Richard Abels, Vegetius, Sallust, battle, the Anglo-Saxon Chronicle, Sherston, medieval military history, Orderic Vitalis, Carolingian, Bernard S. Bachrach, Stephen Morillo, Ashingdon, description, John of Worcester, London, King Alfred, twelfth centuries, medieval chroniclers, Rumble, Agrimensores, De re militari, Worcester, Bertram Colgrave, Edmund Ironside, Bachrach, Middle Ages, Roman land, Ealdorman Byrhtnoth, Anglo-Saxon, Anglo-Saxon England, Archaeologist T. J. O'Leary, Alexander R. Rumble, Companion to Historiography, Professor Bachrach, Roman military tactics, Military Organization, Richard P. Abels, Marjorie Chibnall, William Stubbs, accurate description, extant manuscripts, William the Conqueror, Roman military, William of Poitiers, William of Malmesbury, Royal Frankish Annals, Bede, Ecclesiastical History, manuscript copy, classical Roman practice, R.A.B. Mynors, Oxford, Asser, Roman authors, classical learning, Alfred P. Smyth, Stephen Morillo John, Roman practice, Roman standards, battle of Ashdown, Guy Halsall, Dorothy Whitelock, Charles R. Bowlus, Roman military standards, Classical terminology, David Hill, Burghal Hidage, Charlemagne, closer to Caesar, Charles Bowlus, armed forces, Roman empire, administrative capabilities, military terminology, Anglo-Saxon Fortifications
1 A Lying Legacy? A Preliminary Discussion of Images of Antiquity and Altered Reality in Medieval Military History Richard Abels and Stephen Morillo Introduction In 1990 one of the co-authors of this article, Richard Abels, was asked by Donald Scragg to contribute a chapter on late tenth- and early eleventh-century English tactics and strategy to a volume of essays marking the millennium of the battle of Maldon. He agreed readily, though he was concerned about the paucity of source materials describing battles. Other than the poem of the battle of Maldon, he knew only two extended battle narratives for this period that might shed light on English tactics, John of Worcester's accounts of Edmund Ironside's victory over Cnut at Sherston and his subsequent defeat at Ashingdon in the year 1016.1 To the Anglo-Saxon Chronicle's accounts John added telling details. At Sherston, he tells us, when Edmund drew up his army according to the terrain and the forces he had, he moved the best soldiers into the front line
, placed the rest of the army in reserve [in subsidiis], and addressing each man by name, exhorted and entreated them to remember that they strove for country, children, wives and homes, and with these most inspiring words he fired the soldiers' spirits. Then he ordered the trumpets to sound, and the "cohorts" [cohortes] to advance gradually. The enemy army did the same. When they arrived at the place where they could join battle they rushed together with their hostile standards and with a great shout. They fought with spear and lance, striving with all their might. Meanwhile, King Edmund Ironside made his presence felt in fierce hand-to-hand fighting in the front line. He took thought for everything; he himself fought hard, often smote the enemy; he performed at once the duties of a hardy soldier and of an able general.2 1 The Chronicle of John of Worcester, vol. 1, ed. R. R. Darlington and P. McGurk, trans. Jennifer Bray and P. McGurk, Oxford Medieval Texts (Oxford, 1995), pp. 486 and 490 (s.a. 1016). 2 John of Worcester, Chronicle, 2:4867: Vbi exercitum pro loco et copiis instruit, optimum quenque in primam aciem subducit, ceterum exercitum in subsidiis locat unumquenque nominas appellat, hortatur, rogat, ut meminerint se pro patria, pro liberis, pro coniungibus, atque suis domibus certare, et optimis sermonibus militum animos accendebat, deinde tubicines canere et cohortes paulatim incedere iubet. Idem facit, hostium exercitus. Vbi eo uentum est ubi ab illis prelium committi posset, maximo clamore cum infestis signis oc currant, lanceis et gladiis pugna geritur, maxima ui certatur. Interea Eadmundus Ferreum Latus in prima acie comminus acriter insta bat, omnia provide bat, multum ipse pugna bat, sepe hostem feri bat, strenui militis et boni imperatoris officia sumul exequebatur. Italicized passages are from Sallust, Catiline, 59.160.4.
