem' within the context of expectations raised by the preceding discourse, to which the following material (often many clauses) provides a solution, while the final purpose clause plays the much more local role of stating the purpose for which the action named in the immediately preceding clause is performed. This is illustrated in the following clauses (the purpose clauses are underlined): 4-106. Eng 3 (27) To understand this paradox, we must first examine how these disasters ... 4-107. Eng 2 (115) and the pot, the raft, and the wheel combined to provide the means to transport grain ... . In analyzing Fries and Thompson's positions, it can be seen that both the Theme and the Rheme have local and global functions, but they carry these out differently. For Fries, the Theme of the clause has the role of orienting readers to the message which follows in the rest of the clause. In example 4-106, the reader is expected to interpret the Rheme of the clause as the way of understanding the paradox, as that is the orientation provided in the Theme. This concords with Thompson's view of things. However, she takes the notion further in terms of discourse. As she points out, initial purpose clauses often initiate a discourse span, whereby a number of subsequent clauses provide the solution to the problem stated in the purpose clause. Indeed, the scope of the initial purpose clause in example 4-106 stretches from Clause 27 to Clause 51. Therefore, it is possible to say that the initial purpose clause has a two-fold role: global at the level of discourse, in that its role 161
is to link what has been happening previously in the text with what is to follow, and local at the level of the clause, in that it provides the point of departure for that particular clause, it provides a frame in terms of why "we must first examine how these disasters affected Europe". We have seen in section 3.2 that Berry delineates two roles for Theme, one at the level of discourse and one at the level of clause. Not all Themes carry out the former role, but they do the latter. The discourse role is the one that is global in terms of text and cohesion, while its role in the clause is local. Likewise, the Rheme also has a dual role, as it has concerns within the clause, in its function of expanding on or providing more information about the Theme, and it has concerns with the whole of the text. In example 4-107, the purpose clause must be interpreted in the light of its role of expanding on the Theme. In Thompson's vision, then, the final purpose clause states the purpose for which the pot, the raft, and the wheel combined; thus its scope is limited to this clause, and, therefore, it is acting locally. Its purpose scope ends with that clause, unlike example 4-106, whose initial purpose clause initiates a span of many subsequent clauses. In stating the purpose, however, the Rheme also acts to move the text along with regards to information, in this case information about how a trading economy came to exist. It is in this sense that Fries says that the Rheme advances the global concerns of the text. Obviously, Themes and Rhemes look both inwards and outwards: Themes seem to do the latter more in terms of connecting with what has come before in the text, while Rhemes seem to do so more in terms of furthering the discourse. In sum, there are several different reasons for circumstantials to appear in Theme or in Rheme position, and these reasons hold across both corpora. The similarity in the genre and in the general function of circumstantials in clauses and in discourse, then, goes a long way in explaining the similar configuration of circumstantial Adjuncts across Themes and Rhemes in this study. The percentage of clauses containing subordinate clauses which express circumstantial meaning are laid out in Chart 4-14. As was the case with subordinate clauses as Theme, while there appear to be differences between the two corpora, the total number of clauses in the Rhemes is too low to be able to assert that, for example, there are 162
more expressions of Cause and Contingency in Spanish history textbooks than in the American. Also here there are more differences amongst texts of the same language than across the two corpora. The number of clauses which express cause in English ranges from 0 to 5 instances, while in Spanish it ranges from 0 to 6 instances.
Chart 4-14: Circumstantial Clauses in Rheme
2%
2%
1%
1%
1%
1%
English
1%
Spanish
0%
0%
0%
Acc Angle Cau Cont Ex Loc Man Matt Other
In combining the results of circumstantial elements, non-finite clauses, prepositional phrases and adverbials functioning as Adjuncts with clauses as Theme, (Chart 4-15), the results of the frequency of the Adjunct types is repeated, which underscores the relative unimportance of the differences found in the frequency of the types expressed through subordinate clauses.
Chart 4-15: Combined Results: Circumstantial Clauses and Adjuncts in Rheme
20% 18% 16% 14% 12% 10% 8% 6% 4% 2% 0%
English Spanish
Acc Angle Cau Cont Ex Loc Man Matt Other
163
The same phenomenon occurs with the circumstantial subtypes. While Chart 4-16 shows a higher preference in the Spanish data for expressing the notions of reason and concession in the Rhemes of the clause than does English. Chart 4-16: Clause Circumstantial Sub-types in Rheme 2% 2% 1% 1% 1% En g lis h 1% S p a n is h 1% 0% 0% 0%
Mean Com Qua Beh Res Pur Rea Conc Cond Spa Temp
However, if we look at the combined results of the sub-types (Chart 4-17), this difference is minimized. And again, the configuration of the circumstantial Adjunct sub-types is mirrored here in the combined results, confirming the notion that the Spanish texts do show a greater predilection for characterizing some aspect of the process through a quality.
Chart 4-17: Combined Results: Circumstantial Clauses and Adjunct Sub-types in Rheme
16% 14% 12% 10% 8% 6% 4% 2% 0%
En glis h S pa n is h
Mean Com Qua Beh Res Pur Rea Conc Cond Spa Temp
164
4.7 Conclusions: Transitivity Analysis This chapter has looked at the history textbook corpus from the point of view of transitivity, by analyzing the processes, participants and circumstances involved in the Themes and the Rhemes of the clauses. Some interesting conclusions have been drawn from the analysis. First of all, with regards to Theme, there is a preference across the history textbooks to thematize a circumstance, especially one of location, be it temporal or spatial. Secondly, due to the nature of its grammar, Spanish prefers to thematize the process itself, above all in situations where the process adds little newsworthy information to the clause. This, of course, cannot be done in English, except in very few contexts. We have also seen that there is a great deal of nominalization across both corpora, in the form of Carriers, Identifieds, and Factors. Chart 4-18 shows the most frequent circumstances and participants across the whole corpora, in the Themes and the Rhemes of the clauses combined. This shows that the most popular feature of the textbooks is attribution, with a great portion of both of the texts made up by participants in attributive or identifying relational clauses. Indeed, authors thematize Carriers and Identifieds much more than any other participant (see Chart 4-2 and Chart 4-3), making them the point of departure. Circumstantial expressions of manner are also high on the Chart 4-18; however, they are low in count as Theme of the clause (see Chart 4-4), as are Attributes. These findings manifest the descriptive and explanatory nature of the textbooks: Carriers are taken as points of departure, and Attributes elaborate on events and phenomena in the Rhemes of the clauses. Circumstances of manner develop concepts further in the Rhemes of all types of processes. This tendency towards description and explanation are slightly higher in the Spanish corpora. Also highly prevalent in both corpora is the rooting of the reader in time and space and is of equal frequency across the two languages. 165
Chart 4-18: Frequent Circumstances and Participants 45% 40% 35% 30% 25% 20% 15% 10% 5% 0%
English Spanish
Attribute, Identifier Carrier, Identified Loc Factor, Affected, Result, R.A., Cau Man Actor Goal Phenomenon Existent
Cause is definitely an active participant in the history textbooks, as was hypothesized. Chart 4-18 shows the combined results of those participants and circumstances which indicate cause: Factor, Affect, Result, Resulting Attribute, and Circumstantial Adjuncts and subordinate clauses which express cause (labeled Cau). This combined group shows the prevalence of cause/effect relationships across the two corpora, and, again, this is seen to occur with a highly similar frequency. It is also interesting to note that neither Cause circumstantials nor Factors are thematized as frequently as many of the other categories. Thus, Cause is contained more in the Rhemes of the clauses, where the overall meanings of the texts are advanced. The English corpus shows a higher percentage of Actors and Goals. This, along with the greater tendency in Spanish towards attribution and circumstantial elements of manner, and along with the higher percentage of material clauses in English (10% more than in the Spanish corpus) may show a slightly greater tendency in English to give a more dynamic account of history through narration of actions and of individuals, and a slighter greater tendency in Spanish to provide the reader with a more static account through structural explanation of events. 166
Existents, with a slightly higher frequency in Spanish, and Phenomena, make up the final two most popular categories of participants/circumstances in the corpora. It is logical that in history textbooks there is frequent mention of existence of events, places, phenomena and so on, in terms of their coming into existence or their changing shape of existence. Phenomenon in mental processes is an interesting case, as it is not coupled with the same frequency of mention of the accompanying participant type, Senser. This is because quite often the writers postulate that something is known, or understood, without attributing that knowledge or understanding to a specific Senser. In sum, the analysis has shown a very definite character of history textbook writing, which holds for the most part across the two languages. The Theme analysis has shown the most common points of departure, and the Rheme analysis the most common ways of expanding on this information. However, it is difficult to know whether this is a composite picture of a history textbook genre, as there is little to compare it with besides the studies mentioned, for example those by Francis (1989, 1990), Whittaker (1995) and Ghadessy (1995). Clearly more studies are needed to compare and contrast the results arrived at here, to determine if the findings are tied to the genre of history textbooks, to history writing in general, or perhaps to textbook writing in general. 167
5. Thematic Patterning 5.1 Thematic Structure and Information Structure The previous chapter focused on the experiential content of the Themes and the Rhemes of the clauses; in this chapter, the focus is on the textual. For Halliday, textual meaning "is relevance to the context: both the preceding (and following) text and the context of situation" (Halliday, 1985a: 53). Here, the focus is particularly on the role played by Theme in helping to provide texture. The role Theme plays in the textual metafunction of language, which means "creating relevance to context" (Halliday, 1994: 36), is not all that straightforward, although, indeed, Halliday (1994) centers his discussion of Theme on this metafunction. The textual metafunction, which construes the clause as message, encompasses not only thematic structure, but also information structure. These two structures together "constitute the internal resources for structuring the clause as message" (Halliday, 1994: 308). The composition of thematic structure has been discussed in Chapter 3, so we move here to discussion of information structure. While thematic structure is reflected in the word order of the clauses, information structure is not realized by the sequence of elements in the clause, as the focal point of information may appear at any point of the clause. Information structure is composed of the two functions Given and New. Given refers to information which is presented as recoverable, either in the context of the situation or in the surrounding text, in what has been mentioned previously. New refers to something that is presented as non-recoverable, in that it has not been mentioned before, or it is unexpected or surprising or in contrast to previous information and/or expectations. According to Halliday, the relationship to the textual metafunction of thematic structure and information structure is that, while both are speaker-selected, Theme + Rheme is speaker-oriented, in that it is what the speaker decides to take as the point of departure of the message, while Given + New is listener-oriented, in that it is presented as what listeners already know or have at their disposal at any given point in the discourse. In other words, "given-new is a discourse feature, while theme-rheme is not" (Halliday, 1976: 179) because Given-New, or Information, "structures the item in such a way as to relate it to the preceding discourse, while thematization structures it in a way that is independent of what has gone before" (ibid: 180). 168
The purpose of the above discussion is to bring in the notion of thematic progression. Thematic progression refers to the way Themes interact with each other and with Rhemes in order to provide continuity in discourse and to organize the text. Danes defines thematic progression as follows: ...the choice and ordering of utterance themes, their mutual concatenation and hierarchy, as well as their relationship to the hyperthemes of the superior text units (such as the paragraph, chapter...), to the whole text and to the situation. Thematic progression might be viewed as the skeleton of the plot. (Danes, 1974: 114). Thus, for Danes, thematic progression, along with other structures, represents text connexity. This would seem to clash with Halliday's notion that the Theme-Rheme construct is independent of discourse. This clash can be explained by the difference in Halliday and Danes' view of Theme; for Halliday Theme is what comes first in the clause, while Danes' employs a procedure using wh-questions in order to elicit the Rheme of an utterance (more along the lines of Firbas' Communicative Dynamism; see section 3.1.3). Danes' view of Theme, then, reflects more Halliday's Given-New structure, which is indeed considered to be a discourse structure. As the present study centers on the Theme-Rheme structure, and not on Given-New, thematic progression would not seem to have a bearing here. However, as we have seen (section 3.3), Halliday points out: "Other things being equal, a speaker will choose the Theme from within what is given, and locate the focus, the climax of the New, somewhere within the Rheme" (1994: 299). This point was touched on in Chapter 3, where it was stated that discourse concerns are often brought to bear in what the speaker chooses to place in Theme position. Indeed, in textbooks, we can presume that at least one of the purposes of the author(s) is to present information in a way that facilitates comprehension on the part of students, or novices in many cases to the subject field. One way of doing this is to present information in such a way that readers can connect with relative ease what they are reading with what has come before, thus building up a clear frame of reference of, in this case, the historical experiences presented in the text, as "other things being equal, reading is easier when the New information is presented in the Rheme position" (Bloor and Bloor, 1992: 35). Ventola & Mauranen (1991: 469) also make this point: "Thematic development is important for the readability and clarity of a text. In a well-formed text, the thematic patterns should reflect the organization of text-content and facilitate the reading process." Vande Kopple (1986) carried out experiments in reading by using texts which followed the `Given first' 169
principle and others which did not, and found a significant difference in increased readability with the former. For him, given information first "makes good sense because it is easier to make a connection to what one knows and then add some new information to it than to receive new information, store it, and then learn what it connects to" (Vande Kopple, 1991: 326). Richards (1993) offers thematic progression as a means distance materials writers can use in order to make their texts more comprehensible to readers and in order to avoid a major problem of writing, which, as expressed by Bloor and Bloor (1992: 34) is: "how to carry the reader along with the writer's assumptions about what is and what is not shared information in the absence of the possibility of negotiation of meaning such as is available to participants in face to face interaction". Indeed, in Richards' analysis of a set of materials which students had found difficult to follow, thematic progression was difficult to establish and confusing, as at times, for example, the writer placed New information in Theme position. 5.2 Thematic Progression In light of the above discussion, the history textbook corpus was analyzed for thematic progression in order to determine whether the writers did indeed provide for ease of reading and assimilation of material by placing Given material in Theme. Placing Given material in Theme implies that the Theme is connected with something that has come before, thus bringing about different types of thematic progression, depending on what the Theme of any given clause may be connected to previously. According to Danes (1974), in Czech scientific writing, as well as other professional texts, there are three main types of thematic progression. In the first, the Rheme portion of each sentence becomes the Theme of the following sentence. This is called simple linear progression. and, according to Danes, (1974) is the most elementary, or basic, thematic progression. This is illustrated here with an example from the history textbook corpus: 5-1. Eng 2 (60) so a combination of assertiveness by males and acquiescence by females may have pointed toward social divisions based on gender. (61) One result of such social divisions has been a comparative lack of information about the role of women in history; (62) the reconstruction of this role, the restoring of women to history has been a leading theme of historical research in the present generation. In example 5-1, each subsequent Theme picks up on information provided for in the previous Rheme; this can be mapped as: 170
Figure 5.1: Mapping of Simple Linear Progression (60) so a combination of assertiveness by males and acquiescence by females may have pointed toward social divisions based on gender. (61) One result of such social divisions has been a comparative lack of information about the role of women in history; (62) the reconstruction of this role, the restoring of women to history has been a leading theme of historical research in the present generation. T60 +R60 T61 +R61 T62 +R62 In another pattern, successive sentences share the same Theme, which is termed constant Theme. Again, an illustration is provided from the corpus: 5-2. Eng 7 (68) Homer employs a particular event, the quarrel between an arrogant ... (69) Homer grasps that there is an internal logic to existence. (70) For Homer, says British classicist H.D.F. Kitto, "actions must have their consequences: ill-judged actions must have uncomfortable results". In example 5-1, the Theme of the three clauses is Homer, and the thematic progression can be mapped as seen in Figure 5.2, below: Figure 5.2: Mapping of Constant Theme Progression Homer Homer For Homer T68 = T69 = T70 Dubois (1987: 93) characterizes these two types as "canonical". Danes goes on to posit a third pattern, which is called derived Theme; in this pattern, the topics of each sentence are individually different, but are all derived from the same overriding theme, or hypertheme or overall theme of a paragraph or text. Dubois (1987) raises objections to considering derived Theme as a different type of progression, basing her objections on Danes' point indicating that givenness can be indirect as well as direct, through semantic inference or semantic implication, including relations obtained through hyponymy and hyperonymy, and through associative relations, which Danes' (1974: 110) exemplifies with restaurant associated with lunch, summer with vacations, and science with investigator. This obliges researchers "to treat instances of indirect constant as hypertheme - which they do not do consistently" (Dubois, 1987: 108). The problem here may be related to the notion of what is inferrable, which can be connected to the notion of subject specialist knowledge. Prince (1979, 1981; mentioned above in section 3.1.2) describes an entity as inferrable "if the speaker assumes that the hearer could have inferred it, via logical ­ or, more commonly, plausible--reasoning, from entities already Evoked, 171
or from other Inferrables" (Prince, 1979: 271). To illustrate this, an example from the history
textbook corpus is used here:
5-3. Eng 7
(115) The Greeks also valued free citizenship. (116) An absolute king, a tyrant who ruled arbitrarily and by decree and who was above the law, was abhorrent to them. (117) The ideals of political freedom are best exemplified by Athens.
The Rheme of 115 in example 5-3 introduces the notion of free citizenship. In 116 an entity is
introduced through the use of the indefinite article "an", which, in effect, introduces the entity as
new. However, we can surmise that the author presumes on the part of the reader a knowledge
of the world in which absolute kings, or tyrants who rule arbitrarily, are antonymous with the
concept of free citizenship. As the Theme in 117 continues on the concept of political freedom,
this set of clauses has been analyzed as follows:
Figure 5.3: Mapping of example 5-3: (115) The Greeks also valued free citizenship. (116) An absolute king, a tyrant ... (117) The ideals of political freedom T115 + R115 T116 T117
In other words, clauses 115 and 116 form a simple linear thematic progression chain, while clauses 116 and 117 form a constant Theme chain. This has been arrived at through inference, and through what we can assume the reader knows about the world. This is not to say that these presumptions are correct; in fact, there could feasibly be readers of these texts who do not make these inferences, who may not connect these notions in their cognitive schema of the political world. I, as the reader-analyst, may have more world experience than the age group for which this text is targeted. Indeed, this shows the problem of the derived Theme analysis; it means that the analyst is making the decision as to whether the Theme in question is included in the text through semantic inference, thus placing it in the categories of simple linear or constant Theme. It is interesting to note that Danes (1974) says that he is working mainly with scientific and other professional texts, and, in fact, states that the "choice and sequence of the derived utterance themes will be controlled by various special (mostly extralinguistic) usage of the presentation of the subject matter" (Danes, 1974: 120). Nwogu (1990), in his analysis of three medical genres, found examples of the derived Theme pattern only in the most specialized of the genres. Perhaps the notion of derived Theme can be thought of as less "given" and more indirect
172
to a `lay' reader; thus one would not expect to find it in texts where the readers do not share the same degree of knowledge shared by writers and readers of very specialized texts. Thus, writers of non-specialized texts for non-specialized audiences may need to infuse their texts with clearer links of meaning, thereby producing links which are closer to constant Theme or simple linear Theme rather than to derived Theme. However, at the same time, it must be noted that Dubois' (1987) study is based on the thematic progression of talks given in conjunction with a slide presentation at a meeting of experimental biologists, whom we can assume share a high degree of specialized knowledge. Perhaps the lack of derived Theme here can be attributed to the spoken nature of the presentation; while it is true that the audience and the speaker share a similar degree of knowledge, the spoken format makes information processing more difficult than would a written format, so it may be that the speaker makes even indirect links clearer, thus prompting Dubois to analyze Theme chains in certain cases as constant Theme rather than as derived Theme. It is difficult to come to a conclusion on this point without comparing and contrasting data which these researchers have analyzed as pertaining to the derived Theme progression. Dubois does include one instance in her analysis: Figure 5.4: Derived Theme Example (Dubois, 1987: 94) (1)Herman Rahn and his colleagues (2) Recent work with some lizards Dubois identifies, with some difficulty, as she points out, as the hypertheme for these two Themes "something like `sources of information for the hypothesis of constant relative alkalinity'"(ibid: 94). She goes on to say that these two Themes would appear to be totally unrelated if it were not for the hypertheme. The example of derived Theme which Danes includes is not from a specialized text, but from a description of New Jersey, perhaps taken from an encyclopedia. The example is (Themes have been italicized for ease of discussion): Figure 5.5: Derived Theme Example (Danes, 1974: 120) New Jersey is flat along the coast and southern portion; the northwestern region is mountainous. The coastal climate is mild, but there is considerable cold in the mountain areas during the winter months. Summers are fairly hot. The leading industrial production includes chemicals, processed food, coal, petroleum, metals and electrical equipment. The most important cities are Newark, Jersey City, Paterson, Trenton, Camden. Vacation districts include Asbury Park, Lakewood, Cape May, and others. Dubois' objection could perhaps apply in this case. Leaving aside the there Theme of the fourth clause, all of the other Themes can be thought of as carrying an implied "in New Jersey" (e.g. the 173
northwestern region in New Jersey; The coastal climate in New Jersey,...); therefore this passage could be analyzed as employing mainly a constant Theme progression based on the notion of semantic inference in givenness. Another analyst, Weissberg (1984), take hypertheme to mean a whole/part relationship, which can include specific mention of the overriding theme in the text, as can be seen from his example: Figure 5.6: Derived Theme Example (Weissberg, 1984: 490) The reflector was protected from the weather by an outer window of 0.10 mm tedlar. The focal length of the reflector was 22.8 cm. The back of the reflector was protected from the weather with a black polyethylene cover stapled to the frame. The reflector rack was mounted as ... Given the repetition of the word "reflector" in all of the clause Themes, it would seem more consistent to label this progression as constant Theme. Indeed, in the history textbook corpus, it was easier to be consistent by analyzing these types of relationships as such, following Dubois. Another example in addition to the one given above (5-3) from the corpus of possible derived Theme in Danes' terms, but which was analyzed as constant Theme, can be seen in Figure 5.7: Figure 5.7: Possible Derived Theme (Analyzed as Constant Theme): Eng 8 (13) Beautiful homes for the well-to-do, constructed around gracious open patios decorated with fountains and Greek statues, were built on the Palatine Hill. (14) Delicate temples in the Hellenistic style were erected beside the river. (15) Warehouses and shops expanded to deal with the rising commerce of the empire. (16) More ominously, the multi-storied apartment houses of the city's slums expanded to accommodate the thousands of dispossessed farmers... Here we see several examples of "buildings", a concept which has been mentioned previously in the text and which could be taken to be the hypertheme of this section. However, these four clauses have been analyzed as forming a constant Theme chain, as their Themes are all hyponyms of building. Therefore, for the sake of consistency in analysis, derived Theme has not been taken into account as a thematic progression type in this study. This does question the notion of the derived Theme chain, as Danes' and Weissberg's examples, taken from specialized texts, also can be analyzed as constant Theme through inference. This does not solve the problem of inference entirely, however. It is still up to the analyst to decide that a link occurs, rather than analyzing a Theme as new information. As Vande Kopple (1991: 317) points out, "...the issue of whether some inferences from a text are valid will probably always remain a matter of interpretation, subject to influences from different cultures, subcultures, times, kinds and amounts of readers' background knowledge, readers' 174
purposes and questions about writers' intentions". This has been kept in mind during the analysis, and where the links are thought of as difficult, especially when keeping in mind a younger reader, these will be mentioned, or they will be analyzed as outside the thematic progression patterns (see section 5.3.2, below). Returning to the different types of thematic progression, Danes also includes a progression sub-type which is called split Rheme, where the Rheme of a clause contains two ideas which are developed in subsequent clauses, e.g.: Figure 5.8: Example of Split Rheme Progression Eng 10 (183)He also enlisted the aid of two able adventurers, Alberoni and Ripperda, ... (184)Alberoni, the son of an Italian gardener, was successively a cook,... (185)Ripperda, a Dutch business expert and diplomat, ultimately lost the favor ...
which can be mapped as:
Figure 5.9: Mapping of Split Rheme Progression
(183)He also enlisted the aid of two able adventurers, Alberoni and Ripperda, ...


(184)Alberoni... (185)Ripperda,...
T183 + R(i) T184
+ R(ii) T185
To the above patterns of constant Theme, Simple Linear, and Split Rheme, I have added the pattern Split Theme. Split Theme means that a Theme may contain more than one idea, and these ideas are developed in different subsequent clauses, as exemplified in Figure 5.10:
Figure 5.10: Split Theme Progression Eng 3 (122) The upward movement of wages and the downward price of cereals led ... (123) Better wages in both town and countryside enabled the population to ... (124) While the price of wheat fell, wine, beer, oil, butter, cheese, meat, fruit, ...
Which can be mapped as follows:
Figure 5.11: Mapping of Split Theme Progression
(122) The upward movement of wages
and the downward price of cereals


(123) Better wages in both town and countryside (124) While the price of wheat fell,
T122i T123
T122ii T124 175
Dubois (1987) adds a further point in her discussion which is that the above types of thematic progression can be either contiguous or gapped. Some of the examples she gives have a gap between them of as many as 12 clauses (in Dubois' derived Theme example cited above in Figure 5.4, there is a gap of 80 clauses). As the purpose of this chapter is to determine how the history textbook writers package the information in order to make it easy for students to unpackage the text, it is felt here that such gaps between referents are not optimal. Therefore, gaps of more than three clauses were not counted as chains. In other words, Themes for which the reader has to go back more than two clauses to find a previous mention of a concept were not counted as part of a thematic progression chain. To sum up this section, the history textbook corpus has been analyzed for four different types of thematic progression, which can be grouped into two overall types: Theme progression, which includes constant Theme and Split Theme patterns, and Rheme progression, which takes in Simple Linear and Split Rheme development. These four types can also be gapped in the case that the referent of the Theme lies in a clause not more than three clauses away. In the next section, I outline how patterns were determined. 5.3 Method of Analysis 5.3.1 Thematic Progression Chains All of the clause Themes were analyzed to determine how they related, if indeed they did so, to the previous discourse, whether to a previous Rheme or a previous Theme. This was determined on the basis of the presence of cohesive devices such as exact lexical repetition, synonyms, pronouns, substitutions, and ellipsis, (Halliday and Hasan, 1976), which can be expanded to include paraphrase and semantic inference (Nwogu & Bloor, 1991), and relationships involving antonymy, and hyponymy. Semantic inference has already been mentioned above, along with shared knowledge between the writer and the reader. Determining whether that can be said to exist is not always a straightforward process, as we have seen in example 5-3 above. There are more examples in the corpus that cause problems in determining whether the Theme of the clause forms part of a thematic progression chain or not. In example 5-4 below, the Theme in clause 80 is especially problematic: 5-4. Eng 8 (78) Finally, in 12 B.C., he became pontifex maximus, or supreme priest, (79) and took over supervision of the state religion. (80)The forms were undisturbed, 176
At first glance, it is tempting to analyze "the forms" as referring to "the forms of state
religion". However, if we look at the clauses subsequent to that:
5-5. Eng 8 (81)but Augustus in fact held the reins of command. (82)Any spark of independence in the senate quickly died out.
it would seem more likely that the author is referring to "political forms" rather than to religious
ones, making this Theme a summative expression based on a number of previous clauses which
discuss Augustus' political activities and how he maintained at least the forms of the political
structures. In fact, there is a mention in the Rheme of clause 67 of "old constitutional forms".
Although no subsequent mention is made of forms until clause 80, it is to be presumed that the
material between clause 67 and clause 80 is an elaboration of these forms. At any rate, this is
difficult for the reader to work out as the author does not specify the summative nature of the
expression by saying something like "Overall, the forms were undisturbed".
Example 5-4 brings in a phenomenon about which a decision had to be made: this is with
respect to verb-initial constructions and how they were analyzed. Usually, and as in the case of
Example 5-4, clause 79, verb-initial constructions with ellipsis, or pro-drop, were analyzed as
forming part in a simple linear progression chain; in other words, they were taken to have come
from the previous Rheme. This makes sense, as the verb in one Rheme, for example in clause 78,
forms a chain with the verb in the subsequent clause, as in 79. So "took over" is seen as forming
part of a chain with "became". This analysis is possible, given that the Subject for the verbs is the
same, as is implied through the ellipsis, and the verbs have the same inflection for tense.
This point warrants more development. In order to provide for this, a more extended
example is used here from the Spanish corpus:
5-6.Spa 5
(26) Este crecimiento demogrбfico preocupу tempranamente a ciertos pensadores, como Thomas Robert Malthus (1766-1834). (27) En su Ensayo sobre el principio de la poblaciуn (1798) examinу, con visiуn derrotista, el fenуmeno de la expansiуn demogrбfica en relaciуn con las limitadas posibilidades de producciуn de bienes. (28) En sus teorнas partнa de la idea de que la poblaciуn crece mбs que los medios de subsistencia. (29) Consideraba que la poblaciуn aumentarнa en progresiуn geomйtrica (2,4,8,16...) mientras que las subsistencias sуlo lo harнan en progresiуn aritmйtica (2,4,6,8...). (30) Buscaba la soluciуn en la limitaciуn voluntaria de la poblaciуn basada en el retraso de la edad nъbil y en la continencia periуdica. (31) Abogaba tambiйn por mejoras en la productividad agraria mediante la introducciуn o extensiуn de cultivos mбs rentables para la alimentaciуn humana.
Clause 26 forms a simple linear chain with Clause 27, as the Theme in 27 refers to Malthus'
work, "In his Essay.... This then forms a constant Theme chain with Clause 28, "In his
177
theories...". Then follow 3 clauses, 29, 30 and 31, which are all verb initial through ellipsis. The verb in Clause 28 (underlined) has also undergone ellipsis, thus the thematized verb in 29 forms a simple linear pattern with "partнa" in Clause 28. As we have seen above, one of the criteria for establishing relationships between Themes and previous bits of text is through ellipsis (see section 5.3.1). Thus, to some extent, it could be said that the referent for Consideraba in Clause 29 is Malthus, which would mean then that it forms a constant Theme chain with Clause 28, not a simple linear chain. However, the verbs are also chained together through their inflection: нa (for verbs which end in -ir and -er) and aba (the inflection for -ar verbs) denote past time. Thus, in addition to all having the same referent, they all refer to the same past time; hence, partнa - consideraba buscaba - abogaba form a chain; consideraba forms a simple linear chain with partнa, given that this latter verb is encoded in the Rheme of its clause, and then consideraba forms a constant Theme chain with buscaba. This sample is analyzed as follows:
Figure 5.12: Mapping of Example 5-6: (28) En sus teorнas partнa (29) Consideraba (30) Buscaba (31) Abogaba T(28) + R28 T29 T30 T31
This also happens in the English corpus, with coordinated clauses in which the Subject has been
elided:
5-7.Eng 10 (190) They improved the tax system, (191) encouraged industry, (192) built up the navy, (193)and fortified strategic points in the Spanish empire in America.
Improved - encouraged - built - fortified form a chain due to their common referent and their
past tense inflection.
There was one exception to this; clause 4 in Spa 7, in which the dropped pronoun has no
definite referent. This is illustrated in example 5-8 below: (the verb Theme with pro-drop in
question is underlined as well as italicized):
5-8.Spa 7
(1) La recuperaciуn iniciada en las dos ъltimas dйcadas del siglo XVII se afirmу progresivamente en la centuria siguiente, llegando, a travйs de altibajos y notables divergencias nacionales y regionales, a un saldo global sumamente positivo. (2)Al comenzar las crisis revolucionarias Europa tenнa una poblaciуn mбs numerosa, mбs rica y mбs instruida que en ningъn perнodo anterior 178
(3)y, lo que es mбs importante, no volverнa a sufrir estancamiento o regresiуn en lo sucesivo. (4)Era el alba de la mutaciуn mбs importante en la historia humana.
From the Rheme, it is clear that "Era el alba" refers to the dawn of an age, but it is not at all clear
from the previous clauses precisely what that age is. The reader can piece together from the first clause that the writer is referring to sometime after the last two decades of the 17th century and
sometime during the following century, and, from clause 2, to a time when the revolutionary
crises were beginning, and when Europe had a more numerous, richer and more educated population. From the title of the chapter, the reader can infer that "Era" refers to the 18th century;
thus, the process of figuring out the referent here is complex, and, therefore, it is counted outside
the Theme chains, but it is the only instance of this type. All other instances of ellipsis can be
clearly chained to a previous verb, through their common referent and by their common
inflection.
Examples 5-6 and 5-7 above bring in another aspect of the analysis: it is usually the case
that clauses, especially pivotal ones such as 29 in example 5-6 and 191 in example 5-7, take part
in more than one thematic progression chain, connecting to a previous Rheme in a simple linear
progression and then connecting with a subsequent Theme, thus forming part in a constant
Theme chain. Dubois (1987: 95) terms this type of complex development "multiple", and she
finds it to be "the rule rather than the exception", as indeed has also been the case in the history
textbook corpus. In these cases, clauses are counted twice in the analysis, but this is taken into
account in the final tally, as will be seen later in section 5.5.
Another type of elliptical expression causes problems in making connections. The ellipsis
in this type is pragmatic, rather than grammatical, in nature, as it involves the elision of concepts.
An example of this is the Theme in clause 185 in example 5-9:
5-9. Eng 6
(183) Economic historians and demographers sharply dispute the impact of the plague on the economy in the late fourteenth century. (184) The traditional view that the plague had a disastrous effect has been greatly modified. (185) The clearest evidence comes from England, where the agrarian economy showed remarkable resilience.
