Home          Transcriptions     Contact Us                                        Malory                Morte           Bibliography

The Morte Darthur



The History, Provenance, and Discovery of the Winchester Manuscript


In 1934, the discovery of the Winchester Manuscript changed the way people studied and read the Morte Darthur. Hidden away in a forgotten corner of the Fellow’s Library at Winchester College, W.F. Oakeshott first announced the discovery in the Daily Telegraph on June 25, 1934 and in The Times on June 26, 1934. For nearly four-and-a-half centuries, the manuscript had remained unknown, appearing in the library’s catalogue only in 1839 and then only acknowledged as “The History of Prince Arthur. Imperfect” and believed to be another copy of Caxton’s text which the library already owned a copy of. Ironically, evidence in the form of printing ink marks and a paper repair made with a printed Indulgence by Pope Innocent VIII has been found that confirms that the manuscript was present in the print shop of Caxton himself while he was preparing his own edition. However, the Winchester is not believed to be the original source text used by Caxton. Some clues to the manuscript’s provenance can be found in the pages themselves though. The Followell family is believed to have owned the manuscript prior to its acquisition by the Fellow’s Library. A signature of Richard Followell can be found on folio 348. The manuscript has since been acquired by the British Library, and there is where it currently resides.


Before the discovery of the Winchester Manuscript, the 1485 edition of the Morte Darthur printed by William Caxton had been considered the earliest and most authoritative text. No other text existed to challenge that of Caxton’s. All other editions of the Morte were simply derivatives of the original text published by Caxton or his apprentice Wynkyn de Worde. Almost as soon as the manuscript was discovered however, Eugene Vinaver began to prepare it for publication, an edition that has long been considered one of the best and most authoritative since its first publication in 1945. (Vinaver had originally been intending to publish a new edition of the 1485 Caxton text when the manuscript was discovered.) 


Since Vinaver’s edition, multiple editions based on the Winchester have appeared with the most recent being that of Stephen H.A. Shepherd printed by Norton & Company in 2004. Each of these various editions has been a hybrid text, using the text of the Caxton in those places where the Winchester has missing pages. Each of these texts has also been edited by adding modern punctuation marks and deleting scribal errors. A facsimile edition of the Winchester was finally published in 1976 by the Early English Text Society, but until now no unpunctuated, non-hybrid transcription has been available. (This text contains links to the missing sections of Caxton’s text, but these are not included in the main representation of the transcription itself. The text of Caxton’s is only included for those attempting to read the complete story of Arthur here.)


The Manuscript’s Features




Condition of the Manuscript:


As has already been mentioned, the Winchester manuscript is incomplete, missing the first quire (or first eight pages) and the final quire, as well as f.32-33v and f.252-252v. Two pages, f.192 and f.400 also have large pieces missing from their corners. Most of the rest of the text remains complete with the exception of a few letters missing from wormholes. The rest of the manuscript contains 473 leaves of good French paper. The manuscript consists of sixty quires, 56 of which are ‘regular’ or contain four sheets folded to make eight leaves. Catchwords exist at the end of each quire to indicate to the binder and the scribes the beginning and ending of each section. The current binding on the manuscript dates from 1948 in alum tawed goatskin. 


The Scribes:


Two very distinctive ‘hands’ appear in the manuscript. These two different styles indicate the presence of two different scribes, A & B, who originally came from Warwickshire where the manuscript is believed to have been originally commissioned. This evidence comes mainly from the style of scripts chosen. Scribe A wrote from f.9-34, from f.35v until f.44v, from f.191v until 229, and from 349-484v. Scribe B wrote from f.45-191 and from f.229v-346v. Two pages contain the handwriting of both scribes, that of f. 35 and f.45. Scribe A’s script overall is less regular and more formal, much like the anglicana script used by lawyers of that period. He switches back and forth from this more formal hand to a secretary script with a much looser, informal appearance. Scribe B has a much more predictable informal secretary script introduced into France a century earlier. 


Rubrication & Initials:


Most of the manuscript’s decoration and section separation involve Lombard initials of varying sizes, some with highly elaborate ‘pen flourishings’ added by the scribes. In addition, the proper names of both people and places have been written in red ink, or rubricated, perhaps for ease of reference and certainly to add some variety and beauty to the page. Overall, though, the limited decoration, lack of illumination, and informal hands suggest that this manuscript was likely commissioned as a presentation copy for someone in the middle class.




Evans, Murray J. “The Explicits and Narrative Division in the Winchester MS: A Critique of Vinaver’s Malory.” Philological Quarterly 58.3 (Summer 1979): 263-281.


---. “The Two Scribes in the Winchester MS: The Ninth Explicit and Malory’s‘Hoole Book.’” Manuscripta 27.1 (March 1983): 38-44.