Richard Abels and Stephen Morillo
John's description of Ashingdon is similar. Edmund, we are told, "drew up his battle line with three lines of reserves" [triplicibus subsidiis aciem instruit] and attacked in that order Cnut's troops, who had deployed "on level ground."3 In comparison with the vague accounts in the Anglo-Saxon Chronicle of these and other battles, John's narratives are a treasure trove of detail about preparations for battle and deployment of troops, derived, Abels then believed, from either a lost recension of the Chronicle or some other near contemporary source.4 Edmund's approach to battle, as presented by John, was strikingly similar to that of the ancient Romans, and the resemblance was made even clearer by John's use of classical phrases such as copiis instruit and words like cohortes. After writing a draft of the article, Abels discovered why. John had lifted his accounts, almost word for word (though in highly edited form), from, respectively, Sallust's Catiline and his Jugurthine War.5 Chagrined, Abels removed his extended analyses of Sherston and Ashingdon, added a discussion of "military organization" to fill out the article, and hoped his readers would not notice how little there was in it on battlefield tactics.6 Before he did this, however, he asked our friend and colleague Professor Bernard S. Bachrach for advice. Bachrach did not see a real problem. That John had selected these particular passages from Sallust was to him highly significant, for John undoubtedly had chosen them because he deemed them to be accurate characterizations of what had actually occurred in Edmund's battles. Abels did not agree then nor do we agree now. Rather, we think that all that John knew about these battles came from his main source, the Anglo-Saxon Chronicle, and that he added the details from Sallust to spruce up the narrative
3 John of Worcester, Chronicle, 2:4901: "triplicibus subsidiis aciem instruit, dein singulas turmas circumiens, monet atque obtestatur uti memores pristine uirtutis atque uictorie sese regnumque suum a Danorum auaritia defendant cum iis certamen fore quos antea uicerunt. Intera Canutus paulatim in equum locum suos deducit . At contra rex Eadmundus aciem sicuti instruxerat uelociter movet, et repente signo dato Danos inuadit. Italicized words are from Sallust, Jugurtha, 49. 250. 3. Cf. also Sallust, Catiline, 59: "paululum commotus, ... instructos ordines in locum aequum deducit." 4 C. R. Hart argues that the Worcester Chronicle up to the year 1016 was compiled between that year and c.1020 by the monk Byrhtferth of Ramsey, to whom he attributed the composition of the later annals, including these battle descriptions. C. R. Hart, "The Early Section of the Worcester Chronicle," The Journal of medieval history
9 (1983), 251215. More cautiously, Michael Lapidge, noting the resemblance between the language of the Worcester Chronicle and Byrhtfeth's Life of St. Oswald, has suggested that a lost version of the Anglo-Saxon Chronicle underlies the entries for 958 through 992, and that this chronicle could possibly have been composed at Ramsey, perhaps even by Byrhteferth. Michael Lapidge, "Byrhtferth and Oswald," in St. Oswald of Worcester, eds. N. Brook and C. Cubitt (Leicester, 1996), p. 76. See also Alfred P. Smyth, The Medieval Life of King Alfred the Great (Basingstoke, Hampshire and New York, 2002), pp. 6971. Cf. P. McGurk, "Introduction" to The Chronicle of John of Worcester, 2:lxxix-lxxxi, which rebuts Hart's case. 5 R. R. Darlington and P. McGurk, "The `Chronicon ex Chronicis' of `Florence' of Worcester and Its Use of Sources," Anglo-Norman Studies 5 (1983), 18596, at 193, n. 37; C. R. Hart, "Early Sections of the Worcester Chronicle," Journal of Medieval History 9 (1983), pp. 3034. 6 Richard Abels, "English Tactics, Strategy and Military Organization in the Late Tenth Century," in Donald Scragg, ed., The Battle of Maldon, AD 991 (Oxford, 1991), 14355. John of Worcester's accounts of Sherston and Ashingdon are discussed on p. 153, n. 17.
Images of Antiquity and Altered Reality
and to demonstrate his own erudition. To be sure, John's compression and editing of the Sallust texts may be significant. Among the details that he chose not to include in his narrative of the battle of Sherston are Catiline's decision to drive away his horses so that his men would have to stand and fight, descriptions of topography and terrain clearly inappropriate for Sherston, and references to centurions and subordinate commanders on the wings. John also changed Sallust's "pila omittunt, gladius res geritur" to "lanceis et gladiis pugna geritur." Similarly, in his account of the battle of Ashingdon, John edited out Sallust's description of how the Roman commander Metellus deployed his slingers (funditores) and archers (sagittarios) between the companies of infantry, and placed his cavalry on the wings. These omissions may reflect John's awareness of differences between armies of his day, the first decades of the twelfth century
, and those described by Sallust. (In this respect, John's decision to change Sallust's description of soldiers throwing javelins [pila] to soldiers thrusting with lances [lanceae] is suggestive.) Or John may have simply wished to condense the narratives. In either case, John probably used Sallust's Cataline and Jugurtha not because Edmund Ironside actually commanded and fought like a first century BC Roman general but to demonstrate his familiarity with a classical authority then in vogue.7 Bachrach's interpretation is indeed possible, but strikes us as inherently less likely.8 It is less likely unless, of course, one begins, as does Bachrach, with the assumption that "the picture of medieval military history which is emerging today" is that of "continuity between the ancient and medieval periods."9 Bachrach's thesis is founded, in part, upon his careful study and