The reader (and analyst, for that matter) might ask "The clearest evidence for what?" Following
from the previous Rheme in Clause 184, it is tempting to think that the author is referring to the
clearest evidence for the fact that the traditional view has been greatly modified; however, in
reading the Rheme in 185, it becomes clear that the evidence is for the notion that the plague did
not have such a disastrous effect. The reader must read the Rheme in 185 in order to be able to
179
process the Theme. Here the author would have done well to have written something along the lines of "The clearest evidence that the plague's effects were not as disastrous as previously thought comes from England...". One that stretches this even further is: 5-10. Spa 2 (148) y adquiere un volumen espectacular en los primeros aсos del siglo XX, hasta 1914. (149) La Primera Guerra Mundial pone freno casi definitivo a esta difusiуn europea... In this example, the Rheme of Clause 148 develops the notion that the volume of world migration was spectacular until 1914. Connected to this is the Theme in 148, the First World War. It may seem simply to be a brand new Theme (Bloor and Bloor, 1992), especially if the reader does not know what year the First World War began. It seems to me that the writer has assumed that the reader knows this, and has thus chained the Theme in 149 to the Rheme in 148. These examples are the clearest illustration of some of the problems inherent in establishing links; in most of the cases, it is fairly easy to make the links and establish if the Theme of a clause proceeds from a previous Theme or a previous Rheme. However, this is not to say that new entities are not introduced into the discourse. In fact, by no means do all of the Themes fit into the thematic progression patterns. Those that do not are explained in the next section. 5.3.2 Peripheral Themes As mentioned above, in no way are all of the Themes in the corpus chained to a previous Theme or Rheme in close proximity. However, this does not necessarily mean that they are difficult to process, only that they are difficult to analyze in terms of linking them to one previous point in the discourse in order to say that they belong to one or another chain. These Themes are important to the organization of the discourse; they are only labeled as peripheral here in that they are peripheral to the thematic progression patterns. They can be divided roughly into four categories: pragmatic, grammatical, extralinguistic and metatextual. These will now be discussed in that order. In the course of a given text, very often an entity would be thematized quite regularly, either repeated quite often as the Theme of clauses in a section or a whole chapter. Because of the repetitions, it became difficult to ascertain whether one of these types of Themes proceeded from a preceding Theme or from a Rheme, in order to identify it as belonging to one pattern or another. This is due to the fact that these Themes have been mentioned many times throughout the previous discourse, thus allowing the writer to insert 180
them at any given moment, without necessarily providing a specific link to an immediately preceding clause. These types of Themes are somewhat similar to 'discourse Themes' in Berry's (1995: 18) terms as they are "a priority set of types of meaning that reflects his /her [the writer or speaker's] underlying concerns for the duration of the text or large section of text". Here, these repeated Themes have been labeled key Theme. An example of this can be seen in both Eng 1 and Spa 1, both texts on the study of history. "History" becomes key Theme in these texts, as it is constantly repeated as Theme throughout both chapters. There are other minor key Themes running through these chapters, such as "the book" and "readers". Another type of pragmatic Theme which was not analyzed as forming part of a chain involves summative expressions. These are very similar to what Francis (1994: 85) terms "retrospective labels", which serve to "encapsulate or package a stretch of discourse" [emphasis original]. Retrospective labels do not have as a referent a single nominal group, i.e. they are not repetitions or synonyms of any preceding element. Indeed, in the corpora there are Themes which encapsulate a stretch of preceding text (anywhere from two previous clauses to twenty-five), hence they do not proceed from just one previous Theme or Rheme. I have preferred here not to borrow Francis' term, as it refers to any anaphorically cohesive nominal group, and here I wish to narrow the focus to only those Themes which function as retrospective Themes, or summative Themes. Thus, they are labeled simply sum. The summary Theme does just what its name implies: it summarizes a part of the text in the Theme, and then the Rheme serves either to wrap up the section with an overall comment or it serves to pave the way for a subsequent Theme. They often serve as transitional points in the text, e.g.: 5-11. Eng 6 (198)Even more significant than the social effects were the psychological consequences. As the content of the Theme implies, it sums up a section on the social effects of the plague (beginning at clause 183), and the Rheme presents the new material which the author will next develop: the psychological consequences. Another type of retrospective Theme is that which refers to an entire previous clause, rather than only to a previous Theme or Rheme, e.g.: 5-12. Spa 1 (35)Faltaba evidentemente un modo diferente de hacer historia que, al tiempo que nos expusiera lo que ocurriу, buscara el comprender por quй habнa sucedido. (36)Esto dio lugar a diversos intentos que finalmente cuajaron en una nueva clase de Historia. 181
In clause 36, the demonstrative pronoun "esto" refers to the whole of the previous clause, to the notion that lacking was a different way of doing history. It would be impossible to say that the Theme in 36 comes from only the Rheme, as the notion of lack, encapsulated in the verb Theme in 35, is crucial to the unpacking of "Esto". Therefore, this kind of Theme has been labeled previous clause. Another type of pragmatic Theme is one which is labeled back. As explained earlier, thematic progression links were only established over a maximum of three clauses, i.e. a gapped development was identified if a Theme related to a maximum of two clauses prior to its immediate predecessor. Therefore back refers to links that go farther back than three clauses. Themes can refer back from four clauses to twenty. At times, a key Theme is dropped for a while, and then it is brought back in suddenly through a Theme, which would then be labeled as back. At other times, it is a Theme which has been mentioned perhaps only once before, and thus might not be as readily available in the reader's framework of the text. An example of this is the Theme of clause 40, bad weather and the Theme of clause 41 poor harvests (underlined here for ease of reference): 5-13. Eng 6. (29)Then the fourteenth century opened with four years of bad harvests. (30)Torrential rains in 1310 ruined the harvest (31)and1 brought on terrible famine. (32)Harvests failed again in 1322 and 1329. (33)In 1332 desperate peasants survived the winter on raw herbs. (34)In the half-century from 1302 to 1348, poor harvests occurred twenty times. (35)The undernourished population was ripe for the Grim Reaper, who appeared in 1348 in the form of the Black Death. (36)These catastrophes had grave social consequences. (37)Population had steadily increased in the twelfth and thirteenth centuries, (38)and large amounts of land had been put under cultivation. (39)The amount of food yielded, however, did not match the level of population growth. (40)Bad weather had disastrous results. (41)Poor harvests meant that marriages had to be postponed. Bad weather was mentioned in the previous paragraph, in clause 30 through torrential rains in 1310; poor harvest were also mentioned a number of times in the previous paragraph. I find as a reader that these two Themes divert the readers' attention away from social consequences, the topic of this section, back to the previous paragraph, and thus slow text processing. While these Themes do refer back to either a previous Theme or a previous Rheme, it is felt that, given the break through distance of the chaining, they are not analyzed as being either constant or linear mentions. There is another use of back Theme which allows for organization of text over several paragraphs, in that it is used to chain paragraphs. At times, a Theme of the first clause in a 182
given paragraph refers back to a Theme mentioned perhaps a paragraph earlier. An example of this is: 5-14. Spa 9 (25)Tres corrientes culturales afectan a la Penнnsula Ibйrica... (28)Una es de origen atlбntico, (36)Otra corriente cultural procede del otro lado de los Pirineos (40)Una tercera corriente cultural la representa el Mediterrбneo, ... Here, three cultural currents are posited in clause 25 as affecting the Iberian Peninsula. These are then named in three subsequent paragraphs, beginning with clauses 28, 36, and 40. The Themes of each of these clauses, then, were labeled back. Perhaps it would have been better to give these Themes, which do guide the reader, a different name from the type in the previous example. However, they are similar in that they do require going back in the text in order to find the referent. The above-mentioned Themes which fall outside the thematic progression patterns, key Theme, back, sum, and previous clause, do provide for thematic continuity in the text as the referents of these Themes are all clearly available to the reader in the text. Somewhat more difficult than the others, perhaps is back, as the analyst sometimes has to do some work to make the connection. However, for readers, the referent may still be available in the cognitive framework they have been building up of the text. With the next set of Themes, this is not always the case. We begin this set with the one which I consider pivotal, in that it is somewhere along a cline of being present in the text and of being new to the discourse. This type is called related. It may in some sense thought to be similar to Danes' derived Theme, in that these Themes are related to topics that recur in the history textbook but they may not always be thought to be specifically derived from some kind of hypertheme. These are Themes such as "historians", "scholars", "chroniclers", or words which refer to historical evidence, e.g. "muchos datos nuevos", "archeological evidence"; it is interesting to note that these become key Theme in Eng 1 and Spa 1, the introduction to history chapters, but their mention in other chapters is very sporadic. Readers should have no problems with most of these related Themes, as their frame of reference falls clearly within the scope of the subject matter. However, there are some which are a stretch, e.g. Mommsen in example 5-15. Spa 6 (41) Con Mommsen la historiografнa se desprende de la filologнa, de la que hasta entonces dependнa. In this section, the discussion is on how history has been carried out with respect to Ancient Rome. It is doubtful that many students would know that Mommsen is a historian (indeed the 183
analyst did not!); it can only be inferred from the Rheme of the clause, so there was a temptation to label this Theme as new. Indeed, in Prince's (1979, 1981) taxonomy, she includes under new information a category called "unused". Unused entities are not inferrable in the text, but it is assumed that these entities are available to the text receiver. This often makes it difficult to draw the line between what we can assume a text producer includes as given, or available, to the reader; as Virtanen (1992: 102-104) notes, "The assumptions that the text-producer is making about the text-receiver's consciousness and the knowledge they share may not actually tally with the state of things in the receiver's actual text world and universe of discourse". Also, the writer may assume that the skilled reader will deduce that Mommsen is a historian, although perhaps not all textbook readers have these pragmatic skills. This is simply a difficult call to make. However, I believe that in example 5-15, the textbook writer is presenting the Theme as given (it is interesting to note the inclusion of only the last name - suggesting familiarity; at the same time, perhaps Mommsen has been mentioned in a previous section or chapter - this point is touched upon again below), thus perhaps assuming a greater shared knowledge than an average reader might have. This same author does the same thing only a few clauses later with historians Montesquieu and Gibbon:: 5-16. Spa 6 (44) asн, Montesquieu y Gibbon se ocupan de la tan controvertida cuestiуn de la decadencia del Imperio, ... These, then, have been analyzed as related, along with historians, chroniclers, and so on. Moving along the cline of what can be considered as readily available to the reader in terms of connections, the next Theme type consists of those elements which are new to the discourse, and are, naturally, labeled new. One type of new Theme goes against the perceived convention for academic writing put forward by Bloor & Bloor (1992: 38): that of not treating a topic as Given which has been mentioned previously only in a section heading; in other words "if a topic is introduced for the first time in a heading or title, it needs to be re-introduced in the body of the text"; thus, readers of academic texts do not seem to take as acceptable anaphora in the first sentence of a section for an entity in a subject heading. This happens on very few occasions, but it does occur, e.g. in Spa 1 (where it happens on three occasions): 5-17. Spa 3 3) La filosofнa de la Historia (37)Aunque el tнtulo es desafortunado y falso, la costumbre lo ha consagrado. This example is particularly interesting, because the author uses in this clause, which immediately follows the title, the noun phrase el tнtulo to refer to the title, without introducing the notion of 184
the philosophy of history in the clause itself. The text preceding this section hints at a new way of looking at history, without mentioning what it is. So, the only mention of the philosophy of history is in this title. Even four lines later, in clause 40, the author again refers to the philosophy of history through another noun phrase: 5-18. Spa 3 (40)El movimiento no cristalizу hasta la йpoca de la Ilustraciуn, and, in fact, "la filosofнa de la Historia" is not mentioned until the next clause, clause 41. Therefore, this author takes the title as central in providing the semantic referent to the discussion. Another example is in Spa 5: 5-19. Spa 5 La experiencia rusa (108)Tambiйn fue distinta. Here, in clause 108, the author refers to the Russian experience through pro-drop. Given the scarcity of instances of this type of new Theme, it can only be put forward as a feature of these particular texts. It would be interesting to find out if it is more acceptable in Spanish, however, through further investigation and analysis. At any rate, the Themes of these clauses have been analyzed as new. There is often a very fine line between some of the previous categories and new. For example, it was posited above that "historians", "scholars" and the like could be considered related to the subject matter, and thus not necessarily "new" in the discourse (or, perhaps better expressed in Prince's terms, not "not given"). But there are examples such as the following: 5-20. Spa 2 (68) un autor francйs ha asegurado que sуlo el 10% alcanzaba los 20 aсos. There were no previous mentions of authors, thus, this has been analyzed as new. It may be a bit of a stretch, given that "author" may refer to a historian; clearer, in this sense, are the following examples: 5-21. Eng 2 (23) The astronomer Carl Sagan has reckoned that, 5-22. Eng 4 (8)As William Shakespeare observed, ... Astronomers and writers are not as readily available as historians are in a conceptual framework on history. In example 5-21, there has been a mention of the age of humans in comparison to the age of the world, and the writer calls on Carl Sagan to illustrate his point. Example 5-22 follows a discussion of the importance of learning about history in order to know who we are today, and the author sums this up by giving an observation by Shakespeare. Thus both Sagan and 185
Shakespeare are new Themes. By analogy, the French author mentioned in example 5-20 above has also been taken as new. Other new Themes include the first Theme of each chapter analyzed, as is to be expected. Also, there are some concepts which most likely have been introduced in a previous chapter of the book, especially in those cases where the chapter has been taken from the middle or the end of a book. They have been analyzed here as new, given that the chapters were analyzed in isolation from the rest of the book. I feel that this is justified in part as for student readers it might help to have some specific reference made to a previous chapter to facilitate comprehension and connection of new ideas. Other new Themes just appear "out of the blue", so to speak. Example 5-23 below illustrates this. A new Theme is introduced in Clause 138. (I have included here a rather long stretch of the text in order to give a feel for the previous flow of information: 5-23. (129)The social and intellectual consequences of the Agricultural Revolution were far-reaching. (130)The typical village, with two or three hundred inhabitants, required greater control and organization than the earlier hunting packs of twenty to thirty persons. (131)Authority no doubt continued to rest in the hands of the male elders, who elected a chief to lead them. (132)The farmlands, which were usually considered common property, were worked cooperatively. (133)Since individual and community survival depended on closely coordinated efforts, a premium was placed on consensus and conformity. (134)In such a society there was little room for dissent and individualism. (135)Ideas or actions that interfered with community patterns were viewed as "wrong" [immoral) (136)and were punished accordingly. (137)Thus, the foundations of both "ethics" and "law" were laid in those early agrarian societies. (138)Religious beliefs and rituals were closely associated with food production. The Theme in 138 is brand new (Bloor and Bloor, 1992) in that there has been no previous discussion of religion or religious beliefs, with one minor exception. In clause 69, we see: (69)Language also aided in the creation of abstract ideas - such as the concepts of guardian, spirits, magic, and life after death (religion). However, this one mention, and parenthetical at that, cannot be thought of as a precursor to Religious beliefs and rituals in clause 138. One wonders if the author connects religious beliefs and rituals to the foundations of both ethics and law in clause 137. However, this is very questionable; also, the Rheme in 138 goes on to develop an idea which is completely different from the preceding paragraph. It is interesting to note that the new Theme in clause 138 is paragraph initial; Virtanen (1992) finds that clause initial adverbials which contain brand-new information function "as a crucial marker for the most high-level segmentation in the narrative". 186
Thus, in clause 138 in example 5-23 above, the insertion of a brand-new Theme coincides with a paragraph marker, thus moving on to a new topic. However, not all of the new Themes coincide with new paragraphs or sections (and, indeed, many thematic progression chains flow over paragraphs and even section markers). For example, Spa 6 includes a new Theme in the middle of the first paragraph, in clause 4:: 5-24. Spa 6 (1)La historia de la antigua Roma se distingue por su complejidad en cuanto que exige un trabajo y un mйtodo mбs riguroso de investigaciуn que cualquiera otra йpoca debido a diversos factores: (2)para su estudio se necesita un completo dominio de las ciencias auxiliares; (3)nos hallamos, muy a menudo, en el terreno de las conjeturas, de las teorнas indemostrables, de las hipуtesis...; (4)la problemбtica religiosa se encuentra exenta de racionalismo, siendo, por tanto, muy distinta a la moderna; ... It is difficult to see any connection between the complexity of studying ancient Rome and the religious problem. In sum, new Themes in this study are considered to be those entities whose connection to the discourse is opaque and often quite puzzling. They cause the analyst, and thus, perhaps, the reader, to pause and try to figure out what the connection between the Theme and the rest of the discourse might be. Weissberg (1984) draws on Clark and Haviland's (1977) notion of "inferential bridging" to discuss this notion of obscure linkage, pointing out that what may be easily linked for one reader may be quite difficult for another, thus requiring "more extensive bridging and more reading effort" (Weissberg, 1984: 491). A subcategory of pragmatic Themes, and one which does not pose the above problem for the analyst/reader, is Adverbials. Adverbials of time especially appear in Theme position without necessarily being connected to a previous Theme or Rheme. The subject of History is in a sense time, in particular the narration of events over time. And we have seen in Chapter 4 the prevalence of temporal elements as Theme. Therefore, time adverbials can be thought of as "given", as they are at the heart of the subject. Less frequent are Adverbials of place, but these are also used in Theme position. Virtanen (1992) argues from data that adverbials tend to contain information which are in the middle of Prince's taxonomy, somewhere between evoked, or "on the counter" either textually or situationally (Prince, 1979: 270), and brand-new. At times, adverbials, especially temporal ones, form part of a constant Theme chain, thus becoming a text-strategy at a given point, where the writer takes as the point of departure a point in time in a number of subsequent clauses, e.g.: 5-25. Eng 1 (11) Today we are likely to ask what industrialization did to the environment; (12) in the 1950s scholars more often asked about the conditions that led to economic... 187
(13) a generation earlier the central question was how workers had gained ... Here, the writer takes as the point of departure time, and moves back through time in the Themes of the clauses to show how different periods embody different concerns. However, this type of chaining is not very frequent; when it does take place, it tends to occur over just 2 clauses. In fact, time adverbials are much more likely to occur at the start of a Rhematic theme chain (either a simple linear or a split Rheme progression pattern). There seems to be no overall pattern to their occurrence; they can occur as paragraph initial or they might not; they occur before new Themes, in the middle of gapped Themes, and before sum Themes, as well as prior to other types. Therefore, it is felt here that they are thought of by writers as simply "free", in that they are readily available to readers of history. For this reason, they are considered as a separate category here. Circumstances of manner are on the borderline between the pragmatic and grammatical Theme types, specifically attribution appositives (see section 4.4.5). These two types can be best explained through example: 5-26. Example of a manner Circumstance: Eng 1 (84) In method the new social history borrows theories, terminologies, and techniques.. 5-27. Example of an attribution appositive: Eng 10 (154) Intelligent but timid, lazy, and debauched, Louis XV did not have the interest These two types are similar in that they serve to throw the attention of the reader onto the Subject appearing in the Rheme of the clause. They do not function to link back to something previous in the text; therefore, their function is much more local to the clause. However, I have not wanted to categorize them as new, given that their referents follow so close behind, and indeed they are forward-looking, or cataphoric, in nature, and they do not interfere with text-processing. They are labeled here fronted. Related to this notion of cataphora is the Theme type which is labeled "inversion", where the Subject and verb are inverted, thus putting the verb into Theme position. These will be discussed more fully in the results section, section 5.5, below. Spanish also has impersonal constructions (see Chapter 3) including one type of passive expression, in which the initial reflexive "se" is Thematized. Other types of grammatical Themes which lie outside the thematic progression patterns are extraposed and existential constructions and thematic equatives in both Spanish and English. Also, in interrogative clauses, the "Wh" element is the Theme. It is felt that, while these constructions do not add to thematic progression, they do not cause difficulty in text processing for readers. 188
Another category of peripheral Themes is extralinguistic. This includes "we" and two instances of the imperative: "imagine" and "recuйrdense". These are taken as given, specifically "situationally evoked" in Prince's (1979, 1981) terms, and, once again, they are analyzed outside the thematic progression chains, but they do not hinder text processing. A final category of peripheral Themes is labeled metatextual. These are Themes which refer to the text itself, yet, unlike the textual Themes looked at in Section 5.6 below, the reference is encoded as a prepositional phrase, a clause or as a nominal group, and often refers to other parts of the text, as well as to the authors' purpose. Examples are: 5-28. Eng 2 (17) In this first chapter our aim will be to look at the life of the earliest human beings ... Eng 3 (11) As Chapter 12 will show, there were vigorous developments in philosophy, ... Eng 6 (13) This chapter will focus on these questions. Eng 9 (25) As Figure 26.3 shows, more than eleven million left in the first decade ... Spa 9 (10) Por ello, aquн se ha optado por una fуrmula mбs equilibrada... Spa 10(59) En el capнtulo tercero nos ocuparemos de ellos. 5.4 Expectations on Thematic Progression What can be expected of history textbook writers in terms of textuality? We have already seen in section 5.1 that Richards (1993) advises writers of distance learning materials to provide for ease of reading through theme progression patterns. What advice does the history textbook writer receive? Marwick (1970: 59) counsels history textbook writers in the following way: ...the historian should communicate his information in a clear and efficient manner. At best his manner of presentation should conform to the highest canons of literary style, showing that history is art as well as science...Depending on the subject of study, and bearing in mind both that history is concerned with change through time, and with explanation and interconnections, he should establish a reasonable balance between narrative, analysis and description. Clarity, efficiency, and style are of importance, as well as a mixture of narrative, analysis and description. The bearing the first three have on thematic progression has been discussed in previous sections in this chapter; the bearing of the last three are discussed subsequently. In order to put forward some expectations with regard to results on thematic progression, it is necessary to look to other analysts' work in this area. The obvious can be stated here, based on the authors mentioned in section 5.1 above (Bloor and Bloor, 1992; Ventola & Mauranen, 1991; Vande Kopple, 1986, 1991; Richards, 1993), and on Marwick's concern for clarity and efficiency: authors of the history textbooks will provide links for their student readers by chaining Themes in thematic progression patterns. As 189
mentioned above in section 5.2, Nwogu (1990) found instances of derived Theme progression only in the highly specialized texts in his corpus, and this mainly occurs in the Discussion sections; in addition to that finding, his corpus shows that simple linear thematic progression patterns tend to predominate in stretches of text involving explanation and exposition. This seems to be confirmed by Weissberg's (1984) study of scientific writing, in which the simple linear pattern occurs more frequently than the constant Theme pattern, especially in the Discussion and Introduction sections of research papers. As well as to the use of this pattern with the rhetorical patterns of explanation and exposition, Nwogu attributes the use of the simple linear pattern to the nature of readers and writers and to the relationship between them. The level of shared knowledge in texts which were written for a general audience, in which Nwogu finds a high incidence of the simple linear pattern, is unequal, as is the case in the history textbook corpus in this study. Nwogu indicates that the difference in shared knowledge between the writer and the reader limits the range of options from which the writer can choose a point of departure for a clause; therefore, "the foundation for a subsequent utterance can only be established by selecting from the range of information contained in the rheme of the preceding context to which the reader and writer can hold as shared knowledge" (Nwogu, 1990: 244). In using the simple linear pattern, then, writers can ensure that the readers are constantly "with them" in terms of points of departure, thus elaborating on concepts in a way which allows readers to optimally build up the conceptual framework. In a spoken corpus, the Lancaster Spoken English Corpus, on the other hand, Gуmez (1995) found a predominance of the constant Theme pattern, which she attributes to the preponderant use of subject pronouns, common nouns and proper names in spoken text. This is in contrast to written texts, such as research articles and textbooks, where the heavier use of nominalization would seem to require greater use of the simple linear pattern. The constant Theme pattern, on the other hand, was found by Nwogu (1990) to predominate in specialized texts written for specialists in his corpus. He attributes this in part to the type of text, or schematic unit, involved: descriptions tend to organize information through constant Theme patterns. But he finds a fuller explanation for the higher use of the constant Theme pattern in specialized texts to lie in the relationship existing between the writer and the reader, and on the author's assumption of the reader's knowledge. The relationship here is equal ("specialist-specialist"), and the author can 190
assume a considerable amount of shared knowledge on the part of the reader. Therefore, in this case it is not necessary for the author to continually draw on information presented in the previous Rheme, as Nwogu found to be the case in the asymmetrical texts; instead, authors in a symmetrical relationship with their readers are more likely to draw on the stock of shared knowledge as the point of departure of the clauses. If this element remains constant over a series of clauses, the result is the constant Theme pattern, which is advantageous in that "the task of information processing is made relatively easier because it ensures that the reader can devote greater attention to the information being communicated in the rheme" (Nwogu, 1990: 252). Based on the above discussion, the following was hypothesized: 1. From the above, it would seem that the simple linear thematic pattern would want to predominate in the history textbook corpus, given the asymmetrical writer/reader relationship. 2. However, rhetorical patterns do play a role in the use of one pattern over another; as already mentioned, exposition and explanation tend to prefer the simple linear pattern, while description tends to prefer constant Theme. The latter also seems to hold in the case of narrative, as narratives "tend to relate a sequence of events ... involving a common character or set of characters" (Fries, 1983: 124), or have as a point of departure, in addition to characters, a setting of time or place. Given that history textbooks include a mixture of narration, description and analysis, then constant Theme patterns are predicted to occur as well, but in lower frequency than simple linear patterns, given the pedagogical nature of the texts, which would call for a good amount of explanation. 3. If we add to this the finding mentioned in the previous chapter with respect to the experiential content of the Themes, which showed a slight tendency in the English text towards causal explanation, which is associated with narrative, and a slight tendency in the Spanish texts towards providing a structural explanation for historical events, it is hypothesized that the ratio of simple linear over constant Theme would be smaller in English than in Spanish; in other words, there will be more constant Theme patterns in English than in Spanish. This is predicted to hold over the other patterns, split Theme and split Rheme as well, as split Theme is seen 191
to be a thematic development similar to constant Theme, while split Rheme can be classified as rhematic with simple linear. 4. In addition to providing for ease of reference through thematic progression, it is expected that authors will also repeat key Themes throughout their texts; Martin indicates that the method of development of a text, i.e. the information contained in its Themes (Fries, 1983), will pick up on "just a few [of all the experiential meanings available in a given field]. and weave them through Theme time and again to ground the text - to give interlocutors something to hang onto, something to come back to an orientation, a perspective, a point of view, a perch, a purchase" (Martin, 1992b: 489). 5. Summative expressions are also expected to appear in Theme position with a relative amount of frequency, given that these are very helpful in summarizing a stretch of text and linking it to a new stretch. 6. It is expected that rel and new Themes will be kept to a minimum. 5.5 Results of Thematic Progression 5.5.1 Results of Thematic Progression Patterns The first result of interest is that of the similarity/difference between thematic development (constant Theme and split Theme) and rhematic development (simple linear and split Rheme) across the two languages. Before giving those results, an explanation of the presentation of the results is warranted. The number of chains in each text was counted (chains consisted of two clauses or more) and compared to the total number of clauses, which gave the percentage of times a chain type appeared in a text. This necessarily means that if percentages are added together of thematic chains, rhematic chains, and Themes not counted as part of a chain, the result will add up to more than 100%. This is because many Themes were counted twice, as they often were linked to a previous Rheme and then subsequently began a thematic chain, or vice versa. Indicated in Table 5.1, then, are total Themes in thematic chains (TC), total Themes in rhematic chains (RC), Themes which were not counted, and Themes which were counted twice. As further explanation, by adding up the first two columns, the total number of Themes involved in chains is arrived at. By subtracting the figures in column 3 (Counted Twice) and then adding those outside the 192
chains, in column 4 (Not counted), the total is 100% (numbers have been rounded to the
nearest whole number)
Table 5.1: Overall Thematic Progression Results
Corpus Tot. Themes Tot. Themes Tot. in
in TC
in RC
Chains
Counted Not Twice Counted
English
43%
62%
105%
41%
36%
Spanish
33%
61%
94%
37%
43%
Key: Tot. = total; TC = thematic chains; RC = rhematic chains
As predicted above, while the number of Themes in rhematic progression chains is virtually
identical across the two languages, the ratio of Themes in continuous Theme progression
patterns is lower in Spanish than it is in English. The first finding may be attributed to a
fairly equal amount of exposition and explanation, or analysis, in the history textbooks,
while the second may be attributed to the higher amount of narrative in the English corpora,
which was perceived in the previous chapter. However, this needs to be analyzed in more
depth, as there are also 7% more Themes not counted in chains in Spanish than in English.
Before attributing the greater amount of thematic progression to a greater amount of
narrative, other explanations must be considered. As explained in section 5.3.2 above on
peripheral Themes, there are a number of Themes which do not chain to previous Themes or
Rhemes because of their grammatical nature. It is worth, then, examining the categories into
which the Themes which fall outside of the progression patterns in terms of their origin are
divided. This can be seen in Table 5.2 below; the peripheral Themes are ordered from most
to least frequent based on the results from the Spanish corpus:
193
Table 5.2: Theme Types outside Progression Patterns
Theme Type
English Spanish
back
7.41% 7.80%
key Theme
6.37% 6.12%
new
4.87% 5.87%
sum
4.51% 4.58%
time
4.11% 4.06%
inversion
0.00% 3.07%
existential
0.80% 2.43%
se
0.00% 2.41%
extraposition
0.85% 1.99%
rel
1.24% 1.26%
impersonal
0.00% 0.74%
we
2.91% 0.64%
manner
0.43% 0.64%
previous clause
0.25% 0.52%
place
0.52% 0.28%
fronted
0.47% 0.26%
metatextual
0.61% 0.20%
wh
0.45% 0.19%
recuйrdense
0.00% 0.10%
passive
0.00% 0.07%
thematic equative
0.06% 0.06%
ellipsis
0.00% 0.06%
imagine
0.05% 0.00%
The results are strikingly similar across the two languages in terms of key Theme, back references, summative expressions, time adverbials, new information and related expressions. The major difference between Spanish and English here seems to lie in the grammatical Themes. If we add up impersonal expressions, inverted Subject/verb constructions, se, and existential constructions, the total is 8.65%, which goes a long way in explaining the difference between the two languages in terms of Themes falling outside the patterns. This grammatical difference would be one way of explaining away the discrepancy between English and Spanish in terms of number of constant Theme chains. There are fewer of these in Spanish, but it may be because they are broken up by the grammatical nature of the language, especially with regards to Subject/verb inversion (as explained before in section 5.3, ellipsis is analyzed as forming part of simple linear chains). However, this is canceled out by gapped chains: it is to be remembered that in the analysis, links were counted with gaps of up to three clauses. This is illustrated in the following examples: 5-29. Spa 3 (151) Cuando Alfonso XI comenzу a gobernar (1325) el reino estaba ... (152) y parecнa imposible poder reducir a la nobleza. 194
(153) Alfonso tuvo que utilizar para ello todos los medios...
5-30. Eng 7 (23)Minoan civilization lasted about 1,350 years (2600-1250 B.C.) (24)and reached its height during the period from 1700 to 1450 B.C. (25)The centers of Minoan civilization were magnificent palace complexes, ...
Example 5-29 shows a constant Theme chain broken by extraposition in the second clause.
Clauses 151 and 153 in that example form a constant Theme chain, and this is reflected in
the counts. Example 5-30 shows a constant Theme chain broken by ellipsis; clauses 23 and
25 in that example form a constant Theme chain, while, at the same time, clauses 24 and 25
form a simple linear chain through ellipsis. Thus, the grammatical nature of Spanish would
not necessarily mean a reduction in the count of constant Theme chains. Simply put, in
English there are a greater percentage of clauses involved in constant Theme progression
patterns. Perhaps this can be explained by the expectation that this would occur because of
a greater amount of narrative included in the English texts, as was evidenced by the results
of the transitivity analysis, and of description. Further evidence can be provided for this if
we look at the distribution of the length of chains in the constant Theme progression
patterns; the percentages of distribution of the different chain lengths can be seen in Table
5.3 ("2" refers to two clauses involved in the chain; i.e. two clauses begin with a Theme
having the same referent, gapped or contiguous; "3" means three clauses, and so on.)
Table 5.3: Distribution of Constant Theme Progression Patterns (according to length)
English
Spanish
Chain length No. of clauses % of clauses No. of clauses % of clauses
involved
involved
involved
involved
CT 2
316
21.12%
252
19.89%
CT 3
141
9.43%
93
7.34%
CT 4
72
4.81%
52
4.10%
CT 5
45
3.01%
5
0.39%
CT 6
18
1.20%
6
0.47%
CT 7
21
1.40%
0
0.00%
CT 8
8
0.53%
0
0.00%
Table 5.3 shows that the English corpus consistently shows a higher percentage of clauses which are involved in constant Theme progression chains than does the Spanish corpus; what is of special interest is the difference in the longer chains, those involving 5 clauses or more. In the Spanish corpus, there are only 11 clauses involved in longer chains; there is 1 instance of a constant Theme chain of 5 clauses, and 1 of chain of 6 clauses. The English corpus has 9 instances of constant Theme chains of 5 clauses, 3 of 6 clauses, 3 of 7, and 1 of
195
8 for a total of 92 clauses involved in long constant Theme chains. It is worth looking at the
distribution of these Theme chains in the English corpus to analyze further the notion of
greater use of narrative as an explanation for a higher count of constant Theme progression
patterns.
Table 5.4: Distribution in English Corpus of Long Constant Theme Progression Patterns
Chain Eng 1 Eng 2 Eng 3 Eng 4 Eng 5 Eng 6 Eng 7 length
Eng 8 Eng 9 Eng 10
CT 5
3.70% 3.76% 0.00% 0.00% 3.03% 0.00% 8.29% 8.82% 0.00% 0.00%
CT 6
0.00% 0.00% 4.44% 0.00% 0.00% 0.00% 6.63% 0.00% 0.00% 0.00%
CT 7
0.00% 0.00% 0.00% 0.00% 4.24% 0.00% 3.87% 0.00% 0.00% 3.50%
CT 8
0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 4.42% 0.00% 0.00% 0.00%
Total 3.70% 3.76% 4.44% 0.00% 7.27% 0.00% 23.20% 8.82% 0.00% 3.50%
It is interesting to note in Table 5.4 that there are texts in the English corpus which have no long Theme chains, as happens in the Spanish corpus. Thus, any comment that is made regarding the texts as a whole corpus keeps this clearly in mind. Eng 7 is by no means the text with the most clauses involved in constant Theme patterns, if we include those chains of 2, 3 and 4 clauses. However, it jumps way to the forefront with a very high count of 23.20% of clauses involved in long constant Theme chains. The topic of Eng 7 is Ancient Greece, and we can see that some of these constant Theme progression chains are quite "narrative" in character; that is, they narrate the deeds of an individual, and that individual is a central character in those passages of the text: 5-31. Eng 7 (159) In 594 B.C., Solon [c. 640-559 B.C.), a traveler and poet with a reputation for wisdom, was elected chief executive. (160) He maintained that the wealthy landowners, through their greed, had disrupted ... (161) Solon initiated a rational approach to the problems of society ... (162) he sought practical remedies for these ills; (163) and he held that written law should be in harmony with Dikй, ... (164) At the same time, he wanted to instill in Athenians of all classes a sense of ... (165) Solon aimed at restoring a sick Athenian society to health by restraining ... 5-32. Eng 7 (169) but he refused to confiscate and redistribute the nobles' land ... (170) He permitted all classes of free men, even the poorest, to sit in the Assembly, ... (171) He also opened the highest offices in the state to wealthy commoners, ... (172) Thus, Solon undermined the traditional rights of the hereditary aristocracy (174) Solon also instituted ingenious economic reforms. We can see in these two examples how the author relates historical events through the use of narrative: a central historical character, Solon, is used as the focal point of these passages, and his deeds form the Rhemes. However, not all of the passages in which long
196
constant Theme chains are used are like these examples. The longest one, the constant
Theme chain involving 8 clauses, centers around Homer, but the Themes are about Homer's
works, rather than about him as an individual character. Thus, this passage is a description
of the works of Homer, as can be seen in example 5-33:
5-33. Eng 7 (76) In Homer, we also see the origin of the Greek ideal of aretй, excellence. (77) The Homeric warrior expresses a passionate desire to assert himself,... (78) In the warrior-aristocrat world of Homer, excellence was principally interpreted ... (79) Homer's portrayal also bears the embryo of a larger conception of human excellence, (82) Thus, in Homer we find the beginnings of Greek humanism - a concern with man ... (83) Essentially, Homer's works are an expression of the poetic imagination ... (84) But his view of the external order of nature and his conception of the individual form the foundations of the Greek outlook. (85) Although Homer did not intend his poetry to have any theological significance, his ..
This use of a long constant Theme chain for purposes of description can be seen in the same
text in the following example:
5-34. Eng 7 (92) The city-state based on tribal allegiances was generally the first political association during the early stages of civilization. (94) The scale of the city-state, or polis, was small; (95) most city-states had fewer than 5,000 male citizens. (96) Athens, which was a large city-state, had some 35,000 to 40,000 adult male citizens... (98) The polis gave individuals a sense of belonging, (99) for its citizens were intimately involved in the political and cultural life of ... (101) The mature polis was a self-governing community that expressed the will ...
And the following chain from Eng 7 shows the constant Theme progression being used for a
combination of narration and description:
5-35.
(127) The Spartans learned only one craft, soldiering, (129) They were trained in the arts of war (131) Military training for Spartan boys began at age seven; (132) they exercised, drilled, competed, and endured physical hardships. (134) Spartan soldiers were better trained and disciplined (136) But the Spartans were also criticized for having a limited conception of aretй.
In this example, the Spartans provide the constant Themes of the clauses; sometimes their
deeds are narrated, and at other times, the author describes either the Spartan soldiers or
their military training. These examples from Eng 7 seem to confirm the notion that constant
Theme progression chains do seem to indicate the presence of narration and description.
