1 Introduction

Since the emergence of the scholarly field of book history, it has been customary to see the printed book as an agent of change. By being typeset, printed, and issued in large numbers – by being published – texts can penetrate society and change how its members perceive the world. In considering the preprint literary world, our analytical approach has tended towards the opposite direction. We look at the numbers of surviving manuscript copies of a text from the Middle Ages and we typically assess them as signifiers of the interests of the times, as cultural objects and not as evidence of an attempt to distribute a given text. Whereas Aldus Manutius, publishing classical texts in Venice around 1500, is seen as an actor who made a contribution to the diffusion of humanistic modes, the proliferation of copies of Cicero’s works in the twelfth century is taken instead to indicate a growing interest in classical Latin literature. The printed book sets change in motion; the manuscript book is merely the effect of change.

Analogically, when we reflect on the literary success of a given work in the world of print, we almost reflexively think about the role publishing played in its making. In a manuscript culture, by way of contrast, the dissemination of a work is typically understood as an organic process, taking place outside the author’s control and often extending far beyond his or her lifetime. To a point, this difference in perspective is valid and natural. The printing press makes of publishing a clearly defined action taking place at a specific moment; furthermore, owing to the financial burden involved in printing an edition, it also makes publishing inherently commercial and promotional. In medieval manuscript culture, publishing is not an act similarly defined by any mechanical operation or investment in large-scale production belonging to a particular point in time. As Daniel Hobbins has emphasised, publishing in manuscript culture is always a diffuse process.Footnote ¹

We should not, however, assume that this process was typically innocent or void of authorial intentions. What is the point of writing in any context – print, manuscript, or digital – if not for an audience? Printing may have raised the stakes by greatly enlarging the potential audience and making a wide distribution much easier to achieve but it did not change the fundamental rationale of publishing as an interface between an author and an audience. What is more, while a limited audience of close associates was enough for some medieval authors, narrative evidence clearly shows that others intended to make their works available in an open literary public sphere, the existence of which seems to have been generally acknowledged. That no single actor could play the distributive role of the printer should not obscure the fact that authorial publishing did happen in the manuscript context and that its fundamental goal was the same as in the age of print: making a new text available to an audience.

In what follows, I present both a methodological suggestion about how authorial publishing can be conceptualised in manuscript culture and an empirical application of this proposed conceptual framework. My key analytical proposition, around which this study is organised, is that authorial publishing in manuscript culture was about creating a task force – a publishing circle – to drive forward the distribution of a text. The publishing circle, again, operated within a wider network of literary connections and modalities of book production which provided the intellectual and material infrastructure for dissemination: in brief, a ‘publishing framework’. Methodologically, I suggest that these twin concepts of publishing circle and publishing framework can provide a helpful pair of tools for empirical analysis of authorial publishing in any manuscript context.

The substance of this study consists of an application of this approach to three twelfth-century Anglo-Norman historians. The works which I will examine are the Historia Anglorum by Henry of Huntingdon (c.1088–1157), Gesta regum Anglorum by William of Malmesbury (b. c.1090, d. in or after 1142) and De gestis Britonum by Geoffrey of Monmouth (d. 1154/5). Besides offering a contribution to the study of publishing in manuscript culture per se, I aim to provide a better vantage point for the appreciation of these works in their historical context. All these texts attained canonical authority soon after their appearance and they came to provide the starting point for all later medieval histories of Britain, whether in Latin or in the vernacular. Indeed, they are still among the most widely read and studied twelfth-century texts from England. My analysis of their publishing histories suggests that this success was not a simple consequence of having written attractive texts but that these authors were actively seeking an audience and literary fame. They were, it seems, also keenly aware that, by so doing, they were participating in a public discourse, one with potentially political implications.

To clarify how my approach differs from previous attempts to conceptualise publishing in a manuscript context, I shall begin by discussing previous work done on this topic. I shall then provide a survey of the wider historical framework of publishing, the key components of which were social networks and the infrastructure of book manufacture. Then, more briefly, it will be necessary to review the nature of the evidence on which this study rests. The reconstruction of the publishing histories themselves, which follows, constitutes the main body of this study, in which the authors are discussed separately. This straightforward narrative structure is necessitated by the complicated nature of the source work involved. While publishing histories are often intricate in the case of print, the disentangling of the messy threads of evidence that allows us to see how publishing took place in manuscript culture is a yet more complicated affair.