7 The growing popularity of Sallust is suggested by the rise in the number of copies of extant manuscripts of his work between the tenth and twelfth centuries: four in the tenth, thirty-three in the eleventh, and fifty-eight in the twelfth. L.D. Reynolds, ed., Texts and Transmission. A Survey of the Latin Classics (Oxford, 1983), pp. xxvi-xxvii. 8 Some obvious questions suggest themselves regarding Bachrach's interpretation that John used Sallust because the description fit what he knew of the battle. Why Sallust? Did John search through any number of classical sources until he found an appropriately parallel battle? What other descriptions did he have available? Given that John edited the accounts in Sallust in ways already noted, why did he not borrow shorter phrases and descriptions from several sources, surely a technique more likely to produce an accurate description than wholesale appropriation of a single battle-piece from a single source? All medieval chroniclers' heads were full of classical words and short phrases, many of which regularly found their way into descriptions of warfare see for example the discussion of Orderic Vitalis below, pp. 1011 and notes 1820. But when they had clear and abundant information about a battle, either from witnessing it or from other eyewitnesses, they had no trouble creating original descriptions in which classical words and phrases provided some of the bricks of the structure, so to speak, rather than the entire architecture. See, for example, William of Poitiers' description of the Hastings campaign and battle, which positively invited borrowings from Caesar, but whose account of Hastings is clearly his own: The Gesta Guillelmi of William of Poitiers, ed. and trans. R.H.C. Davis and Marjorie Chibnall (Oxford, 1998), hereafter WP; see further discussion below, pp. 56 and notes 1516. 9 Bernard S. Bachrach, "Medieval Military Historiography," in Michael Bentley, ed., Companion to Historiography (London, 1997), p. 206. On Charlemagne's study of Roman military tactics and strategy, see Bachrach, Early Carolingian Warfare: Prelude to Empire (Philadelphia, 2001), p. 162.
Richard Abels and Stephen Morillo
intimate knowledge of the literary and documentary source
s that survive from the early and central Middle Ages
and, in part, upon his complete rejection of the old historical paradigm of Rome's "Fall" and the West's descent into "the Dark Ages." early medieval
texts, as he takes pains to point out, abound with classical allusions and often employ technical Roman military and administrative terminology in describing events. Bachrach contends, moreover, that the preservation, transmission, and popularity (as measured by numbers of manuscripts) throughout the early Middle Ages of Roman military handbooks, notably Vegetius's De re militari and Frontinus's Strategemata, attests to the Roman foundations that underlay military organization in the West between the fourth and twelfth centuries. "[F]rom the efforts of Diocletian and his successors during the fourth and fifth centuries," Bachrach asserts, "until the development of gunpowder toward the end of the Middle Ages, the essentials of military organization, relative effective troop strengths, strategy, and tactics demonstrate startling continuity in Rome's successor states Byzantium and the kingdoms of the medieval West."10 But how much of this apparent continuity is real and how much is the consequence of the classicizing tendencies of medieval chroniclers? All historians when assessing sources face the problem of representation versus reality. For military historians of the Middle Ages, this problem takes the special form of classicizing sources versus the realities of early medieval warfare. This article considers this problem. Our consideration cannot rest on an exhaustive survey of all the possible evidence for this problem. Rather, we will examine a few selected and representative pieces of evidence in order to make points about the evidence as a whole, points that are as much philosophical as they are specific historical analysis. Indeed, this article and the rejoinder by Professor Bachrach that will follow in the next volume of this journal are attempts to distill and thus clarify the various issues methodological, historiographical, terminological and philosophical implicit in divergent interpretations of early medieval warfare contained in the works each of us has published on specific subjects.11
Written Sources The problem of classicizing terminology is a pervasive one in the sources for medieval military history. John of Worcester's use of Sallust to describe the battles 10 Bachrach, Early Carolingian Warfare, p. 51. Bachrach has developed and advanced this thesis in three monographs, Early Carolingian Warfare; The Anatomy of a Little War: A Diplomatic and Military History of the Gundovald Affair: 568586 (Boulder, CO, 1994); Fulk Nerra, the Neo-Roman Consul, 9871040: A Political Biography of the Angevin Count (Berkeley, CA
, 1993); in dozens of articles, many of them gathered in two Ashgate Variorum volumes, Armies and Politics in the Early Middle Ages (London, 1993) and Warfare and Military Organization in Pre-Crusade Europe (London, 2002); and in several critical book review
s. 11 Directly relevant to the terminological difficulties of the sources for early medieval warfare is S. Morillo, " Milites, Knights, and Samurai: Medieval Military Terminology and the Problem of Translation," in The Normans and Their Adversaries at War: Essays in Memory of C. Warren Hollister, ed. Richard P. Abels and Bernard S. Bachrach (Woodbridge, 2001), 16784.