Other texts in the English corpus with higher percentages of long constant Theme chains
197
also bear this out, as can be seen in the following examples taken from Eng 8, a text on Ancient Rome: 5-36. Eng 8 (104) Augustus slept in the same bedroom in this house for forty years, ... (105) and when he wanted to work on secret state papers or to be alone, he retreated ... (106) His way of life was as simple as his house. (107) His clothes were usually woven and sewn by the womenfolk of his house; (108) he drank little 5-37. Eng 8 (163) Nero [reigned 54-68) was dissatisfied with everything he inherited, his palaces... (164) He murdered his wife and mother and Claudius' son, (166) He cast away the last remnants of Roman dignity (168) and he compelled senators to do the same. (169) He threw off the guidance of his tutor, the philosopher Seneca, .. There is also an example in Eng 10, a text with only one long Theme chain: 5-38.Eng 10 (40) Under the first two Hanoverian kings the cabinet was only starting to accumulate this immense authority. (41) George I and George II by no means abdicated all the royal prerogatives. (42) They took a direct interest in the South Sea Bubble and other financial matters (43) and they intervened in the conduct of war and diplomacy to a degree ... (44) George II was the last English monarch to command troops in person ... (45) The two Georges chose their cabinet ministers from the Whig party, (46) They did so, however, not because they were required to, ... In the Spanish corpus, the same phenomenon can be observed; long Theme chains are used for descriptive passages. Spa 10, on the eighteenth century, contains a 5 clause constant Theme chain which describes the Illustration: its origins, ideals, followers' activities, major characteristics, and the theories elaborated by its philosophers: 5-39.Spa 10 (2) La Ilustraciуn constituyу el fermento ideolуgico de las revoluciones polнticas (3 Su ideario puede resumirse en estos principios: libertad, igualdad y economнa libre. (4) Sus seguidores proclamaron y difundieron en sus escritos la libertad de pensamiento ... (5) El rasgo fundamental de aquel movimiento fue el espнritu crнtico, aplicado no sуlo ... (6) Los filуsofos de la Ilustraciуn elaboraron teorнas que afectaron a todos los planos... Example 5-40 below is a passage which centers around the new type of gold work which appeared in Spain some time during the Bronze Age, taken from Spa 9, a text on prehistoric Spain. The new gold work is mentioned in the Rheme of the clause immediately previous to this passage, in clause 92, and this passage describes when it was introduced, how its prototypes are similar to an earlier era, what its decorative techniques consisted of, how the 198
clasps used were similar to types found in other areas, its area of dispersion and how its characteristics showed the role played by locally produced gold: 5-40.Spa 9 (93)Su introducciуn puede remontarse a la fase anterior, (94) pues sus prototipos parecen ser ejemplares bretones de fines del Bronce Medio (95) pero las tйcnicas decorativas, probablemente, son locales (96) y sus cierres machi-hembrados ofrecen los mejores paralelos en el occidente ... (97) Su dispersiуn se centra en la zona entre el Sistema Central y el Guadiana, ... (98) y sus caracterнsticas evidencian el papel que el oro de producciуn local ... However, it will be remembered from Chapter 4 that Spa 9 is a highly specialized text, and develops concepts about the prehistoric past through quite a bit of description of tools used during past ages. It was noted then that the high use of relational clauses pointed to a large amount of description. Yet, Spa 9 is low in constant Theme progression patterns: while it is the only text in the Spanish corpus with a constant Theme chain of at least 6 clauses, only 23.30% of its clauses are involved in constant Theme patterns, while 73.79% of its clauses are involved in simple linear progression pattern. That is an incredibly high percentage, much higher than any other text in the whole corpus, both English and Spanish. Therefore, instances of constant Theme progression patterns may not be an indication of a great deal of description, nor may their absence be an indication of a small amount. Before looking at the role played by simple linear progression patterns in description, as well as in explanation and analysis, which were two expository patterns in which simple linear patterns were posited to be prominent in section 5.4 above, we look at one final example here of a long constant Theme chain, taken from Eng 5, a text on pre-history: 5-41. Eng5 (54)Homo sapiens sapiens, the sole surviving form of the human species, made its first known appearance at about the same time as H.s. neanderthalensis was dying out, ... (55) It seems to have arrived in these areas as an immigrant, perhaps from Africa... (56)Its evolutionary origins and precise relationship to neanderthalensis remain a mystery, though it is generally agreed that they were both subspecies ... (57)The advantage of the new subspecies over the older one seems to have lain... (58)H.s. sapiens inherited the cultural accomplishments of preceding species. (59)This type knew how to use fire, a skill that had appeared many centuries earlier. (60)It also became increasingly sophisticated in making and using tools; Here, homo sapiens sapiens is the central character. As it is not an individual, the narration of its "deeds" seems more of a description. And here lies one of the problems in connecting the Theme chains to one or another type of expository pattern: categories of exposition overlap (Henner-Stanchina, 1985). Is the above example mainly a description of homo sapiens sapiens, or is it an explanation of why it prevailed over other human species? It seems that through the description, the latter explanation is arrived at. Van Lier (1994: 330) 199
points out that, in research, "a distinction between description and explanation is never clear cut: whenever we describe, we explain, and whenever we explain, we describe. This is because description implies interpretation, and interpretation is part of explanation". In Spa 9, the text mentioned above in example 5-40 above, which deals with prehistoric Spain mainly through the tools used during that time, relies on, as also mentioned, a lot of description. It has a very high number of relational clauses, which are related to description. Yet it is the text with the greatest amount of simple linear patterns in the whole corpus, and a low frequency of constant Theme clauses. Below is an example of one of the longer simple linear chains (6 clauses involved): 5-42.Spa 9 (46) La valoraciуn de las peculiares caracterнsticas tipolуgicas y tecnolуgicas de las armas del Bronce Final de todo el occidente de Europa respecto a las de la cultura centroeuropea de los C.U., ha permitido, a partir de los aсos 1940, identificar paulatinamente un "Cнrculo Atlбntico", que constituye un complejo tecnolуgico y de elementos de cultura material de gran personalidad, especialmente bronces y orfebrerнa, que se extiende por las regiones marнtimas de todo el occidente de Europa. (47) Estos elementos no constituyen una cultura uniforme, (48) pero demuestran la existencia de elementos culturales comunes, bбsicamente relacionados con actividades minero-metalъrgicas, que se explican por la identidad y complementariedad en recursos minerales y tйcnicas metalъrgicas de las tierras baсadas por el Atlбntico, en un momento en que los metales constituyen el elemento cultural mбs esencial para el avance tecnolуgico de la sociedad, lo que repercute directamente en las estructuras econуmicas y sociales. (49) El oro de Irlanda y del occidente de la Penнnsula Ibйrica, el estaсo de esta ъltima, de Bretaсa y de Cornualles, el cobre de Irlanda y de La Penнnsula Ibйrica, explican la apariciуn de crecientes contactos comerciales que desarrollan paralelamente intercambios tecnolуgicos y de ideas, facilitados ademбs por el carбcter ribereсo de todas estas regiones. (50) Esto indica una doble relaciуn de causa a efecto en el desarrollo de la navegaciуn, como elemento de comunicaciуn y en la difusiуn de las ideas. (51) En este aspecto destacan las relativas a los campos tecnolуgicos, los mбs relacionados en dichos contactos culturales de mayor transcendencia, como los socio-econуmicos e incluso los ideolуgicos. The author begins with a material clause in 46, the only non-relational clause in this chain, by thematizing the evaluation of the typological and technological characteristics of the weapons of western Europe with respect to those of Central Europe, and in the Rheme, develops the notion that this evaluation has allowed an "Atlantic Circle" to be identified, which constitutes a technological complex of elements of `cultural material of great personality". Indeed, this taking the characteristics of tools and weapons and using them to establish trends, areas, and ages is a main theme of this whole text. Theme 47 picks up on the elements mentioned in the Rheme of 46, positing that they are not culturally uniform. 200
Theme 48 is chained to Rheme 47 through ellipsis, and, also in a relational clause, states that they do show the existence of some common cultural elements, mainly related to mining, which can be explained because of the complementary nature of mining resources and technology in the areas bathed by the Atlantic at a time when metals are the most essential cultural element for the technological advancement of society, which has a direct effect on the economic and social structures. These common mining elements are picked up on in Theme 49, through the gold of Ireland and of the western part of the Iberian peninsula, the tin of the latter with that of Brittany, the copper of Ireland and that of the Iberian peninsula, which explains the appearance of growing commercial contact which develops parallel to the exchanges of technological ideas, made easier by the seaside nature of these regions. This whole Rheme is picked up on by the demonstrative pronoun "esto", in the Theme of 50, and then the Rheme develops the notion that this (the appearance of the growing commercial contact) shows a double relationship between cause/effect in the development of shipping, as both an element of communication and of diffusion of ideas. Then this last bit of the Rheme is picked up on in Theme 51: within the diffusion of ideas, Rheme 51 develops those related to technological fields. It is clear, then, that this chain of ideas develops a more dynamic type process, where one piece of evidence can be explained through some wider phenomenon, which leads to further explanation, and so on. This type of chained explanation can also be seen in the English texts (underlining is used to illustrate the links in the chain): 5-43. Eng 10 (29) for they remembered the uses that Cromwell and James II had made of this weapon. (30)The Glorious Revolution, which in its preliminary stage had done so much to confirm distrust of the army, had also confirmed Britain's unique and greatest asset - the supremacy of Parliament. (31)Parliament had approved the accession of William and Mary in place of James II. (32)When Anne, Mary's sister and the last Stuart monarch, died in 1714, Parliament had already arranged for the succession of the House of Hanover. (33)Under the first kings of the new house - George I (1714-1727) and George II (1727-1769) - the institution that would ultimately ensure the everyday assertion of parliamentary supremacy was beginning to mature. (34)This was the cabinet. Through a chain of events, the author leads the reader to the establishment of the cabinet, in Rheme 34. The interesting aspect of this chain is that, unlike the previous one in example 5- 42, it is not developed through description, but through narration of events. There are characters that carry out deeds, and here there is only one relational clause: 34. 201
Spa 3 is another interesting text in this respect. It is the text in the Spanish corpus with the highest frequency of Actors as Theme, and I had expected a greater frequency of constant Theme chains, as it seems "narrative" in nature, due to the number of Actors: specific individuals, mainly kings and nobles and many of their deeds and events are narrated throughout the chapter; however, it has 20% more simple linear chains (54.55%) as compared to constant Theme chains (34.55%). To illustrate, example 5-44 shows a long (6 clause) simple linear chain from Spa 3: 5-44. Spa 3 (90) Sancho IV (1284-1295), apellidado "el Bravo", que equivalнa a colйrico e irascible, se proclamу rey, a pesar del testamento de su padre, mientras el infante de la Cerdo lo hacнa ayudado por el rey de Aragуn, en cuyos estados estaba refugiado. (91) Como tambiйn Francia, cuyo rey era pariente del infante, lo apoyaba, tratу Don Sancho de conseguir la amistad de uno de los dos estados frente al otro, (92) pero este arreglo diplomбtico resultу difнcil, porque los principales personajes de la corte estaban divididos respecto a la alianza que habнa de escogerse. (93) La reina, Doсa Marнa de Molina, se inclinaba a Francia, mientras que el infante Don Juan y el poderoso seсor de Vizcaya, Don Lope de Haro, preferнan la amistad de Aragуn. (94) Don Lope, que gozaba respecto al rey de la privanza mбs absoluta, se enemistу con йste cuando Don Sancho decidiу por fin pactar con Francia, (95) y tras una ruidosa entrevista que tuvo lugar en la villa de Alfaro, el vizcaino fue muerto por el propio rey Don Sancho [Tragedia de Alfaro, 1288). In this example, we can see quite a few "characters", Sancho IV, the infante of Cerdo, Doсa Maria de Molina, Don Lope de Haro, about whom events are narrated. The narration moves in a chain from Sancho IV, in Theme 90, who proclaimed himself king, to the infante of Cerdo, in Rheme 90 through a subordinate clause, who also proclaimed himself king, with the help of the king of Aragon. Theme 91 picks up on the infante, also through a subordinate clause, one which indicates cause: that France also supported the claims of the infante, given that the French king was a relative of the infante, thus setting up the reason for which Don Sancho tried to gain the friendship of one of the kingdoms, France or Aragon, which forms the Rheme of 91. This attempt at a diplomatic arrangement then becomes the Theme of 92, and the Rheme of that clause explains why it was difficult, due to the division amongst the principal members of the court. Then the Theme of 93 thematizes one of the members of the court, Doсa Marнa de Molina, who was more inclined towards France, and brings into the Rheme through a subordinate clause another member, Don Lope de Haro, who preferred Aragуn. He then becomes the Theme of 94, and his enmity with Don Sancho is brought up in the Rheme, and is given a time framework: the enmity began when Don Sancho signed a treaty with France. The time reference is picked up in the Theme 202
of 95: after a clamorous interview that took place in Alfaro, the Rheme of 95 displays the death of Don Lope by the hand Don Sancho himself. While examples 5-43 and 5-44 show narration of events, this narration is different from that of examples 5-32, 5-33, 5-36, and 5-37, which were developed through constant Theme chains. The constant Theme chains simply list a set of events, usually in chronological order, while the simple linear chains establish relationships, usually implying cause-effect: one link in the chain leads to another. Thus, while saying that constant Theme chains imply more narration of events is to state the matter too simply, perhaps particularly in the field of history, it does seem to be the case that constant Theme chains imply a simple telling of events, without establishing relationships between events beyond that of time. On the other hand, the use of narration in simple linear chains implies more of an analysis of events. The same can be said of description. In a constant Theme chain, description centers around a single event, item, or person, and supplies the reader with a fuller picture of that event, item or person; it is more static in nature. In a simple linear chain, description moves the reader from, for example, an item or a person to a fuller event or events, thus providing connections between phenomena, and is more dynamic. Thus, as Lotfipour-Saedi and RezaiTajani (1996: 243) point out, "there is no one-to-one correspondence between a specific mode of thematization and a single genre". This, then, can be extended to rhetorical modes, such as narrative. In sum, perhaps it is not surprising that in both corpora there is a greater amount of simple linear chaining than there is of constant Theme. This indicates the analytical, or explicative, nature of the texts, which was expected due to the pedagogical purpose of the texts and the asymmetrical relationship between the writer and the reader. Furthermore, while the two corpora are equal in the amount of rhematic chaining, the American history textbook corpus shows a greater frequency of constant thematic chaining. This may provide more support for the tentative conclusion reached in chapter 4 following the transitivity analysis, that there may be more of a tendency for the American texts to show more of a narrative bent to writing up history, while the Spanish texts show more of a tendency towards structural explanations, of explaining events through webs of relationships. As stated before, this tendency is by no means specific to each of the corpora, as again here there are as many differences between texts of the same corpus as there are across corpora. 203
5.5.2 Results of Problematic Peripheral Themes
The above discussion mainly relates to the expectations as regards the type of
thematic progression patterns, and whether they proceed from previous Themes or Rhemes.
Another expectation posited earlier was that writers would indeed provide connections for
their readers either through chains, or through the various connecting type Themes which
fall outside the patterns: key Themes, sum Themes, previous clause Themes, and "available"
Themes such as time or place references. We have seen in Table 5.2 above that they do
make use of these Theme types: just under 16% of the Themes in the English corpus fall
into one of these categories, and just over 15% in the Spanish corpus. The discussion here
now moves to the other type of Themes: those which mean work for the reader in terms of
connecting them to the discourse and to the conceptual framework which they build up as
they read the text. The results of the Themes which do not form part of the thematic
progression patterns were presented in Table 5.2, above. For purposes of the discussion
here, the results of those categories which call for a greater effort on the part of the reader,
back, rel, new, inversion, are repeated in Table 5.5 and Table 5.6, along with the breakdown
of how these categories are represented in the individual texts:
Table 5.5: Themes outside Progression Patterns in English:
Theme type
Eng 1 Eng 2 Eng 3 Eng 4 Eng 5 Eng 6 Eng 7 Eng 8 Eng 9 Eng 10 Total
back inversion new rel
2.22% 0.00% 2.22% 1.48%
8.27% 0.00% 2.26% 1.50%
9.56% 0.00% 7.35% 4.41%
2.94% 0.00% 5.88% 0.00%
6.63% 11.00% 0.00% 0.00% 5.42% 5.26% 0.60% 3.35%
7.22% 0.00% 4.44% 0.00%
6.43% 0.00% 9.94% 0.58%
5.93% 0.00% 2.96% 0.00%
13.93% 0.00% 2.99% 0.50%
7.41% 0.00% 4.87% 1.24%
Total
5.93% 12.03% 21.32% 8.82% 12.65% 19.62% 11.67% 16.95% 8.89% 17.41% 13.53%
Table 5.6: Themes outside Progression Patterns in Spanish:
Theme type
Spa 1 Spa 2 Spa 3 Spa 4 Spa 5 Spa 6 Spa 7 Spa 8 Spa 9 Spa 10 Total
back inversion new rel
3.76% 0.75% 5.26% 1.50%
6.08% 0.00% 8.11% 0.00%
8.54% 3.66% 8.54% 0.00%
7.89% 1.97% 8.55% 1.97%
3.25% 0.00% 9.76% 0.00%
9.79% 12.05% 7.14% 0.70% 0.60% 14.29% 1.40% 6.02% 7.14% 4.90% 2.41% 0.00%
7.84% 11.65% 3.92% 4.85% 1.96% 1.94% 0.98% 0.97%
7.80% 3.07% 5.87% 1.27%
Total
11.27% 14.19% 20.74% 20.38% 13.01% 16.79% 21.08% 28.57% 14.70% 19.41% 18.01%
The two corpora are different in totals, but this can be explained mainly by Subject/verb inversion in Spanish. While inversion does break up thematic progression patterns in that it presents the verb first, usually with no attachment to a previous verb, as occurs with prodrop, does it cause assimilation problems for the reader? Many of the instances of inversion pose no problem, as they are similar to cases of extraposition or existential verbs, which also 204
often present cases of inversion, as was mentioned in Chapter 4, but which are not included in the count of "problematic" Themes. This is because usually the existential verb presents a new Subject, so inversion in these cases actually contributes to easier information processing on the part of the reader. A few of the cases of inversion which do not involve existential verbs also allow newer information to be presented in the Rheme, e.g.(new information is underlined): 5-45. Spa 10 (47) Es muy significativa a este respecto una frase del Emilio: [thereafter follows a quote]. 5-46. Spa 10 (68) A este respecto jugу un destacado papel la Masonerнa, aparecida en Gran Bretaсa,... In these two cases, the new information, which in both cases is encoded in the grammatical subject of the verb, is allowed to appear late in the Rheme because of inversion. From the data in the corpus, this functional explanation of inversion would also seem to fit with de Miguel's (1988; see also section 4.6.4 above) explanation of inversion as taking place with what she calls `ergative verbs', or anticausatives as well as those which semantically express a process leading to a change of state rather than an action. In many cases, where these latter type of verbs occur, the Subject of the verb, as well as expressing a changed state, is also new information: 5-47. Spa 5 (113) Tambiйn se introdujeron reformas administrativas, ... This section of the text discusses a series of reforms in property and land use, and then Clause 113 moves to administrative reforms, which is brand new in the discussion. Therefore, inversion here serves the function of allowing this new information to be placed in the Rheme. There is one instance of inversion, however, which was mentioned in Chapter 4, and which may be of a type which causes some problem in comprehension: 5-48. Spa 8 (15) Pero no debe aplastarnos el peso de estos fondos ingentes. Here, the verb is more informationally weighted, as the reader might ask why these "fondos ingentes", referring to the vast amount of written historical documents, must not crush us. The reader must wait until Clauses 16 and 17 to find out that the analysis of these documents will help in this matter. So, in this case, the reader "must hold the new information in abeyance while he waits for the given information and searches for its antecedent. This increases the load on memory and makes comprehension less than optimal" 205
(Clark & Haviland, 1977: 13). At any rate, this is one of the very few instances where inversion is not optimal in this sense. Therefore, if we set aside inversion, the greater number of total Themes in the Spanish corpus (given in Table 5.6 above) which may cause comprehension problems is not markedly different from the English corpus. If we revise Table 5.5 and Table 5.6 above by deleting the category "inversion", the results are:
Table 5.7: Themes outside Progression Patterns in English (revised):
Theme Eng 1 Eng 2 Eng 3 Eng 4 Eng 5 Eng 6 Eng 7 Eng 8 Type
Eng 9 Eng 10 Total
back new
2.22% 8.27% 9.56% 2.94% 6.63% 11.00% 7.22% 2.22% 2.26% 7.35% 5.88% 5.42% 5.26% 4.44%
6.43% 5.93% 13.93% 7.41% 9.94% 2.96% 2.99% 4.87%
rel
1.48% 1.50% 4.41% 0.00% 0.60% 3.35% 0.00% 0.58% 0.00% 0.50% 1.18%
Total
5.93% 12.03% 21.32% 8.82% 12.65% 19.62% 11.67% 16.95% 8.89% 17.41% 13.47%
Table 5.8: Themes outside Progression Patterns in Spanish (revised):
Theme Spa 1 Spa 2 Spa 3 Spa 4 Spa 5 Type
Spa 6 Spa 7 Spa 8
Spa 9
Spa 10 Total
back new rel
3.76% 6.08% 8.54% 7.89% 5.26% 8.11% 8.54% 8.55% 1.50% 0.00% 0.00% 1.97%
3.25% 9.79% 12.05% 9.76% 1.40% 6.02% 0.00% 4.90% 2.41%
7.14% 7.14% 0.00%
7.84% 1.96% 0.98%
11.65% 1.94% 0.97%
7.80% 5.87% 1.27%
Total 10.53% 14.86% 15.85% 18.42% 13.01% 17.49% 20.48% 14.29% 10.78% 14.56% 15.03%
The Spanish corpus shows slightly higher counts in back references, related references, and new entities. The main point here is the fact that there are these percentages of Themes in both corpora, which are connected with difficulty to the flow of discourse. It is difficult to say what an acceptable percentage might be, if any, given that there are no other studies of which I am aware that analyze this in textbooks. However, given the difficulty students often have in grappling with reading and comprehending history textbooks, it would probably be more optimal if textbook writers would attempt to relate clause Themes more through thematic patterns, or by picking up on a key Theme. However, writers can provide specific links for their readers through other means. We have seen in Table 5.2 above that there are low percentages of metatextual Themes (e.g. "As we have seen in Chapter 3...") in both corpora. However, this does not take into account another explicit way writers have of signalling relationships to readers: textual Themes. The next section is devoted to these.
206
5.6 Textual Themes 5.6.1 Textual Theme Types As explained above, thematic progression based on linking ideas through synonymy, substitution, ellipsis, and so on, help the reader to follow the flow of the text. In addition to the ideational Themes of the clauses, which were analyzed above in order to establish the thematic progression patterns in the two corpora, there is another Theme type which is considered in this chapter on textuality: textual Themes (cf. discussion on Multiple Theme in Chapter 3, sections 3.5.1.2 and 3.5.2.2). Textual Themes are links which specify the relationship of the clause to the surrounding text and context. Textual Themes, according to Halliday (1994) include continuatives, structural Themes, and conjunctive Adjuncts. Continuatives are a small set of discourse items which "signal that a new move is beginning", such as yes, no, well, oh, now. Continuatives are not considered separately in this study, as Halliday seems to posit them as signalers in spoken discourse, since he describes them as a "response, in dialogue, or a move to the next point if the same speaker is continuing" (Halliday, 1994: 53). Indeed, none of these markers occur in the English corpus in the sense Halliday ascribes to them as continuatives; the only instances of now are temporal and external to the text (they refer to nowadays or at the present time), and these are considered ideational Themes. In Spanish, continuatives can be expressed by the words sн, no, ah, ahora bien and pues. There are 4 instances of ahora bien in the Spanish corpus, and these are considered under conjunctive Adjuncts, simply because they are too small a set to count separately. There are two instances of sн in the Spanish corpus. However, they are not included as responses, but in order to underscore the polarity of a statement, as in: 5-49. Spa 4 (137)pero sн encaja perfectamente aquн lo que Tucнdides nos dice de йl: Given its role in this statement, it is considered in Chapter 6, with Modality. Structural Themes are coordinators (and, or, nor, neither, but, yet, so, then) and subordinators. The former are considered here in this chapter in their role in providing texture, while the latter are not, for the simple reason that subordinated clauses, as explained in Chapter 3, are not considered separately from the independent clause they accompany for their thematic configuration. The final type of textual Theme is that of conjunctive Adjuncts, which relate the clause to the preceding text in some way, e.g. through apposition (in other words), or 207
addition (moreover). In this study, both coordinators and conjunctive Adjuncts are considered; the latter are analyzed in both the Themes and the Rhemes of the clauses. This is because they can appear after the ideational Theme, while coordinators cannot. 5.6.2 Expectations on Textual Themes As seen above, thematic progression chains show that ideational Themes of clauses often share propositional content with preceding Themes or Rhemes, or both. Mauranen (1993c: 105), in a comparative study of Theme in English and Finnish journal articles, hypothesizes that when this does not occur, when there is no link in propositional content then "an orienting theme constituent must be present to mark the intended connection". In orienting Themes, Mauranen includes connectors, adverbials, Complements, and modal and reporting clauses. This section of this chapter focuses on connectors. Given Mauranen's hypothesis, then, it was expected that textual Themes would have a tendency to appear in the history textbook corpus in particular with those clauses whose Themes are problematic for text comprehension, such as example 5-48 above, repeated here for ease of reference: 5-50.Spa 8 (15) Pero no debe aplastarnos el peso de estos fondos ingentes. Indeed, in this example, the reader is informed of the contrastive relationship of Clause 15 to the preceding discourse. Before looking at other studies, whose results can help in predicting the use of textual Themes in this corpus, there is an aspect of conjunction which needs to be discussed: that of internal/external relationships. Martin describes internal relations as "more "rhetorical" than experiential" (Martin, 1992b: 178), as they "obtain in the organisation of the text itself rather than in the organisation of the world the text describes" (ibid: 180). Martin states that the underlying opposition between the two "is that of text vs "reality"" (ibid: 180), or, stated another way, internal relations "structure semiosis; external ones code the structure of the world" (ibid: 180). He further provides a distinction through the notions of field and genre: external relations are mainly oriented to field, internal ones to genre, that is "they encode the organisation of the text as it is formulated to construct our culture" (ibid: 180). In their studies of Theme across genres, both Whittaker (1995) and Francis (1990) discuss internal relationships encoded in textual Themes. For Francis, internal relations include examples such as "in this context" and "therefore" (Francis, 1990: 62); Whittaker contrasts internal textual Themes with external ones; internal Themes give information about "the organization or function of part of the text", and include "e.g." and 208
"but", while external logical connectors express logical relationships which hold in the world outside the text, and include "thus" and "then" (Whittaker, 1995: 113). Martin (1991) describes "thus" as an internal conjunctive structure, which clashes with Whittaker's gloss on "thus" as external. Halliday (1994: 338) points out that it is often hard to tell the two apart, especially in cases of clause relations of elaboration and some types of extension. Perhaps the problem is that it is difficult to pinpoint what is meant by logical relationships outside the text. This can be illustrated by using the following example: 5-51. Eng 9 (72) Russia's five million Jews were already confined to the market towns and small cities of the so-called Pale of Settlement, where they worked as artisans and petty traders. (73)Therefore, when Russian Jewish artisans began in the 1880s to escape both factory competition and oppression by migrating - a migration that eventually totalled two million people -, this was basically a once-and-for-all departure. There are two events here: that Russia's five million Jews were confined to market towns and that once they migrated, they did not return. The author establishes a cause/effect relationship between the two, and perhaps it is tempting to think of that relationship as external to the text, as a fact of the real world. However, given that there could have been other reasons for which the Jews did not return, the relationship is in effect more internal to the text itself, or, perhaps it would be possible to say that the relationship exists externally but in the author's perception of the world (and in that of other people who agree with the author's interpretation). This would accord more with Martin's point above with regard to the text's role in the construction of culture; in the previous chapter to this one, it was pointed out that the role of the historian is to establish links between events. A relevant quote from that discussion to the discussion of internal/external relations is Tilly's (1985: 13): "European social history' s central activity... concerns reconstructing ordinary people' s experience of large structural changes". That reconstruction is construed by the text, and thus the relationships are internal. Vasconcellos (1992) distinguishes between the cognitive and noncognitive use of Adjuncts such as at the same time. The cognitive refers to a specific time, and she gives the example: The Jesuits were the first priests to come to stay and by 1561 they were already building their first colйgio in the city of Salvador. They were followed by the Benedictines (1581), the Carmelites (1586), and the Franciscans (1587). At the same time, a sizable secular clergy ... was becoming established. (Example taken from Vasconcellos, 1992: 155) 209
Thus, she takes At the same time in this example as the cognitive, or topical/ideational, Theme of the clause. Her example can be compared with the following two, taken from the history textbook corpus: 5-52. Eng 7 (161) Solon initiated a rational approach to the problems of society by deemphasizing the gods' role in human affairs and attributing the city's ills to the specific behavior of individuals; (162) he sought practical remedies for these ills; (163) and he held that written law should be in harmony with Dikй, the principle of justice that underlies the human community. (164) At the same time, he wanted to instill in Athenians of all classes a sense of working for the common good of the city. 5-53.Eng 1 (9) We are shaped by values, interests, and customs that (like our churches, universities, and political parties) have been formed by past experiences. (10) At the same time, the way we look at the past and what we see there -the questions we ask and the problems we pose- arise out of current concerns. With example 5-52, it is tempting to gloss "at the same time" as referring to at the same point in time as the other events which were occurring, as with Vasconcellos' example, and thus clearly an external relation. The narrative nature of those two texts lends itself to this gloss. However, example 5-53 would not seem to lend itself to a gloss which interprets at the same time as referring to a real point in time. It seems to mean rather "equally", clearly an internal relation. Nonetheless, and at the same time, this gloss causes one to re-interpret at the same time in example 5-52, which also seems to be able to take on the interpretation of "equally". Indeed, Martin (1991: 324) in comparing science and history textbooks, finds that the conjunctive structure of historical explanations through exposition "foregrounds internal, not external relations". Indeed, while there are a few instances of "finally" and "finalmente" in the corpora which could be considered external, the bulk of conjunction is internal, and thus this distinction has not been analyzed for exact numbers. With regard to conjunction in expository texts, Martin (1991: 324) indicates that "[a]s is typical of Exposition, very little of this internal conjunctive structure is made explicit". Whittaker (1995) corroborates this: in her study of linguistics and economics articles argumentative paragraphs were found to contain many more textual Themes than expository ones. This can give some indication upon which to base predictions of the amount of conjunction which might appear in the history textbook corpora. It can also be indicative to look at other corpus analyses carried out in different genres. Unfortunately, there is not much in the literature on Theme upon which to make predictions. And a further problem in comparing studies is that, as mentioned in Chapter 3, different analysts take as 210
their basic unit of analysis a different construction: the sentence, the clause, or the clause complex. However, at least an idea of what might be expected can be approached by looking at other results. Whittaker (1995), taking the orthographic sentence as the unit of analysis in a study of economics and linguistics academic articles, found roughly 15% of the sentences in each set of texts to contain a textual Theme. She found that two texts, one in linguistics and one in economics, stood out with respect to the others, in that they used textual Themes in nearly 25% of the clause complexes analyzed. She attributes this greater use in these texts to the fact that they had a larger proportion of argument as opposed to exposition, and posits that perhaps density of textual Themes may be one mark of argument. The content of the history textbooks is not overtly argumentative in nature; thus, perhaps the density of textual Themes in this study can be expected to be lower than that of Whittaker's study. Francis (1990) found similar percentages of Themes with a textual component across three news genres (news, editorials, and letters), but she does not include those percentages, which would allow for a comparison with Whittaker's data. Ghadessy (1995: 134) feels that in most texts "there are plenty of conjunctions functioning as textual Themes", and the written sports reports which comprise his data use textual Themes in 45% of the clauses; his unit of analysis is based on the clause, except minor and rankshifted clauses. He contrasts his high results with Xiao's (1991) study, unfortunately with no mention of the basic unit of analysis, where recipes use textual Themes in 11% of the clauses, and fables, 18%. Dubois (1987: 99), who analyzes independently conjoined clause complexes, as does the present study, finds textual Themes in 20%6 of the clauses in a slide talk from a biology conference. In Bдcklund's (1992), study of telephone conversations, where the unit of analysis is the same as for Dubois, nearly 44% of the clauses have a textual Theme. These high percentages contrast with Taylor's (1983) study of science and history textbooks, the latter naturally being of special interest to this study. Taylor's unit of analysis is similar to that of Dubois, except that he analyzes non-restrictive relatives and reported clauses separately. He finds that 14.06% of the clauses have a textual Theme in science textbooks, and 12.26% in history. In order to present a clearer picture of the comparison/contrast amongst the different studies, Table 5.9 presents the different results: 6 No exact percentage is given, but she provides the counts of "the four conjunctive adverbials" of one particular slide talk, which includes however, in fact, so, and now. In this study, in fact is included in modal Themes, so the totals of the other 3 were taken and divided by the total number of clauses analyzed by Dubois to arrive at the percentage. 211
Table 5.9: Textual Theme Results from other Studies
Genre
% of Tex.
Themes
Sports reports
45.00%
Telephone conversations
44.00%
Biology talk
20.00%
Fables
18.00%
Economics texts
15.00%
Linguistic texts
15.00%
Science textbooks
14.06%
History textbooks
12.26%
Recipes
11.00%
The studies have been ordered purposely from the one with the most textual Themes to the one with the least. It seems surprising that the textbook genres are amongst those having the fewest, given that one might expect educational texts to provide clear signposting for their readers. Ventola and Mauranen (1991: 463) point out the usefulness of coordinators and conjunctive Adjuncts for the reader in that they "can greatly facilitate the reader's task of decoding the organization [of a text] by marking logical relations explicitly with certain linguistic signals which have a linking or connecting function." Vande Kopple (1985) reports on research by Meyer (1982), in which she tested what effects the presence or absence of text signalling devices had on ability to recall content. While she found that there was no difference for the good and poor readers, average readers (from ninth grade and junior college) demonstrated better recall if the passages contained signalling devices. However, Taylor (1983) holds that there is a great deal of asyndeton in formal school writing. And, given the genre-based hypothesis upon which the present study is based, his figures may be closer to the mark for the corpus analyzed here. This may be related to Whittaker's (1995) finding of higher proportions of textual Themes in argument as opposed to exposition, and Martin's (1991) affirmation that exposition leaves a lot of the conjunctive relationships implicit. It may also be related to Mauranen's point above, in that perhaps relationships are implicit in instances involving thematic chaining. Thus, perhaps textual Themes will not be highly prevalent, except in cases where the relationships are not explicit. Therefore it was predicted that the number of textual Themes in the corpus would appear in a frequency similar to that of Taylor's findings. Also, there was an expectation that those ideational Themes which are more difficult to link to the rest of the discourse will tend to appear with a textual Theme. And, finally, given the genre-based hypothesis of this
212
study, it was predicted that there would be similarity across the two languages with respect to the use of textual Themes. This prediction is very speculative, because it is difficult to see from the different studies previously mentioned if indeed there is any relationship between genre and use of textual Themes. In a different genre, that of economics articles, ValeroGarcйs (1996) compares native English speakers and Spanish speakers, both writing in English. She finds that the native-English texts use more connectors than do the SpanishEnglish texts (the two Spanish-English texts use 20.5% and 12.7% respectively, and the native-English texts use 28.4% and 17.6%, respectively). While it is interesting to note that her findings are somewhat similar to Whittaker's (1995) for economic texts (the average in Whittaker's corpus was around 15%; however, some of the texts had 25%), thus, perhaps, suggesting some relation to genre, Valero-Garcйs finding clashes with that of Moreno (1997) on causal metatext in Spanish and English research articles in business and economics. The latter study shows similar strategies for expressing cause-effect intersentential relations, and a similar distribution of these. A final point is that the Rhemes of the clauses of the history textbooks were also analyzed for conjunctive Adjuncts. The expectation here is that, if there is a conjunctive Adjunct connecting the clause to a previous clause, it will be more likely to appear in the Theme. Halliday (1994: 49-50) states: It is not difficult to see why modal and conjunctive Adjuncts tend to come at the beginning of the clause: if one of them is present at all, then in a sense it is a natural theme...if there is some element expressing the relationship to what has gone before, by putting this first we thematize the significance of what we are saying: `I'll tell you how this fits in'.
5.6.3 Textual Theme Results
The results for textual Themes appear in Table 5.10:
Table 5.10: Textual Themes Results
Text
Text. Themes Text
#
%
Eng 1
31 22.96% Spa 1
Eng 2
36 27.07% Spa 2
Eng 3
38 28.15% Spa 3
Eng 4
7 20.59% Spa 4
Eng 5
24 14.55% Spa 5
Eng 6
41 19.71% Spa 6
Eng 7
44 24.31% Spa 7
Eng 8
44 25.88% Spa 8
Eng 9
39 28.89% Spa 9
Eng 10
46 23.00% Spa 10
213
Text.Themes
#
%
42 31.58%
18 12.08%
59 35.76%
29 18.95%
19 15.57%
36 25.17%
28 17.07%
8 28.57%
45 43.69%
18 17.48%
Total English 350 23.40% Total Spanish 302 23.91% There are two noteworthy aspects of the results. first of all, if we look at the results of the two corpora together, the total amount of textual Themes is virtually identical. This is as predicted above, and seems to confirm that there may be genre constraints in effect across the two languages. However, given that these results are much higher than Taylor's for history textbooks, this obviously needs more corroboration. We will return to a comparison with Taylor's study in this respect below, in looking at conjunctive Adjuncts and coordinators separately. The above point made about genre constraints in place can be set aside also by the second noteworthy finding, which is that, as with the other findings in this study, there exist as many differences between texts of the same language as exist across the two corpora. It has been previously suggested in this study that this could be due to either author preference or the subject matter of the chapter in question, or some combination of both. Therefore, the results from Table 5.10 above are repeated below, in Table 5.11, along with the authors of the texts and the subject matter of each one. The table has been ordered from the text having the highest percentage of textual Themes, to that having the lowest, across both corpora. This allows for comparison of subject matter across the two languages. 214
Table 5.11: Textual Themes Percentages with Author and Subject Matter
Text
Author
Subject
%
Spa 9 Cerdб, et.al
Prehistory
43.69%
Spa 3 Ballesteros & Alborg 12th cent France & 12th & 13th cent Spain 35.76%
Spa 1 Ballesteros & Alborg Introduction to the study of history
31.58%
Eng 9 McKay, et.al
19th century population
28.89%
Spa 8 Fernбndez
Preface
28.57%
Eng 3 Chambers, et.al
14th & 15th centuries
28.15%
Eng 2 Chambers, et.al
Prehistory
27.07%
Eng 8 Roy
Ancient Rome
25.88%
Spa 6 Santos
Ancient Rome
25.17%
Eng 7 Perry
Ancient Greece
24.31%
Eng 10 Brinton, et.al
18th century
23.00%
Eng 1 Chambers, et.al
Introduction to the study of history
22.96%
Eng 4 Greer & Lewis
Preface
20.59%
Eng 6 McKay, et.al
13th & 14th centuries
19.71%
Spa 4 Mangas
Ancient Greece
18.95%
Spa 10 Palacio, et.al
18th century
17.48%
Spa 7 Domнnguez
18th century
17.07%
Spa 5 Palacio, et.al
14th & 15th centuries
15.57%
Eng 5 Greer & Lewis
Prehistory
14.55%
Spa 2 Fernбndez
19th century population
12.08%
There does seem to be some tendency for texts written by the same authors to cluster together in terms of percentage of textual Themes used: the two Ballesteros & Alborg chapters both have a high percentage of textual Themes, 35.76% and 31.58%. Two of the Chamber's texts are very close together, with 28.15% and 27.07%, and the other is not too far behind with 22.96%. Greer and Lewis show a slightly higher difference between their preface and their chapter on prehistory, with 20.59% and 14.55% respectively. This tendency is not continued with two sets of authors: the texts by the same authors which have the greatest amount of difference are the two McKay chapters, with 28.89% and 19.71%, and the Fernбndez chapters, with 28.57% and 12.08%. Thus, author preference, or, perhaps, the preference of editors, may explain the difference in part. As to the other speculation, that subject matter may have an influence on textual choices, this is born out in part by three of the texts on Ancient Greece and Rome, and two of those on the 18th centuries. However, it is contradicted by the findings from the two introductory and preface texts, and especially by the chapters on prehistory. One of the latter, Spa 9, shows the highest count, with 43.69%, another shows a somewhat higher than
215
average count, with 27.07%, and one shows one of the lowest counts, with 14.55%. These
differences can be explained in part by author preference, and in part by the density of
information presented in Spa 9, which the author helps the reader to digest through the
inclusion of textual Themes. Obviously, then, these results need to be contrasted with
further results from the same authors, on the one hand, and other studies on the same
subject, on the other.