2 William of Malmesbury

William of Malmesbury was one of the most learned Latin scholars of his day. We know very little about his early life but he was born not far away from the Benedictine abbey of Malmesbury (Wilts), which he entered at an early age and from where he may have been sent to Glastonbury, Canterbury, or both, for further education.Footnote ³⁴ William’s literary output was prolific, consisting of history, hagiography, florilegia, and biblical exegesis.Footnote ³⁵ This study will concentrate on the publishing history of his most successful work, the Gesta regum Anglorum (GRA), but it will also examine more briefly those of his other historical compositions – the Gesta pontificum Anglorum (GPA) and the Historia novella (HN) – to help us better understand William as a publisher.

The genesis of the GRA went back to William’s encounter with Matilda (d. 1118), queen of King Henry I. According to William, she had first asked William for an account of the family history of St Aldhelm, founder of Malmesbury and her own relative. Once this had been composed, Matilda requested that William proceed with a more comprehensive history of the English.Footnote ³⁶ The substance of this project, which became the GRA, was finished c.1125, and chronologically William never continued this narrative any further, even though he lived until (at least) 1142.

Very soon after the end of his history, in 1125, a fair copy (T) was produced from William’s draft (W1) and, from this copy, descended the earliest textual family of the GRA, the so-called T version.Footnote ³⁷ The manuscript of T itself (which does not survive) did not receive any integral dedication but two letters were written to accompany it and they, together with the manuscript evidence, allow us to see how the first steps in the publishing of the work were taken.Footnote ³⁸ These letters, both of which accompanied T, announced the work to King David of Scotland and Matilda, Henry I’s daughter, usually called Empress Matilda because of her previous marriage to Henry V of Germany and to differentiate her from her mother, Queen Matilda (d. 1118), King David’s sister and the original commissioner of the GRA.

This curious set of letters allows us to fix the presentation of the T manuscript and the beginning of the circulation of the GRA to a specific historical moment. The letter addressing David asked him to accept and authorise the GRA and to take care that both the book and the messenger from Malmesbury accompanying it be conducted to the empress.Footnote ³⁹ One may wonder whether this authorisation and such help in delivery were really needed and, indeed, David was not addressed simply so that he could add his stamp of authority. Malmesbury Abbey had been without an abbot since 1118 and the second half of the letter begged David to remedy this situation.Footnote ⁴⁰ A Scottish king had no prerogative to appoint abbots to a Wiltshire monastery, however, and the monks must have hoped that he would exercise his influence in the English court, where he had spent a lengthy period (from Michaelmas 1126 until summer of 1127) just after the GRA had been finished.Footnote ⁴¹ This hypothesis is compatible with what is known about the movements of Henry I’s court. The first documentary trace of Henry’s arrival in September 1126 is a charter issued at Portsmouth. This is followed by a series of charters given at Rockingham, in the East Midlands.Footnote ⁴² King David does not appear as a witness in any of these, even though one of the Rockingham charters (Regesta, no. 1459), confirming Walter Espec’s gifts to Kirkham Priory, has a long witness list with all the important names of the court present. However, we find King David at the head of the witness list in a charter issued at Woodstock, around fifty miles from Malmesbury and evidently familiar to William himself, sometime in the autumn of 1126.Footnote ⁴³ The likely conclusion is that David was met by the monks of Malmesbury in 1126 as he was making his way towards the meeting of the court and that a representative from Malmesbury joined his entourage with the book.