Hellinga, Lotte. “The Malory Manuscript and Caxton.” Aspects of Malory. Eds. Toshiyuki Takamiya & Derek Brewer. Cambridge, UK: D.S. Brewer, 1981. pp. 127-142


Kelliher, Hilton. “The Early History of the Malory Manuscript.” Aspects of Malory. Eds. Toshiyuki Takamiya & Derek Brewer. Cambridge, UK: D.S. Brewer, 1981. 143-158.


Ker, N.R. “Winchester College MS.13.” The Winchester Malory: A Facsimile. Oxford, UK: Early English Text Society, 1976. ix-xxii.


Oakeshott, W.F. “The Text of Malory.” Times Literary Supplement. September 27, 1934. Accessed 2/28/08. <http://www.english.uga.edu/~jdmevans/public/september.html>


Vinaver, Eugene. “The Story of the Book.” The Works of Sir Thomas Malory. Vol. I. Ed. P.J.C. Field. Oxford, UK: Clarendon Press, 1993. xxxv-lvi.



The Debate of Caxton vs. Winchester:


Few extant witnesses of the Morte from before 1500 still exist. Only three witnesses have survived: the Winchester Manuscript, Caxton’s incunable of the Morte from 1485, and his successor Wynkyn de Worde’s further adaptation of the Morte in 1498. Around these different witnesses, much of the traditional textual debate has revolved as scholars try to decide on the most ‘correct’ or most ‘authoritative’ text. Yet each of the witnesses for the Morte has its own attractions, its own values, its own interpretations, and certainly its own place in the textual context of the Morte. 


Of these three, though, the first two, the Winchester manuscript and Caxton’s incunable, have provided the most cause for debate. It has become a battle of Winchester versus Caxton, and the two combatants are both heavyweights. Indeed, it seems as if the textual debate for Malorians is actually more heated than debates which rage around other texts with more witnesses. Even more, it seems that with only two viable textual camps for scholars to join, the debate has only become more intense. In spite of this, I believe, as Sue Ellen Holbrook does, that “the Malory incunable and the Malory manuscript are not only ‘witnesses’ to what Malory wrote but also contemporary forms of what people read when they read what he wrote” (355). They both have important places in the Morte’s textual context and, perhaps more importantly, even in the historical context of the movement from manuscript culture to print culture.


Most scholars believe that the earliest, and for them ‘most authoritative,’ witness can be found in the Winchester Manuscript. Discovered in 1934 in a back corner of the Fellows’ library at Winchester College by W.F. Oakeshott, the Winchester Manuscript created a major stir among Malory scholars. For many of them, the manuscript’s physical character granted it authority. “It was, above all, a manuscript; gleaming with the authority of its handmade quality, its essential uniqueness” (Holbrook 324). During its time, though, this would have granted it little authority. Most books at that time were manuscripts; many were more decorated; and many had greater authority. In comparison with most manuscripts of the period, the Winchester is quite plain. In fact, the manuscript is plain enough and damaged enough that for nearly four-and-a-half centuries it remained undiscovered and virtually ignored.


So what makes up this manuscript? What makes it unique? First of all, being a manuscript, some of its most interesting features pertain to its material composition. Written on bifoliated paper and at approximately three inches thick, the manuscript weighs in as one of the largest and heaviest witnesses of any text in English literature. It consists of 473 leaves with the front and back quires missing (this may be one of the reasons why it was ignored for so long). A few other leaves are missing or damaged (missing fols, 32-33 & 252, damaged fols. 192 & 400). Within the surviving leaves, the manuscript can be divided into four main sections (fols. 9-70; 71-346; 349-409; 409-484) marked by major breaks in the text such as blank pages, or partially filled pages, and larger initials.


Furthermore, the bibliographic features of the manuscript also contribute to its uniqueness. The manuscript shows the influence of two different scribes referred to as Scribe A and Scribe B who write mainly in a plain secretary style script. And as Vinaver notes, the manuscript is well written with few errors and few corrections, making for a fairly clean copy. The scribes also use some older alphabetic characters, the thorn (þ) and the yogh (з). In addition, the text contains minimal punctuation with a type of section break which likely signaled a pause in reading or a place for transition, similar to our modern paragraph break, represented by a double slash (//) as well as other pause marks consisting of virgules (/) and the punctus (·). Abbreviations appear frequently. The most notable feature of the manuscript lies in its hierarchy and decoration based on a series of 111 red initials and rubricated proper names. The manuscript also has some irregular marginalia, also in red, added by the scribes. Overall, these features of appearance make for an attractive, but not overly ornate, manuscript which was likely produced for a member of the merchant class. Still, with all this in mind, the most unique feature of the manuscript seems to be the fact that it is the only manuscript of the Morte.