With regard to the difference between conjunctive Adjuncts and coordinators as
Theme, the following results were obtained:
Table 5.12: Conjunctive Adjunct and Coordinator Results
Adjuncts Coordinators
Adjuncts Coordinators
Text
%
%
Text
%
%
Eng 1
4.44%
18.52% Spa 1
13.53%
18.05%
Eng 2
6.77%
20.30% Spa 2
4.70%
7.38%
Eng 3
3.70%
24.44% Spa 3
4.24%
31.52%
Eng 4
14.71%
5.88% Spa 4
2.61%
16.34%
Eng 5
3.03%
11.52% Spa 5
9.84%
5.74%
Eng 6
4.33%
15.38% Spa 6
11.19%
13.99%
Eng 7
4.42%
19.89% Spa 7
6.10%
10.98%
Eng 8
2.94%
22.94% Spa 8
14.29%
14.29%
Eng 9
12.59%
16.30% Spa 9
18.45%
25.24%
Eng 10
3.00%
20.00% Spa 10
7.77%
9.71%
Total Eng
5.01%
18.38% Total Spa 8.31%
15.60%
Also here there are a couple of noteworthy findings. First of all, in comparison with Taylor's (1983) data, there is a similarity if the Adjunct results of the two corpora are taken together. By doing this, by adding the total thematic conjunctive Adjuncts in the English data to those of the Spanish data, and dividing by two, the result is 6.66%. Taylor (1983) found 6.30% of clauses had thematic textual Adjuncts in his history textbook data. Obviously, more studies which measure conjunctive Adjuncts in Theme position are needed to make any kind of generic statement, but the similarity in overall results proves interesting. At any rate, the English data falls somewhat below Taylor's finding, as there seems to be a greater preference in the English data for coordinators over conjunctive Adjuncts than in the Spanish data. Obviously, in both corpora coordinators are more prevalent as Theme than conjunctive Adjuncts, given that their position in the clause is obligatorily initial. In all, the overt use of conjunction is low for a genre type where one would expect clear signposting for the reader. This becomes clearer if we look at the uses to which conjunction is put. First of all, with regard to coordinators, the following table explains how 216
they are used in both corpora with those clauses which in effect do not cause problems for
reading comprehension:
Table 5.13: Coordinator Use with Non-problematic Clauses:
Coordinator Use
Eng Spa
Total coords in chains
149 117
Total coords. w/ellipsis
66
31
Before existential
4
6
Before extraposition
3
7
Before impersonal
0
3
Before inversion
0
6
Before manner
3
0
Before passive
0
1
Before sum
7
0
Before key Theme
6
6
Before Time
12
6
Before we
3
3
Before wh
2
0
Total
255 186
Given that the total number of coordinators in the English corpus is 275, and in the Spanish corpus, 197, it can be seen that the coordinators are used mainly to reinforce relationships that are often explicit lexically. This is, however, to ignore a difference between a lexical tie and a conjunctive one: a conjunctive tie tells the reader how the writer relates the information in one clause to that given in a previous clause; it tells the reader whether the relationship is one of addition, adversativity, contrast, causality, and so on (cf. Rudolph, 1984), thus allowing the reader to build up the logical relations between ideas in the text. The lexical relationship picks up on something already mentioned in a previous clause and thus adds content to the mental schema which the reader builds up of the text. This use of conjunction to express logical relationships between ideas on top of an already existing lexical tie can also be seen in the case of Adjuncts:
217
Table 5.14: Conjunctive Adjunct Use with Non-problematic Clauses:
Conjuntive Adjunct Use Eng Spa
Total Con. Ads. in chains 40 44
Before existential
0
9
Before extraposition
1
4
Before impersonal
0
4
Before inversion
0
7
Before manner
0
1
Before metatextual
0
1
Before passive
0
0
Before place
1
1
Before previous clause
0
1
Before sum
2
7
Before thematic equative
0
1
Before key Theme
11
5
Before Time
2
4
Before we
3
3
Before wh
1
0
Total
61 92
Given that the total number of conjunctive Adjuncts in the English corpus is 75, and of the
Spanish corpus, 105, again we can see the main function of these Adjuncts is to map a
logical function on top of a lexical tie. There are very few conjunctions left to take care of
those clauses which were difficult to relate lexically to the rest of the text. In going back to
Mauranen's (1993c) point that these clauses would have some sort of orienting Theme, in
which she includes connectors, adverbials, Complements, and modal and reporting clauses,
we can see that this is definitely not the case with the corpus analyzed here. We can
eliminate from the account here adverbials, Complements, and reporting clauses, as these
would be taken to be Theme, thus if they came before a new, back, or rel element, these last
would not be considered Theme. Thus, if we look at those new, back, and rel Themes which
have as an orienting Theme a modal or connecting Theme, we can see that in few cases does
this occur:
Table 5.15: Orienting Themes with Problematic Clauses: English Corpus
English Corpus Total w/conj. w/coord. w/modal
Adjunct
Back
122
9
10
5
Rel
20
1
0
1
New
73
4
10
1
218
Table 5.16: Orienting Themes with Problematic Clauses: Spanish Corpus
Spanish Corpus Total w/conj. w/coord. w/modal
Adjunct
Back
100
7
6
5
Rel
18
2
0
0
New
76
4
5
3
In sum, it would seem that in both corpora, the authors would do well to help their novice readers along the way by providing more orienting Themes, especially those which signal the logical relationships between the different ideas in the text. A couple of examples from the corpus illustrate this: 5-54.Eng 2 (9) Permanence, complexity, and sophisticated social organization transformed the agricultural village into something we can recognize as urban, the form of habitation crucial for the development of civilization. (10) Expansion caused these early societies to develop still more complicated forms of political and social control. 5-55.Spa 2 (67) En los hospicios la mortalidad de los hijos naturales fue mucho mayor; Spa 2 (68) un autor francйs ha asegurado que sуlo el 10% alcanzaba los 20 aсos. In Example 5-54 above, expansion in Clause 10 is brand new to the text. The author could have initiated the clause with a coordinator (and) or a conjunctive Adjunct (then, at the same time), or a combination (and then). In example 5-55, given that Clause 68 exemplifies the point made in Clause 67, for example would help readers to add the new information to the schema they build up of the text. A final point relates to the use of conjunctive Adjuncts in the Rhemes of the clauses. It was predicted that these would tend to appear in a thematic position more often than in a rhematic one. It must first be pointed out that only those conjunctive Adjuncts which could have appeared in Theme position without any change in the logical relationship expressed were counted. Examples include: 5-56.Eng 8 (18) Rome was not yet a beautiful city, however. 5-57.Spa 4 (29) La historia griega no es, por tanto, una historia lineal como puede ser la de Roma o la de las nacionalidades modernas en los ъltimos siglos. The results appear in Table 5.17:
219
Table 5.17: Conjunctive Adjuncts in the Rheme
Text
Conj.
% Text
Adjuncts
Eng 1
14
10.37% Spa 1
Eng 2
7
5.26% Spa 2
Eng 3
9
6.67% Spa 3
Eng 4
2
5.88% Spa 4
Eng 5
14
8.48% Spa 5
Eng 6
7
3.37% Spa 6
Eng 7
14
7.73% Spa 7
Eng 8
9
5.29% Spa 8
Eng 9
7
5.19% Spa 9
Eng 10
19
9.50% Spa 10
Total English 102
6.82% Total Spanish
Conj. Adjuncts 11 1 9 7 10 8 10 1 2 2 61
% 8.27% 0.67% 5.45% 4.58% 8.20% 5.59% 6.10% 3.57% 1.94% 1.94% 4.83%
While the prediction holds for the Spanish corpus, where 8.31% of the clauses had conjunctive Adjunct in thematic position, in English the amount is higher in rhematic position than in thematic position. At any rate, the number of clauses across both corpora which contain some form of conjunctive Adjunct or coordinator indicating a logical relationship between ideas is low, thus perhaps providing more evidence of the use of grammatical metaphor across the textbooks from both cultures; temporal and consequential logical relationships may find their expression in more incongruent places in the clause, i.e. in the verbal process, such as (verbs expressing cause are underlined): 5-58. Eng 5 (83) These advances in toolmaking led to fundamental alterations in human social organization and way of life. 5-59. Eng 8 (121)Finally, the coming of internal peace brought economic prosperity. This finding can be combined with the finding in the previous chapter with respect to the reliance in the history textbooks on attributive relational processes, thus reflecting a good deal of nominalization. As Martin (1991: 333) states: "In history nominalization is strongly associated with realizing events as participants so that logical connections can be realized inside the clause". 5.7 Conclusions In this chapter, we have seen that, setting aside grammatical constraints, the overall results from the English and the Spanish corpora are very similar with respect to simple linear progression, and with certain Theme types which fall outside of the patterns, including summative Themes, repeated key Themes, and time adverbials, and with respect to "problematic" (in terms of text comprehension) Themes, such as back references, related 220
ideas, and new Themes. The English corpus, however, showed a higher percentage of constant Theme progression chains, especially in the case of long Theme chains, involving 5 clauses or more. It is felt that this backs up to some extent the conclusion reached in Chapter 4: that in the American history textbooks, there is more of a tendency towards narrative than in the Spanish history textbooks in this corpus. However, the conclusion has also been reached in this chapter that it is simplistic to state that constant Theme progression is associated with narration and description. Narrative and description also occur in both corpora with the simple linear pattern, showing a tendency to analyze and explain through that type of narrative or descriptive chain, rather than to list events in simple chronological order or to list the features of a person, place, or thing. Through the analysis of the textual role played by the Theme types which fall outside the thematic patterns, in particular, back, related and new expressions, and through the analysis of textual Themes, the conclusion was arrived at that, in both cultures, there seems to be a curious affluence of difficult Themes, while at the same time there is a small number of textual Themes, especially of conjunctive Adjuncts. The results arrived at were very similar across the two corpora, save in Spanish more conjunctive Adjuncts were used in Theme position than in English, while in English more coordinators were used, and more conjunctive Adjuncts were used in Rheme position than in Spanish. At any rate, the differences are small, and the corpora totals, especially those of conjunctive Adjuncts, are very similar to Taylor's (1983) findings for history textbooks. It is felt that this supports to some extent validity for the genre-based hypothesis upon which this study is based. Obviously, however, more analysis is needed, in comparing same authors with same authors, and same subjects with same subjects, in order for more support to be put forward for the hypothesis. 221
6. The Interpersonal and Theme/Rheme 6.1 Interpersonal Elements in Theme and Rheme 6.1.1 Interpersonal Function The two previous chapters were based on the experiential and textual metafunctions; in this chapter, we move to the interpersonal. The interpersonal metafunction is concerned with social relationships as they are realized in text, in this case the interaction between the authors of the textbooks and the reader/students. Obviously, writer intervention is a constant in the text: simply put, texts are written by writers. However, writers can choose to make their intervention more or less obvious, and they can do so through the interpersonal metafunction in several different ways. The purpose of this section is to discuss interpersonal elements as they appear in the Themes and Rhemes of the corpus. It does not enter into the scope of this paper to give a full account of the clause as exchange, but to look at those elements of the clause which either: a)hint at the writer's opinion or position as to the proposition(s) included in the clause b)show some kind of engagement with the reader. 6.1.2 Writer Intervention in History Textbooks The traditional practice of writing up history, or "normal historical practice" (Berkhofer, 1989: 183), has at its base two practices: first of all, the obtainment of facts about the past from evidence, and secondly, but perhaps more importantly (Berkhofer, 1989), the combination of the facts obtained into a coherent narrative or synthesis. This involves the historian in making generalizations from sources, and in reconstructing/portraying past events as they happened. The assumption between the historian and the reader is that the historian is presenting an accurate picture of the past, and "the central presupposition of idealized, normal history production is transparency of medium, in which the exposition conveys or at least parallels factuality" (Berkhofer, 1989: 184). Berkhofer states that the effect normal historians try to achieve is that of factuality, "to impress the reader that the structure of interpretation is the structure of factuality" (ibid: 190-191). Thus, the authors' task is not as straightforward as narrating a series of events, since they are not direct witnesses to the events they relate; instead, they have to rely on other 222
sources of information, both first hand (historical documents and artifacts) and second hand through other historians' accounts of the past, which the history textbook writers sift through in order to offer the student not simply an account of past events, but an interpretation of those events (Berkhofer, 1989). At the same time, there is a lack of symmetry between the two participant roles involved in this act of communication. The author has knowledge which the student requires or is lacking. Therefore, the author's voice in a history textbook is usually one of authority over the reader, one who is in a position to know. Thus, history textbook authors typically use declarative statements to tell their readers the way things were, to narrate what battle took place where and when, what certain individuals did, and so on. However, their interpretive role may at times take over their authoritative/authorial role. In this case, one would expect their statements to be qualified in some way, for there to be an indication that what they are relating is not a narration of events, but their own interpretation of those events. It is easy to imagine that if a group of eminent historians were discussing together their interpretations of a set of historical events, especially in terms of causal relations, they would probably hedge their statements considerably, more so than when talking with a group of novices. Indeed, it has been noted through discourse analysis that school textbooks feature less hedging than other types of expository writing, for example, reports in specialist scientific journals (Bloor & Bloor, 1995: 54). According to Myers (1989: 12-13) hedging presents a claim to the community as provisional and it "reflects, not the probablility of the claim, and not the personal doubt of [the researchers]..., but the appropriate attitude for offering a claim to the community". Thus, hedging is related to politeness (Brown and Levinson, 1978) and in research articles allows the writer to respect the negative face of other researchers. There are many ways of including a hedge in a text, e.g. through markers of modality and attribution to other authors or to impersonal agents. Myers (1989) looks at hedges in scientific research articles, while Pindi and Bloor (1987) look at hedging and attribution in economics texts. Holmes, in an analysis which includes research articles in history, (1997: 333) mentions the "relative scarcity in history of references to previous research, a scarcity that is plausibly rooted in the limited development of cumulative research programs and the absence of a theoretical consensus in that discipline". In their more authoritative role, authors of history textbooks tend to present information as factual, and to ignore or downplay other possible interpretations of historical 223
accounts. Indeed, Berkhofer (1989: 197) goes so far as to say that "normal history orders the past for the sake of authority and therefore power over its audience" and later goes on to say that "[b]y assuming a third person voice and an omniscient viewpoint, authors... assert their power over readers in the name of REALITY" [emphasis original] (ibid: 197). This is not a feature unique to history writing. Halliday (1996: 364-365) generalizes this phenomenon to school textbooks, when he states: "[if] they [textbooks]are to function effectively, the readers they are addressed to must believe in what they say". He culls information from Luke, Castell and Luke (1989:245ff, in Halliday, 1996: 365) and Olson (1989: 233ff, in Halliday, 1996: 365) which indicates that: textbooks sanctify `authorized (educational) knowledge' simply by authorizing it - what is in the text books is thereby defined as knowledge - and text books maintain this authority by various means such as claiming objectivity and creating distance between performer and reader, and so come to be accepted as `beyond criticism'. ...textbooks strive for clarity, explicitness and an unambiguous presentation of the facts; they seek to `delimit possible interpretations'. (Halliday, 1996: 365) Furthermore, and perhaps less contentiously, it is the author's purpose to inform. Francis (1990) found a higher frequency of occurrence of interpersonal Themes in editorials and letters than in news reports, as the first two are more "personal and evaluative, as well as more concerned with persuasion than information, [so] there is likely to be more direct interaction between writer and reader" (61). Berry (1995: 58) distinguishes between what she terms interactional and informational thematisation; the former refers to Themes which "foreground the interactiveness of the discourse by referring to speaker, hearer, writer, reader," and the latter to Themes which "foreground the organisation of the content by referring to aspects of the topic" (ibid: 82, fn.5). Martin (1986) indicates that thematization along these lines will have different effects: informational thematization makes for texts which are well oriented to helping readers looking for information, as they can focus on thematic content to get the gist of the text. At the same time, however, these texts may seem impersonal and alienating; thus, more interactional thematization increases the `human face' of the text, however, "the gain in humanity is at the expense of clarity and of economy in reading and processing time" (Berry, 1995: 59). Brown and Yule (1983: 143) note that one characteristic of primarily interactional conversational speech is that the interactional aspect is frequently thematized, through, for example, personal pronouns such as I and you. Thus, in more informationally aimed texts, such as textbooks, one would expect a downplay 224
of interpersonal/interactional type Themes. Further evidence for this, specifically in history textbooks, is provided by Eggins, Wignell and Martin (1993: 92). They find a correlation between interpersonal distance and the use of grammatical metaphor: "...[t]he closer the interpersonal distance between interactants (the more feedback is immediate, the more there is visual and aural contact), the more congruent the language used is likely to be." Thus, a greater use of grammatical metaphor implies more interpersonal distance between speaker and addressee. We have seen in Chapter 4 that there is a great deal of grammatical metaphor in the history textbook data looked at in this study. In sum, based on the above discussion, the expectation is that in the corpus analyzed here, that of history textbooks written for first-year university students/final year secondary students, there will be a small amount of overt writer intervention across both languages. Overt writer intervention here is taken to mean hedging, or providing some means of indicating to the reader that there are other possible positions which would interpret historical events in differing ways. However, hedging is just one feature of the more interpersonal side of things. Another aspect is that of engagement with the reader: some kind of indication of the presence of the reader in and through the text. Textbooks, as we have seen above, demonstrate distance between writers and their readers, thus minimizing any kind of exchange. However, at times writers do include readers through questions and direct addresses to the reader. It is more difficult to hypothesize here as to the differences/similarity across the two languages. This is because of the Contrastive Rhetoric hypothesis, explained in section 1.3 above. There, rhetoric was tied to the relationship between speakers/writers and their audiences and the effect that the former would like to achieve over the latter. Indeed, in this regard, in section 2.2.2 above, we have seen studies on student writing and on some professional texts (e.g. business memos) which showed differences in Anglophone and Spanish writing, mainly with respect to greater impersonality and formality in texts written in Spanish. However, it is felt here that the nature of textbooks, with their asymmetrical writer/reader relationship, coupled with a common tradition in history writing in western societies (see section 4.1 above) mean that it is possible to hypothesize that the two corpora will display similar results with respect to the interpersonal elements of the Themes and Rhemes of the clauses analyzed here. At the same time, nonetheless, as with the textual and ideational analyses, there may be differences across texts within each of the corpora, which may be attributable to 225
differences in subject matter: sections of the text dealing, for example, with ancient history may include more hedging and interpersonal comments than sections dealing with more recent historical events. The reason for this is that with more recent events the historian has more "facts" to work from, more documented evidence. In the case of ancient history, the historian has to piece together a picture of "reality" through archeological evidence, and/or rely on other experts' interpretations of the same. In sum, in order to look at these two areas of interpersonal expressions in text, those which highlight the positioning of the author(s) with respect to the statements they make, as well as those which indicate the degree of writer/reader interaction, the texts have been analyzed for features which show modality, on the one hand, and reader/author reference on the other. The English and Spanish texts have been analyzed for the number of these interpersonal features in both the Themes and the Rhemes of the clauses/clause complexes. Before laying out hypotheses based on these notions for the corpus here, a review of other researchers' findings in the area of the interpersonal is warranted. The discussion is brief, as there are not many analyses which specifically report on interpersonal elements in the Themes of the clauses, and none that I am aware of at the time of writing on these in the Rhemes of clauses. In his study of written sports reports, Ghadessy (1995) found that 6.11% of the clauses included an interpersonal Theme. It is not absolutely clear what he includes in his analysis as interpersonal Theme. He states that he is following Halliday (1985) in his analysis, so presumably he includes modal Adjuncts, WH-items in questions, finite modal operators, and vocatives. This could explain the difference between his findings and those of Thomas and Hawes (1997), also in the realm of newspaper discourse, in this case editorials. They note a difference between editorials in The Sun and The Times in modal Adjunct use: 3.6% and 1.8% respectively. In a set of research articles in linguistics, Whittaker (1995) found an average of 8.84%, with one text having a percentage as high as 16.4%. In the same study, Whittaker, who analyzes for a range of interpersonal Themes, found a very similar percentage of interpersonal Theme inclusion in a set of economics texts: 8.56%; in this set also, one text stood out at 13.1%. She attributes the differences in the cases of the salient texts to text type: in each of the sets, these texts were different; the salient linguistic text was a chapter in a book, rather than a research article, and the salient economics text was written to be spoken, although it was published in an academic journal. The differences 226
between her findings and those of Ghadessy and Thomas and Hawes may not be attributable
to generic constraints, however, as Whittaker includes grammatical modality in her account
of interpersonal Theme see section 6.1.3.2 below), thus the range of items analized is wider.
Bдcklund (1992) clearly indicates that she analyzes modal Adjuncts and vocatives as
interpersonal Theme in her study of telephone conversation, in which she finds that 22.43%
of the clauses contain an interpersonal Theme. This would seem to confirm the more
interactional nature of conversation, and the higher percentage may also be attributed to the
use of vocatives, which Thomas and Hawes (1997: 125) point out as "more properly the
domain of spoken discourse or limited kinds of written discourse such as poems and songs".
The study of greatest interest here is that of Taylor (1983), who finds a very low percentage
of modal Themes, by which he seems to mean modal Adjuncts, in the textbooks he looks at:
0.08% in science textbooks, and 0.63% in history textbooks. These findings are laid out in
Table 6.1:
Table 6.1: Interpersonal Theme Results from other Studies
Genre
% of Interpersonal
Themes
Telephone conversations
22.43%
Linguistic texts
8.84%
Economics texts
8.56%
Sports reports
6.11%
The Sun editorials
3.60%
The Times editorials
1.80%
History textbooks
0.63%
Science textbooks
0.08%
It is interesting to note the order from most to least. In light of the previous discussion, it is not surprising that conversation would include more interpersonal Themes. Also, it has been posited that research articles include more hedging than textbooks. The newspaper genres are perhaps the most surprising, especially the fact that the sports reports, which one would expect to be a mirror of "reality" are higher than the editorials. A cursory glance at the modal Adjuncts which Thomas and Hawes (1997: 126) found in their study shows that the majority express probability: of course, certainly, naturally, by no means, surely, indeed. Thomas and Hawes (ibid: 125) note that several of these expressions "reinforce the "selfevident" image of what is being presented, telling the reader their argument is beyond question". At any rate, the low amount in the textbooks serves to reinforce Bloor and Bloor's point mentioned above that textbooks exhibit less hedging than other genres. 227
Thus the hypotheses put forth for this chapter are: 1. Modality in Theme position will be downplayed in the history textbooks in a similar way across the two corpora; thus, there will be a similar, low number of overt textual features expressing modality in the Themes of the clauses. 2. As in number 1 for the Themes of the clauses, the same is hypothesized for the Rhemes. However, the prediction is that overt textual features expressing modality will be higher than in the Themes; in other words, it is expected that if the authors do choose to recognize the existence of divergent positions regarding the propositions they make by including markers of modality, they will downplay this by favoring Rheme over Theme position. 3. Overt author/reader interaction will be hardly evident in the texts; in other words, the number of direct references to the authors or their readers and questions will be low across both corpora. 6.1.3 Modality Modality is the expression of speaker/writer attitudes (henceforth, I shall refer to writers, given that the corpus under analysis here is written) towards a proposition, or, and perhaps more apropos when examining a history textbook corpus, the means by which writers "express attitudes towards the event contained in the proposition" (Downing and Locke, 1992: 381), with regard to possibility, probability, and certainty, as well as some temporal notions such as usuality. A declarative statement with no marker of modality is presented as truth, as reality. As can be expected, such statements abound in the history textbook corpus in both languages, and, indeed, they form the vast majority of the statements made. The following is only a small sample of such statements: 6-1. Eng 1 (1) Everyone uses history, 6-2. Eng 2 (10) Expansion caused these early societies to develop still more complicated forms of political and social control. 6-3. Spa 2 (140) Irlanda tenнa en 1835 ocho millones de habitantes; 6-4. Spa 1 (129) Asн naciу la Historia moderna. At other times, writers insert markers of modality, such as: 6-5. Eng 3 (105) Presumably, they saw no point in working for the future when the future was so uncertain. In this way, they lessen their commitment to the truthfulness of what they are saying; they no longer present it as a "fact", but as something which possibly or probably could be true. 228
With markers of certainty, however, writers actually increase commitment to the truthfulness by emphasizing the conviction they hold for the truth of the proposition. 6-6. Spa 2 (125) Indudablemente las motivaciones econуmicas han jugado en todo momento un papel decisivo. This is one interpretation of modality. We have seen above that modality is also related to the politeness phemenon and hedging. White (1998) also points out that functional analyses of modality show that truth-functional explanations are inadequate in fully explaining its role in texts. He explains that modality allows speakers to "acknowledge the contentiousness of a particular proposition, the willingness ... to negotiate with those who hold a different view, or the deference of the speaker for those alternative views". Thus, in example 6-5 above, "presumably" is a recognition on the part of the other of "divergent heteroglossic positions", or multiple points of view, on the proposed statement. In other words, rather than saying that speakers are lessening their commitment to truth in statements in which they use markers of modality, it could be said that they are showing that the proposition is open to other possible interpretations. On the other hand, the declarative statements in examples 6-1 through 6-4 above (declarative statements with no markers of modality) downplay this possibility of heteroglossic diversity. And finally, for White, a marker of modality such as "undoubtedly", or "indudablemente" in Spanish, is, like "presumably", fundamentally heteroglossic in that this type invokes the diversity, although they then serve to "close down" that heteroglossic diversity. In other words, "undoubtedly" entertains the ghost of the notion that there could be a doubt, although the author is in effect stating that there is no doubt. Thus, in a truth-value interpretation, it could be said that the author, through "indudablemente", shows a high degree of commitment to the truth of the proposition which states that economic motivations have played a decisive role at all times. In White's interpretation, the author is denying the existence of other possible interpretations. This is interesting from the point of view of Halliday's quote above on authors delimiting possible interpretations of texts in order to gain authority combined with Berkhofer's quote which states that historians need to construe interpretation as factuality. Modal meanings have several different realizations in the clause. The main types of realization are modal Adjuncts and finite modal operators. These are discussed below, along with interpersonal grammatical metaphor, which provides another means for writers to interpolate their commitment to propositions. All instances of modal Adjuncts, finite modal operators and grammatical metaphor, which form part of independent and dependent 229
clauses, finite or non-finite, as well as those modifying prepositional phrases functioning as circumstantial Adjuncts, are considered here. Those appearing in rankshifted clauses are not included in the analysis. Nor are modal expressions which form part of nominal groups. This means that intensifiers such as yet when used in constructions like the example below are not included. 6-7. Eng 8 (91) Many a senator, ...on his way to a yet more demeaning debate in the senate, ... 6.1.3.1 Modal Adjuncts Modal Adjuncts "express the speaker's judgment regarding the relevance of the message" (Halliday, 1994: 49). Halliday divides modal Adjuncts into two broad, but overlapping, categories: mood Adjuncts and comment Adjuncts. Mood Adjuncts are so called because they modulate, or "tune", the frequency, generality, or probability of the verbal process expressed in the proposition. They can do so through expressions of polarity, modality, temporality or mood. Polarity refers to the positiveness/negativeness of the proposition: simply whether it is "yes" or "no", affirmative or negative. An analysis of polarity is not really of interest to this study, and it has already been explained in section 3.5.2.2 that the negative expressed in the process of the clause as part of the structure of the verbal group is not considered here as a separate constituent. Nonetheless, there are occasions when authors emphasize the polarity of their declarations by inserting a separate, overt marker, as exemplified by: 6-8. Spa 4 (137) pero sн encaja perfectamente aquн lo que Tucнdides nos dice de йl: Here, the "sн" serves to reinforce that `yes what Tucidides has to say about him fits perfectly here'. This kind of polarity marker is counted here as a modal Adjunct. Modality refers to probability and usuality in propositions, temporality includes typicality and time, and mood concerns expressions of obviousness, intensity, and degree. Comment Adjuncts, on the other hand, give the speaker's judgment about the whole of the proposition, are less tied to the verbal process, and thus are not integrated to the same degree into the mood structure of the clause. However, as comment and mood Adjuncts overlap, they are considered together here, under the heading of Modality. Thus, three types are counted here: Polarity, Temporality, and Modality. Systemic functional accounts of Theme (Halliday, 1994; Bloor and Bloor, 1995; Eggins, 1994; Downing and Locke, 1992, inter alia) all include modal Theme as a 230
category of interpersonal Theme. Interpersonal Theme makes up one of the three categories of general Theme types, the other two can also be expressed in terms of metafunction: textual and ideational (see sections 3.5.1.2 and 3.5.2.2 above). In addition to modal Themes under types of interpersonal Themes, there are two other types: vocative and moodmarking. Vocatives are direct addresses, such as a personal name, which are clause initial but are not part of the grammar of the clause (e.g. "John, why are you bothering me?"). There are no examples of vocatives in the corpus, so they do not enter into this discussion. Under mood-marking Themes, Halliday includes finite verbal operators, WH-interrogatives, and imperative "let's", when these are not preceded by another experiential element. As to finite verbal operators, Halliday (1994) distinguishes two categories: temporal and modal. It is unclear in which contexts these would appear in Theme position: presumably in questions in English. Questions are considered separately in this section on the interpersonal function in Theme and Rheme, and so are not included here with modal Themes. One reason for which they are considered separately is that in Spanish the finite can come first in declarative clauses, again, in more environments than it can in English, which, for grammatical reasons, would cause an imbalance in this comparative analysis if all of the clauses were analyzed for finite verbal operators as Theme. The focus of this section is modality, so only finite modal operators are considered here. Modal Adjuncts, one of the main realizations of interpersonal Theme, are of special interest in considerations of Theme as they are flexible as to their placement in the clause: writers can choose to place them at the beginning of the clause, at the end, or somewhere in between (e.g. after the Subject). According to Halliday (1994: 49), they tend to come at the beginning of the clause, as it is natural for speakers to thematize their own angle on the matter, to make this the point of departure. Many of the studies using Theme mentioned already in this study include analyses of interpersonal Themes; often the type and/or frequency is related to issues of genre (e.g. Ghadessy, 1995; Francis & Kramer-Dahl, 1992; Whittaker, 1995; inter alia). Given their importance in the literature on Theme, the frequency of modal Adjuncts has a place of special importance in this study. 6.1.3.2 Modality through Grammatical Metaphor In addition to modal Adjuncts, there are a myriad of other ways for writers to express opinions on or their degree of commitment to a proposition, or their willingness to admit other possible interpretations, sometimes perhaps hiding to some degree that an 231
opinion is what is being put forward or downplaying the fact that there are other possible interpretations. For example, Whittaker (1995) takes into account cases of extraposition where impersonal "it" appears in Subject position as a place marker for the extraposed clause, followed by an adjective which indicates the writer's judgment of the proposition encoded in the clause. However, as the clause is projected, as it were, through the impersonal "it", the writer's degree of commitment actually appears to be neutral, "dissimulating the fact that they are expressing their opinions" (Halliday, 1994: 355). This idea of dissimulation mirrors Berkhofer' s observation on normal history writing that "most (all?) of what is presented as (f)actuality is a special coding of the historians' synthetic expository texts, designed to conceal their highly constructed basis". (Berkhofer, 1989: 194). Thus, this type of extraposition may be a tool which history writers use to add to the apparent factuality of their narratives. An example of this type of clause from the English corpus is: 6-9. Eng 2 (103) It is possible that, while males in a group were hunting, women played a crucial part in the discovery and nurturing of early agriculture. Unlike in Whittaker's analysis, however, (cf. also Francis, 1990: 61: Martin, 1995), here the "it" clause is not considered as interpersonal Theme. There are two reasons for this analysis. First of all, the clause following the "it" clause is embedded; thus, if the "it" clause were considered in its entirety as an interpersonal Theme, the clause complex as a whole would then, in a sense, have no Rheme. Here, following Halliday (1994: 60), in English the dummy Subject "it" is considered the Theme of the clause if it is the first element (see Chapter 3, section 3.5.2). It has already been pointed out in section 3.5.1.1 that empty "it" is not a participant in the transitivity of the clause. Indeed, one motivation for extraposition may be end-focus (Downing and Locke, 1992): extraposition allows the writer to place in the Rheme information, e.g. in this case both the circumstance while males in a group were hunting and the whole proposition women played a crucial part in the discovery and nurturing of early agriculture, which may be the focus of the clause and which may have to do with the global concerns of the text. At the same time, the expression of modality, possible in the example above, through extraposition, appears in a more neutral point of the clause, neither in the Theme, which is one point of prominence in the clause, nor in the final position of the Rheme, which is another point of prominence in the clause (Halliday, 1994). The second reason for not considering the "it" clause as modal Theme is connected to the discussion above of apparent factuality in history writing: I wish to make a distinction here 232
between more overt interjections of modality through the thematic prominence of modal Adjuncts, and this more dissimulating means of modality. Therefore, while these constructions are not considered as interpersonal Theme in English, they have been taken into account below as one of the means of expressions of modality in the Rhemes of the clauses. Extraposition is also possible in Spanish (see sections 3.5.2.1.1 and 3.5.2.1.2 above) although the grammatical construction is somewhat different in that there is no empty subject pronoun: the morphology of the verb indicates the third personal singular in the case of impersonal expressions. 6-10. Spa 7 (126) Es indudable que algъn progreso existiу; <> In these cases, the finite is taken as ideational Theme, while the modalized adjective indudable appears in the Rheme. There are a few exceptions to this: cases of extraposition where an element of modality appears in the finite and there is no adjectival modification, for instance: 6-11. Spa 1 (112) No puede decirse que no exista la Historia en la Edad Media, <> It cannot be said that History does not exist in the Middle Ages. In Spanish, given that pro-drop takes place in impersonal expressions, the finite forms part of the Theme. In an account of modality, these could also be considered under finite modal operators, rather than in this section on grammatical metaphor through extraposition. There are only four in total of this type, so, overall, it does not really matter where they are considered in this account of modality. However, for the sake of consistency, they have been included here under grammatical modality in Theme position. There is a further caveat with regard to extraposed constructions. Halliday (1994: 266-267) distinguishes between non-modalized versions of these types of clauses (it is the case that...) and modalized versions ("it may be the case that...). This idea has been extended here to the type of impersonal clause illustrated above, which are both modalized. There are a few extraposed clauses in the corpus which are not modalized, e.g.: 6-12. Spa 5 (61) Se calcula que en Inglaterra... 6-13. Eng 7 (14) Now it is known that two civilizations preceded Hellenic Greece: the Minoan and the Mycenaean. In these statements, the authors present the process as reality, they do not conjecture that things might, could, or can be calculated or known, but that they are calculated or known: 233
it is a reality, i.e. there are no other interpretations. According to Downing and Locke (1992: 382, emphasis original), "modality is said to express a relation to reality, whereas an unmodalized declarative treats the process as reality". As this is what is happening in the two examples mentioned above, and in a few more instances in the corpus, they are not included in the account of modality. Another way in which writers can present their stance on a proposition in a seemingly neutral way is by introducing a nominalized modal structure by means of an existential structure (Whittaker, 1995). This can be seen in the following existential clauses (the Theme is in italics; the nominalized structure is underlined, and the modalized element is in brackets): 6-14. Eng 2 (130) but there is no (clear) evidence that women exercised political rule. 6-15. Spa 4 (110) y no hay indicio (alguno) que permita hablar de "una nueva y mбs elevada civilizaciуn traнda por el pueblo griego perteneciente al grupo indoeuropeo" Along these same lines, Whittaker (1995) also includes in her account of interpersonal Themes projecting clauses of mental and verbal processes. Tadros (1994: 70) states that "[t]he writer detaches him/herself from propositions by attributing them to others". Thus, with these constructions, writers can either distance themselves more from the content of the proposition while at the same time acknowledging the possibility of its having differing interpretations by encoding some third party as the Subject of the projecting clause, i.e. through attribution of the proposition to a Sayer external to the text. 6-16. Eng 2 (117) Some archaeologists even hold that some cities were formed not for the sake of local agriculture but as trading centers. or they can ally themselves more closely by using first person, e.g.: 6-17. Eng 2 (112) and we may assume that this craft was largely practiced by women. These are also considered in the analysis of interpersonal elements. The question in both cases, that of existential structures and that of verbal/mental projection, is whether to count the instance as occurring in the Theme or in the Rheme of the clause. In English, the case of existential constructions is clear cut: "there" functions much the same as the impersonal "it" mentioned above, as a place marker for ideational content which the author wishes to postpone until the Rheme. Therefore, in those cases, the interpersonal element is counted as occurring in the Rheme. The case of projection is more difficult, however. Rationale for considering the first ideational element in the projecting clause as Theme, e.g. some archaeologists in example 6-16 and we in 6-17, and the rest of the clause along with the 234
projected clause as Rheme was explained earlier in section 3.4.4. Therefore, for consistency, here the interpersonal commitment will be considered as part of the Rheme through the projecting verb. However, in Spanish, given pro-drop, often the projecting verb is thematic, e.g.: 6-18. Spa 5 (29) Consideraba que la poblaciуn aumentarнa en progresiуn geomйtrica... This could, then, make for a difference between the two corpora with respect to counts of modality in Theme position. We have seen with the experiential metafunction that it is difficult to decide when to draw the line in analyzing for grammatical metaphor (see section 4.4.3 above); with interpersonal elements, the same holds. The interpersonal is always present in the clause, and it is simply a matter of to what degree: if it is more or less overt. If one considers cases of extraposition, such as the modalized examples mentioned above, one then wonders whether cases such as the following, where the necessity of something in the first case, and the importance in the second, are predicated, should not be included: 6-19. Spa 7 (83) Por consiguiente un considerable incremento en la producciуn agropecuaria fue necesario para hacer frente al progreso demogrбfico, 6-20. Eng 1 (65)At the same time one of the most important aspects of the Western tradition is a general sense that certain great thinkers and artists, certain leaders, certain movements and events have been particularly important. Another type of case which could be considered is non-finite extraposed clauses, such as: 6-21. Spa 3 (152) y parecнa imposible poder reducir a la nobleza. However, as Whittaker (1995: 112) points out, while these constructions "express the author's evaluation, here they are clearly not comments on the proposition - they are the proposition". Therefore, neither these latter construction types nor the former are considered in the account of Modality. In sum, grammatical metaphor, like more congruent means of expression such as modal Adjuncts, allows authors to express their opinion on, indicate their commitment to, or acknowledge varying interpretations of a proposition. The types of realizations considered here under the heading of "Grammatical Modality" are: - modalized extraposed constructions - existential clauses with nominalized modal structures - mental and verbal projecting clauses 235
6.1.3.3 Finite Modal Operators As for finite modal operators, there are major differences between the two languages as to how these are encoded. English uses a wide variety of modal operators which are attached to the verbal groups as separate words. Spanish also does this to some extent, but it also employs inflections which indicate modality. For example, in English, we can say:
We may go. We might go. We could go. We can go Spanish would encode all of these meanings either with the modal operator "poder" or by using the conditional form, changing the inflection according to the degree of possibility:
Puede que vayamos. Podrнamos ir. Irнamos Podemos ir.