In the context of this particularly momentous assize of the royal court, the petition for assistance towards an introduction to Matilda does not seem in the least absurd. It was David’s sister (and Empress Matilda’s mother) who had commissioned the GRA; and David himself, having spent his youth in Henry I’s court, was a relatively familiar figure to the Malmesbury monks. This was not the case with Empress Matilda. She had been in Germany since 1114, when she left England as a child bride of eight. With her husband, Emperor Henry V, recently deceased, she was returning to England in the autumn of 1126, escorted by her father King Henry. She was virtually an alien and, yet, in the Christmas court of 1126, she was publicly named heir to the English throne and oaths of allegiance were sworn by the magnates.Footnote ⁴⁴ In the various arrangements made to pave her way, her uncle, King David, played a central role. He was the first lay magnate to swear loyalty to Matilda and the symbolic value of his commitment was high.Footnote ⁴⁵ The presentation of the GRA to her, via David, should undoubtedly be placed in this context.Footnote ⁴⁶ From the perspective of the Malmesbury monks, it would have made sense to try to establish close ties with the empress by way of a historical work which her own mother, Queen Matilda, had commissioned. From David’s perspective, playing intermediator might well have been a welcome tool in the, no doubt, somewhat awkward process of establishing a rapport with the female outsider who was going to be the future ruler of England and, therefore, the single most important external element affecting the future of his own kingdom. Symbolically, the book gift had potency. It introduced Matilda to the past of her future realm, just as she had been introduced to the past of the empire through a presentation copy of Ekkehard of Aura’s chronicle on the occasion of her marriage to the emperor.Footnote ⁴⁷ That political potency of the gift would only have been emphasised by the continuum between mother as commissioner and daughter as dedicatee.

This presentation, datable on reasonable grounds to the second half of 1126, was, from what we can tell, the beginning of the publishing history of the GRA. While it is theoretically possible that the work was simultaneously (or even slightly earlier) available by other means, there is no evidence to support the suggestion. The language of the dedicatory letter to Matilda suggests that the work had been recently finished and, indeed, the gap of time between the end of the narrative and the (probable) moment of presentation was not long.Footnote ⁴⁸ More importantly, such an interpretation receives support from the manuscripts. The T version survives in nine copies, of which seven date from the twelfth or the beginning of the thirteenth century. They are all continental and, apart from one manuscript from Clairvaux, their medieval provenance is in north-eastern France or Flanders. Matilda returned to the Continent after her visit to England (1126–7) and she spent most of her time there until the end of the 1130s.Footnote ⁴⁹ Given that not a single British descendant of the T manuscript presented to her is known, the odds are that she took the presentation copy to Normandy with her.Footnote ⁵⁰ The dedicatory letters written for the presentation survive in only a single manuscript, from the second half of the twelfth century, now Troyes, Médiathèque du Grand Troyes, MS 294bis. This manuscript bears a late medieval Clairvaux shelfmark and, given the similarity of its monochromatic decoration to other Clairvaux books of the period, it appears likely that it was produced there as well.Footnote ⁵¹ The absence of the dedicatory letters from all other manuscripts suggests that the letters moved with the presentation copy as separate documents, accessed by the copyists working at or for Clairvaux in the second half of the twelfth century.

If we restrict the word ‘publishing’ to mean the first entry of a written work into circulation with its author’s permission to make more copies of it, this, then, was the instant of publication: the monastery of Malmesbury presenting William’s work to Empress Matilda. However, if we think of publishing as the author–audience interface, this was just the beginning of the process. There are reasons to think that the initial book launch, despite its auspicious setting, turned out to be a disappointment. Certainly, the T manuscript given to Matilda completely failed to initiate any circulation of William’s magnum opus in England, having left no textual trace in the insular transmission. The initial success on the Continent was apparently limited as well. We only find copies of the T version dating from the middle of the twelfth century onwards and they all stem from Flanders or eastern France, i.e. from outside Matilda’s continental heartlands in the Rouen region and Anjou. Given that the dedicatory letters were copied at Clairvaux, it rather appears that the T manuscript started to produce offspring only once it had, for some reason, left Matilda’s orbit.

Whether William understood the disappearance of the T manuscript to the Continent as a publishing failure we cannot know but his next move certainly appears to have been an attempt to make the work more widely available by extending the publishing circle. Sometime in 1127–9 a new copy (the archetype of the second or A version) was produced from William’s draft, now somewhat revised, and given to Robert, earl of Gloucester, King Henry I’s influential illegitimate son. William did not explain why he had composed a new dedication but he acknowledged the previous one with his choice of words and, between the lines, one may read possible disappointment with it: ‘I have thought it right to dedicate especially to you (uobis potissimum) the History of the English Kings which I have lately put out (quem nuper edidi).’Footnote ⁵² The dedicatory letter is not actually found in any of the copies of the second version, only in copies of still later ones, but the epilogue of the second version makes it clear that Robert was already its dedicatee.Footnote ⁵³ The logical conclusion is that the second presentation copy was accompanied by a physically separate cover letter, just as apparently was the case with the first one delivered to Matilda.