Images of Antiquity and Altered Reality
of Sherston and Ashingdon, in other words, though unexpected, is far from isolated. Some are explicit; others may be hidden. Classical models, for example, may have influenced the Maldon-poet's account of that battle. Take the poet's detailed description
of Byrhtnoth's deployment of troops (ll. 1721), with its emphasis upon the ealdorman's "instruction" of his troops about how to stand fast and hold their shields: Рa южr Byrhtnoр ongan beornas trymian, Rad and rжdde, rincum tжhte, hu hi sceoldon standon and юone stedan healdan, and bжd южt hyra randan rihte heoldon, fжste mid folman, and ne forhtedon na. (Then Byrhtnoth set about drawing up the men there, he rode and instructed, he told [literally, taught] the soldiers how they should form up and hold the position, and he asked that they should hold their shields properly, firmly with their fists, and not be at all afraid.)12 Although these lines have often been read as evidence for the lack of training of the Essex fyrd, they may simply be the poet's attempt to render and explain the Latin stock phrase for drawing up battle lines, instructio aciem (or copias), as Abels suggested in the aforementioned article.13 Indeed, the Maldon-poet's representation of Ealdorman Byrhtnoth as having ordered his men to drive away their horses may have drawn upon the same battle narrative from Sallust's Catiline that John of Worcester later used in his account of Sherston.14 If so, the poem tells us less about late Anglo-Saxon military practice and the actions of Byrhtnoth in command than many have thought. William the Conqueror's crossing of the English Channel positively invited comparisons with Caesar's invasion of Britain in 55 BC, and William's hagiographer William of Poitiers, a clerk from a military background who might be expected to provide informed contemporary descriptions of William's invasion, readily and openly obliged. And while the list he provides of the ways in which William was greater than Caesar15 is relatively easy to deal with, more problematic are the ways in which his account of the Conqueror's crossing may
12 The Battle of Maldon, ll. 1721, ed. and trans. Donald Scragg, in Scragg, ed., Battle of Maldon, AD 991, pp. 1819. 13 Abels, "Tactics," in Scragg, ed., Battle of Maldon, p. 153, n. 10. Compare, by contrast, the detailed description by William of Malmesbury of Henry I instructing his English troops (presumably also the fyrd) in 1101: ...docebat quomodo, militum ferociam eludentes, clypeus objectarent et ictus remitterunt; ("he taught them how, in meeting the attack of the milites, to defend with their shields and return blows"): William of Malmesbury, De Gestis regum Anglorum, ed. William Stubbs (Rolls series
, 188789), p. 472. 14 Cf. Sallust, Catiline, 59: "Dein, remotis omnium equiis, quo militibus exaequato periculo animus amplor esset, ipse pede exercitum pro loco atque copiis instruit." This was first suggested by C. R. Hart, "Maldon," in An Essex Tribute: Essays Presented to Frederick G. Emmison, ed. Kenneth Neale (London, 1987), pp. 756, reprinted in Cyril Hart, The Danelaw (London, 1992), pp. 5423. 15 WP, pp. 16870.
Richard Abels and Stephen Morillo
have been embellished to make it closer to Caesar's. Did William really become separated from his fleet early in the crossing, dropping anchor to wait for the rest to catch up, just as Caesar had?16 It is possible, of course, but also suspicious. Suspicion, however, is useful, because when classical episodes and motifs appear, they can be more or less totally discounted, leaving us with less information (which is bad), but not misleading information (which is worse). A more dangerous problem arises with the classicizing hidden in basic terminology. The Annals of Metz is typical of Carolingian sources, for example, in consistently describing Charlemagne's forces as legiones, as for instance the four groups of troops the King sent against the Saxons in 774 that the Royal Frankish Annals simply calls "detachments" [scarae].17 Similarly, Orderic Vitalis uses legiones to describe forces of troops and cohors to describe smaller units, but the examples show how far from the classical meaning of these terms the medieval terminology had wandered. In addition to using legio of Crusader forces, for example,18 he also used it to describe units of Turks [legio Turcorum], and noted that Peter the Hermit had fled back to Constantinople before Kilij Arslan's attack at Civetot because sua cohors no longer obeyed him.19 For Orderic, clearly, cohors did not carry any implication of Roman-style organization, but rather was it simply a good classical term for a group of soldiers; we know this because he had already characterized Peter's followers as an imbellam catervam an unwarlike rabble.20 Nor could legio carry implications of Roman organization when applied to Turks. But what about Charlemagne's army? Here we come face to face with the central interpretive problem facing military historians of early medieval Europe: how we read the military terminology depends crucially on how we conceptualize the economic, social and administrative worlds that produced Armed Forces
, it being a given that armies are shaped by the worlds that produce them, reflecting the material capabilities and cultures of those worlds. But armies and warfare, so central to the narrative sources for early medieval history, are by their very centrality a large part of the evidence we have for the economic, social and administrative capabilities of those worlds. In this situation, pervasive classical terminology creates a great danger of tautology: How do we know early medieval kingdoms were administratively sophisticated? Because they produced legiones. How do we know these legiones were up to Roman military standards? Because they were produced by administratively sophisticated governments operating on Roman models.