Carretero (1991:92) gives the following finite modal operators, among other modal marking types, as epistemic modal markers: - futuro imperfecto de indicativo: Marнa estarб jugando ahora. - expresiones modales: Marнa debe de estar jugando, Marнa tiene que estar jugando - ciertos verbos: Creo que Marнa esta jugando.
To these I have added the conditional suffix "нa". Thus, the total list of modal operators which appear in the English corpus is:
Figure 6.1: Modal Operators in the English Corpus
appear
may
shall
can
might
should
could
must
tend
had to
seem
will
would
And in the Spanish corpus:
Figure 6.2: Modal Operators in the Spanish corpus

hay que
-бn
necesita
- нa
poder
soler
deber
parecer
236
If there are two instances of modality, e.g. a finite modal operator and a projecting clause, as in the following example (the finite modal operator is underlined, the projection is in brackets): 6-22. Eng 2 (48) The evidence from some nonindustrial societies today may women provided most of the food supply through gathering, these are tabulated twice in this account of Modality: once under grammatical modality and again under finite modal operators. However, this only occurs in a very small handful of cases. 6.1.3.4 Modality in Theme 6.1.3.4.1 Modal Adjuncts in Theme It was hypothesized in section 6.1.2 that overt textual features expressing modality would be low in number across both corpora. This is indeed the case with respect to modal Adjuncts, as can be seen in Table 6.2:
Table 6.2: Modal Adjuncts in Theme
Text
Modal Themes Percentage Text
Eng 1
4
2.96% Spa 1
Eng 2
9
6.77% Spa 2
Eng 3
4
2.96% Spa 3
Eng 4
0
0.00% Spa 4
Eng 5
1
0.61% Spa 5
Eng 6
5
2.40% Spa 6
Eng 7
5
2.76% Spa 7
Eng 8
5
2.94% Spa 8
Eng 9
4
2.96% Spa 9
Eng 10
3
1.50% Spa 10
Total English
40
2.67% Total Spanish
Modal Themes Percentage
7
5.26%
6
4.03%
2
1.21%
9
5.88%
2
1.64%
4
2.80%
5
3.05%
1
3.57%
1
0.97%
2
1.94%
39
3.09%
The incidence for modal Adjuncts in Theme position is only slightly lower in English than in Spanish. In fact, there are greater differences amongst texts of the same language, especially in the Spanish corpus, rather than overall across the two. It was conjectured above that, when writing about more ancient civilizations and about prehistoric events, authors may feel the need to modulate declarations with a modal expression, and, if so, would perhaps thematize this. Indeed, the two chapters with the greatest percentage of modal Adjuncts in Theme position are Eng 2 and Spa 4, which both deal with ancient history: the "first civilizations" and ancient Greek civilization, respectively. However, this expectation does not hold for all of the texts on ancient history. Eng 5 (Prehistory), Eng 7
237
(Ancient Greece), Eng 8 (Ancient Rome), Spa 6 (Ancient Rome), and Spa 9 (Prehistory) do not show the relatively high percentages that these two texts do. Eng 7 and Eng 8 show slightly higher percentages than average in modal Themes, and Spa 6 is slightly lower than average. However, Eng 5 and Spa 9 each show only one instance of a modal Adjunct in Theme position (although it must be kept in mind that the mean for all of the texts is only 3.9). An explanation of the difference thus must go further than subject matter. For example, Spa 9, written for university students of history, is a highly specialized text in comparison to the others in the corpus. The section analyzed concerns the end of the Bronze Age in Spain, its characteristics and division into different cultural areas. The authors discuss the contentious nature of the delimitation of the cultural areas (as they also do of the delimitation of the different periods of the Bronze Age), and defend their propositions early on with the following declaration: El procedimiento seguido ha sido el de procurar delimitar cada una de las culturas particulares por la dispersiуn geogrбfica de sus elementos mбs caracterнsticos, valorando su secuencia cultural y las relaciones de unas culturas con otras. Esto supone una primera aproximaciуn a la delimitaciуn de las бreas culturales del Bronce Final, no basada en generalizaciones o en meras opiniones subjetivas. (Cerdб, et.al, 1989: 342) Here, it is clear that the authors are of the opinion that their interpretations are based on more than "subjective opinion" and "generalizations". In order to back up their statements about the different phases and cultural areas, they give extensive descriptions of tools which were used during that period. Given that the text is highly specialized, they can give space to this detailed evidence, thus providing more support for their interpretations and lessening the need for mitigation, or, at least, lessening the need to thematize the mitigation. Eng 5, on the other hand, is similar in topic to Eng 2: both deal with the first known civilizations of man in their prehistoric period. However, they are at opposite ends of the scale in terms of number of modal Adjuncts, with Eng 2 having the most and Eng 5 having the least (after Eng 4). Thus, the conjecture that subject matter might be a determining factor in amount of modal Adjuncts in Theme position does not hold entirely. The other factor that may be at work here is author style. Eng 4, with no modal Adjuncts in Theme position, and Eng 5 were written by the same authors, and they are the texts having the fewest thematic modal Adjuncts. Eng 4 is a textbook preface, a place where presumably authors could insert their opinion with less restraint perhaps than in the textbook proper. 238
Yet Spa 8 is also a textbook preface, and it is higher than average in percentage of modal
Adjuncts. One of the problems here is that both prefaces are very short. Spa 8 has only 1
modal Adjunct in Theme position, yet given that it consists of only 28 clauses complexes,
the resulting percentage is high.
In order to make discussion of this section clearer, the modal Theme results in
percentages are reproduced below in Table 6.3 along with the content matter of the chapter
and the author(s). They are displayed in descending order, from the text having the greatest
number of modal Adjuncts to that having the least.
Table 6.3: Modal Theme Percentages with Subject and Author
Text ref. Percent Subject
Author
Eng 2
6.02% Prehistory
Chambers, et.al
Spa 4
5.88% Ancient Greece
Mangas
Spa 1
5.26% Introduction to the study of history
Ballesteros & Alborg
Spa 2
4.03% 19th century population
Fernбndez
Spa 8
3.57% Preface
Fernбndez
Spa 7
3.05% 18th century
Domнnguez
Eng 3
2.96% 14th & 15th centuries
Chambers, et.al
Eng 9
2.96% 19th century population
McKay, et.al
Eng 1
2.96% Introduction to the study of history
Chambers, et.al
Eng 8
2.94% Ancient Rome
Roy
Spa 6
2.80% Ancient Rome
Santos
Eng 7
2.76% Ancient Greece
Perry
Eng 6
2.40% 13th & 14th centuries
McKay, et.al
Spa 10
1.94% 18th century
Palacio, et.al
Spa 5
1.64% 14th & 15th centuries
Palacio, et.al
Eng 10
1.50% 18th century
Brinton, et.al
Spa 3
1.21% 12th cent France & 12th & 13th cent Spain Ballesteros & Alborg
Spa 9
0.97% Prehistory
Cerdб, et.al
Eng 5
0.61% Prehistory
Greer & Lewis
Eng 4
0.00% Preface
Greer & Lewis
Furthering the line of investigation with regard to author preference, Eng 1, Eng 2, and Eng 3 were all written by the same group of authors, and are, indeed, from the same book. While Eng 1 and Eng 3 are identical in their use of modal Themes, Eng 2 uses more than double this number. The greater use in Eng 2 has already been attributed possibly to subject matter (prehistory). Eng 1 is an introduction to the study of history: as with a preface to a book, again here it is possible to conjecture that authors may feel less restraint in using modal Themes. Indeed, Spa 1, also an introduction to the study of history, has one
239
of the highest percentages in the Spanish corpus, while Eng 1 is higher than average, although not as high as Spa 1. It is interesting to note that the highest averages in the English corpus belong to Eng 2, Eng 1, and Eng 3 (along with Eng 9), which could indicate that this particular group of authors have a tendency to thematize modal Adjuncts. This could also be the case for Eng 9 and Eng 6, also written by a same set of authors, while, at the same time, these two texts are also more similar in subject matter: both of these texts involve discussion of changes in population size; Eng 6 deals with the 13th and 14th centuries, and Eng 9 with the 19th century. On the other hand, Spa 1 and Spa 3 are written by the same authors, yet they are at the opposite ends of the scale in terms of percentage of modal Adjuncts in Theme position. The differences here may be attributable again to the difference in subject matter: Spa 1 is an introduction to the study of history, and, as already mentioned, the authors may feel less restraint here in thematizing modality. Spa 3 has as its Subject 12th and 13th century France and Spain. There is a lot more narrative here than in the other texts, as evidenced by the greater number of Actors, especially specific individuals, (see sections 4.6.2 and 5.5 above). The authors, then, stick very closely to widely accepted accounts of "facts", or historical events, to what happened, rather than to interpretation. Another set of texts written by same authors is comprised by Spa 2 and Spa 8. Here, to the contrary of the previous set of texts, the amount of modality is similar (relatively high), although the content matter of the chapters is highly different: Spa 2 concerns the 19th century and Spa 8 is a textbook preface. While the case of the latter has been discussed above, where it was mentioned that the high percentage could be misleading, as there is only one instance in a relatively short text, Spa 2 goes counter to the suggestion that chapters dealing with more recent history might thematize modality less than chapters about ancient history. Therefore, perhaps the similarity between Spa 2 and Spa 8 can be explained by the fact that they were written by the same authors. Spa 5 and Spa 10 were also written by a same set of authors, and are similar in their use of modality in Theme position. They both also deal with relatively more recent periods of history, 14th and 5th centuries in Spa 5 and the 18th century in Spa 10. In sum, it is difficult to consistently explain the differences amongst the texts in terms of modal Adjuncts in Theme position. It seems that at times the content may have an influence, and at other times it may simply be a matter of stylistic preferences of the authors. Obviously, more data would be needed, both in terms of same authors and same content, in 240
order to reach a more tenable conclusion. Still, the most interesting finding for purposes of this study is the relative similarity overall across the two languages in the infrequent use of modal Adjuncts. It is also interesting to compare the findings in this section to those of Taylor in the history textbooks he looked at, where he found that modal Adjuncts were thematized in only 0.63% of the clauses. The textbooks he looked at were from Australia and were targeted at high school year 7 students, an age level somewhat below the targeted age level of the textbooks in the present study. The percentage of modal Adjuncts in the present study is also low, yet perhaps it could be conjectured that the higher age of the targeted audience puts some pressure on the writers to open up slightly more vistas of other possible interpretations, as the asymmetrical gap between authors and readers is ever so slightly smaller than in the history textbooks in the Taylor study. This is pure conjecture, however, and more studies which measured the author/reader gap and the use of modal Adjuncts would be of interest to uphold or refute the conjecture. 6.1.3.4.2 Other Types of Modality in Theme Position As mentioned above, in addition to modal Adjuncts, taken into account in this analysis of writer intervention through the interpersonal function are extraposition, existential constructions with nominalized modal elements, and impersonal and personal clause complexes involving projection (all of which are considered under the heading of Grammatical Metaphor), as well as finite modal operators. Rather than consider each of these categories separately here, they are combined along with modal Adjuncts in order to present a full picture of the interpersonal in Theme position for a more coherent discussion. For each category, the number of instances and the percentage of occurrence is given. 241
Table 6.4: Total Modality in Theme Position, English Corpus
Mod. Adjuncts Gram Mod
FMO'S
Total Mod. in Theme
Text
#
% #%
#
%
#
%
Eng 1
4 2.96% 0 0.00% 2 1.48%
6
4.44%
Eng 2
9 6.77% 0 0.00% 0 0.00%
9
6.77%
Eng 3
4 2.96% 0 0.00% 0 0.00%
4
2.96%
Eng 4
0 0.00% 0 0.00% 0 0.00%
0
0.00%
Eng 5
1 0.61% 0 0.00% 0 0.00%
1
0.61%
Eng 6
5 2.40% 0 0.00% 0 0.00%
5
2.40%
Eng 7
5 2.76% 0 0.00% 0 0.00%
5
2.76%
Eng 8
5 2.94% 0 0.00% 1 0.59%
6
3.53%
Eng 9
4 2.96% 0 0.00% 0 0.00%
4
2.96%
Eng 10
3 1.50% 0 0.00% 0 0.00%
3
1.50%
Total English 40 2.67% 0 0.00% 3 0.20% 43
2.87%
Key: Mod. Adjuncts = modal Adjuncts; Gram. Mod = grammatical modification; FMO's= finite modal operators; Total Mod. in Theme = Total modality in Theme position
Table 6.5: Total Modality in Theme Position, Spanish Corpus
Mod. Themes Gram Mod
FMO'S Total Mod. in Theme
Text
#
%
#
%
#
%
#
%
Spa 1
7 5.26%
3
2.26%
6 4.51%
16 12.03%
Spa 2
6 4.03%
0
0.00%
1 0.67%
7 4.70%
Spa 3
2 1.21%
0
0.00%
2 1.21%
4 2.42%
Spa 4
9 5.88%
1
0.65%
2 1.31%
12 7.84%
Spa 5
2 1.64%
1
0.82%
0 0.00%
3 2.46%
Spa 6
4 2.72%
1
0.68%
1 0.70%
6 4.20%
Spa 7
5 3.05%
1
0.61%
1 0.61%
7 4.27%
Spa 8
1 3.57%
0
0.00%
1 3.57%
2 7.14%
Spa 9
1 0.97%
1
0.97%
4 3.88%
6 5.83%
Spa 10
2 1.94%
0
0.00%
1 0.97%
3 2.91%
Total Spanish 39 3.09%
8 0.63% 19 1.50%
66 5.23%
Key: Mod. Adjuncts = modal Adjuncts; Gram. Mod = grammatical modification; FMO's= finite modal operators; Total Mod. in Theme = Total modality in Theme position
The Spanish corpus now shows almost double the amount of modality in Theme position that the English corpus shows. This is not surprising, however, taking into account the fact that Spanish allows for verb-initial constructions in more environments than does English. We have seen in example 6-18 above that, given the pro-drop nature of Spanish, the verbal process itself may be thematic in cases of projection, thus meaning that the instance of interpersonal interpolation occurs in Theme position. Also, in the case of finite modal operators, the verb-initial constructions occur for several different reasons. In addition to elision in the case of an already mentioned Subject, another reason is inversion, where the Subject follows the verb. This allows the writer to place the Subject in the Rheme, in the same way as the impersonal "it" place marker allows this to occur in English, for reasons 242
which may be related to notions of end-weight or end-focus, style (see section 3.1.3) or thematic patterning (see section 4.6.4), as exemplified by (the finite modal operator is underlined): 6-23. Spa 8 (15) Pero no debe aplastarnos el peso de estos fondos ingentes. Also, impersonal expressions, as already mentioned, are subjectless in Spanish, as the third person singular is included in the morphology of the verb thus again placing the finite modal operator (underlined) in Theme position:
6-24. Spa 7 (148) habrнa que esperar todavнa un siglo para que se pudiera disponer de abonos sintйticos y de maquinaria adecuada;
The instances in English are limited to the following type of construction:
6-25. Eng 8 (153)and would sit in the Temple of Castor and Pollux, between the statues of the divine twins, "offering himself for the adoration of visitors to the Temple."
where due to the coordinated clause complex, the Subject may be suppressed.
While the differences between the two corpora are attributable, then, to grammatical
differences between the two languages, it is interesting to note what happens with the
individual results as a consequence of adding on these expressions of modality to the
previous category of modal Adjuncts on their own. While the English texts do not show
much of a difference, as already explained, the Spanish texts, especially Spa 1, show
interesting differences as a result of adding on the other categories, as can be seen in Table
6.6:
Table 6.6: Differences in Percentages: Modal Adjuncts in Theme/Total Modality in
Theme
Text ref. Percent Percent Subject M.A. Tot.
Author
Spa 1
5.26% 12.03% Introduction to the study of history
Ballesteros & Alborg
Spa 4
5.88% 7.84% Ancient Greece
Mangas
Spa 8
3.57% 7.14% Preface
Fernбndez
Spa 9
0.97% 5.83% Prehistory
Cerdб, et.al
Spa 2
4.03% 4.70% 19th century population
Fernбndez
Spa 7
3.05% 4.27% 18th century
Domнnguez
Spa 6
2.80% 4.20% Ancient Rome
Santos
Spa 10
1.94% 2.91% 18th century
Palacio, et.al
Spa 5
1.64% 2.46% 14th & 15th centuries
Palacio, et.al
Spa 3
1.21% 2.42% 12th cent France & 12th & 13th cent Spain Ballesteros & Alborg
Spa 1 has now jumped way to the forefront in terms of interpersonal elements in the Themes of the clauses. This is mainly due to finite modal operators: these are thematized in 6 clauses, which is a greater number than in any of the other texts in the Spanish corpus (the 243
other being Spa 9, which thematizes them in 4 instances). In Spa 1, an introduction to the study of history, the authors often use future or conditional tenses to speak of what will happen or what would happen if history were interpreted one way or another; they also speak of what historians need to take into account in their interpretations of history. An extract from this chapter illustrates this (the modalized verbs are underlined; when they are thematic, they are also in italics, as all Themes in this study): ...(80)Sin embargo, el concepto "bosque" es en sн imprescindible (81)y necesita ser considerado, pero despuйs de investigar las peculiaridades referentes a los бrboles en concreto. (82)Estб es la labor de la sнntesis. (83)La sнntesis no es, pues, como creen muchos, "extracto", sinopsis seca, sino el intento de captar el concepto "bosque". (84)Un estudio de todos los sabios de los siglos XV y XVI en Europa serнa muy provechoso, monogrбficamente hecho, (85)pero dejarнa sin especificar el movimiento renacentista en sн, las tendencias y razones que los movieron a actuar a todos en un sentido determinado y similar. Here, the authors first state the importance of being able to see the complete picture, or the whole "wood". They then go on to say that this must be considered; the verb is thematized here due to pro-drop. They later go on in clauses 84 and 85 to explain that a study of one "tree", that of all of the wise people from the 15th and 16th centuries would be very advantageous, but would leave unspecified the renaissance movement itself; again, in this latter case the verb is thematic through pro-drop. The introduction to the study of history chapters (also Eng 1, as will be seen when looking at modality in the Rhemes of the clauses) are those which are least tied to "facts", to a narration of real events. This goes some way in explaining the greater use of modalized expressions. In adding thematized finite modal operators and instances of grammatical metaphor to the discussion, the Spanish corpus now shows an interesting cluster of topics towards the top of the list, where expressions of modality are greater. It was speculated earlier that chapters on prehistory, ancient history, prefaces, and introductory chapters would show more hedging, and the results here seem to bear this out. The greatest move upward in this sense has been Spa 9, which, as already explained, is a highly specialized text, and includes lengthy descriptions of tools from prehistoric periods. It was the lowest on the list in use of modal Adjuncts, yet it jumps up to position number 4 in the list of texts from the Spanish corpus when adding on finite modal operators and grammatical metaphor. They use the expression "pueden considerarse" (`they can be considered') in 3 instances in theme position (Subject elision through pro-drop places the verb in thematic position). This 244
expression is used to sum up discussions of tools and to tie them to interpretations of their importance in shaping culture or in defining the different cultural areas, or to introduce new considerations of divisions of cultural areas. These forms of modality are much more subtle than modal Adjuncts; thus, while allowing a perhaps necessary insertion of modalized expression, in that they are dealing with prehistory, about which concrete documentary evidence is not abundant, at the same time they allow the text to achieve its objective demeanor, which, as was seen in the quote by the authors from the foreword in the previous section, is an important goal of the text. Thus, for the grammatical reasons mentioned above, the Spanish corpus allows for greater instances of modality in Theme position than does English. Given that these reasons are grammatical, and especially given that in many of the cases in Spanish there is no option but to apply pro-drop, i.e. in impersonal expressions and in unmarked cases of Subject elision, the cross-cultural generic constraints argued for in this study may be in place, simply overridden by grammatical forces. In order to balance out this grammatical difference, the Rhemes were examined for instances of modality using the same methodology as in the case of the Themes. The results of this analysis follow. Again here special mention is given to modal Adjuncts, due to their flexibility of position. 6.1.3.5 Modality in Rheme As were the Themes, all of the Rhemes in the corpus were analyzed for the same grammatical realizations of modality: modal Adjuncts, grammatical metaphor, and finite modal operators. Modal Adjuncts are considered first, and their instances are compared with modal Adjuncts in Theme position. Then a full account of modality in Rheme is considered by combining those results with results from grammatical metaphor and finite modal operator. 6.1.3.5.1 Modal Adjuncts in Rheme Position Table 6.7 gives the total instances and percentages of Modal Adjuncts in Rheme position: 245
Table 6.7: Modal Adjuncts in Rheme
Text
MA's in Rheme Percent Text
Eng 1
17 12.59% Spa 1
Eng 2
9 6.77% Spa 2
Eng 3
13 9.63% Spa 3
Eng 4
2 5.88% Spa 4
Eng 5
17 10.30% Spa 5
Eng 6
8 3.85% Spa 6
Eng 7
4 2.21% Spa 7
Eng 8
18 10.59% Spa 8
Eng 9
13 9.63% Spa 9
Eng 10
19 9.50% Spa 10
Total English
120 8.02% Total Spanish
MA's in Rhemes Percent 13 9.77% 9 6.04% 6 3.64% 16 10.46% 2 1.64% 7 4.90% 11 6.71% 1 3.57% 12 11.65% 2 1.94% 79 6.25%
The first striking finding is that there are many more modal Adjuncts in Rheme position than in Theme, almost triple the percentage in the English corpus, and slightly more than double in the Spanish corpus. This is contrary to Halliday's claim, mentioned above in section 6.1.3.1, that speakers tend to thematize modal Adjuncts. Also, while there are other studies, as we have seen, which show frequency of modal Adjuncts in Theme, I know of none which looks at modal Adjuncts in Rheme. Thus, it is difficult to know whether the results here, with a greater number in the Rhemes of the clauses, are significant in that perhaps authors prefer not to include such an overt type of interpersonal expression under thematic prominence, and therefore prefer to include them in non-prominent positions in the Rheme. Fries (1983) gives an example of thematized perhaps in a clause which shows the obvious vulnerability of a statement an author makes in a text. He points out that placing perhaps later in the clause makes the possibility, the "perhapsness" of the statement seem less important and more tentative, while the statement "calls for the author to acknowledge that vulnerability in a clear and obvious way" (ibid: 138). Also, while modal Adjuncts could feasibly be placed in clause final position, which is the place of prominence normally reserved for the most newsworthy information of the clause, it is interesting to note that in this corpus there are no modal Adjuncts in clause final position. Whether this is a feature of written language, or it is a generic feature attributable to a desire to de-emphasize modality, would need bearing out in further study. With regard to the results in Table 6.7, as with modal Theme results displayed in Table 6.2, there are great differences amongst texts of the same language. Some of the texts in the English corpus are remarkable in the difference between modal Adjuncts in 246
Theme and in Rheme. For example, Eng 5 had a very low percentage of modal Adjuncts per Theme: 0.61%, while the Rheme percentage is much higher: 10.30%. This would perhaps provide evidence for the notion that authors of history textbooks prefer not to give thematic prominence to modal Adjuncts. Also with this higher frequency of modal Adjuncts in Eng 5, we can return to the conjecture made above with modal Themes: perhaps chapter content does have an effect on the amount of modality. It was speculated that chapters on more remote times from the present, given the relative lack of concrete documentation available as is the case with later periods, would present a greater amount of modality. Looking first at the three chapters in the corpus on pre-history (Eng 2, Eng 5 and Spa 9), it is interesting to note that Eng 2 had the highest percentage of modal Adjuncts in Theme position (6.77%), and it maintains this exact same percentage of modal Adjuncts in the Rheme. Eng 5 and Spa 9 both had very low counts of modal Adjuncts in Theme (0.61% and 0.97% respectively). Yet, in the case of modal Adjuncts in Rheme, they have 10.30% and 11.65%, respectively, which represent some of the highest percentages in that category. It seems to be more of interest, then, to look at these two categories of modal Adjuncts together to check this content-based conjecture. As with Table 6.3 above, the percentages (total number of modal Adjuncts divided by the number of clauses) appear in Table 6.8 below, along with the subject and author(s) of the text. 247
Table 6.8: Total Modal Adjunct (Themes and Rhemes) Percentages with Subject and Author
Text ref. M.A's Subject
Author
Spa 4
16.34% Ancient Greece
Mangas
Eng 1
15.56% Introduction to the study of history
Chambers, et.al
Spa 1
15.04% Introduction to the study of history
Ballesteros & Alborg
Eng 2
13.53% Prehistory
Chambers, et.al
Eng 8
12.94% Ancient Rome
Roy
Spa 9
12.62% Prehistory
Cerdб, et.al
Eng 3
12.59% 14th & 15th centuries
Chambers, et.al
Eng 9
12.59% 19th century population
McKay, et.al
Eng 10 11.00% 18th century
Brinton, et.al
Eng 5
10.91% Prehistory
Greer & Lewis
Spa 2
10.07% 19th century population
Fernбndez
Spa 7
9.76% 18th century
Domнnguez
Spa 6
7.69% Ancient Rome
Santos
Spa 8
7.14% Preface
Fernбndez
Eng 6
6.25% 13th & 14th centuries
McKay, et.al
Eng 4
5.88% Preface
Greer & Lewis
Eng 7
4.97% Ancient Greece
Perry
Spa 3
4.85% 12th cent France & 12th & 13th cent Spain Ballesteros & Alborg
Spa 10
3.88% 18th century
Palacio, et.al
Spa 5
3.28% 19th century
Palacio, et.al
In Table 6.8, the configuration is different from that of Table 6.3. In Table 6.8 there are more texts of the type predicted in the top positions, which are the positions reflecting the greatest amount of modality. Texts on prehistory, ancient history, and on the introduction to the study of history now occupy the top 6 slots, while in Table 6.3 they occupied the top 3 slots only. However, there are still texts of these types which are low in the table: Spa 6 on ancient Rome and Eng 7 on Ancient Greece, while Eng 5 on prehistory is in the middle of the chart. Again, author style may be in effect in some of these cases, although analysis of more texts would be needed to bear this out. With the Themes of the clauses, other types of modality have been analyzed above. We now turn to an analysis of other types of modality in the Rhemes of the clauses.
6.1.3.5.2 Other Types of Modality in Rheme
Here, as in the case of modality in Theme, the three different types of modality, modal Adjuncts, grammatical modality, and finite modal operators, are considered together.
248
The frequency in numbers and percentages of their appearance in the Rhemes of the clauses
appear in Table 6.9 for the English corpus and Table 6.10 for the Spanish corpus:
Table 6.9: Total Modality in Rheme, English Corpus
M.A's in Rheme
Gram Mod
FMO'S Tot. Mod. in Rheme
Text
#
%
#
%
#
%
#
%
Eng 1
17 12.59% 0
0.00% 21 15.56% 38
28.15%
Eng 2
9 6.77% 14 10.53% 27 20.30% 50
37.59%
Eng 3
13 9.63% 5
3.70% 7 5.19% 25
18.52%
Eng 4
2 5.88% 0
0.00% 1 2.94%
3
8.82%
Eng 5
17 10.30% 8
4.85% 10 6.06% 35
21.21%
Eng 6
8 3.85% 12
5.77% 5 2.40% 25
12.02%
Eng 7
4 2.21% 8
4.42% 0 0.00% 12
6.63%
Eng 8
17 10.00% 9
5.29% 9 5.29% 35
20.59%
Eng 9
13 9.63% 0
0.00% 0 0.00% 13
9.63%
Eng 10
19 9.50% 2
1.00% 3 1.50% 24
12.00%
Total English 119 7.95% 58
3.88% 83 5.55% 260
17.38%
Table 6.10: Total Modality in Rheme, Spanish Corpus
M.A.'s in Rheme
Gram Mod
FMO'S
Text
#
%
#
%
#%
Spa 1
13 9.77% 5
3.76% 15 11.28%
Spa 2
9 6.04% 3
2.01% 5 3.36%
Spa 3
6 3.64% 2
1.21% 11 6.67%
Spa 4
16 10.46% 13
8.50% 10 6.54%
Spa 5
2 1.64% 3
2.46% 5 4.10%
Spa 6
7 4.90% 1
0.70% 5 3.50%
Spa 7
11 6.71% 9
5.49% 4 2.44%
Spa 8
1 3.57% 2
7.14% 6 21.43%
Spa 9
12 11.65% 2
1.94% 9 8.74%
Spa 10
2 1.94% 2
1.94% 6 5.83%
Total Spanish 79 6.25% 42
3.33% 76 6.02%
Total Mod. in Rheme
#
%
33
24.81%
17
11.41%
19
11.52%
39
25.49%
10
8.20%
13
8.84%
24
14.63%
9
32.14%
23
22.33%
10
9.71%
197
15.60%
There is a slight difference here between the English and Spanish corpus in terms of total modality in Rheme, which complements the greater modality that appears in Spanish in Theme. In fact, if the totals are added up for each of the corpora for total modality in Theme and total modality in Rheme, the results are: for the English corpus, 20.25% of the clauses contain some textual feature of modality (a modal Adjunct, a finite modal operator, or one of the types of grammatical metaphor), and for the Spanish corpus, the result is 20.82%. Hence, the difference between the two corpora in terms of the greater amount of modality in Theme position in the Spanish corpus can be explained by grammatical constraints. Thus, this finding supports the generic hypothesis which underlies this study.