Manuscript evidence indicates that this presentation worked better in advancing the distribution of the work in Britain than the previous one and that its success was at least somewhat more immediate. For the second (or A) version of the GRA, we have ten manuscript witnesses, nine of them dating from before the middle of the thirteenth century, five from or before c.1200 and two probably from William’s and Robert’s lifetime. Apart from one manuscript described as showing English and French scribal hands, their origins have all been assigned to England.Footnote ⁵⁴ While there is no evidence that would directly inform us of Robert’s own contribution to this dissemination, there is evidence at least to show that William had managed to foster a close relationship with the earl. William’s last historical composition, the Historia novella, unfinished at the time of his death, was not only dedicated to Robert but also deeply partisan towards him.Footnote ⁵⁵ Furthermore, it has been plausibly suggested that the systematic toning down of material that denigrated William I and William II – Robert’s grandfather and uncle – in the third version of the GRA resulted from Robert’s feedback.Footnote ⁵⁶

Most tangibly, Robert’s sustained interest in William’s literary activities is indicated by a manuscript that postdates both of them, London, BL, MS Royal 13 D. II, from the end of the twelfth century. This manuscript contains three texts which were dedicated to Robert: Geoffrey of Monmouth’s De gestis Britonum, William’s GRA, and William’s Historia novella. They all have features indicating closeness to the dedicatee. The text of the De gestis is textually the single best witness of the version dedicated to Robert.Footnote ⁵⁷ That of the Historia novella is a special ‘Robertian’ recension, with revisions and rubrics emphasising his achievements.Footnote ⁵⁸ The text of the GRA is based on the third version but it has been cautiously corrected against the second version, which was the version of the presentation copy given to Robert.Footnote ⁵⁹ Like the text of the De gestis, it too is highly accurate and has been given a prominent place in the editions. There is no independent evidence on the origin of the manuscript but its early thirteenth-century home was Margam Abbey (OCist, Glam), one of the few monastic foundations of Robert of Gloucester. The probable interpretation of this assemblage of evidence is that the book’s exemplars were the presentation copies which Robert had received and which his descendants had preserved. The textual state of the GRA indicates that this literary patrimony contained (at least) two copies of it, i.e. both the original presentation copy with the second version text and a copy of the third version, possibly revised according to Robert’s suggestions. Mainly because of the ‘Robertian’ nature of the text of the Historia novella, it has been suggested that it was revised by Robert’s son, Roger of Worcester and, indeed, it is a possibility that all these books had belonged to him before they were used in the making of the Margam book.Footnote ⁶⁰ Whether or not this was the case, the textual ensemble demonstrates that Robert took care to preserve the books he had received and that the collection was considered a patrimony worth curating by his descendants.

Manuscript evidence suggests that, from the second version onwards, the GRA was also being distributed from Malmesbury through other channels than Robert. One of the copies of the second version, London, British Library, MS Arundel 35, datable to William’s lifetime and possibly quite close to the presentation of this version, allows us a glimpse of the mechanisms at work. Arundel 35 is textually the purest copy of the second or A version.Footnote ⁶¹ Thomson and Winterbottom have suggested that this manuscript was copied at Winchester from William’s draft which had been sent there precisely for this purpose.Footnote ⁶² Certainly no fewer than four twelfth-century copies descend from it.Footnote ⁶³ The contact between Malmesbury and Winchester could be framed in institutional terms but it is not difficult to name an individual who might very well have been involved. From 1129 onwards, the bishop of Winchester was Henry of Blois, to whom William was already connected. Henry had, since 1126, been abbot of Glastonbury, where William did much research work, and, ultimately, William dedicated his history of Glastonbury Abbey to him. What is more, we know that Henry had a copy of the GRA produced for the library at Glastonbury.Footnote ⁶⁴