In contrast, William Caxton finished printing his version of Le Morte Darthur, sixteen years after Malory finished his original (Blake “Caxton at Work” 223), on July 31, 1485, at what he calls the demand of “many noble and dyuers gentylmen” (Sommer 1). Two different versions of Caxton’s text exist, one at the Pierpoint Morgan Library in New York (of which a facsimile exists from Scolar Press) and the other at the John Rylands University Library of Manchester (Holbrook 325). The material features of the Caxton differ little from the Winchester. In fact,


They both use paper as the material . . . The pages have similar dimensions, are written in a single column, and have almost exactly the same size of writing space/print block . . . They both exhibit numbers in the lower right-hand margin beneath the text block for the first four leaves of each quire . . . They are close to the same weight, height, length, and weight . . . neither has pictures, borders, headlines, or folio numbers; neither is divided physically into volumes; and in neither are the narrative units prepared as detachable booklets. (Holbrook 344)


As this description shows, in regard to most material features, Caxton’s text resembles the Winchester more than it differs from it. The main differences in the appearances of the two texts are bibliographic and concern the appearance of Caxton’s text due to the presence of books and chapters, which Caxton freely admits that he adds (Sommer 4), and the lack of the use of decorative red ink. Caxton’s text also has more punctuation than can be seen in the manuscript, most of which is in the form of the virgule or slash (/) with the occasional punctus added (·). The punctuation of the Caxton, then, appears more frequently and divides clauses more regularly than the punctuation found in the Winchester, which varies in its division of the text and of clauses.


Obviously, then, if the main material features of these two texts do not differ much, the main differences must lie in the actual lexical content, in the words themselves and the way they are interpreted. In this respect, the Winchester and Caxton differ quite a bit more. Outside of the expected variants between the two based on scribes’ mistakes or compositors’ mistakes, the most significant divergence occurs in Caxton’s book five, or what has come to be referred to as the Roman War episode. Indeed, Caxton’s version of this episode is only half the length of that found in the Winchester. As a result, this episode has become a main cause of debate between scholars as they try to determine which is the better version. In my opinion, both episodes are equally valid and equally important within themselves; any truly contextual resource should present them both as separate entities. They need not be combined. 


Other lexical differences also exist, though, mostly between different scribal errors/emendations and compositor’s errors/emendations. However, these small lexical differences have taken on more substance in instances where different words have been deliberately chosen by Caxton to replace those found in the Winchester. One well-known example can be found in the much-debated Roman War Episode when Arthur dreams of a battle between two beasts. In the Winchester, the two beasts are a dragon and a bear (f. 75-75v). In the Caxton, the two beasts are a dragon and a boar (Sommer 165). Some scholars have suggested that Caxton may have altered the text in response to the late fifteenth-century political situation. 


In C[axton], the bear is (six times) turned into a boar. The change must have been deliberate, and it created a bold political allusion: the boar was the badge of King Richard III and the dragon that of Henry Tudor. The change would only have made sense in or just before 1485, and it is difficult to see who could have been responsible for it but Caxton himself. (Field “Caxton’s Roman War” 133)


Whatever the reason for the lexical differences between the Caxton and the Winchester, the differences themselves still provide clues to the time period in which the Morte was written and later received. In the context of the Morte, Caxton has not been the only person accused of letting political agendas affect the texts. The Morte itself, both in the Winchester and the Caxton, has long been thought to have strong political undertones and references to the War of the Roses, and these would have been added by Malory himself. Caxton’s changes simply carry on this political theme. Nonetheless, the differences in both texts highlight the importance of their separate being. Readers and critics should not read the two different texts, and certainly the two different Roman War episodes, the same way. Both are important as separate pieces of evidence in the medieval context and specifically in the textual context of the Morte. 


As a final note, the apprentice of Caxton, Wynkyn de Worde, also printed his own copies of the Noble and Joyous History of the Book Entitled Le Morte Darthur, first in 1498 and again in 1529, basing his copies on the 1485 printed copy of Caxton’s. In fact, Kevin T. Grimm notes that it is de Worde’s copy, not Caxton’s, which affected interpretations of the Morte, and most especially its layout, for the next four-and-a-half centuries (135), and what many Malorians have forgotten to recognize is that after 1498, most of the time a reader’s ‘Caxton’ was usually a ‘de Worde.’


In respect to the actual features of the text and the changes made by de Worde, according to Grimm, de Worde’s mainly regularizes and clarifies some of Caxton’s earlier textual divisions through several different means.