249
As with the other categories in this chapter (and indeed throughout the thesis) there
are major differences within the two corpora between the different chapters in terms of total
modality in Rheme, as can be seen in Table 6.9 and Table 6.10 above. Rather than look at
the implications of that in terms of chapter topic and authors based solely on modality in the
Rheme, I have combined all of the instances of modality in the Themes and the Rhemes of
the clauses. Thus, throughout this chapter, a build-up towards a complete picture of
modality has been taking place. The results of all of the categories of modality in the
Themes and the Rhemes of the clauses in terms of frequency in percentages are laid out in
Table 6.11. Again, the percentages are given in descending order, from most at the top of
the table and least at the bottom:
Table 6.11: Total Modality Percentages with Subject and Author
Text ref. Percentage Subject
Author
Eng 2
44.36% Prehistory
Chambers, et.al
Spa 8
39.29% Preface
Fernбndez
Spa 1
36.84% Introduction to the study of history
Ballesteros & Alborg
Spa 4
32.68% Ancient Greece
Mangas
Eng 1
32.59% Introduction to the study of history
Chambers, et.al
Spa 9
28.16% Prehistory
Cerdб, et.al
Eng 8
24.12% Ancient Rome
Roy
Eng 5
21.82% Prehistory
Greer & Lewis
Eng 3
21.48% 14th & 15th centuries
Chambers, et.al
Spa 7
18.90% 18th century
Domнnguez
Spa 2
16.11% 19th century population
Fernбndez
Eng 6
14.42% 13th & 14th centuries
McKay, et.al
Spa 3
13.94% 12th cent France & 12th & 13th cent Spain Ballesteros & Alborg
Eng 10
13.50% 18th century
Brinton, et.al
Spa 6
13.29% Ancient Rome
Santos
Spa 10
12.62% 18th century
Palacio, et.al
Eng 9
12.59% 19th century population
McKay, et.al
Spa 5
11.48% 19th century
Palacio, et.al
Eng 7
9.39% Ancient Greece
Perry
Eng 4
8.82% Preface
Greer & Lewis
In the first check, Table 6.3, which looked at modal Adjuncts in Theme position, 3 of the top slots were taken by those texts which were conjectured to include more modality: ancient history, prehistory, and introductory chapters. In the second check, Table 6.8, which looked at modal Adjuncts in Theme and Rheme position, the first 6 slots were occupied by these texts. In this third check (Table 6.11), we can see that the first 8 slots are taken up by
250
these texts. In other words, as we have added on the categories of modality, several of these texts have moved up to positions of showing a greater number of textual features of modality. This would seem to indicate that Theme position is not at all the favored position for expressing modality. Eng 2 is the text which is most consistent in that it has the highest amount of overall modality, and it had the highest amount of modal Adjuncts as Theme as well, thus, perhaps, manifesting more of a willingness to show tentativeness and openness to other possible interpretations of history. Eng 2 is a text which moves from making non- modalized declarative statements about what is commonly accepted about prehistoric periods, e.g.: (3)Human beings began to abandon a nomadic existence and live in settled agricultural communities only about 10,000 years ago. (4)The earliest farming villages so far identified appeared in the hills of Asia Minor. (5)Some 5000 years later, settlements grew up along the banks of the Tigris, Euphrates, Nile, and Indus. (6)The lands irrigated by these rivers supported many more people than could the highlands (7)and did not require the effort of the whole population to guarantee a stable supply of food. (8)This margin of freedom and leisure made possible a diversity of occupations, specialization of skills, and experiments in building, thought, law, government, and military organization. (9)Permanence, complexity, and sophisticated social organization transformed the agricultural village into something we can recognize as urban, the form of habitation crucial for the development of civilization. to making modalized conjectures in interpreting the way things may have been, e.g.: (47)Men and women must have shared most of the tasks in the earliest human societies. (48)The evidence from some nonindustrial societies today may suggest that women provided most of the food supply through gathering, even performing at the same time another task, that of nursing children. (49)Thus, by maintaining a stable food supply, and obviously by caring for the new generation of gatherers, women may have been the dominant members in the economies of such gathering bands. (50)If there was a division of labor by gender, we might guess that the men did most of the hunting and killing of animals, though this can hardly be proved. (51)When a large animal was finally killed, it provided food and thus relief from gathering for several days. (52)The trapping and killing of an animal must have been less common than the daily gathering of food. (53)If the model we are constructing is valid, the more specialized work of hunting and killing may have enabled men to claim for themselves a dominant social position. The conditional statement in clause 53 further serves to underscore the interpretive nature of the speculations, and shows that the authors are comfortable with making the conjectural nature of their interpretations obvious to the reader. This greater amount of modality in Eng 251
2 can be contrasted with Eng 7, a chapter on Ancient Greece. We have seen in section 5.5.1 that Eng 7 contains long constant topic chains of narration and description; for this reason, it seems to be a text with less interpretation than, for example, Eng 2, thus perhaps meaning that the authors feel less of a need to modalize their propositions. As for Eng 2, it is also of interest to note that the introductory chapter by the same authors also contains a high frequency of modalized features, yet their text on the 14th and 15th centuries, Eng 3, is much lower with 21.48%. This finding could confirm the supposition that more remote content historically, and something like an introductory chapter, would contain more modality than would a chapter on more recent history, for which there is more concrete evidence. Interviews with the authors would help to shed more light on this finding. Another interesting set of texts to compare are Eng 7 and Spa 4, both on ancient Greece, but at opposite ends of the scale in terms of the modalized features analyzed here. Eng 7, with only 9.39%, as has already been explained, is a text with a lot of description and narration of events, thus it is low in interpretation and analysis, which may explain the low amount of modality seen in Table 6.11. Spa 4, on the other hand, is much more analytical, evidenced by the fact that the word "tesis" is repeated 7 times; that is, the authors look at different theses, or hypotheses, put forward by different historians and, by producing evidence in favor or against, weigh their validity. Thus, they, like the authors of Eng 2, also portray history as allowing more than one interpretation, as evidenced by the following extract, in which we see the controversy over dating stone tablets from Cnosos (modalized elements are underlined): (97)La tesis generalmente admitida sostenнa que las tablillas de Cnosos pertenecнan a fines del siglo XV, йpoca de la destrucciуn del Palacio y que las de Pilos serнan unos dos siglos posteriores. (98)Palmer propone una tesis que se va generalizando cada dнa mбs, y que baja en dos siglos la fecha de las tablillas de Cnosos. (99)Este dato evidentemente complica la exposiciуn histуrica que se daba sobre la base de la diferente fecha de las tablillas. Thus, to some extent the differences in frequency of modalized elements can be explained by the subject matter of the text, or, perhaps more accurately, by how the subject matter is viewed, whether it is something open to interpretation or whether it is presented as a "true" account of reality, thus linking subject matter and author style, i.e. different authors may view the field differently and thus make textual choices which portray the field differently. But, as already stated, this would need more corroboration in further studies of same authors, on the one hand, and same subject matter on the author, along with 252
corroboration from interviews with the authors. At any rate, one set of texts seems to go against this trend: the two prefaces, Spa 8 with a high count of 39.29%, and Eng 4 with the lowest amount at 8.82%. Motivation for this possible difference will be left for the next section on reader/writer interaction, the second type of interpersonal element under analysis here. As a final note to this section, the first two hypotheses put forward in section 6.1.2 have been borne out by the results to some extent: modality in the Themes of the clauses is certainly low, as predicted by the first hypothesis. The second hypothesis predicted greater amount of modality in the Rhemes, and this has been borne out, perhaps not surprisingly, given that the more grammatical means of modality tend to appear in the verbal processes of the clauses, which are normally rhematic. At the same time, the second hypothesis receives support from the greater frequency of modal Adjuncts in the Rhemes of the clauses as opposed to the Theme. Still, at any rate, the average amount of modality even in the Rhemes is low, although more studies of this type from other genres are needed in order to bring to light whether this finding has any generic significance. 6.1.4 Reader/Writer Interaction The second type of writer intervention measured here is reader/writer interaction. This includes direct addresses to the reader and questions, as these overtly include the reader in the text. It also includes specific references to the authors, as by this means authors intervene directly in the text, and show an overt alignment with the propositions they make. This is in accord with Berry's (1995: 64) view; in her study of thematic success in school writing, she regarded a Theme as interactional if it "included a word/phrase which referred to to [sic] the writer or reader(s) of the passage, or to a group of people which included the writer or reader(s)". In this category, all instances of mentions of the reader and writer are tabulated, including those which appear in rankshifted clauses. 6.1.4.1 Reader Reference Reader reference includes all instances of first person plural pronouns in which the reader is obviously included, any second person pronouns, and mentions of "the reader", which also includes "students", as the textbooks are directed towards students. With the case of first person plural pronouns, often the instance is inclusive of the reader and the writer, and at other times it is exclusive, i.e. writer only. Obviously, in this category which 253
looks at reader reference in Theme only instances of the former are counted. With regard to referring to the reader, in this corpus there are no examples of second person pronouns. The reader is always referred to in third person, except indirectly in one case in each of the corpora where an imperative is used. This serves to underscore the distancing of the author with the addressee, and would also seem to move the texts farther towards the information end of an informational/interaction axis. Examples of reader references in Theme position are:
6-26. Spa 4 (95) las que nos han llegado corresponden, por tanto, a un sуlo momento 6-27. Eng 8 (103) We can still see its small painted rooms set around a little courtyard, 6-28. Eng 1 (48) Readers of the fifth edition of The Western Experience will find themselves using this book in all these ways: to learn history as a subject in itself, to enrich their culture, and to test their ideas of social behavior, And in Rheme: 6-29. Spa 7 (14) Partiendo de la cifra, poco segura, de 180-190 en 1800 nos hallamos ante un crecimiento del 70-75%, 6-30. Eng 7 (76) In Homer, we also see the origin of the Greek ideal of aretй, excellence. 6-31. Spa 8 (8) En tal supuesto no debe recluirse el estudiante en una posiciуn repetitiva,
As hypothesized in section 6.1.2, low reader inclusion is expected across both corpora. The
exception for this could be the two prefaces, as those are aimed directly at readers. The
frequency of reader mention in Theme can be seen in Table 6.12, and for mentions in the
Rheme in Table 6.13:
Table 6.12: Reader References in Theme
Text: English RRinT Percent Text: Spanish RRinT Percent
Eng 1
19 14.07% Spa 1
3 2.26%
Eng 2
3 2.26% Spa 2
1 0.67%
Eng 3
0 0.00% Spa 3
0 0.00%
Eng 4
0 0.00% Spa 4
5 3.27%
Eng 5
5 3.03% Spa 5
0 0.00%
Eng 6
2 0.96% Spa 6
1 0.70%
Eng 7
0 0.00% Spa 7
1 0.61%
Eng 8
1 0.59% Spa 8
1 3.57%
Eng 9
0 0.00% Spa 9
1 0.97%
Eng 10
1 0.50% Spa 10
1 0.97%
Total English
31 2.07% Total Spanish
14 1.11%
Key: RR in T = reader references in Theme
254
Table 6.13: Reader references in Rheme
Text
RRinR Percent Text
RRinR Percent
Eng 1
7 5.19% Spa 1
8 6.02%
Eng 2
4 3.01% Spa 2
1 0.67%
Eng 3
5 3.70% Spa 3
2 1.21%
Eng 4
2 5.88% Spa 4
10 6.54%
Eng 5
4 2.42% Spa 5
4 3.28%
Eng 6
0 0.00% Spa 6
2 1.40%
Eng 7
4 2.21% Spa 7
1 0.61%
Eng 8
1 0.59% Spa 8
4 14.29%
Eng 9
1 0.74% Spa 9
0 0.00%
Eng 10
3 1.50% Spa 10
0 0.00%
Total English
31 2.07% Total Spanish
32 2.53%
Key: RRinR = reader reference in Rheme
For the most part, as expected, the frequency of reader mention is very low across both corpora, and the differences in the totals are minimal. In Theme position, there is a slightly larger difference between the two corpora, and one text stands out in this respect, which accounts for the discrepency: Eng 1, an introduction to the study of history. The authors here thematize all-inclusive "we" in 11 instances in the first part of the text, in talking about the importance of history to all of us and about how we tend to look at history. In the second part of that chapter, they thematize "the reader", and discuss how readers will want to approach the textbook. Spa 1, also an introductory chapter, plays this down, and focuses much more on different philosophies of history, on different ways of studying history. Thus, Eng 1 introduces the reader more specifically to their textbook, while Spa 1 introduces the reader to the study of history in general, and hence downplays the author/reader interaction. The Spanish texts overall leave more reader references for the Rheme of the clause, for positions of non-prominence in that they are neither thematic nor are they Rheme final. Spa 8, for example, the textbook preface, has a high amount of reader reference in the Rheme. It was predicted that perhaps this chapter, and its counterpart, Eng 4, would show higher frequencies of reader reference in the Themes of the clauses, and also in the Rhemes. This expectation was not borne out at all by Eng 4. As just mentioned, it was confirmed in Spa 8, but only in the Rhemes of the clauses. Further comment on these two chapters is left for later discussion in the next section on author reference.
6.1.4.2 Author Reference Author reference includes all explicit mentions of the author(s) of the text. These are mostly realized by first person plural pronouns, although in some cases the author(s) refer
255
to themselves in third person singular or plural (e.g. "the author"). Some of these instances have already been included in the analysis above, under grammatical metaphor, in cases where authors align themselves specifically with a proposition, e.g.: 6-32. Eng 2 (112) and we may assume that this craft was largely practiced by women. Therefore, there will be some repetition here. However, the majority of instances are not of authors aligning themselves with a projected proposition. Instead, the instance forms part of the proposition, where the authors are a direct participant in the process expressed in the proposition, e.g.:
6-33. Eng 4 (10) Following this aim, we have focused on the outstanding institutions, ideas, and creative works that have formed (and expressed) Western civilization. At other times, the references are meta-textual, and include descriptions of authors' aims through relational processes:
6-34. Eng 2 (17) In this first chapter our aim will be to look at the life of the earliest human beings and to say what we can of their social structure.
Or the authors are Actors in material processes in which they project what they will do in the text:
6-35. Eng 5 (22) Drawing upon nonwritten evidence from geology, paleontology, anthropology, and archeology, we will first look briefly at the long prehistoric period - from the first appearance of human beings to the dawn of civilization.
As with reader reference, author references may occur in either the Theme, as in example 632, or in the Rheme, as in examples 6-33 - 6-35. Frequencies in both number and percentages of author references in Theme position appear in Table 6.14:
Table 6.14: Author reference in Theme:
Text
ARinT Percent Text
ARinT Percent
Eng 1
3 2.22% Spa 1
5 3.76%
Eng 2
7 5.26% Spa 2
2 1.34%
Eng 3
0 0.00% Spa 3
1 0.61%
Eng 4
7 20.59% Spa 4
4 2.61%
Eng 5
0 0.00% Spa 5
2 1.64%
Eng 6
1 0.48% Spa 6
0 0.00%
Eng 7
0 0.00% Spa 7
0 0.00%
Eng 8
0 0.00% Spa 8
2 7.14%
Eng 9
0 0.00% Spa 9
0 0.00%
Eng 10
0 0.00% Spa 10
0 0.00%
Total English
18 1.20% Total Spanish
16 1.27%
Key: ARinT= Author reference in Theme
256
Overall, the texts downplay any overt appearance of the author in Theme position. However, Eng 4 has an extremely high amount of author reference as compared to the other texts in Theme position. While the number of instances is only seven, it must be remembered that this text is a preface, and the total number of clauses is only 34. That it is a preface may explain the high amount of author reference, as Spa 8, also a preface, is relatively high in thematized author references in comparison with the other texts in the corpus, although nowhere near as high as Eng 4. Before commenting further on this finding, it is worth looking at author reference in Rheme, in Table 6.15 below, as the finding is repeated there. Discussion of other results from the above table will also be left for discussion following the presentation of Table 6.15.
Table 6.15: Author reference in Rheme:
Text
ARinR Percent Text
ARinR
Eng 1
0
0.00% Spa 1
2
Eng 2
6
4.51% Spa 2
0
Eng 3
1
0.74% Spa 3
1
Eng 4
6
17.65% Spa 4
2
Eng 5
5
3.03% Spa 5
2
Eng 6
1
0.48% Spa 6
2
Eng 7
0
0.00% Spa 7
0
Eng 8
2
1.18% Spa 8
0
Eng 9
0
0.00% Spa 9
0
Eng 10
0
0.00% Spa 10
3
Total English 21
1.40% Total Spanish 12
Key: ArinR= author reference in Rheme
Percent 1.50% 0.00% 0.61% 1.31% 1.64% 1.40% 0.00% 0.00% 0.00% 2.91% 0.95%
Again, here Eng 4, the preface, has a very high amount of author reference, especially in
comparison with all of the other texts, and in combining the results from the Themes and the
Rhemes, the percentage is extremely high. The authors in Eng 4 present themselves as
Actors in material processes, writing mainly about what they have done in the book. Not
only that, the things they have accomplished are presented as highly positive, e.g.:
6-36. Eng 4 (10) Following this aim, we have focused on the outstanding institutions, ideas, and creative works that have formed (and expressed) Western civilization. 6-37. Eng 4 (22)And throughout the book we have provided numerous cross-references to help point out the interconnections [emphasis original] of ideas and events that occur in various times and places. 6-38. Eng 4 (25) However, we have added information on a number of topics, in response to recent scholarly findings and the developing interests of teachers and students. 6-39. Eng 4 (29) In addition, we have enlarged our treatment of Eastern Europe. 6-40. Eng 4 (31) We have also enlarged our account of social, economic, and technical developments occurring at key periods in the history of civilization.
257
Even those involved in relational processes show this same characteristic: 6-41. Eng 4 (9) Our aim in this book is to present a clear, concise account of truly meaningful human experiences - relevant to the society in which we live. These instances are all mentioned more frequently in the Themes of the clauses than in the Rhemes. It seems as it the authors take the preface more as a selling device, rather than a note to students. This would explain the low amount of modality in this chapter: the authors may downplay tentativeness and other forms of modality in order to present their book as something sound and positive. Spa 8 is also a textbook preface, yet author references are nowhere near as high as in Eng 4. Spa 8 is much more modest in its claims. It even downplays the role of the textbook, any textbook, in the class: Un libro no es nada mбs que un instrumento de trabajo. De ninguna manera puede sustituir las directrices del profesor, del que en definitiva aspira a ser un colaborador eficaz; es el profesor de historia contemporбnea quien seсala los niveles, elige el mйtodo de trabajo y formula los planteamientos de los grandes problemas. Menos todavнa puede desplazar el trabajo del alumno, pieza clave de cualquier enseсanza que merezca este nombre. (Eng 2) Even when in the case of specific author reference, the author is modest: 6-42. Spa 8 (21) El autor de estas lнneas confнa - tiene fe - en no formular planteamientos utуpicos. In fact, there are only 2 author references in Spa 8, and they both refer to the author in the third person, although he did write the preface; indeed, the preface is followed by the words EL AUTOR (although this mention is not included in the analysis). And, as we have seen, Spa 8 also has far more reader references in the preface than Eng 4, which has very few, and none in Theme position, while Spa 8 had the second highest amount of thematized reader references across the two corpora. This may be attributable to a cultural difference: perhaps American textbooks use a the preface as a selling move, directing it at textbook choosers such as teachers or school administrators, while Spanish textbooks use it as space to guide the reader in the study of history. On this point, it is interesting to note a parallel between Spa 8 and Eng 1, which is an introduction to the study of history and is a guide to the reader on how to study history. Eng 1 has only 3 author references and these are all in the third person, although the authors also wrote this chapter. In fact, Eng 1 and Spa 8 are the only texts which refer to the authors in the third person. Thus they distance themselves from the text, making their intrusion into the text somehow more modest. This similarity of Spa 8 with Eng 1 serves to highlight the difference between it and the more actively 258
involved tone of Eng 4, which makes it seem more of a sales pitch. Obviously, these are only speculations, as they are based on only two prefaces. Many more would need to be studied in order to make any kind of real claim. The only other text with a relatively high amount of author reference is Eng 2, with 7 thematized author references, and 6 rhematized ones. As we have seen, Eng 2 is a chapter on prehistory, the first civilizations. Here, author references are used similarly in one way and differently in another from Eng 4. There are some metatextual references; however, these are far less grandiose than those of Eng 4: 6-43. Eng 2 (17) In this first chapter our aim will be to look at the life of the earliest human beings and to say what we can of their social structure. 6-44. Eng 2 (18)We shall then trace the development of an ever more complex life style. The other uses actually show the tentative nature of some of the statements made by the authors, as usually author references are coupled with finite modal operators. This backs up the notion that perhaps chapters on more remote eras will employ a greater amount of modality, and that the authors of this particular text are disposed to present history as open to different interpretations. Examples of this type of statement in Eng 2 include: 6-45. Eng 2 (34) Thus it is safest, for our purposes, to survey only the latest stages in the rise of humanity. 6-46. Eng 2 (50) If there was a division of labor by gender, we might guess that the men did most of the hunting and killing of animals, though this can hardly be proved. 6-47. Eng 2 (101) We cannot write a social history of any of these communities, Except for these two cases, overall author reference is low in the corpus. Mainly, authors across both cultures remove themselves from the text, thus reducing the `human face' of the texts and making them more informational than interactional. Another possible form of interaction, of involving the reader in the text, is to ask the reader a question. Thus, the final section on reader/writer interaction is that of questions, to which I shall now turn. 6.1.4.3 Questions In this study on Theme and Rheme, clauses in the interrogative mood are considered separately from clauses in the declarative mood, which make up the vast majority of the texts, except one instance in each of the two corpora of imperative clauses. The decision to treat questions separately here has two motivations. First of all, I wanted to include them in this section on interpersonal elements in the texts, as the asking of a question could imply a sort of interaction between the reader and the writer. Searle (1969: 66) distinguishes between two types of questions: "real questions", where the speaker wants to find out the 259
answer to something, and "exam questions", where the speaker wants to know if the hearer knows the answer. Questions in textbooks seem to fit neither of these two categories. It is obvious that questions used in this context, where there is no real time interaction between authors and readers, are not included with a congruent interactional function; i.e. authors are not asking questions of readers with the intent of finding out information, nor do they have any means of checking out whether their readers know the answers to the questions they ask. At the same time, while it is the case that questions are used for reasons other than information exchange, this "does not call into dispute the observation that the basic meaning of a question is a request for an answer" (Halliday, 1994: 45). Tadros (1994) includes questions as a category of predictive devices in economics textbooks. Her analysis underscores the interpersonal nature of questions: "The writer detaches himself from the resolution of the disjunction of the proposition posed by the question he asks, and this detachment predicts that he will be involved at some later point to declare his state of knowledge as regards the question" (ibid: 78). Thus, by raising expectations in the reader's mind for an upcoming answer, they are used in textbooks as rhetorical devices: to focus the reader's attention on some particular aspect of the subject. This relates to the second motivation for including them in this study. The Theme/Rheme configuration of questions is a tool authors can use to place in thematic prominence a starting off point for a stretch of text; this starting off point is encapsulated in the WH-word of the question. Thus, questions can be used to introduce a change in focus or topic, and this new discourse topic is compressed into the WH-word. As an illustration, the following extract follows on the heels of a description of the agricultural revolution. The questions signal a move to an interpretation, and analysis of why it took place: 6-48. Eng 2 (78)But why did this revolution take place, (79) what caused people to turn from the gathering pattern that had lasted hundreds of thousands of years? (80) Probably no single cause can provide the answer, (81)but we cannot neglect the question. (82)For millennia the glaciers, which had long before advanced from both polar regions, were shrinking back toward the poles..... Another way of doing this is to simply introduce the notion of cause through a declarative statement, e.g. There are several reasons why this revolution took place. 260
However, this places "several reasons" in a position of non-prominence, and is therefore perhaps not as effective as the question in changing the focus. Another option would be to make "several reasons" Rheme final: This revolution took place for several reasons. However, the first two clauses, 78 and 79, have a mini-development of WH-questions as Theme, and these serve to focus the attention of the reader squarely on the shift from a descriptive method of development of ideas to an analytical one. Thus, the questions signal a textual shift. Clauses 80 and 81 are interesting in this respect, as they could feasibly be eliminated; the authors could ask the first question and then launch right into their answer. Yet they suggest that no single cause can provide the answer, thus reinforcing the possibility of other answers to the question, or of other possible interpretations of historical events, a notion also reinforced by the thematic modal Adjunct in Clause 80. We have already seen previously in this chapter that Eng 2 offers history as a field open to much interpretation, and in providing answers of a tentative nature to the questions they ask, they reinforce this notion. Tadros (1994: 79) includes a limit of a two sentence succession for a question to be considered as predictive in her model, as more than two interrogative sentences in succession implies "`not now, but later'". Eng 1 includes a set of questions to which no answers are provided. They are there simply to show readers the range of questions that historians consider, and, thus, perhaps, to arouse their readers' interest in the study of history. The notion of "question" itself they find central to the study of history, as evidenced by several of the clauses included early on in the chapter: 6-49. Eng 1 (10)At the same time, the way we look at the past and what we see there -the questions we ask and the problems we pose- arise out of current concerns. (11)Today we are likely to ask what industrialization did to the environment; (12)in the 1950s scholars more often asked about the conditions that led to economic growth; (13)a generation earlier the central question was how workers had gained a larger share of industrial wealth. (17)We, who have been formed by the past, ask of history questions from our own time in order to learn more about both past and present. (18)Some questions must be asked repeatedly, (19)some issues arise again and again; They later provide a list of questions, which serve to illustrate the types of issues history addresses. Thus, the chain of WH- questions allow for a thematic configuration which 261
signals that the whole of the textbook is an exploration of answers to the kind of information required in the Themes of the clauses. The chain of questions is then followed by two clauses which both thematize these types of questions, which thus serve to underscore the prominence of questioning as the method of development of this part of the text: 6-50. Eng 1 (37) We also use history to discover about human behavior by discovering its variety, its similarities, and its limitations in different times and circumstances. (38) In what kind of God or gods did people believe? (39) What customs and institutions sustained those beliefs? (40) How did people obtain food and protect human life? (41) What tasks in each society were assigned to men and to women, to young and old, to rich and poor? (42) Did formal beliefs, economic activities, established institutions, and social classes sustain each other (43) or did they conflict? (44) Can we estimate for any particular moment the countervailing pressures for change and for continuity? (45) These types of questions concern all the social sciences, (46) and all of them make use of historical examples in seeking answers. Eng 4 also uses questions to set up the rest of the textbook as an answer to these questions. In fact, the first two clauses are questions: 6-51. Eng 4 (1)What is the nature of the Western world? (2)How has it shaped the men and women who are its heirs? Eng 6 uses questions in the introduction to the chapter on the 13th and 14th centuries to give readers an overview of the topics they will discuss in the chapter, and then cap the list off by indicating that the function of the chapter is to "focus on these questions": 6-52. Eng 6 (7) The miseries and disasters of the later Middle Ages bring to mind a number of questions. (8) What economic difficulties did Europe experience? (9) What were the social and psychological effects of repeated attacks of plague and disease? ... (11) Does this theory have validity for the fourteenth century? (12) What political and social developments do new national literatures express? (13) This chapter will focus on these questions. The rest of the questions in the corpus are of the first type, mentioned above and illustrated by example 6-48, except that the majority of the other texts do not indicate that there are other possible answers to their questions; they embark upon a straightforward answer to the question they pose. Thus, the questions serve to direct the readers' attention to a shift in focus, often to the causes or effects of the phenomenon under discussion: e.g.: 6-53. Spa 2 (63)їCuales son estos frenos? 262
(64)La mortalidad infantil sigue siendo muy elevada, aunque comienza a descender en el ъltimo cuarto de siglo....
6-54. Eng 5 (15) Why was civilization so long in coming to this planet - some billions of years? (16) An answer takes form as we peer back through those thousands of centuries.
6-55. Eng 9 (43) What kind of people left Europe, (44) and what were the reasons for doing so? (45) Most were poor people from rural areas, though seldom from the poorest classes.
It was hypothesized in section 6.1.2 that questions would not be frequent in the two
corpora, and the results in Table 6.16 bear this out:
Table 6.16: Frequency of Questions
Text
Questions Percent Text
Eng 1
6 4.44% Spa 1
Eng 2
3 2.26% Spa 2
Eng 3
2 1.48% Spa 3
Eng 4
2 5.88% Spa 4
Eng 5
1 0.61% Spa 5
Eng 6
8 3.85% Spa 6
Eng 7
1 0.55% Spa 7
Eng 8
0 0.00% Spa 8
Eng 9
2 1.48% Spa 9
Eng 10
1 0.50% Spa 10
Total English
26 1.74% Total Spanish
Questions Percent 0 0.00% 3 2.01% 1 0.61% 1 0.65% 0 0.00% 0 0.00% 2 1.22% 0 0.00% 0 0.00% 0 0.00% 7 0.55%
The use of questions appears with greater frequency in the English corpus, although clearly the use is infrequent in both corpora. The major difference between the two corpora is that the English texts, when they use questions, are more inclined to include a list of questions, several, such as in example 6-50 (7 questions) or 6-52 (4 questions with one gap), or at least two (such as examples 6-48, 6-51, and 6-55). In fact, in only 5 instances are single questions included in the English corpus, while in the Spanish corpus, only in one instance is a chain of two questions included; the other instances are all single questions, designed to introduce a new stage of development of the text.
6.1.4.4 Total Reader/Writer Interaction As with modality, here a buildup of the frequency of the more `interactive' features of the texts is considered.
263
Table 6.17: Total Interactive Features, English Corpus
Interactive feature:/ RRinT RRinR ARinT ARinR Ques Total Text:
Eng 1
14.07% 5.19% 2.22% 0.00% 4.44% 25.93%
Eng 2
2.26% 3.01% 5.26% 4.51% 2.26% 17.29%
Eng 3
0.00% 4.44% 0.00% 0.74% 1.48% 6.67%
Eng 4
0.00% 5.88% 20.59% 17.65% 5.88% 50.00%
Eng 5
3.03% 2.42% 0.00% 3.03% 0.61% 9.09%
Eng 6
0.96% 0.00% 0.48% 0.48% 3.85% 5.77%
Eng 7
0.00% 2.21% 0.00% 0.00% 0.55% 2.76%
Eng 8
0.59% 0.59% 0.00% 1.18% 0.00% 2.35%
Eng 9
0.00% 0.74% 0.00% 0.00% 1.48% 2.22%
Eng 10
0.50% 1.50% 0.00% 0.00% 0.50% 2.50%
Total English
2.07% 2.14% 1.20% 1.40% 1.74% 8.56%
Key: RRinT = reader reference in Theme; RRinR = reader reference in Rheme; ARinT = author reference in Theme; ARinR = author references in Rheme; Ques = questions
Table 6.18: Total Interactive Features, Spanish Corpus
Interactive feature:/ RRinT RRinR ARinT ARinR Ques Total
Text:
Spa 1
2.26% 6.02% 3.76% 1.50% 0.00% 13.53%
Spa 2
0.67% 0.67% 1.34% 0.00% 2.01% 4.70%
Spa 3
0.00% 1.21% 0.61% 0.61% 0.61% 3.03%
Spa 4
3.27% 6.54% 2.61% 1.31% 0.65% 14.38%
Spa 5
0.00% 3.28% 1.64% 1.64% 0.00% 6.56%
Spa 6
0.70% 1.40% 0.00% 1.40% 0.00% 3.50%
Spa 7
0.61% 0.61% 0.00% 0.00% 1.22% 2.44%
Spa 8
3.57% 14.29% 7.14% 0.00% 0.00% 25.00%
Spa 9
0.97% 0.00% 0.00% 0.00% 0.00% 0.97%
Spa 10
0.97% 0.00% 0.00% 2.91% 0.00% 3.88%
Total Spanish
1.11% 2.53% 1.27% 0.95% 0.55% 6.41%
Key: RRinT = reader reference in Theme; RRinR = reader reference in Rheme; ARinT = author reference in Theme; ARinR = author references in Rheme; Ques = questions
The English texts are over 2% more "interactive" than the Spanish texts. This could reflect a cultural difference, a difference which enters the realm of rhetoric, as explained in section 1.3. It is a small difference, however, and one which would need more corroboration from other studies. It is interesting here also to examine the differences amongst the texts. Again, it is easier to portray these differences in a table together with the information on the subject matter of the chapters, along with the authors' names. Thus, Table 6.19 presents the percentages of interactive features per total clauses of the texts, which are ordered from the highest to the lowest:
264
Table 6.19: Total Interactive Features Percentages with Subject and Author
Text ref. Percentage Subject
Author
Eng 4
50.00% Preface
Greer & Lewis
Eng 1
25.93% Introduction to the study of history
Chambers, et.al
Spa 8
25.00% Preface
Fernбndez
Eng 2
17.29% Prehistory
Chambers, et.al
Spa 4
14.38% Ancient Greece
Mangas
Spa 1
13.53% Introduction to the study of history
Ballesteros & Alborg
Eng 5
9.09% Prehistory
Greer & Lewis
Eng 3
6.67% 14th & 15th centuries
Chambers, et.al
Spa 5
6.56% 14th & 15th centuries
Palacio, et.al
Eng 6
5.77% 13th & 14th centuries
McKay, et.al
Spa 2
4.70% 19th century population
Fernбndez
Spa 10
3.88% 18th century
Palacio, et.al
Spa 6
3.50% Ancient Rome
Santos
Spa 3
3.03% 12th cent France & 12th & 13th cent Spain Ballesteros & Alborg
Eng 7
2.76% Ancient Greece
Perry
Eng 10
2.50% 18th century
Brinton, et.al
Spa 7
2.44% 18th century
Domнnguez
Eng 8
2.35% Ancient Rome
Roy
Eng 9
2.22% 19th century population
McKay, et.al
Spa 9
0.97% Prehistory
Cerdб, et.al
Here, the interactive nature of the preface from the English corpus, Eng 4, is highlighted. A possible explanation for this has been put forth earlier in the chapter: that perhaps the authors view the preface as a selling place, which would indeed call for them aligning themselves with their readers and presenting themselves as an important part of the interaction. Eng 1, an introduction to the study of history, and Spa 8, a textbook preface, but one which is presented as an introduction to the textbook, are similar in their interactive features, perhaps because both want to interest their readers in learning more about history. Spa 1 is also an introduction to the study of history, and does include a fair number of interactive features, although we have seen that they are more interested in analyzing the different ways in which history can be studied, rather than raising the interest of their readers in the study of history. Eng 2 follows its trend of presenting history (or, more specifically, prehistory) as something open to interpretation, a trend which was evident in its high amount of modality; this is underscored here by its use of interactive features. This can be contrasted with the highly specialized text on prehistory, Spa 9, with its extensive descriptions of tools; hence, it is a much more information-oriented text, but which, judging 265
from the fairly high amount of modality, does provide some indication of the tentativeness of the interpretations made. 6.2 Conclusions The purpose of this chapter has been to examine the role of the interpersonal in the Themes and the Rhemes of the clauses of the history textbooks. It was hypothesized that the interpersonal aspect of texts would be played down in the interest of focusing on the presentation of information, and on presenting this information as "reality". As Theme is a position of thematic prominence in the clause, it was expected that modal Themes would not be prevalent across both corpora, and this indeed was proven to be the case. The history textbooks have a lower frequency of modal Themes than studies of newspaper reports, research articles, and conversations. Then, more grammatical forms of modalizing clauses were investigated. This showed a difference between the English and Spanish corpora, 2.87% of Themes in the English corpus contained a modalized feature, as compared to 5.23% in the Spanish corpus, but this difference can be explained on grammatical grounds, given that Spanish allows for verb-initial constructions in more clause types than English does; hence, initial finite modal operators make for more modality in Theme in Spanish. The clauses were then analyzed for modality in the Rheme. Modal Adjuncts were seen to be more prevalent here than in the Themes, which can be explained by two factors: one is their tie to the verbal process, and another is that perhaps at times history textbook writers prefer not to include their modalized angle on the proposition in the position of thematic prominence. Then, the categories of grammatical modality were calculated for the Rhemes of the clauses in the corpora. The slight difference between the two corpora, a little under 2%, can be explained again by the grammatical discrepancies between the two languages. Throughout the analysis presented in this chapter, the buildup of modality was checked. Of interest to the generic hypothesis upon which this study is based is the overall findings on modality: for the English corpus, 20.25% of the clauses contain some feature(s) of modality, and for the Spanish corpus, 20.82%. The differences within each corpus across the different texts can perhaps be explained by how the authors portray the field of history, as more or less open to differing interpretations. 266
Finally, the interactional nature of the texts was checked through reader/author references and the use of questions. The two corpora had similar, very low frequencies in both the Themes and the Rhemes of the clauses, with a couple of exceptions. The first is that of Eng 1, an introduction to the study of history, which contains a relatively high number of reader references in both the Themes and the Rhemes of its clauses, but which shows well over double the number in the Themes. They involve their readers to a great extent in the study of history, emphasizing how it is important to all of us, and then guiding their readers to the use of their textbooks. It also uses questions to awaken the interest of the reader. It is very low in author reference, however. Eng 4, on the other hand, is low in reader reference, and very high in author reference, with almost equal frequencies in the Themes and the Rhemes of the clauses. Eng 4 is a preface and highlights the authors' deeds in producing a fine textbook. It puts this text at the top of the list in terms of use of interactive features. In sum, overall this chapter has served to reinforce the notion that the forces of audience, content, and purpose will produce similar texts with regard to Theme/Rheme across cultures. A slight rhetorical difference in the use of questions has been noted, however. Also, the differences amongst the different texts with regards to interpersonal elements and the possible explanations for those differences underscore the notion that content, or how the field is portrayed, can produce different thematic choices. That is, historians who view their field, or perhaps more accurately, construe their field as something open to differing interpretations will perhaps imbue their texts with more overt interpersonal elements. 267
7. Conclusions and Implications 7.1 Introduction It has been attempted in this study to establish whether similarities in the purpose, audience and content of history textbooks engender similarities in textual choices in the three metafunctions, i.e. in the experiential, the textual, and the interpersonal, across two cultures, specifically the U.S. and Spain. It also explores whether or not thematic and rhematic choices can indeed be linked to the context of situation and culture within which the texts were written, and thus give a clear reflection of the field, mode and tenor of the texts. Furthermore, inasmuch as the Hallidayan notion of Theme has been conceived with the English language in mind, this study has stretched the application of Theme to a language which shows some grammatical dissimilarity with English. Thus, it was questioned early on whether it could be applied to Spanish, and whether it would allow for comparison of equal elements across two languages. Also, it was posited in the first chapter that the analyses carried out in this study using the Theme/Rheme construct would allow for a composite picture of the texts in terms of their communicative potential, in terms of seeing linguistic choices as representative of the field, mode and tenor of the situation of context. The purpose of this chapter is to probe the aims and procedures of the study. We will begin by examining the application of the Theme/Rheme construct. 7.2 The Theme/Rheme Construct: A Viable Tool 7.2.1 Theme in Spanish This study has throughout kept very close to a Hallidayan definition of Theme. Halliday (1994) makes very clear that his description of Theme is for English, although he does state that the formal realization of Theme, i.e. first position, holds for other languages as well. During the course of this study, a number of problems related to applying this to Spanish have been considered, not least of which is the unmarked option of pro-drop and of SV inversion with certain verbs (e.g. existential verbs). However, Theme here is seen as "the point of departure" of the clause, and it accords with Travnicek's (1962, in Vasconcellos, 1992: 149) view that Theme may take the form of a verb, and even a grammatical particle. Indeed, Halliday's (1970: 164) conception of Theme as "the peg on 268
which the message is hung" "takes the theme beyond nounhood and permits it to be expressed by any element of language whatsoever" (Vasconcellos, 1992: 149). Halliday (1967b: 238) also points to the possibility of verbal Themes in the Celtic dialects of English. It has been seen throughout the study that verb Themes do not cause a problem in the analysis. In fact, often, they lexically match the subject matter of a given section or a chapter, as in Spa 2, a chapter on 19th century population growth and movement: 7-1. Spa 2 (33) Crecen de prisa los continentes que reciben poblaciуn europea ... (37) Aumenta la poblaciуn de los paнses con cambio social... (123) emigran los habitantes de paнses superpoblados, como Inglaterra o Alemania, Here, notions of growth, rise, and emigration are encoded in the verb initial Themes, and they thus help in showing the overall concerns of this chapter. Another problem for the analysis with respect to Spanish is se (see section 3.5.2.1.2). One way to deal with se in a thematic analysis is to also include the finite verb of the verb phrase it forms part of as Theme, given that se does not realize a participant function in the clause; thus the finite verb would fulfill the criteria for the ideational Theme. However, for several reasons this latter procedure has not been adopted here, and se has been considered as Theme. We have just seen that Travnicek (1962, in Vasconellos, 1992: 149) considers that even grammatical particles can be considered Theme, as in Travnicek's view, Theme occurs without fail in initial position. In cases of extraposition, Halliday considers empty "it" as Theme, and in existential constructions, there, which also has no participant function, is considered as Theme. Given these precedents in English, se has also been taken as the Theme of the clause. It also fits in with the description of Theme provided in Chapter 3 (see section 3.2). It is a jumping off point which indicates impersonality or the passive reflexive nature of the verbal process, thus allowing for constructions which attribute no agency of the verbal process. This absence of agency corresponds in many instances to the notion, which seems to underlie much historical explanation, that things simply come about, (a notion expressed in Halliday's analysis through the system of ergativity (see section 4.3.1.1), as in the following example: 7-2. Spa 2 (50)se produce un impulso demogrбfico de compensaciуn, Here, the se refers to a demographic impulse of compensation, which is left for the Rheme final position of prominence. The `peg' on which this message is hung is that of no agency, the demographic impulse simply comes about. 269
Thus, these two seemingly problematic areas in applying a Hallidayan analysis to Spanish are actually unproblematic as they serve to reinforce the function of Theme as described in this study. However, these two aspects of Spanish grammar can cause problems for thematic progression, to which we now turn. 7.2.2 Thematic Progression Hakulinen (1989: 62) points out that if "we follow the principle of picking out the initial elements and labeling them as themes, we get absurd results for the description of thematic progress in the text as a whole". Indeed, it was seen in Chapter 5 that a rather large percentage of clauses, 36% in English and 43% in Spanish, did not fit into any of the thematic patterns suggested by Danes (1974). This is not a problem with the patterns themselves, as Danes employs a criterion for Theme specification which accords more with the notion of Given. This procedure ensures that the Theme of a clause will almost necessarily (with the exception of a clause in which all of the concepts are New to the text) link back either with a previous Theme or a previous Rheme. Thus, Danes' (1974) theory needs to be modified in order to apply it to an analysis which applies a different criterion for Theme specification. In sum, in this study an application of Danes' thematic progression patterns has been applied using a procedure for Theme identification which does not accord with Danes' procedure. However, this was felt to be opportune, as it was explained in section 5.1 that it is optimal for writers to choose some element from the previous discourse to take as the point of departure for any given clause, as this allows for ease of discourse processing, an aim which textbook writers would presumably keep in mind. It needs to be clarified here that only two of Danes' progression patterns were investigated in this study, as his third pattern, derived Theme, was seen to conflate with constant Theme (see section 5.2; also Dubois, 1983: 108). At any rate, this application of the patterns would have no difference on results, as the two categories conflate; thus the elimination of one of Danes' categories does not in any way imply that fewer clauses were categorized. Hence, the fact that 36% of the clauses in the English corpus and 43% in the Spanish corpus did not fit into either of the patterns could serve to support the claim made by Hakulinen, included in the previous paragraph. The patterns simply do not hold over a rather large proportion of the clauses in both corpora. Dubois (1983), who also works from a Hallidayan conception of Theme, expanded Danes' thematic progression scheme to include `gapped' progressions (where the Theme of a clause refers either to a previous non- 270
contiguous Theme or Rheme) and `multiple' progressions, where a Theme refers back to more than one Theme or Rheme. Dubois (ibid: 109) further states: "That these subtypes exhaust the possibilities I strongly doubt, given, at a minimum, the enormous potential for register and channel differences". Indeed, in this study, it was felt best to consider a number of Themes as taking part in none of the progression patterns, given that the purpose of the thematic progression analysis was to investigate whether textbook writers did indeed provide clear links for their readers. Therefore, for example, while Dubois analyzes gapped development over relatively large distances, it is felt here that textbook writers would do well to keep their thematic links closer and clearer for their readers. It is worth inserting a reminder here: not all of the clauses with Themes that did not fit into one of the patterns are problematic for text processing, as many of the Themes refer to entities which are readily available given the subject matter (e.g. temporal adverbials and what have been termed key Themes here), and other Themes are lexically null (e.g. se, empty `it' and existential `there') yet pragmatically necessary in allowing for a Theme/Rheme + Given/New configuration of the clause, a configuration which SV inversion also provides for.. Thus, the application of an analysis using thematic progression patterns can reveal absurd patterns in a text. However, if what the analyst is looking to elucidate are precisely those cracks in the patterning of text, perhaps in order to suggest a more optimal patterning and to show Theme choices which are meaningful to the specific genre under analysis (such as temporal adverbials in history texts), then thematic patterning is a useful tool. In this study, indeed, the analysis of thematic patterning has allowed for a picture of the way the texts are organized in both corpora, and this has then allowed for a clear comparison of text organization in Spanish and English history textbooks. 7.3 Summary of Results: A Cross-Cultural Picture of History Textbooks The overall results of this study have on the whole confirmed the cross-cultural genre-based hypothesis set out in section 1.6, which postulated that the Spanish and U.S. history textbooks in the corpus, given the similarity of their audience, content and purpose, would display similarities with respect to thematic content, i.e. textual, interpersonal and experiential elements, and thematic progression patterns. There are, nonetheless, some deviations from the expectations. 271
With regards to the analysis of process, participants and circumstances functioning as Theme, a composite picture of history textbooks writing shows an inclination across both corpora to take as the most popular participant point of departure a Carrier or Identified, while Attributes and manner circumstances are popular features in the Rhemes of the clauses, thus indicating a high amount of description and explanation, which was expected. This finding can be coupled with a finding from the thematic patterning analysis, that of the greater frequency of the simple linear progression pattern in both corpora. While the simple linear pattern is not often associated with narration and description, it was demonstrated in Chapter 5 that the simple linear pattern is indeed used to these purposes, thus providing for narrative and descriptive patterns which at the same time explain and interpret. However, there were differences between the two corpora with respect to both of the above analyses. In the transitivity analysis, the English corpus showed a significantly greater frequency of Actors, while in the thematic pattern analysis, this same corpus showed a higher frequency of constant Theme progression than did the Spanish corpus. Added to this is the fact that the Spanish corpus showed slightly higher instances of Attributes and manner circumstances, a finding which becomes more significant when seen in the larger picture together with the lower amount of constant Theme patterning. It has been posited from these findings that perhaps the American history textbooks employ a method of writing up history which is somewhat more narrative in nature, while the Spanish history textbooks employ a method which is more structural in nature, in that they show more of an emphasis on events and the relationship of events to the systems in which they are seen to function. It was suggested that this perhaps reflects a desire on the part of the writers to provide a framework which is more readily available to student readers, as students seem to prefer narrative explanations (Halldйn, 1997; cf. Bloor, 1996, for reference to the basic human need for narrative). This finding is very tentative, as, while it can be applied to the corpus as a whole, the individual texts in the corpus do not always serve to confirm it. It is also interesting to combine findings from the analyses related to the textual and interpersonal metafunctions. With regard to the former, it was seen that there is a similar percentage across the two languages of peripheral Themes which can be problematic for text processing (13.47% in the English corpus and 15.03% in Spanish). It was also seen that textual Themes are not often used to help the reader with these more difficult Themes. With regard to interpersonal elements, especially in the area of reader/writer interaction, it was 272
seen that the texts are very much informationally oriented, and that little interaction takes place, with some exceptions, e.g. introductory chapters or prefaces. These findings from the two different areas serve to portray the textbooks as perhaps very distant from the reader. At the same time, the English corpus is slightly more "interactive", in that it employs 2% more interactive features, i.e. references to the reader and questions, than does the Spanish corpus. Furthermore, the results of thematic progression patterns (5.5.1) showed that in the English corpus 10% more clauses were chained to nearby clauses than in the Spanish corpus, through either a thematic or a rhematic link. Most of this difference can be explained by the grammatical nature of Spanish, especially by verb-initial constructions, yet there is still a small difference of 1.5% which remains, meaning that the English corpus provides for thematic chaining slightly more than the Spanish corpus. Furthermore, while the two corpora are very similar in textual Theme use, the English corpus adds to that by employing almost 2% more conjunctive adjuncts in the Rhemes of the clauses. While all of these differences are very slight, if we add them to the slightly more narrative tendency of the English corpus, we see a textbook corpus which may support the notion suggested in section 2.2.2 that Spanish writers expect more of their readers (Valero-Garcйs, 1996), while English may belong to a more writer-responsible culture (Hinds, 1987). This may also be explained by the contexts in which the texts are used. It will be remembered that history teaching at Spanish universities is specialized, i.e., it is targeted towards students who choose history as their degree, while history teaching at American universities, in addition to the degree program, also involves students from other degree programs who are required to take a general history course in Western Civilization. Thus, the slight differences in the two corpora may hint at a tendency on the part of the Spanish history textbooks to attempt to inculcate the readers into the discourse of the historian, especially through a leaning towards the preference of structural explanation (this point is returned to in section 7.4 below). Another finding which adds to the composite picture of history textbooks across the two cultures is that of the low use of modal Adjuncts as the point of departure of the message, i.e. on very few occasions do the authors of the textbooks in the corpus (2.67% in English and 3.09% in Spanish) hang their clauses on the peg of opening their statements up to discrepant interpretations, especially when compared with other genres, i.e. research articles. This has been attributed to the lack of symmetry between the reader and the writer: 273
the writer is in possession of information which the reader does not have. Thus, the writers are not attempting to persuade the reader to agree with their statements; they are presenting the bulk of their statements as factual. It has been seen that the authors of the textbooks then do build up throughout the clause to a greater amount of modality than that which appears in the Theme, through finite modal operators, modal Adjuncts in the Rheme, and various forms of grammatical modality; however, the percentage of the clauses of the texts which use these methods of expressing modality is still low (17.38% in English and 15.60% in Spanish), thus presenting a picture of the texts as expressing objective reality. Thus, through the various linguistic analyses carried out in this study, the history textbook corpus presents a composite picture as to its objective, informationally-oriented, non-interactive explanatory and interpretative discourse. This picture is similar across the two cultures, except for a slight difference in the way the reader is viewed. At the same time, this statement ignores the fact that there were some wide differences amongst texts from each of the corpora. These differences seem to be related to some extent to differences in the subject matter of each of the chapters, again, to the field, and to a smaller extent to author style. 7.4 Limitations and Theoretical Implications In the introduction, it was stated that it was the intent of this study to contribute to systemic-functional theory, genre analysis and contrastive rhetoric. It is the purpose of this section to explore those contributions and their implications, as well as the limitations of the study. In wanting to provide for a composite picture of history textbooks for the uppersecondary/tertiary level across two cultures, and in order to take into account variables such as author style and subject matter in a cross-linguistic study, many times observations made are based on a small amount of data, e.g. two prefaces or two introductory chapters. The overall corpus size is slightly over 55,000 words, with sections from 20 chapters, on a breadth of topics and with a number of different authors. This was considered necessary in order to carry out the different types of analyses applied throughout for several reasons. First of all, the analyses of thematic progression and transitivity roles involve a very close scrutiny of the texts; thus, they cannot be carried out automatically across large corpora. Secondly, a range of authors was needed, as if the study were limited to one author or set of 274
authors, any differences found could be as easily attributed to individual idiosyncrasies as to
cultural differences. Finally, a pilot analysis and comparison of 3 texts from each of the
languages showed overall similarities in the total results across the two languages, yet
showed differences between individual texts which seemed to relate to the subject matter of
the texts, and this warranted further exploration.