16 WP, p. 110 and n. 3, where the editors also point out the many Vergillian passages in WP's description of the crossing. 17 Annales Mettenses Priores, ed. B. von Simson (Hannover: MGH SUS, 1905), s.a. 774; Annales regni Francorum 741828, ed. F. Kurze (Hannover: MGH SRG, 1895), s.a. 774. See the dis- cussion in John France, "The Composition and Raising of the Armies of Charlemagne," Jour- nal of Medieval Military History 1 (2002), 6182 at 73 and n. 37: "The use of such classical terms to describe quite different realities is one of the problems of using Carolingian sources." 18 The Eccelsiastical History of Orderic Vitalis, ed. and trans. Marjorie Chibnall, 6 vols. (Oxford, 196980), 5:78 (hereafter OV). 19 OV, 5:38. 20 OV, 5:32.
Images of Antiquity and Altered Reality
There is also broader linguistic and representational evidence for continuity, as in titular nomenclature and pictures on coinage that take Roman models. On the basis of such evidence, Roman continuity maximalists such as Bachrach, Charles Bowlus and K. F. Werner argue that any general transformation of society, economy and government from late Roman times into the early Middle Ages was minimal.21 Thus, while Bachrach admits briefly in his tome on Carolingian warfare that Western Europe
was less wealthy than its Muslim neighbors,22 he argues that "the militarization of the civilian population for local defense [occurred] within the institutional structures of the later Roman empire"23 and that the obsequia the military households of the powerful men of Francia were the functional and state-directed equivalent of the "professional element of armies that had flourished during the later Roman empire."24 Classical terminology and the classicizing world view of our early medieval sources, in other words, if taken at face value, create a world that will admit of some change, but constrains it within institutional structures and mental outlooks whose viability and influence assume central importance in this world. In short, the very coherence of the literary sources creates a powerful lens. If one believes, as we do, that it is a distorting lens, and that the underlying realities of early medieval warfare had fundamentally changed, making a lie of the legacy of antiquity found in the sources, how do we step outside that lens? Where can we turn to decide between opposed positions on Roman continuity in military organization and practice?PHYSICAL EVIDENCE
In some cases one may test the hypothesis of military continuity with Rome against physical evidence. Alfred the Great's system of burhs in the late ninth century is one such case. King Alfred's connections with Rome are numerous and clear. He made two childhood pilgrimages to the Eternal City, produced coinage closely modeled on Roman originals, and interpolated references to the
21 The results of this approach are epitomized by the title of Bachrach's study of the eleventhcentury ruler of Anjou, Fulk Nerra, the Neo-Roman Consul, 9871040: A Political Biography of the Angevin Count. For maximalist views of Carolingian warfare, both in terms of Roman administrative continuity and especially in terms of demographic stability and Big Numbers for Carolingian armed forces, see Charles R. Bowlus, Franks, Moravians and Magyars. The Struggle for the Middle Danube 788907 (Philadelphia, 1995); K. F. Werner, "Heeresorganisation und Kriegfьhrung im deutschen Kцnigreich des 10. und 11. Jahrhunderts," Settimane de Studi de Centro Italiano sull'alto Medioevo 15 (Spoleto, 1968). France, "Armies of Charlemagne," provides an excellent balanced overview of the issue of numbers in Carolingian armies; his conclusions tend to the Small Numbers side. 22 Bachrach, Early Carolingian Warfare, p. 175. 23 Bachrach, Early Carolingian Warfare, p. 53. 24 Bachrach, Early Carolingian Warfare, p. 167; see also 210.