First, de Worde adds book titles to each of the first ten books . . . . The continuity of the volume is visually reinforced by printing the book number in the top margin of ever page. Further, in addition to retaining Caxton’s table of rubrics preceding the text, de Worde reprints each rubric at the head of the appropriate chapter in the text. Finally he also adds twenty-one woodcut illustrations, placing one at the beginning of each of the books, and one at the beginning of Chapter 4 in Book XXI.” (138)


Basically, de Worde’s text has the most altered appearance of the Morte when compared to the Winchester manuscript. De Worde’s copy is also a large bifoliate book, but unlike the Winchester and Caxton it has double columns of text, consistent running heads and chapter rubrics, and most noticeably the presence of the illustrations. Due to these distinctions, modern scholars have considered de Worde’s copies some of the most beautiful copies of the Morte but also the least ‘authoritative’ of the fifteenth-century witnesses. 


As far as actual lexical changes, de Worde made quite a few changes to Caxton’s copy, most of which many scholars have considered “incidental” to the basic meaning of the text, but he did make one quite substantive addition at the beginning of chapter twelve in Book XXI (Grimm 134). In this section, de Worde adds “a lengthy moral commentary on Lancelot’s grief at the tombs of Arthur and Guenevere” (Grimm 134). De Worde’s changes, both bibliographic and lexical, make his copies important witnesses in the history of the Morte. Thus, this witness too plays an important role in the reception and textual context of the Morte and should be included in any resource that aims to show that context. 


But what are the actual relationships between these three witnesses? What has led to their similarities and their differences? In truth, all three witnesses have a close relationship, but the Winchester and Caxton have a particularly close one. In fact, research with the use of special lights on the Winchester manuscript has shown that the manuscript was likely in Caxton’s printing shop sometime between 1480-3 and again in 1489 (Hellinga 133). Various traces of ink and print offsets from Caxton’s type 2 and type 4 fonts can be dimly seen on various folios of the manuscript from its time spent in Caxton’s shop. A vellum fragment from an indulgence printed by Caxton in 1489 was also used to repair the corner of a torn page of the Winchester manuscript. This physical evidence places the manuscript in Caxton’s shop for nearly a decade and shows how Caxton and his compositors almost certainly used the manuscript as a type of reference text for corrections. However, scholars do not believe that Caxton actually used the Winchester for his copy text, not only because it lacks the marks usually made by compositors, but also because of the range of differences between the Roman war episode. Instead, the ink stains on the manuscript likely mean that the Winchester was used to correct or check Caxton’s text but was not the actual copy text marked up by compositors. Nonetheless, these two texts are quite closely related; the Winchester doubtless influenced the development of the Caxton in some ways, and sometime between 1480-9 these two texts existed in close proximity.


At some point, the manuscript left the shop, as did Caxton’s text; the two, which once existed in harmony in a single shop, have become the mascots in a textual war between scholarly camps. Among Malorian scholars, one must be either in favor of the Winchester or the Caxton. The history and features of both texts have become lost in the petty details of lexical differences and opinions on authority. We have forgotten how the dialogue between these two texts, the relationship established in the 1480s between them, can help us understand something more about the historical period, about the late fifteenth-century, and about English manuscript and print cultures.


The hermeneutic gap created by the two documents can be utilized to gain insight into late manuscript and incunabula reading communities, into scriptorium and printing house practices, into changing lexical patterns in the fifteenth century, and into a literary community in which the production of literary texts was very much in flux. (Roland 316)


We must not forget, as Meg Roland points out, the important conclusions and historical contexts that can be established by reestablishing the former relationships between the texts, by placing them in harmonious dialogue yet again. These two texts are not the only contemporary fifteenth-century witnesses, though. The final and simplest relationship exists between Caxton’s text and de Worde’s. After all, de Worde adapted and changed the copy made by Caxton to produce his own text. The differences between Caxton’s and de Worde’s text, and especially between the manuscript and de Worde’s text, should not be used to question their authoritative status but rather to reveal, in even greater detail how, in just two decades, the demands and opinions of the reading public and of printers in general changed. 


For a long while, the Winchester manuscript is the earliest extant witness of the Morte; evidence from watermarks and printing ink stains has shown that the manuscript must have been commissioned sometime between 1477 and 1483 (Ker x, Hellinga 133). 


See Oakeshott’s “The Finding of the Manuscript” in Essays on Malory, Ed. J. A. W. Bennet (Oxford: Clarendon, 1962): 1-6.


In 1839 or 1840 when W.T. Alchin cataloged the library at Winchester, the manuscript was simply listed as “The History of Prince Arthur. Imperfect” (Ker xx). Before that its location is uncertain, though a few recent studies have proposed it was possessed by the Malory family in Litchborough and later the Risleys where at some point it ended up in the schoolroom at Winchester. (See Paul Yeats-Edwards)


Many modern editions, such as those of Vinaver and Shepherd represent these marks by starting a new paragraph.


The presence of different scribal/compositor errors involving homeoleuton or eye skip also show this.