Putting the first and second versions of the GRA into circulation seems to have involved at least three separate authorial distribution channels: Matilda, Robert, and Winchester. The manner of dissemination of the third (or C) version, datable to c.1135, suggests that, by then, this publishing circle had succeeded in creating an interest in the text. The third version survives in nine English manuscripts for which it has not been possible to establish a clear stemma. The editors have concluded that these manuscripts most likely reflect a developing exemplar, i.e. William’s draft of the third version, which was copied at various moments during its composition.Footnote ⁶⁵ This interpretation of the textual evidence, while it may seem audacious, finds support in the odd placement of the dedicatory epistle in these manuscripts. This is not found at the beginning of the work, which would be its natural position, but between the end of Book III and the start of Book IV.Footnote ⁶⁶ Such a position seems unlikely for a copy intended to function as the official exemplar of a new version. However, in William’s working copy the letter could easily have been placed there for practical reasons. The GRA consists of five books and we know from William’s own testimony that there was originally a hiatus in writing after Book III was finished.Footnote ⁶⁷ This kind of compositional history could have led to the draft existing as two separate codicological units, i.e. one including Books I–III and the other Books IV–V. Since the dedicatory epistle to Robert was composed later than most of the work itself, it could not be easily inserted in its proper place at the beginning of the working copy. At the same time, seeking to preserve its text on a separate leaf would have left it prone to loss. An empty leaf at the end of William’s working copy would have been a natural place for making an archive copy of the dedicatory text.

Indeed, one later manuscript hints at this kind of an arrangement. Paris, Bibliothèque nationale de France, MS lat. 6048, from the second half of the twelfth century, contains a text of the third version but, untypically, omits the dedication.Footnote ⁶⁸ In this manuscript, the third book, after which the dedicatory letter is found in all other copies, ends mid-page (fol. 66v), with seven ruled lines after its last words. The last word of Book III (‘modum’) is written in display script, an effect often reserved for the last words of a work or a codex and not used at the end of the first (fol. 7 r) or second (fol. 48 v) book of the GRA in this manuscript. The preface to the fourth book begins a new folio (67 r), with an elaborate initial, and the text of the fourth book itself is opened by an even grander, seven-line initial, followed by a line in extra-large display script. The other changes of books are not similarly treated.Footnote ⁶⁹ This treatment of the beginning of the fourth book would be compatible with the exemplar having existed as two separate codices, one containing Books I–III and the other IV–V. As to the absence of a dedicatory letter in this manuscript, it bears noting that the opening folios are missing. In other words, the scribe of this manuscript may well have moved the letter to its ‘correct’ place at the beginning of the whole work.

Whatever the precise story behind the idiosyncratic Paris manuscript, the other representatives of the third (or C) version apparently descend from a series of copies made of William’s draft of the GRA over the period c.1126–35.Footnote ⁷⁰ During this period, William was working at least in part at Glastonbury, undertaking research for his work De antiquitate ecclesiae Glastoniensis. Traces of the Glastonbury period are observable in the content of the third version of the GRA as well, and the editors have suggested that it was in fact produced for the Glastonbury community.Footnote ⁷¹ Certainly, the manuscript presenting the latest state of its text was copied by a monk of Glastonbury in 1411,Footnote ⁷² and its exemplar was an early Glastonbury copy.

Besides being copied for the Glastonbury community in its finished state, the evolving draft of the third version was also copied during its process of composition. Most of the copies descending directly from the working copy of William’s Glastonbury period are unlocalised but one of them allows us to reconstruct one of the directions in which the text travelled. This is Cambridge, University Library, MS Ii. 2. 3, from the third quarter of the twelfth century. Its medieval provenance is Buildwas (OCist, Salop) and it has elaborate monochromatic initials, often found in Cistercian books of the period, although these are unique in their decorative language. It was very likely produced for the Buildwas library around 1170.Footnote ⁷³ However, its text has an interpolated passage, which led William Stubbs, the nineteenth-century editor of the GRA, to think that it was made for St Peter’s Abbey, Gloucester.Footnote ⁷⁴ This passage narrates how a certain Roger of Gloucester donated a manor to St Peter’s on his deathbed and it additionally contains the text of the donation document and a later court decision defending St Peter’s title to this manor.Footnote ⁷⁵ While St Peter’s was not the place of origin for this manuscript, Stubbs was correct in assuming that the text of the GRA with such an interpolation must have been originally made for use at St Peter’s. The most plausible explanation for the presence of this passage in the Buildwas book (CUL Ii. 2. 3) is that it was copied from an exemplar belonging to St Peter’s. Two as yet unlocalised twelfth-century copies, Oxford, All Souls College, fragment b. 32, no. 22 (two bifolia) and Oxford, All Souls College, MS 33 also contain the interpolation and, thus, descend from the same St Peter’s copy. St Peter’s Abbey, renovated under its post-Conquest abbot, Serlo (d. 1104), was flourishing in William’s time and its book collection evidently functioned as another important supply centre for the GRA, although whether William was directly involved in making the text available at Gloucester is difficult to know. However, the fact that William’s unfinished text was copied several times during the Glastonbury period indicates that the GRA was now subject to considerable attention. One suspects that the extension of William’s publishing circle to include Robert of Gloucester, Henry of Blois, and the Winchester and Glastonbury communities, two focal points in the ecclesiastical landscape of south-western England, was an important factor in its increasing appeal.Footnote ⁷⁶