Richard Abels and Stephen Morillo
city and its history into his translations.25 Most dramatically, Alfred in the 880s and 890s created a defensive network of fortified towns and cities
, at least five of which, Bath, Chichester, Exeter, Portchester, and Winchester, were refurbished Roman foundations, and restored the Roman City of London
.26 What is interesting is how little continuity can be found between the Roman civitates and the Alfredian burhs. By Alfred's time most Roman cities in England were like London; the enclosed Roman city lay deserted and derelict, while a commercial wic had sprung up outside its walls. The reoccupation and refurbishment of deserted Roman walled towns had less to do with a memory of Rome than with the pragmatic realization that these sites were strategically well sited, since they were lined up with the Roman road system that was still the major conduit for transportation in ninth-century Wessex, and possessed the remnants of formidable defenses (which Alfred's surveyors often chose to ignore; see below). Although Alfred had lived in Rome for a year during his childhood, soon after the completion of the walled "Leonine City" to defend St. Peter's and the pilgrim communities against raiders from the sea, the king and his builders made no effort whatsoever to emulate Pope Leo IV and his predecessors and build in stone. Alfred's burhs were defended by outer and inner ditches and earthen ramparts. This was true even for those that had Roman walls. At Winchester an impressive new double ditch was dug and traces of its trench survive, indicating an original width of 8.2 m and depth of 1.7 m. As with Offa's Dyke, dump construction was the rule, though the earthen walls were often reinforced with turf and timber revetments, and, in some cases, crowned with wooden palisades.27 There is no evidence that the stone walls of Winchester or the other burhs of Roman origin were restored before the end of the tenth century. Discrepancies between the actual measurements of the circuits of the Roman walls of Exeter and Bath and the predicted measurements for their defenses based on the Burghal Hidage suggest that the Alfredian and Roman defenses did not in fact coincide. Alfredian Exeter, whose 734 hides imply a wall length of 3028 ft but whose Roman walls have a circuit of 7600 ft, seems to have occupied only a small portion of the Roman city, perhaps Rougemont Hill where the later Norman
25 Richard Abels, Alfred the Great: War, Kingship and Culture in Anglo-Saxon England (London, 1988), pp. 5777, 212. King Alfred's Version of St. Augustine
's Soliloquies, ed. T. A. Carnicelli (Cambridge, MA, 1969), p. 97: "nat ic no thi hwa Romeburh timbrede the ic self gesawe" ("I know not who built Rome on account that I myself saw it"). 26 Hastings and Worcester may also have begun as Roman sites
. John McN. Dodgson, "The Burghal Hidage Place-Names," in David H
ill and Alexander R. Rumble, ed. The Defence of Wessex: The Burghal Hidage and Anglo-Saxon Fortifications (Manchester and New York, 1996), p. 99; David Hill, "Gazetteer of Bughal Hidage Sites," in Hill and Rumble, Defence, pp. 190 1, 1956, 2024, 205, 21415, 225, 226. 27 Abels, Alfred the Great, p. 206. Vegetius discusses the construction of ditch-earthen wallrampart field fortifications (Vegetius, Epitoma rei militaris, 3.8), but the Alfredian burhs were conceived as permanent fortresses more akin to the late Roman Saxon Shore Forts. Nor do Vegetius's prescriptive depths and widths for ditches and walls match well the actual measurements for the surviving Anglo-Saxon defenses. Cf. discussion by Bachrach, Early Carolingian Warfare, pp. 2334.
Images of Antiquity and Altered Reality
castle was built.28 Bath's 1000 hides indicate defenses measuring 1375 yds, some 125 yds more than the measure of the Roman wall. Archaeologist T. J. O'Leary's excavations have revealed Anglo-Saxon defenses lying outside the wall, which suggests that "the Burghal Hidage assessment was being made for the perimeter of the outworks, since their circumference, as opposed to that of the surviving Roman wall, approximates closely to the 1,375 yards of the Burghal Hidage."29 Nor did Alfred's or his son Edward's fortress builders employ Roman surveying techniques or manuals (Agrimensores or Gromatici). Manuscripts of gromatic texts circulated in ninth-century Francia but were used mainly for teaching geometry rather than for practical surveying.30 The most useful of these surveyor's manuals for military planners, pseudo-Hyginus's De Munitionibus Castrorum, was not among those preserved in Carolingian gromatic collections and was only rediscovered in 1493 by the humanist Galbiato,31 and it was not until the end of the tenth century that the Agrimensores crossed the English Channel.32 Significantly, the formula appended to the Burghal Hidage calculating how many hides and men are needed for the manning and maintenance of specific lengths of wall is given in native English units of linear measurement, the "pole" (OE gyrd) of 16 ft 6 in, "acre" (66 ft), and "furlong" (660 ft), rather than in the Roman pedes and actus used by Roman land surveyors.33 Equally telling is the complete lack of correspondence between the medieval street planning of these Alfredian burhs and the underlying Roman street grids, seen most dramatically at Winchester and London.34 Despite Alfred's issue of a Roman "London" coin to celebrate his restoration of that city in 886, the actual street planning for Alfred's London ignored completely whatever traces of the Roman street system then remained. In short, there is little if any real continuity to be found between the defensive system of Late Roman Britain and the Alfredian network of burhs. Alfred's Burghal network may indeed have been designed strategically to be a defense-in-depth system, but if so, Alfred and his advisors did not draw their
28 Hill, "Gazetteer," in Hill and Rumble, Defence, p. 204. 29 T. J. O'Leary, "Excavations at Upper Borough Walls, Bath," Med. Arch. 25 (1981), p. 22, cited by Hill, "Gazetteer," in Hill and Rumble, Defence, 191. 30 P. K. Marshall, "Agrimensores," in Reynolds, Texts and Transmission, 16, at 45. On Roman surveying, see O.A.W. Dilke, The Roman Land Surveyors: an Introduction to the Agrimensores (Newton Abbot, 1971). The gromatic texts have been translated by B. Campbell, The Writings of the Roman Land Surveyors: Introduction, Text, Translation and Commentary (Hertford, 2000). 31 Marshall, "Agrimensores," 2. 32 Marshall, "Agrimensores," 5. Cambridge, Trinity College 939 (belonging to St. Augustine's, Canterbury) and Oxford, Bodl. Douce 124 (St. Swithun's, Winchester). 33 Rumble, ed., "An Edition and Translation of the Burghal Hidage," in Hill and Rumble, De- fence, pp. 30 and 34. See discussion by Rumble, "The Calculation," in Hill and Rumble, De- fence, pp. 701. For a brief discussion of Roman measurement, see O.A.W. Dilke, Mathemat- ics and Measurement (Los Angeles
, 1987), pp. 267. That the Old English
words for measur- ing, gemetan, and for a measurer, metod, apparently derive from Latin metior suggests, how- ever, that the early Anglo-Saxons associated measures and surveying with Rome. 34 M. Biddle and D. Hill, "Late Saxon Planned Towns," Antiquaries Journal 51 (1971), 7085; Robert Cowie and Robert Whytehead, "Lundenwic. The archaeological evidence
for Middle Saxon
London," Antiquaries Journal 63 (1989), 70618.