Halliday's notion of Theme was chosen to analyze the data given that its usefulness
in genre characterization had been posited by other studies. The ideational Themes were
analyzed for their transitivity roles and their contribution to thematic progression patterns,
and the clauses were analyzed for textual and interpersonal Themes. The Rhemes were also
analyzed for transitivity roles, conjunctive adjuncts, and interpersonal elements. This has
given a composite picture of the texts in terms of the linguistic realizations of the field,
mode and tenor of the communicative act. However, this picture could benefit from analysis
at other levels, such as text macrostructure, i.e. how the texts are divided in chapters and
sections, or use of iconic devices, such as maps, charts, diagrams, and pictures.
As a practical application of part of Halliday's systemic-functional theory of
grammar, this study has tested and proven the applicability of the theory to extended text
produced in real-life contexts across two languages. The practical applicability of the
construct in this study serves to add to the growing literature on the analysis of Theme and
its relationship to genre that have been mentioned at length throughout.
It is felt, then, that this study adds to the theory with reference to the function of
Theme. Throughout this study, Theme choices have been analyzed in terms of their
experiential, interpersonal, and textual role, along with their grammatical function.
Lowe(1987: 7) describes the function of the point of departure, or Theme, as carrying out
any one or any combination of the following:
either (i) or (ii) or (iii)
set the spatial, temporal, situational or individual framework within which the ensuing discourse holds (experiential component) give the direction in which the next part of the argument is going to go (logical component) such as with conjunction give interpersonal information on the ensuing discourse, such as indicating it [sic] illocutionary force (i.e.. whether statement, question, command, etc.), or setting the mood (with manner adverbials) or giving an evaluation of the importance or reliability of the information in the ensuing discourse;
275
This description of the function of Theme is partial in that it leaves out its discoursal role. In section 3.2, it was seen that Berry (1996) assigns a dual function to Theme, discoursal and clausal. Lowe's functions focus on the clause level function, so his framework needs to be expanded to include discourse concerns, such as thematic patterning and summative Themes. Added to these discoursal and clausal functions of Theme are grammatical constraints on word order. Thus, it can be concluded that Theme choice is governed by the combined forces of the following concerns: - textual concerns: text continuity and textual iconicism - experiential iconicism - speaker/hearer roles (e.g. informational vs. interactional) - grammatical constraints This helps to explain the difficulty of defining Theme, in that it carries out a multi-functional role. Metaphorical descriptions of Theme function abound in the literature, as has been reflected throughout this study: e.g. "point of departure", "peg on which the message is hung". Yet, in their flexibility, these do serve to sum up the different functions of Theme. This study has also analyzed Rheme content in terms of participant roles, conjunctive adjuncts, and interpersonal elements. In many studies on Theme, not much scrutiny has been given to the Rheme, with the exception of those carried out by Fries (1992, 1993, 1994, 1995b). For Fries (1995b), the hypothesis which connects the experiential content of Theme with genre cannot be explored by simple counts of, for example, how many Themes refer to spatial location. This is because of the different functions of the Theme and the Rheme of the clause: elements which have more of an orienting role in a text (e.g. spatial location references in a tour guide) will often be placed in Theme, while elements which have a focal role (e.g. spatial location references which form part of a description of events in a narrative) will appear in the Rheme. Thus, "... meanings are not randomly distributed in different positions in the sentences and clauses of a text" (Fries, 1995b: 326). Fries further points out that the genre-Theme hypothesis cannot be explored with a simple raw count of how many Themes refer to spatial location, but "when comparing texts we need to look at all references to a type of meaning (such as spatial location) and then compare the proportions of thematic versus non-thematic occurrences" (ibid). 276
In the analyses carried out in this study, there has been an attempt to compare these proportions. It was seen in Chapter 4, for example, that there are differences in the experiential content of the Themes and the Rhemes. With regard to circumstances, by comparing Charts 4-9 and 4-15, it can be seen that circumstances of manner, especially quality, are rarely thematic, yet they are the most common circumstance type in the Rhemes of the clauses. Also, while temporal circumstances are fairly equal in the Themes and the Rhemes, spatial references are higher in the Rhemes than in the Themes, thus confirming the dominance of temporal circumstances over other circumstances as Theme, while their popularity is equaled or surpassed by manner and spatial circumstances in the Rheme. This does seem to lend some support to the notion that thematic selection is influenced by the contextual variable of field, as history revolves around the occurrence of events in time. Furthermore, the predominance of manner circumstantials and Attributes in the Rhemes of the clauses shows the global concern of the texts, concerns which Fries (ibid) shows correlate with the Rheme, with explanation, of explaining how things were. Also, it was seen that the occurrence of modal Adjuncts was much higher in the Rhemes than in the Themes of the clauses, which can also perhaps be explained by contextual configuration, in that authors may wish to downplay their interpersonal angle on the proposition by including it in a position of non-prominence. However, these findings would need to be compared with comparable findings in other genres in order to ascertain whether the explanation for the differences in Theme/Rheme occurrences is indeed due to genre, or whether there is another explanation, such as maybe the thematization of manner circumstances and/or of modal Adjuncts is a marked option in the language as a whole. The evidence provided throughout the study from other Theme-based genre studies, with which the results from the Themes of this study have been compared, do seem to provide support for genre influence on Theme choice, and thus the explanations for Rheme choices in this study have been influenced by that. Nonetheless, the results from the Rheme analyses given here could serve as a base for comparison with results of rhematic choices from other genres, which might then serve to provide further support for the notion that thematic and rhematic choices respond to generic pressures. This study has also served to add to the literature on transitivity roles in texts. Halliday's model was seen as lacking in delicacy with regard to participants in material 277
processes, specifically in categories related to causation. Thus, categories were added, such as the category of Factor, which refers to a participant, usually inanimate and usually a historical event or process, that brings about a phenomenon or an entity, labeled Result, or that brings about a change in a phenomenon or an entity, labeled Affected. These categories seem to be especially related to nominalization, which is quite frequent in history textbooks. However, it is not felt here that the model per se needs changing; rather, the model may need to undergo slight adaptations depending on the field of the text type under consideration. For example, Bloor (1971) points out that in Fillmore's case system another case is needed to show natural cause (e.g. storms) as their role is neither Instrumental nor Agentive. In text types relating to natural phenomena, e.g. weather reports or natural history texts, it might prove worthwhile for a transitivity analysis to reflect this by including a category which would reflect natural phenomena as cause. The label of Factor may be useful for that type of phenomena, or, again depending on the text type, another label could be used in order to distinguish natural phenomena which causes an effect from historical phenomena which does so.7 Thus, the analysis carried out here on participant roles serves to underscore both the simplicity and the flexibility of a model of language in use. The transitivity analysis has served to show the problems which arise in analysis due to grammatical metaphor, such as with nominalization, and with other examples of language which could often be analyzed at two levels, i.e. as more or less congruent. For example, "suggest" in its congruent meaning represents a verbal process, yet it is used to embody a relational process in the following example: 7-3. Eng 3 (131)By the middle of the fifteenth century, agricultural prices tended to stabilize, (132) and this suggests a more dependable production. This study has demonstrated the difficulty inherent in applying the analysis of transitivity roles due to metaphor. One suggested way of dealing with this is to analyze it at two levels: one at the level of the wording and another at the level of meaning (White, 1999; Martin, 1999). This would prove to be an interesting study for the future, and a dual analysis over an extensive corpus such as this would allow for insights into typical wording-meaning 7 Indeed, there were some instances of natural phenomena occurring the history textbook data, e.g.: Eng 6 (17)Severe weather, ... made a serious situation frightful. Eng 6(30)Torrential rains in 1310 ruined the harvest However, there were not enough of these types in the data to warrant a separate category, and thus they were labeled Factor. 278
relationships in different genres. However, this proved to be beyond the limitations of this study. Finally, this study has served to add to the literature on genre analysis and contrastive rhetoric, especially to the literature in which the former informs and serves the latter. Contrastive rhetoric "posits that speakers of two different languages will organize the same reality in different ways" (Kaplan, 1995: 38). This is evident because "different languages will provide different resources for organizing text" (ibid). Genre theory posits that the communicative purpose of a social event will have an effect on the contextual configuration, i.e. on the variables of field, mode, in tenor, and this is then reflected in textual choices. Furthermore, texts written with a similar communicative purpose will show similarities in the textual choices made, and thus genre characterization can be achieved through analysis of linguistic variables. Indeed, this study has shown strong similarities across two cultures in history textbook genre. Therefore, genre theory and contrastive rhetoric would seem to be at loggerheads, and one aim of this study has been to shed light on this clash in order to delineate how resources for organizing text might differ, on the one hand, and, on the other, to what extent generic pressures might be in place across two different languages and cultures. It is worth pointing out here that this study deals with two related languages, in that they are both, historically speaking, European languages. Furthermore, reading in the philosophy of history shows trends that span Europe and the U.S. At the same time, they are different cultures and different languages, and to some extent the writing of history, and of textbooks, has evolved separately in each country. In other words, it is doubtful that there has been as much influence on history textbook writing across cultures as there has been, say, in writing in the sciences and economics. Also, there are differences in the educational systems at the tertiary level, from which the bulk of the texts analyzed were chosen, such as the inclusion of a general history course at American universities for all degree programs, while history study at Spanish universities is much more specialized. As we have seen, the results of this study show similarities with respect to experiential, interpersonal, and textual content of the Themes. Differences were found in some of the results, specifically in a greater frequency in the English corpus of Actors as Theme, a higher frequency of the constant Theme pattern, and a slightly higher percentage of interactive features. This brief summary of the results ignores grammatical differences 279
between the two languages, mainly that in Spanish there is a much higher frequency of process as Theme, given pro-drop and SV inversion. It is difficult to know if this is the kind of difference Kaplan refers to when he explains that different languages have different resources for text organization. It has been pointed out elsewhere (Thompson, 1978; Payne, 1990) that verb-initial constructions allow for a more pragmatic construal of the clause; in other words, they allow for a New Subject to be placed after the verb, and thus in the Rheme. English does not have this resource available in its grammar (except in very restricted cases). Thus, the results in this study seem to show that there are overall similarities due to the field, mode and tenor of the context of situation in the history textbooks in Spanish and in English. There are some differences which can be explained in terms of the grammatical natures of the two languages , and there are slight differences which can be explained either by tenor or mode. That is, the differences mentioned above seem to reflect a slightly more narrative bent to the American history textbook corpus, and a slight leaning towards structural explanations in the Spanish corpus. This difference may reflect a difference in the audience and in the way writers react to that audience. Halldйn (1997) explains that historians prefer structural explanations, while secondary school students prefer narrative ones. The fact that Spanish history textbooks at the tertiary level are written for students whose degree is history, i.e. who are fledgling historians, may go part of the way in explaining why there is more of a structural bent reflected in Theme selection. Another way of explaining it is that Spanish may be a more reader-responsible language, i.e. it is not as much the responsibility of the text producer to ease the way for the reader as it is in a writer-responsible language. This accords with other studies carried out between writing in Spanish and English (see section 2.2.2). However, there could be another explanation. Perhaps American historians prefer a causal, or narrative, explanation because they construe history as a series of events in relationship to other events, while Spanish historians construe history as a system of events functioning within a system. This would then make the difference between the corpora one of field, and the difference would exist at a level above that of the textbook, i.e. before the consideration of the student reader. The point here is not to determine which of these explanations is "right", as it goes beyond the scope of this paper, and would necessitate further analysis of texts written for 280
professional historians and further investigation as to the nature of history in both cultures. The point is that any explanation seems to point to the variables of field, mode, and tenor. Thus, genre analysis allows us to explain linguistic choices in terms of the contextual configuration of the text, in terms of the context of situation and culture. Thus, genre analysis goes beyond mere description of differences and attempts to explain differences in terms of the way fields are viewed in different cultures and the differing ways in which writers relate to their readers. Therefore, genre analysis can help to identify how it is that cultures view the world and how social relationships are constructed. It is felt, then, that genre analysis moves the field of contrastive rhetoric beyond a description of similarities and differences across languages into explanation of why these similarities and differences may exist. This leads to final questions raised in section 1.6, with respect to whether a crosscultural Spanish and American history textbook genre could be posited, and, if so, could differences be explained by a difference in rhetoric. The history textbooks were classified as belonging to the same genre by dint of their similar communicative purpose. The results in this study have confirmed that this similar communicative purpose is indeed reflected in the textual choices made with respect to the Themes and the Rhemes of the clauses, thus reinforcing the assumption that indeed they do belong to the same genre. The slight differences (beyond grammatical differences) which are manifest across the two cultures can be explained by a difference in the way the field is viewed and/or by a difference in the writer/reader relationship. Thus, rhetoric may be seen to be influenced by the linguistic resources, field construals, and the social construction of the relationships between the interactants which exist in the culture. These exist prior to the specific communicative event which brings about the text, and thus inform the textual choices made within a genre. Given the relative cultural similarities between the U.S. and Spain (historically European cultures and languages), further research into more widely differing cultures would go farther in answering whether generic constraints can be said to exist across cultures where there are more linguistic differences and greater differences in how a specific field is construed and how social relationships are constructed. From the results of this study, it can be said that the history textbook genre in Spain and the U.S. bring about similarities in textual choices in history textbooks due to their common communicative purpose. 281
7.5 Practical Implications This study was not designed with a set of wider practical applications in mind, as its purpose was to test a genre-based hypothesis vis-а-vis the contrastive rhetoric hypothesis, and to test the applicability of the Theme/Rheme construct to a corpus of published texts in a language such as Spanish, which allows for non SVO word order. However, there are some practical applications which have come to mind throughout. The first involves teachers of students of history in their native language. It would serve the students well to study notions of grammatical metaphor and their application in history textbooks. It could be an interesting pedagogical task to take a chunk of text from a history textbook and turn it into a linguistic exercise involving students in converting nominalizations into full clauses and discussing the implications of expressing a clause using grammatical metaphor through nominalization more congruently, i.e. where nominal groups refer to things, verbs to actions, and where temporal and consequential logical relations are realized through conjunctions. This would serve to raise students' awareness of how processes are turned into events, are reified into events, and would allow them to explore and question in greater depth causal relationships. It would also be beneficial to students to raise their awareness of the use of modality in textbooks, to look at statements which are presented as objective mirrors of reality and to contrast them with statements that make clear their subjective position. In working with students at the first year university level, I find that they tend to take what they read in textbooks as undisputed, not always taking on board the idea that claims need evidence. Helping them to see behind unmodalized declarative statements can aid students in critical thinking skills. Peccei and Eiseling (1996, in Nash, Crabtree and Dunn, 1997) feel that one of the objectives to achieve in teaching history is for students to be able to "learn to identify their assumptions, think logically and consider alternative explanations". Secondly, teachers of second language learners who will be studying history in the second language can also help their students by comparing/contrasting the process of nominalization in the two languages, and highlighting how nominalizations are encoded in history textbooks. Furthermore, they can analyze with students instances of grammatical metaphor, and thereby expand the students' receptive repertoire of the expression of logical relations. 282
Finally, history textbooks writers can learn from linguistic analyses of their texts. They may wish to search for ways of making their texts somewhat more interactional, without losing sight of the informational purpose they embody. They also may benefit from incorporating in their writing clearer thematic progression chains, thus allowing for ease of text processing on the part of their readers. Finally, they may wish to contemplate the implications of presenting unmodalized declarative statements in terms of opening doors of differing interpretations to their readers. Indeed, there are lamentations about history teaching across both the U.S. and Spanish cultures, with wars going on in the U.S. as to what and how history should be taught (Nash, Crabtree, and Dunn, 1997; Hitchens, 1998) and with the proliferation in Spain of history textbooks portraying historical events differently in each of the autonomous communities, rewritings which Prades (1997) claims have as their purpose the serving of the nationalistic interests of the different communities. Hitchens (1998) would probably suggest that the best service teachers can do given such a state of affairs is to offer the different versions to their students and then to explore with them why the differing interpretations exist. Halldйn (1997: 201) argues that "learning often means aquiring [sic] an alternative way of conceptualizing the world, not in order to replace a conception that has already been established, but rather to increase one's total repertoire of conceptualizations of the surrounding world". Thus, if textbook writers continue to prefer an unmodalized declarative style of writing, which presents their view of history as factual, again teachers can help students increase their repertoire of interpretations by comparing differing historical accounts of the same event or events. Thus, a final implication of a linguistic study such as the present takes us back to the first one mentioned in this section: to suggest to teachers of history that they be well aware of the language used to encode historic events. According to Blanco and Rosa (1997: 196), one of the purposes of history "should be that of empowering students to defend themselves from ready-made stories and their implications; that is, to provide them with resources for untangling the fabric of the historical stories they encounter". Perhaps some awarenessraising on the part of teachers with respect to nominalization, grammatical metaphor, the encoding of events and individuals and an understanding of their roles in clauses, and modality use can help in untangling the fabric of history writing. As Rosen (1972: 124) states, "A serious concern with the language of the textbook, the language of a subject, is a 283
proper and central concern for all teachers, for it involves the vital participants in most learning - words". 7.6 Implications for Future Research The analysis of corpora across two languages and cultures, history textbooks written by a number of different authors and on a variety of different topics, has illuminated a number of areas for future research. These include applying any one or all of the analyses applied in this study to: 1. a mono-lingual/cultural study of a set of chapters on the same subject matter (e.g. pre-history, 19th century population movement, etc.) from a range of different authors. This would shed light on the similarities/differences based on author preference. 2. a study of a textbook by one author or set of authors. This would provide insights into whether there are differences in thematic content and thematic progression depending on subject matter, by eliminating the variable of author style. 3. a cross-linguistic/cultural study of a number of chapters by different authors on the same subject matter. This would control to a greater extent the variable of field, although it may also serve to highlight the different ways of viewing the field, as in 1, above. This could involve a range of linguistic/cultural backgrounds. 4. a study involving triangulation, in which any of the linguistic analyses are complemented by, for example, interviews with textbook authors and by protocol studies involving both readers and writers. 5. a mono-lingual study in which history textbook writing is compared with history research writing, i.e. writing by historians for other historians, to test the author/reader relationship variable and its effect on the linguistic features of the text, e.g. the use of textual and modal Themes. 6. a more delicate analysis of the transitivity roles, especially of participant roles, e.g. classifying Actors as representing institutions, individuals, etc., which may illuminate different ways of viewing the field, i.e. different philosophies of history, perhaps, within the same culture or across cultures. 284
The results from the present study would provide an apt basis for comparison with any or all of the above suggested investigations. The results and data could also serve as a base for the type of practical activity suggest in the previous section. 285
8. References Alarcos Llorach, E. (1994) Gramбtica de la Lengua Espaсola. Madrid: Editorial Espasa Calpe, S.A. Alonso Belmonte, I. (1997) `Estrategias Textuales y su Realizaciуn Temбtica en Espaсol: Un Estudio de Corpus'. RESLA 12: 55-74. Andrews, R. (1992) `Introduction'. In R. Andrews (ed.) Rebirth of Rhetoric: Essays in Language, Culture and Education. London and New York: Routledge. Aziz, Y.W. (1988) ` Theme-Rheme Organization and Paragraph Structure in Standard Arabic' W. ORD 39/2: 117-128. Babby, L.M. (1980) Existential Sentences and Negation in Russian. Ann Arbor, MI: Karoma Publishers. Bдcklund, I. (1992) `Theme in English Telephone Conversation'. Language Sciences, 14/4: 545-564. Bamforth, R. (1992) `Process Versus Genre: Anatomy of a False Dichotomy'. Prospect, 8/1: 89-99. Barcelona Sбnchez, A. (1986) ` "Tough Movement" en Inglйs y en Espaсol' . InActas del IV Congreso Nacional de Lingьнstica Aplicada: Lenguaje y Educaciуn. Abril 8 al ll de 1986. Vol. I. Universidad de Cуrdoba. Asociaciуn Espaсola de Lingьнstica Aplicada: 26-48. Barcelona Sбnchez, A.(1990) ` Estudio Contrastivo de la Elevaciуn en Inglйs y en Espaсol' . In Actas del VIII Congreso Nacional de Lingьнstica Aplicada, Vigo, 2,3 Y 4 de mayo de 1990. Asociaciуn Espaсola de Lingьнstica Aplicada: 95-108. Bell, R.T. (1991) Translation and Translating: Theory and Practice. London: Longman. Bentham, J.(1839) `Nomography, or the Art of Indicting Laws'. In J. Bowring (ed.) Works of Jeremy Bentham, Vol. III. London: Barton, 231-283. Berkhofer, R.F. (1989) ` The Challenge of Poetics to (Normal) Historical Practice' . In P. Hernandi (ed.) The Rhetoric of Interpretation and the Interpretation of Rhetoric. Duke University Press: 183-200. Berman, R.A., & D. I. Slobin (1994) Relating Events in Narrative: A Crosslinguistic Developmental Study. Hillsdale, New Jersey: Lawrence Erlbaum. Bernstein, B. (1966) `Elaborated and Restricted Codes: An Outline'. Sociological Enquiry, 36: 254-261. 286
Berry, M. (1987) `The Function of Place Names'. In T. Turville-Peter & M. Gelling (eds.) Studies in Honour of Kenneth Cameron. Leeds: School of English, University of Leeds (Leeds Studies in English, New Series XVIII.): 71-88. Berry, M. (1995) `Thematic Options and Success in Writing'. In M. Ghadessy (ed.): 55-84. Berry, M. (1996) `What is Theme? -- A(nother) Personal View'. In M. Berry, C. Butler, R. Fawcett & G. Huang (eds.) Meaning and Form: Systemic Functional Interpretations. Norwood, New Jersey: Ablex Publishing Company: 1-64. Bhatia, V.K. (1993) Analysing Genre: Language Use in Professional Settings. Longman. Blanco, F. & A. Rosa (1997) `Dilthey's Dream. Teaching History to Understand the Future'. International Journal of Educational Research, 27/3: 189-200. Bloomfield, L. (1933) Language. New York: Holt. Bloor, T. (1971) Some Aspects of Transitivity and Passivization in English. Unpublished MA Dissertation. The University of Essex. Bloor, T. (1996) `Three Hypothetical Strategies in Philosophical Writing'. In E. Ventola & A. Mauranen (eds.) Academic Writing: Intercultural and Textual Issues. Amsterdam/Philadelphia: John Benjamins: 19-43. Bloor, T. & M. Bloor (1995) The Functional Analysis of English. London: Edward Arnold. Bloor, M. & T. Bloor (1992) ` Given and New Information in the Thematic Organization of Text: An Application to the Teaching of Academic Writing.' Occasional Papers in Systemic Linguistics Volume 6: 33-44. Brandt, D. (1986) `Text and Context: How Writers Come to Mean'. In B. Couture (ed.) Briggs, A. (1993) `Social History'. The Blackwell Dictionary of Twentieth-Century Social Thought. W. Outhwaith & T. Bottomore (eds.). Oxford: Blackwell Publishers: 595597. Brown, G. & G. Yule (1983) Discourse Analysis. Cambridge, UK: Cambridge University Press. Bolinger, D. (1952) `Linear Modification'. PMLA 67, The Modern Language Association: 1117-1144. Bowers, F. (1988) ` Thematization in Legislative Language: the Observations of Bentham and Coode in Relation to the FG Definition of Theme'. In E.H. Steiner & R. Vetman (eds.) (1988) Pragmatics, Discourse and Text. London: Pinter: 90-98. Burke, P. (1991) New Perspectives on historical writing. Cambridge: Polity. 287
Buteau, M. (1970) `Student Errors and the Learning of French as a Second Language'. International Review of Applied Linguistics 8: 133-145. Butler, C. (1998). `Towards the Millenium: Some Suggestions for the Future of Modern English Language Studies.' Boletнn de la Asociaciуn Espaсola de Estudios AngloAmericanos (AEDEAN), Noviembre 1998. Madrid: Universidad Nacional de Educaciуn a Distancia: 71-85. The Cambridge Dictionary of Philosophy (1995). R. Audi (ed). Cambridge, UK: Cambridge University Press. Carretero, M. (1991-2) `Una Propuesta de la Modalidad: La Aceptaciуn como Categorнa Modal'. DICENDA. Cuadernos de Filologнa Hispбnica, 10: 41-61. Carretero, M., A. Lуpez-Maniуn & L. Jacott (1997) `Explaining Historical Events'. International Journal of Educational Research, 27/3: 245-253. Chartrand, M. & B. Rising (1995) `A Contrastive Analysis of Company Memos in Spanish and English'. In S.Barrueco, E. Hernбndez, L. Sierra (eds.) Lenguas para fines especificos 4. Investigaciуn y Enseсanza. Universidad de Бlcala de Henares Chomsky, N. (1981) Lectures on Government and Binding. Dordrecht, The Netherlands: Foris. Chomsky, N. (1982) Some Concepts and Consequences of the Theory of Government and Binding. Cambridge, Mass: MIT Press. Christie, F. (1985) Language Education. Oxford: Oxford University Press. Cicero (1954) Ad Herennium. Trans. Harry Caplan. Cambridge, Mass: Harvard UP. Clark, H.H. & Haviland, S. E. (1977) `Comprehension and the Given-new Contract. In R.O. Freedle (ed.) Discourse Production and Comprehension. Norwood, NEW JERSEY: Ablex: 1-40. Cloran, C. (1995) ' Defining and Relating Text Segments: Subject and Theme in Discourse'. In R. Hasan & P.H. Fries, (eds.): 361-403. Clyne, M. (1987) `Cultural Differences in the Organization of Academic Texts'. Journal of Pragmatics, 11, 211-247. Cook, M. (1988) The Validity of the Contrastive Rhetoric Hypothesis as it Relates to Spanish-speaking Advanced ESL Students. Unpublished Doctoral Dissertation, Stanford University. Connor, U. (1984) `A Study of Cohesion and Coherence in ESL Students' Writing'. Papers in Linguistics: International Journal of Human Communication 17/3: 301-316. Connor, U. (1996) Contrastive Rhetoric. Cambridge; UK: Cambridge University Press. 288