Richard Abels and Stephen Morillo
inspiration from the Roman past.
Analysis If archaeology leads us away from the Roman world, further insight into the disjunction between classicizing sources and altered realities can be had through closer consideration of the sources themselves from the perspective of the archaeological results and all they imply. In order to illustrate this, let us look closely into one sentence from Orderic Vitalis, describing the indiscipline of the Angevin army that invaded Normandy in 1137. The translation offered here differs slightly from Marjorie Chibnall's, in order to highlight some Key Terms
.35 The optimates ["magnates" in Chibnall], however, who ought to have led separate gatherings [coetus; Chibnall uses "squadrons"] in a lawfully led [legali ductu; Chibnall has "properly levied"] army, were ignorant, unless I am mistaken, of the rigor of Roman discipline in military matters, nor did they conduct their "knightly quarrels" [as Chibnall renders militares inimicicias] in the moderate manner of great men [literally "heroes": haeroum more modeste. Chibnall has "with restraint as lords should"]. The obvious reference here is to Roman discipline and what that implies about the proper method for raising and leading an army. Chibnall notes that this is probably a reference to Vegetius, whose De re militari "was well known in Norman monastic circles."36 Since Orderic is implicitly drawing a contrast between the hated Angevins and his own Normans, one might conclude that Orderic is imputing to the Normans knowledge of Vegetius, Roman methods of raising and training troops, and so on, by contrast with woeful Angevin ignorance. This is indeed exactly the sort of implication Bachrach and others do draw about references to Romanitas in general and Vegetius in particular in Norman and Angevin sources, among others.37 But for us the real emphasis of the source is not on Roman discipline but on a social structure
and a set of social expectations
that distinguish Orderic's world from Rome and link Normandy and Anjou together. The key words are optimates, the tellingly mis-spelled haeroum, and militares. Orderic is describing a world in which great men magnates who should be noble heroes
create armies from among their own followers: optimates leading their own gatherings, and not Roman discipline, is clearly the context for legali ductu. This tells us that the state, as we and probably the Romans understood it, a res publica whose interests were separate from and largely superior to private interests, is not there. These great men dispute prominence among themselves, and settle such disputes according to a set of cultural expectations that surround milites. It is this last 35 OV, 6:472; translation by S. Morillo. 36 OV, 6:472, n. 1. 37 See, e.g., B. S. Bachrach, "The Practical Use of Vegetius's De Re Militari During the Middle Ages," The Historian 47 (1985), 239255. For a critique see S. Morillo, Warfare under the Anglo-Norman Kings 10661135 (Woodbridge, 1994), pp. 118 n. 89, 135 n. 156, 155 n. 68.
Images of Antiquity and Altered Reality
word that bears closest scrutiny, for it still carries the classical meaning "elite soldiers," but the elite is no longer the legionary foot soldiers recruited and trained by the state, but the mounted social elite of optimates and their military households; it is often, though not always accurately, rendered into English as "knights," and certainly designates mounted troops when used in the common phrase milites peditesque: horse and foot.38 Note that Orderic does not imply that milites never quarrel (which might be the expectation if indeed they were part of a military machine run according to principles of Roman organization and discipline), but that they settle their quarrels haeroum more modeste one is tempted to translate "in the discreet way Normans do," since the Normans were Orderic's heroes. This passage probably should not be pushed so far as to become an early reference to chivalry, nor should one argue that the Normans were in fact models of knightly discretion image and reality can cut in many directions, after all. But the clear social and political sense of the sentence, stripped of its misleading classicizing veneer and its erudite allusion to Vegetius, describes a world far from the statedominated universe of the Roman empire, and describes warfare that even Vegetius, with all his complaints about the decline of proper Roman discipline in his own day, might find chaotic. Of course it wasn't chaotic, especially not in mid-twelfth century Normandy and Anjou, but its principles of organization are not obvious to anyone expecting a state-centered system. It is a different world from Rome and, crucially, a different world from our own socially and administratively, and the conclusion we draw is that it was a different world militarily. The implication of this for the classical language in our sources is clear.
Conclusion Early medieval chroniclers and historians were familiar with Roman authors and liked to demonstrate their knowledge. We should not mistake this literary fashion for reality, especially when considering Anglo-Saxon and English medieval military history, and early medieval military history generally. When Asser tells us that Alfred the Great "closed the testudo in proper order," during the battle of Ashdown, "and immediately advanced his standards (vexilla) against the enemy,"39 we ought not to imagine that Alfred had found a copy of Vegetius or Frontinus's Strategemata (neither of which are quoted in an Alfredian text)40 38 For a fuller examination of the word miles and the problems with translating medieval military terms generally, see Morillo, " Milites, Knights, and Samurai." 39 Asser's Life of King Alfred together with the Annals of Saint Neots, ch. 38, ed. W. H. Stevenson (with an Introductory article by Dorothy Whitelock) (Oxford, 1959), pp. 2930. 40 The earliest surviving English manuscript copy of Vegetius'ss De re militari, B.L. Cotton Cleo. D. 1 (s. xi), also containing texts by Vitruvius and Solinus, was probably produced at St. Augustine's, Canterbury in the first half of the eleventh century. Vegetius, however, was certainly known in eighth-century England, since Bede paraphrased him, though whether from a full copy of De re militari or from a florilegia is unknown. See Bede, Ecclesiastical History of the English People, 1.5, ed. Bertram Colgrave and R.A.B. Mynors (Oxford, 1969), p. 26 and n. 1.
Richard Abels and Stephen Morillo
and revived the Roman "tortoise" formation. Rather, he was using or misusing a classical term to express something quite different from Roman practice, a "shield wall" maintained by untrained and, by Roman standards, atrociously undisciplined troops. This tells us nothing about continuity in military practice; only about Asser's own classical learning. The same, of course, is true for terms such as comes and dux. There can be little doubt that when Bede characterized King Penda of Mercia's forces in the Battle of the River Winwaed as thirty legiones led by thirty "most noble dukes" (ducibus nobilissimis) he was using classicizing language to describe warbands of ealdormen and great nobles.41 The comites and duces of the early Middle Ages may have borne the same titles as Roman sector commanders in the Late Empire, but this does not mean that their offices or duties derived in any real sense from the imperial military bureaucracy.42 Similarly, if a source describes a battle or a pre-battle oration in terms that recall classical Roman practice, as Bede does on occasion, it may not mean that the medieval general had studied his Vegetius; it may simply mean that the author had.43 Similarly, the proliferation of early medieval monastic manuscripts of Vegetius's De re militari or of Frontinus'ss Strategemata do not necessarily indicate their practical use by commanders. As Guy Halsall has wisely observed, [W]e cannot be sure that the popularity of Vegetius was not mainly a matter of antiquarianism and the desire to acquire classical learning ... [E]arly medieval people were quite prepared to accord great authority to classical works, even while recognizing that they had nothing to do with the world in which they lived. There is no evidence at all that Vegetius's detailed tactical and organisational recommendations were ever put into practice, and it seems unlikely that his advice on campaigning was followed closely either.44 Though in this article we do not have the space to do more than suggest it, the problem of representation versus reality faced by those who study medieval military history extends well beyond the problem of classicizing chroniclers. Perhaps even more important was the influence of Biblical models upon their writing and thought. Thus those who write on the Crusades tend to take quite seriously the accounts of Fulcher of Chartres and Raymond d'Aguiliers detailing the glorious horrors of the sack of Jerusalem in 1099. But these tales of bloodletting have Biblical echoes in the books of Joshua, Kings, and Revelation that 41 Bede, Ecclesiastical History of the English People, 3.24, ed. Bertram Colgrave and R.A.B. Mynors (Oxford, 1969), p. 290. 42 This is probably true even of seventh-century Visigothic Spain where provincial armies were commanded by generals termed a dux exercitus provinciae . See Guy Halsall, Warfare and Society in the Barbarian West, 450900 (London and New York, 2003), pp. 603 and 250, n. 110. This outstanding book provides the most sophisticated examination to date of the evolution and transformation of military recruitment
, organization, and practice from late Rome into the early middle ages. 43 Cf. Bachrach, Early Carolingian Warfare, p. 162. 44 Guy Halsall, Warfare and Society in the Barbarian West, pp. 145 and 268, n.35. Cf. Bachrach's careful delineation of the differences between Hrabanus Maurus' De Procinctu romanae miliciae and his source, Vegetius, and how he infers from that ninth-century military practices. Bachrach, Early Carolingian Warfare, pp. 84131.
Images of Antiquity and Altered Reality
have not yet been sufficiently explored. In short, the legacy of Antiquity on the study of medieval military history is that of a distorting lens that imposes apparent continuity on changed reality.