Connor, U. & M. Farmer (1990) `The Teaching of Topical Structure Analysis as a Revision Strategy for ESL Writers.' In B. Kroll (ed.) Second Language Writing. Cambridge, UK: Cambridge University Press: 126-139. Connor, U. & R. Kaplan (eds.) (1987) Writing Across Languages: Analysis of L2 Text. Addison-Wesley Publishing Company. Connor, U. & P. McCagg (1993) `Cross-cultural Differences and Perceived Quality in Written Paraphrases of English Expository Prose'. Applied Linguistics 4/3: 259267. Coulthard, M. (ed.) (1994) Advances in Written Text Analysis. London and New York: Routledge. Couture, B. (ed.) (1986) Functional Approaches to Writing, Research Perspectives. Norwood, New Jersey: Ablex. Crookes, G. (1986) `Towards a Validated Analysis of Scientific Text Structure'. Applied Linguistics, 7: 57-70. Cummings, M. (1995) 'A Systemic Functional Approach to the Thematic Structure of the Old English Clause'. In R. Hasan & P.H. Fries, (eds.): 275-316. Danes, F. (1970) `One Instance of Prague School Methodology: Functional Analysis of Utterance and Text'. In P. L. Garvin (ed.) Method and Theory in Linguistics. The Hague: Mouton. Danes, F. (1974) `Functional Sentence Perspective and the Organization of the Text'. In F. Danes (ed.) Papers in Functional Sentence Perspective. Prague: Academia: 106128. Danes, F. (1989) `"Functional Sentence Perspective" and Text Connectedness'. In M.E. Conte, J. Petцfi, & E. Sцzer (eds.) Text and Discourse Connectedness: Proceedings of the Conference on Connexity and Coherence. Urbino, July 16-21, 1984. Amsterdam: John Benjamins: 23-31. Davies, F. (1987) `Reading Between the Lines: Thematic Choice as a Device for Presenting Writer Viewpoint in Academic Discourse'. ELR Seminar Paper, University of Birmingham. Davies, F. (1995) `Marked Theme as a Heuristic for Analysing Text-type, Text and Genre'. Paper presented at the BAAL/CUP Genre Analysis Workshop, Sheffield, July 1995. Davies, M. (1997) `Cohesion in Literary Texts'. Talk given at Saint Louis University, Madrid Campus, March 3rd, 1997. De Castell, S., A. Luke & C. Luke (eds.) (1989) Language, Authority and Criticism: Readings on the School Textbook. London: Falmer Press. 289
De Miguel, E. (1988) `Construcciones Ergativas e Inversiуn en la Lengua y la Interlengua Espaсola'. In J. M. Liceras (ed.) Lingьнstica y adquisiciуn de lenguas segundas: 753-766. Derrida, J. (1990) `Jacques Derrida on Rhetoric and Composition: A Conversation'. Journal of Advanced Composition 10/1: 1-21. Devitt, A.J. (1993) `Generalizing about Genre: New Conceptions of an Old Concept'. College Composition and Communication, 44/4: 573-586. Dik, Simon C. (1980) Studies in Functional Grammar. London: Academic Press. Downing, A. & P. Locke (1992) A University Course in English Grammar. Prentice Hall International. Downing, A. (1991) `An Alternative Approach to Theme: A Systemic-functional Perspective.' WORD, 42/2: 119-143. Dubois, B.L. (1987) `A Reformulation of Thematic Progression Typology'. TEXT, 7/2: 89116. Dudley-Evans, T. & W. Henderson (1990) `The Organization of Article Introductions: Evidence of Change in Economics Writing'. In T. Dudley-Evans & W. Henderson (eds.) The Language of Economics: The Analysis of Economics Discourse, ELT Documents, 134. Modern English Publications/British Council. Eggins, S. (1994) An Introduction to Systemic Functional Linguistics. London: Pinter Publishers Ltd. Eggins, S., P. Wignell & J.R. Martin (1993) `The Discourse of History: Distancing the Recoverable Past'. In M. Ghadessy (ed) Register Analysis: Theory and Practice. London: Pinter Publishers. Eiler, M. (1986) `Thematic distribution as a heuristic for written discourse function. In B. Couture (ed.). Ellis, R. (1985) Understanding Second Language Acquisition. Oxford: Oxford University Press. Enkvist, N.E. (1984) `Contrastive Linguistics and Text Linguistics'. In J. Fisiak (ed.) Contrastive Linguistics: Prospects and Problems. Mouton de Gruyter: 45-67. Fawcett, R.P. (1987). `The Semantics of Clause and Verb for Relational Processes in English'. In M.A.K. Halliday & R. P. Fawcett (eds.) New Developments in Systemic Linguistics. Volume 1. London: Frances Pinter: 130-183. Fernбndez Rodrнguez, S. (1986) Gramбtica Espaсola 4. El Verbo y la Oraciуn. Arco Libros, S.A. 290
Fillmore, C .J. (1968)`The Case for Case'. In Bach, E. & R. Harms (eds.) Universals in Linguistic Theory. New York: Holt, Rinehart & Winston: 1-88. Fillmore, C. J. (1971) `Types of Lexical Information'. In D. D. Steinberg & L. A. Jakobovits (eds.) Semantics. Cambridge University Press: 370-392. Firbas, J. (1986) `On the Dynamics of written communication in the Light of the Theory of Functional Sentence Perspective.' In C.R. Cooper & S. Greenbaum (eds.) Studying Writing: Linguistic Approaches. London: Sage: 40-71. Flynn, S. (1987) A Parameter-Setting Model of L2 Acquisition: Experimental Studies in Anaphora. Dordrecht, The Netherlands: Reidal. Flynn, S. (1989) `The Role of the Head-initial/Head-final Parameter in the Acquisition of English Relative Clauses by Adult Spanish and Japanese speakers'. In Susan Gass & Jacquelyn Schachter, (eds.): 89-108. Francis, G. (1989) `Thematic Selection and Distribution in Written Discourse. Word, 40, No. 1-2: 201-222. Francis, G. (1990) `Theme in the Daily Press'. Occasional Papers in Systemic Linguistics, 4: 51-87. Francis, G. (1994) `Labelling Discourse: An Aspect of Nominal Group Cohesion. In M. Coulthard (ed.): 83-101. Francis, G. & A. Kramer-Dahl (1991) `From Clinical Report to Clinical Story: Two Ways of Writing about a Medical Case.' In E. Ventola (ed.): 339-368. Fries, P. H. (1983) `On The Status Of Theme In English: Arguments From Discourse'. In J. S. Petцfi & E. Sцzer (eds.) Micro and Macro Connexity of Text. 6, 1, 1-38. Hamburg: Helmut Buske Verlag (Papers in Text Linguistics 45): 116-152. Fries, P. H. (1992) `Information Flow in Written Advertising'. Georgetown University Roundtable on Language and Linguistics 1992: 336-352. Fries, P. H. (1993) `The Structuring of Information in Written English Text'. Language Sciences 14/4: 461-488. Fries, P. H. (1994) `On Theme, Rheme and Discourse Goals'. In M. Coulthard (ed.): 229249. Fries, P. H. (1995a) `Patterns of Information in Initial Position in English'. In P. Fries & M. Gregory (eds.) Discourse in Society: Systemic Functional Perspective. Norwood, New Jersey: Ablex Publishing Corporation: 47-66. Fries, P. H. (1995b) `Themes, Methods of Development, and Texts'. In R. Hasan & P. Fries, (eds.): 317-359. 291
Fries, P. H. (1995c) `A Personal View of Theme'. In M. Ghadessy (ed.). Fries, P. H. & G. Francis (1992) ` Exploring Theme: Problems for Research.'Occasional Papers in Systemic Linguistics Volume 6: 45-59. Fuller, G. & K. Gundel (1987) ` Topic Prominence in Interlanguage.'Language Learning 37/1: 1-18. Garcнa Ruнz, V. (1985) `Orden de los Elementos Oracionales en Espaсol'. RILCE, I/1: 91111. Gass, S.M. & J. Schachter (eds.) (1989) Linguistic Perspectives on Second Language Acquisition. Cambridge: Cambridge University Press. Gellner, E. (1973) Cause and Meaning in the Social Sciences. London: Routledge and Kegan Paul. Ghadessy, M. (1995) `Thematic Development and its Relationship to Registers and Genres'. In M. Ghadessy (ed.): 129-145. Ghadessy, M. (ed.) 1995. Thematic Development in English Texts. London: Pinter. Givуn, T. (1995). Functionalism and Grammar. Amsterdam/Philadelphia: John Benjamins Publishing Company. Gуmez, M.G. (1995) `Theme: A Heuristic Method for Discourse Analysis'. Estudios Ingleses de la Universidad Complutense, 3. Madrid: Servicio Pulicaciones, UCM: 43-54. Gosden, H. (1996) A Genre-Based Investigation of Theme: Product and Process in Scientific Research Articles Written by NNS Novice Researchers. Monographs in Systemic Linguistics, Number 7. University of Nottingham: Department of English Studies. Grabe, W. (1987) ` Contrastive Rhetoric and Text-Type Research' . In U. Connor & R. Kaplan (eds.): 115-137. Greenberg, J. H. (1963) Universals of Language. Cambridge, Mass: MIT Press. Gregg, K. R. (1989) ` Second Language Acquisition Theory: The Case for a Generative Perspective' . In S. Gass & J. Schachter (eds.): 15-40. Hakulinen, A. (1989) ` Some Notes on Thematics, Topic and Typology' . In M.E. Comte, J. S. Petцfi & E. Sцzer (eds.) Text and Discourse Connectedness: Proceedings of the Conference on Connexity and Coherence. Urbino, July 16-21, 1984. Amsterdam: John Benjamins: 53-63. 292
Halldйn, O. (1997) `Conceptual Change and the Learning of History'. International Journal of Educational Research, 27/3: 201-210. Halliday, M.A.K. (1967a) ` Notes on Transitivity and Theme in English. Part 1.Journal of Linguistics 3/1: 37-81. Halliday, M.A.K. (1967b) ` Notes on Transitivity and Theme in English. Part 2. Journal of Linguistics 3/2: 199-244. Halliday, M.A.K. (1968) ` Notes on Transitivity and Theme in English. Part 3.Journal of Linguistics 4/2: 179-215. Halliday, M.A.K. (1970) ` Language Structure and Language Function' . In J. Lyons (ed.) New Horizons in Linguistics. Harmondsworth: Penguin: 140-165. Halliday, M.A.K. (1976) ` Theme and Information in the English Clause' . In G. Kress (ed.) System and Function in Language. Oxford University Press: 174-188. Halliday, M.A.K. (1985a) Introduction to Functional Grammar. London; Edward Arnold. Halliday, M.A.K. (1985b) `It's a Fixed Word Order Language is English'. I.T.L. Review of Applied Linguistics, 67-68: 91-116. Halliday, M.A.K. (1987) ` Language and the order of nature' . In N. Fabb, D. Attridge, A. Duran & C. MacCabe (eds.) The Linguistics of Writing: Arguments between language and literature. Manchester: Manchester University Press: Halliday, M.A.K. (1988) `On the Ineffability of Grammatical Categories'. In J. D. Benson, M. J. Cummings & W. S. Greaves (eds.) Linguistics in a Systemic Perspective. Amsterdam: John Benjamins: 27-51. Halliday, M.A.K. (1993a) ` Some Grammatical Problems in Scientific English'. In M.A.K. Halliday & J.R. Martin Writing Science: Literacy and Discursive Power. London: The Falmer Press: 69-85. Halliday, M.A.K. (1993b) `The Construction of Knowledge and Value in the Grammar of Scientific Discourse: Charles Darwin' s "The Origin of the Species"I.n M.A.K. Halliday & J.R. Martin Writing Science: Literacy and Discursive Power. London: The Falmer Press: 86-105. Halliday, M.A.K. (1994) Introduction to Functional Grammar. London; Edward Arnold. Halliday, M.A.K. (1996) `Literacy and Linguistics: a Functional Perspective'. In R. Hasan & G. Williams (eds.): 339-376. Halliday, M.A.K. & R. Hasan (1976) Cohesion in English. Longman. 293
Hartmann, R.R.K. (1980) Contrastive Textology. Comparative Discourse Analysis in Applied Linguistics (Studies in Descriptive Linguistics, vol. 5). Heidelberg: Julius Groos Verlag. Hasan, R. & P. H. Fries (eds.) (1995) On Subject and Theme. Amsterdam/Philadelphia: John Benjamins Publishing Company. Hasan, R. & G. Williams (1996) Literacy in Society. Addison Wesley Longman Higher Education. Hatcher, A. G. (1956) `Theme and Underlying Question: Two Studies of Spanish Word Order'. WORD, 12, supplement, monograph 3. Hawes, T. P. & S. Thomas (1996) `Rhetorical Uses of Theme in Newspaper Editorials'. World Englishes, 15/2: 159-170. Henner-Stanchina, C. (1985) `From Reading to Writing Acts.' In P. Riley (ed.) Discourse and Learning. Longman: 91-104. Hexter, J.H. (1993) `History'. The Blackwell Dictionary of Twentieth-Century Social Thought. W. Outhwaith & T. Bottomore (eds.). Oxford: Blackwell Publishers: 262266. Hinds, J. (1987) `Reader Versus Writer Responsibility: A New Typology'. In U. Connor & R. Kaplan, (eds.) Hitchens, C. (1998) `Goodbye to All That: Why Americans are not Taught History'. Harper's Magazine, 297/1782. Holmes, R. (1997) `Genre Analysis and the Social Sciences: An Investigation of the Structure of Research Article Discussion Sections in Three Disciplines.' English for Specific Purposes, 16/4, 321-337. Houghton, D. (1980) `Contrastive Rhetoric'. English Language Research Journal, 1: 7991. Huebner, T. (1983) A Longitudinal Analysis of the Acquisition of English. Ann Arbor, MI: Koronma Press. James, C. (1994a) `Don't Shoot My Dodo: On the Resilience of Contrastive and Error Analysis'. IRAL XXXII/3: 179-200 James, C. (1980) Contrastive Analysis. London: Longman. Jimйnez, M.L. (1997) Subjects in Spanish: Semantic and Linguistic Conditions on Word Order in Spanish. Unpublished Doctoral Dissertation. Georgetown University. Jimйnez Juliб, T. (1984) `Eje Temбtico y Tema en Espaсol'. In M. Casado Velarde, A. Freire Llamas, J. E. Lуpez Pereira y J. I. Pйrez Pascual (eds.) Scripta Philologica in 294
Memorium Manual Tacloade Cid. Tomo 1. Servicio de Publicaciones: Universidade da Coruсa: 453-492. Jin, H. G. (1994) `Topic-Prominence and Subject Prominence in L2 Acquisition: Evidence of English-to-Chinese Typological Transfer'. Language Learning 44/1: 101-122. Jones, L. K. (1977) Theme in English Expository Discourse. Lake Bluff, Illinois: Jupiter Press. Kaplan, R. (1995) `Contrastive Rhetoric'. In T. Miller (ed.) Functional Approaches to Written Text: Classroom Applications. The Journal of TESOL-France in association with U.S. Information Service: 21-39. Kaplan, R. (1966) `Cultural Thought Patterns in Intercultural Education'. Language Learning, XVI/1&2. Krzeszowski, T. P. (1990) Contrasting Languages: The Scope of Contrastive Linguistics. Berlin: Mouton de Gruyter (Biblioteca Nacional) Lavid Lуpez, J. (1994a) Thematic Development in Texts. Documento interno R1.2.1, ESPRIT Basic Research Action: Dandelion, EP6665, Universidad Complutense de Madrid. Lavid Lуpez, J. (1994b) Theme, Discourse Topic and Information Structuring. Documento interno R1.2.2b, ESPRIT Basic Research Project: Dandelion, EP6665, Universidad Complutense de Madrid. Larsen-Freeman, D. & M. H. Long (1991) An Introduction to Second Language Acquisition Research. Longman. Leinhardt, G. (1997) `Instructional Explanations in History'. International Journal of Educational Research, 27/3: 221-232. Lemke, J. (1998) Contribution to the Sysfling Discussion List ([email protected]). Topic: Genre versus register. 30 September1998. Levinson, S. C. (1983) Pragmatics. Cambridge, UK: Cambridge University Press. Liceras, J. M.(1989) `On Some Properties of the "Pro-drop" Parameter: Looking for Missing Subjects in Non-native Spanish'. In S. M. Gass & J. Schachter (eds.): 109133. Liu, Y. (1995) `The Generic Triad in Classical Rhetoric'. College English, 57/1: 9-26. Lуpez, G. (1982) Article Introductions in Spanish: A Study in Comparative Rhetorics. Unpublished Masters dissertation. University of Aston, UK. Lуpez Garcнa, Б. (1994) Gramбtica del Espaсol. Madrid: Arco Libros, S.L. 295
Lotfipour, K. & F. Rezai-Tajani (1996) `Exploration in Thematization Strategies and their Discoursal Values in English'. Text 16/2: 225-249. Lowe, I. (1987) `Sentence Initial Elements in English and their Discourse Function'. Occasional Papers in Systemic Linguistics, 2: 5-33. Luke, C. S. de Castell, S. & A. Luke (1989) `Beyond Criticism: The Authority of the School Textbook'. In S. de Castell, A. Luke & C. Luke (eds.). Lux, P. (1993) `Identifying Discourse Features Across Language: English and Spanish.' Paper presented at the 1993 TESOL Conference, Atlanta, GA. Lyons, J. (1977) Semantics. Volume II. Cambridge: Cambridge University Press. Mann, W. C., C. M.I.M. Matthiessen, & S. A. Thompson (1992) `Rhetorical Structure Theory and Text Analysis.' In W.C. Mann & S. A. Thompson (eds.) Discourse Descriptions: Diverse Analyses of a Fund-Raising Text. Amsterdam: Benjamins. Marrou, H.-I. (1966)The Meaning of History, translated by R. J. Olsen. Baltimore: Helicon. In Berkhofer, Robert (1989). Martin, J. R. (1993) `Technicality and Abstraction: Language for the Creation of Specialized Texts'. In M.A.K. Halliday & J. R. Martin Writing Science: Literacy and Discursive Power. London: The Falmer Press: 203-220. Martin, J. R. (1991) `Nominalization in Science and Humanities: Distilling Knowledge and Scaffolding Text.' In E. Ventola (ed): 307-337. Martin, J. R. (1992a) `Theme, Method of Development and Existentiality: The Price of Reply. Occasional Papers in Systemic Linguistics, Volume 6: 147-183. Martin, J. R. (1992b) English Text: System and Structure. John Benjamins Publishing Company. Martin, J. R. (1995) `More Than What the Message is About: English Theme'. In M. Ghadessy (ed.): 223-259. Martin, J.R. (1999) Contribution to the Sysfling Discussion List ([email protected]). Topic: Two Questions about Processes. 10 June 1999. Marwick, A. (1970) Common Pitfalls in Historical Writing. The Open University Press. Matthiessen, C. & J. R Martin (1991) `A Response to Huddleston's Review of Halliday's Introduction to Functional Grammar', Occasional Papers in Systemic Linguistics, 5: 5-74. Mauranen, A. (1993a) Cultural Differences in Academic Rhetoric. Peter Lang. 296
Mauranen, A. (1993b) `Contrastive ESP Rhetoric: Metatext in Finnish-English Economics Texts.' English for Specific Purposes 12: 3-22. Mauranen, A. (1993c) `Theme and Prospection in Written Discourse'. In M. Baker, G. Francis, & E. Tognini-Bonelli (eds.). Text and Technology: in honour of John Sinclair. Amsterdam: John Benjamins Publishing Company: 95-114. Mathesius, V. (1975) A Functional Analysis of Present Day English on a General Linguistic Basis. The Hague: Mouton. McCabe-Hidalgo, A. & A. Gomis (1994) `Cultural Input and the Teaching of Composition'. Paper given at Twenty-Eighth Annual TESOL Convention, Baltimore, Maryland, March 8-12. Meyer, B. J. F. (1975) The Organization of Prose and its Effect on Memory. Amsterdam: North Holland Publishing Company. Meyer, B. J. F. (1982) `Reading Research and the Composition Teacher: The Importance of Plans". College Composition and Communication, 33. Montaсo-Harmon, M. R. (1991) `Discourse Features of Written Mexican Spanish: Current Research in Contrastive Rhetoric and Its Implications'. Hispania 74, May. Moreno, A. I. (1997) `Genre Constraints Across Languages: Causal Metatext in Spanish and English RAs. English For Specific Purposes, 16/3: 161-179. Munday, J. (1997) Systems in Translation: A Computer-assisted Systemic Analysis of the Translation of Garcнa Mбrquez. PhD thesis, University of Bradford. Myers, G. (1989) `The Pragmatics of Politeness in Scientific Articles.' Applied Linguistics, 10/1: 1-35. Myers, G. (1994) `Narratives of Science and Nature in Popularizing Molecular Genetics'. In M. Coulthard (ed.): 179-190. Nash, G.B., C. Crabtree & R.E. Dunn (1997) History on Trial. Culture Wars and the Teaching of the Past. New York, NY: Random House. Newsham, G. (1977) The Paragraph in French and English. Unpublished Ph.D. Thesis. Facultй des Sciences et de l'Education, Universitй de Montreal. Nwogu, K. & T. Bloor (1991) `Thematic Progression in Professional and Popular Medical Texts.' In E. Ventola (ed.): 369-384. Nwogu, K. N. (1990) Discourse Variation in Medical Texts: Scheme, Theme and Cohesion in Professional and Journalistic Accounts. Monographs in Systemic Linguistics, Volume Two. University of Nottingham: Department of English Studies. 297
Ocampo, F. (1990) `The Pragmatics of Word Order in Constructions with a Verb and a Subject'. Hispanic Linguistics 4:1 (Spring): 87-127 Oller, J.W. (1971) `Difficulty and Predictability'. Working Papers in Linguistics (Hawaii) 3/4. 79-98. Olson, D.R. (1989) `On the Language and Authority of Textbooks'. In S. de Castell, A.Luke & C. Luke (eds.) Ostler, S. (1987) `English in Parallels: A Comparison of English and Arabic Prose'. In U. Connor & B. Kaplan (eds.) Pak, C. (1995) Editorials in Mexico, Spain and the USA: A Comparative Rhetorical Analysis. Unpublished Doctoral Dissertation: The University of Michigan, Ann Arbor. Payne, D. L. (1990) The Pragmatics of Word Order: Typological Dimensions of Verb Initial Languages. Berlin: Mouton de Gruyter. Peccei, Robert & Fred Eiseling (1996) `Literacy for the 21st Century'. Los Angeles Times, 26 Febuary: 204. Phillips, D. C. (1987) Philosophy, Science, and Social Inquiry. Pergamon Press. Phinny, M. (1987) `The Pro-drop Parameter in Second Language Acquisition'. In Parameter Setting, T. Roeper & E. Williams, (eds.) Dordrecht, The Netherlands: Reidal. Pindi, M. & T. Bloor (1987) ` Playing Safe with Prediction: Hedging, Attribution and Conditions in Economic Forecasting'. In T. Bloor & J. Norrish (eds.). Written Language. London: CILT for BAAL: 55-69. Prades, J. (1997) `La Historia їEra ESO?' El Paнs, November 2, 1997. Prince, E. F. (1979) `On the Given/New Distinction'. Papers from the 15th Regional Meeting, Chicago Linguistic Society: 257-278. Prince, E. F. (1981) `Toward a Taxonomy of Given-new Information'. In Peter Cole (ed.) Radical Pragmatics. New York: Academic Press: 223-255. Purves, A. C. (1988) Writing Across Languages and Cultures: Issues in Contrastive Rhetoric. Newbury Park: Sage Publications, Inc. Quirk, R., S. Greenbaum, G. Leech & J. Svartvik (1972) A Grammar of Contemporary English. Essex: Longman Group UK Limited. Real Academia Espaсola (1979) Esbozo de una Nueva Gramбtica de la Lengua Espaсola. Madrid: Espasa-Calpe. 298
Real Academia Espaсola (1986). Esbozo de una Nueva Gramбtica de la Lengua Espaсola. Madrid: Espasa-Calpe. Rйgent, O. (1985) `A Comparative Approach to the Learning of Specialized Written Discourse'. In P. Riley (ed.) Discourse and Learning. Longman: 105-120. Reid, J. (1988) `Quantitative Differences in English Prose Written by Arabic, Chinese, Spanish, and English Writers. Unpublished Doctoral Dissertation, Colorado State University. Reid, J. (1990) `Responding to Different Topic Types: a Quantitative Analysis from a Contrastive Rhetoric Perspective.' In B. Kroll (ed.) Second Language Writing. Cambridge: Cambridge University Press. Richards, K. (1993)'Direction and Debate in Distance Materials for Teacher Development'. In R. Howard & I. McGrath (eds.) Distance Education for Language Teachers. University of Edinburgh: Institute of Applied Language Studies: 99-112. Ricoeur, P. (1989) `Rhetoric-Poetics-Hermeneutics'. In M. Meyer (ed.) From Metaphysics to Rhetoric. Dordrecht: Kluwer Academic Publishers. Rosen, H. (1972) `The Language of Textbooks'. Language in Education. London and Boston: Routledge & P. Kegan in association with The Open University Press: 119125. Rudolph, E. (1987) `Connective Relations - Connective Expressions - Connective Structures'. In J. S. Petцfi (ed.) Text and Discourse Constitution. Walter de Gruyter. Rutherford, W. (1983) `Language Typology and Language Transfer.' In S.M. Gass & L. Selinker (eds.) Language Transfer in Language Learning. Rowley, Mass: Newbury House. Santana-Seda, O. (1975) A Contrastive Study in Rhetoric: An Analysis of the Organization of English and Spanish Paragraphs written by Native Speakers of Each Language. (Diss. New York University, 1974). Dissertation Abstracts International, 35.6681A. Santiago, R. (1971) A Contrastive Analysis of Some Rhetorical Aspects in the Writing in Spanish and English of Spanish-Speaking College Students in Puerto Rico. (Diss. Columbia University, 1970). Dissertations Abstracts International, 31.6368A. Schlobinski, P. & S. Schutze-Coburn (1992) `On the Topic and Topic Continuity'. In N. Dittmar (ed.) Topic - From Grammar to Discourse. Linguistics, 30/1: 89-122. Searle, J. R. (1969) Speech Acts: An Essay in the Philosophy of Language. Cambridge University Press. 299
Silva Corvalбn, C. (1984) `Topicalizaciуn y Pragmбtica en Espaсol'. Revista Espaсola de Lingьнstica, XIV/1: 1-20. Stearns, P.N. (1993) `The Old Social History and the New'. Encyclopedia of American Social History. Vol. 1. M. K. Cayton, E.J. Gorn, P.W. Williams (eds.). New York: Charles Scribner's Sons: 237-250. Stockwell, R. & J. Bowen (1965) The Sounds of English and Spanish. Chicago: Chicago University Press. Stockwell, R., J. Bowen, & J. Martin (1965) The Grammatical Structures of English and Spanish. Chicago: Chicago University Press. Strei, G. (1971) `A Contrastive Study of the Structure of Rhetoric in English and Spanish Composition'. Unpublished Masters Thesis, McGill University. Suсer, M. (1982) Syntax and Semantics of Spanish Presentational Sentence Types. Washington, D.C. Georgetown University Press. Suzuki, T. (1975) Tozaserta Gengo: Nihongo no sekai [A Bound Language: The world of Japanese]. Tokyo: Shinshiosha. Swales, J.M. (1981) Aspects of Article Introductions. Aston ESP Research Reports No. 1, Birmingham, UK: University of Aston Language Studies Unit. Swales, J.M. (1990) Genre Analysis. Cambridge: Cambridge University Press. Tadros, A. (1994) `Predictive Categories in Expository Text'. In M. Coulthard (ed.) Taylor, C. (1983) `Structure and Theme in Printed School Text.' Text, 3/2: 197-228 Taylor, G. & T. Chen (1991) `Linguistic, Cultural and Subcultural Issues in Contrastive Discourse Analysis: Anglo-American and Chinese Texts'. Applied Linguistics, 3: 319-336. Taylor, H.M. (ed) (1979) English & Spanish in Contrast. Regents Publishing Company. Teskey, P.D. (1976) Theme and Rheme in Spanish and English, Coleraine: New University of Ulster. Thomas, S. & T. Hawes (1997) Theme in Academic and Media Discourse. Monographs in Systemic Linguistics, Number Eight. University of Nottingham, Department of English Studies. Thompson, S.A. (1978) `Modern English from a Typological Point of View: Some Implications of the Function of Word Order'. Linguistische berichte, 54: 19-35. Thompson, S.A. (1985) `Grammar and Written Discourse: Initial vs. Final Purpose Clauses in English'. Text, 5/1-2: 55-84. 300
Tilly, C. (1985) "Retrieving European Lives". In O. Zunz (ed.) Reliving the Past: The Worlds of Social History. Chapel Hill: University of North Carolina Press: 11-52. Travnicek, F. (1962) `O Tak Zvanim Aktualnim Cleneni Vetnнm' [On so-called functional sentence perspective]. Slovo a slovesnost (Prague) 22: 163-171. Translated in Vasconcellos, 1985, (Appendix III: 229-248) by L. Valyiova. Valdйs, G., T. Dvorak & T. Pagбn Hannum (1984) Composiciуn: Proceso y Sнntesis. New York: Random House. Valero-Garcйs, C. (1997) `A Cross-Linguistic Study of the Verbal Syntagm: A Case Study of Economic Texts in English and Spanish'. UNESCO PRSED-LSP Newsletter. 20/1. 25-39. Valero-Garcйs, C. (1996) `Contrastive ESP Rhetoric: Metatext in Spanish-English Economics Texts'. English For Specific Purposes, 15/4: 279-294 Van Dijk, T.A. & W. Kintsch (1983) Strategies in Discourse Comprehension. Academic Press, Inc. Vande Kopple, W.J. (1985) `Some Exploratory Discourse on Metadiscourse'. College Composition and Communication, 36/1: 82-94. Vande Kopple, W.J. (1986) `Given and New Information and Some Aspects of the Structures, Semantics, and Pragmatics of Written Texts.' In C.R. Cooper & S. Greenbaum (eds.) Studying Writing: Linguistic Approaches. London: Sage. Vande Kopple, W.J. (1991) `Themes, Thematic Progressions, and Some Implications for Understanding Discourse'. Written Communication 8 (July): 311-347. Van Lier, L. (1994) `Forks and Hope: Pursuing Understanding in Different Ways'. Applied Linguistics 15/3: 328-346. Vasconcellos, M. (1992) `The Theme as Message Onset: Its Structure and Characteristics'. In N. Dittmar (ed.) Topic - From Grammar to Discourse. Linguistics, 30/1: 147163. Vasconcellos, M. (1985) `Theme and Focus: Cross-Language Comparison via Translations from Extended Discourse'. Unpublished Doctoral Dissertation, Georgetown University. Veel, R. & C. Coffin (1996) `Learning to Think like an Historian: The Language of Secondary School History' In R. Hasan & G. Williams (eds.): 191-231. Ventola, E. (ed.) (1991) Functional and Systemic Linguistics: Approaches and Uses. Mouton de Gruyter. Ventola, E. (1995) `Thematic Development and Translation'. In M. Ghadessy (ed): 85-104. 301
Ventola, E.& A. Mauranen (1991) `Non-native Writing and Native Revising of Scientific Articles.' In E. Ventola (ed): 457-492. Virtanen, T. (1992) `Given and New Information in Adverbials: Clause-initial Adverbials of Time and Place.' Journal of Pragmatics 17: 99-115. von Wright, G.H. (1971) Explanation and Understanding. Ithaca, NY: Cornell University Press. Weissberg, R.C. (1984) `Given and New: Paragraph Development Models from Scientific English.' TESOL Quarterly 18/2: 485-499. White, H. (1973) Metahistory: The Historical Imagination in Nineteenth-Century Europe. Baltimore: Johns Hopkins University Press. White, L. (1988) `Island Effects in Second Language Acquisition'. In S. Flynn & W. O'Neil (eds.) Linguistic Theory in Second Language Acquisition. Dordrecht, The Netherlands: Reidal. White, L.(1989) `The Adjacency Condition on Case Assignment: Do L2 Learners Observe the Subset Principle?'. In S. Gass & J. Schachter (eds.) White, P.R.R. (1998) `Towards a Grammar of Power and Solidarity - Developments in the Linguistics of Evaluation, Inter-Subjectivity and Ideological Position". Paper presented at the 10th Euro-International Systemic Function Workshop. The University of Liverpool, July 22-25. White, P.R.R. (1999) Contribution to the Sysfling Discussion List ([email protected]). Topic: Two Questions about Processes. 9 June 1999. Whitley, S.M. (1986) Spanish/English Contrasts. Washington, D.C. Georgetown University Press. Whittaker, R. (1995) `Thematic Development in Academic and Non-academic Texts'. In M. Ghadessy (ed.): 105-128. Widdowson, H.(1979) Explorations in Applied Linguistics. Oxford: Oxford University Press. Winterowd, W.R. (1968) Rhetoric: A Synthesis. New York: Holt, Reinhart & Winston. Witte, S. P. & R. D. Cherry (1986) `Writing Processes and Written Products in Composition Research.' In C.R. Cooper & S. Greenbaum (eds.) Studying Writing: Linguistic Approaches. London: Sage. Witte, S.P. (1983) `Topical Structure and Revision: An Exploratory Study. College Composition and Communication 34/3. 302
Xiao, Q. (1991) `Toward Thematic Selection in Different Genres: Fables and recipes'. Unpublished manuscript (as mentioned in Ghadessy, 1995). Young, R.E., A.L. Becker, & K.L. Pike (1970) Rhetoric: Discovery and Change. Harcourt, Brace & World, Inc. Texts Analyzed English Corpus: Brinton, C., J. B. Christopher, & R. L. Wolff (1976) A History of Civilization: 1715 to the Present. Chapter 16. The Old Regime and the International Balance. Englewood Cliffs, New Jersey: Prentice-Hall Inc. (Eng 10). Chambers, M., R. Grew, D. Herlihy, T. K. Rabb, & I. Woloch (1991) The Western Experience. Volume I. Chapter 1: The First Civilizations. McGraw Hill, Inc. (Eng 2). Chambers, M., R. Grew, D. Herlihy, T. K. Rabb, & I. Woloch (1991) The Western Experience. Volume II. Chapter 11: The West In Transition: Economy And Institutions 1300-1500. McGraw Hill, Inc. (Eng 3). Chambers, M., R.Grew, D. Herlihy, T. K. Rabb, & I. Woloch (1991) The Western Experience. Introduction. McGraw Hill, Inc. (Eng 1). Greer, T. H. & G. Lewis (1992) A Brief History of the Western World. Chapter 1: The Birth of Civilization in the Middle East. Harcourt Brace Jovanovich, Inc. (Eng 5). Greer, T. H. & G.Lewis (1992) A Brief History of the Western World: Preface. Harcourt Brace Jovanovich, Inc. (Eng 4). McKay, J. P., B. D. Hill & J. Buckler (1995) A History of Western Society. Volume A: From Antiquity to 1500. Chapter 12: The Crisis of the Later Middle Ages. Boston: Houghton Mifflin Company. (Eng 6). McKay, J. P., B. D. Hill & J. Buckler (1995) A History of Western Society: Volume C, From the Revolutionary Era to the Present. Chapter 26 The West And The World. Boston: Houghton Mifflin Company. (Eng 9). Perry, M. (1997) Western Civilization: A Brief History. Chapter 3: The Greeks: From Myth to Reason. Boston: Houghton Mifflin Company. (Eng 7). Willis, F. R. (1981) Western Civilization: An Urban Perspective. Volume I. Chapter 4: The Rome of Augustus. D.C. Heath and Company. (Eng 8). 303
Spanish Corpus: Ballesteros, M. y J. L. Alborg (1973) Historia Universal: Volumen I. Capitulo Primero: Introducciуn al Estudio de la Historia. Madrid: Editorial Gredos, S.A. (Spa 1). Ballesteros, M. y J. L. Alborg (1973) Historia Universal: Volumen II. Capнtulo XXVI: El Siglo XIII Francйs y Los Siglos XIII y XIV Espaсoles. Madrid: Editorial Gredos, S.A. (Spa 3). Dominguez Ortiz, A. (1983) Historia Universal. Volumen III. Capнtulo XXXIII: Economнa y Sociedad en la Europa del Siglo XVIII. Barcelona, Vicens-Vives. (Spa 7). Fernбndez, A. (1986) La Historia del Mundo Contemporбneo. A los alumnos. Barcelona: Vicens-Vives, S.A. (Spa 8). Fernбndez, A. (1986) La Historia del Mundo Contemporбneo. Capнtulo 1: De las Sociedades Agrarias a las Sociedades Industriales. Barcelona: Vicens-Vives, S.A. (Spa 2). Jordб Cerdб, F., M. Pellicer Catalбn, P. Acosta Martнnez y M. Almagro-Gorbea (1989) Historia de Espaсa, Vol. 1. VI. Bronce Final y Edad Del Hierro. La Formaciуn de las Etnias y Culturas Preromanas Madrid: Editorial Gredos. (Spa 9). Mangas, J. (1991) Textos para la Historia Antigua de Grecia. La Grecia del Milenio II a.C. Ediciones Cбtedra, S.A. (Spa. 4). Palacio Atard, V., L. Alvarez Gutiйrrez, y A. Burgoa (1988) Historia del Mundo Contemporбneo. 2.1 La Explosiуn Demogrбfica. Zaragoza: Editorial Luis Vives. (Spa 5). Palacio Atard, V., L. Alvarez Gutiйrrez, y A. Burgoa (1988) Historia del Mundo Contemporбneo. 1.1 La Herencia Ideolуgica. Zaragoza: Editorial Luis Vives. (Spa 10). Santos Yanguas, N. Textos para la Historia Antigua de Roma. Capнtulo Primero. Ediciones Cбtedra, S.A. (Spa 6). 304
1

File: theme-and-thematic-patterns-in-spanish-and-english-history-texts.pdf
Title: Microsoft Word - printed PhD.doc
Published: Sun Jan 1 00:00:00 6
Pages: 305
File size: 1.62 Mb


RE in Secondary Schools, 2 pages, 0.76 Mb

e-Book Collection Title List, 117 pages, 0.54 Mb

, pages, 0 Mb

Maxims for revolutionists, 6 pages, 0.06 Mb

The Courier, 1 pages, 0.38 Mb
Copyright © 2018 doc.uments.com