Finally, the third, or ‘Glastonbury’, version was followed by the fourth (B) version of the GRA. In this version, from which the most critical comments on Robert of Gloucester’s ancestors had been purged, the dedication to him was finally placed in its logical place at the beginning of the work. The textual history shows us that a new master copy had now been made and one suspects it was delivered to Robert. There is, however, little evidence as to the contemporary distribution of the fourth version. It exists in eight manuscripts but, compared to the previous versions, more of these are relatively late copies, the two earliest (unprovenanced) manuscripts datable to the last quarter of the twelfth century.Footnote ⁷⁷ At this stage, the GRA was already available through many channels and, perhaps also because the latest version did not add to its chronological coverage, it did not come to enjoy any privileged position in the later dissemination.

The story of how the GRA entered circulation thus consists of two discrete moments of publishing, in which the author took the initiative, first by delivering a copy to Empress Matilda via King David and then by giving another one to Robert of Gloucester. The dedications, however, show us only one part of William’s publishing circle. The cathedral priory of Winchester and Glastonbury abbey functioned as important textual supply points already during the period of William’s literary activity and their involvement in the distribution was very likely connected to Henry of Blois’s interest in William’s work. While the text of course spread in a more organic way once it was widely enough available, the first stages in the distribution process suggest that the authorial publishing circle was important in creating this initial availability.

The workings of the publishing circle are highlighted by a comparison with William’s other main work, Gesta pontificum Anglorum (GPA). While the GPA is in some sense the ecclesiastical counterpart of the GRA, there were fundamental differences between the two projects. Written over a short period time – perhaps two or three years – the GPA is a less polished literary product than the GRA. Moreover, it never received any dedication and its prefatory material is less comprehensive than that of the GRA.Footnote ⁷⁸ The GPA, a conspectus of the English church and especially of its ‘monastic wing’, was apparently never formally published, perhaps because it was difficult to think of single dedicatee for such a work. This was William writing about and for his own people, the monastic communities of England.

The textual transmission indicates that the early distribution of this text happened differently to that of the GRA, according to a process that would appear to have been fortuitous rather than planned. The GPA exists in two different versions. The first branch of the transmission went back to a single copy of William’s draft. It is preserved in six manuscripts, of which all but one are known to have belonged to religious institutions in Yorkshire or Lincolnshire. It has been plausibly suggested that they originated from a copy made of an early state of the GPA for some northern house.Footnote ⁷⁹ This first version included curiously offensive material and was soon (c.1134) thoroughly revised by William.Footnote ⁸⁰ Most likely, it was never intended for wide circulation and its distribution in Yorkshire seems, from an authorial perspective, like an accident, perhaps the result of a draft given to another for comment and then unintentionally entering wider circulation.

Interestingly enough, we do not find a clear final state of the second version either. Just as was the case with the third version of the GRA, we encounter a textual spectrum resulting from copies having been made of the evolving draft at various stages in its development.Footnote ⁸¹ Some of the manuscripts of the second version fall into geographical groupings. The members of one belonged to religious houses in the Midlands and Yorkshire,Footnote ⁸² and those of another have associations with Kentish institutions.Footnote ⁸³ The impression is that William never issued a finished master copy of the GPA but that some monastic houses, once they knew about his project, arranged for the copying of his working copy and then passed on the text within their own circle. A text that was intended primarily for readers based in religious houses and cathedral chapters perhaps did not need promotion in the same way as one that sought more widespread and mixed readership in the twelfth-century res publica litterarum, aristocratic and scholarly circles included. In other words, the monastic publishing framework functioned, for certain kinds of texts, so efficiently that it was not in equal measure necessary to actively create a publishing circle to achieve distribution.

3 Henry of Huntingdon

William may have travelled across Britain in search of sources but he remained a monk, whose life unfolded within the stable framework of a Benedictine community. Henry of Huntingdon (c.1088–1157) played many more varied roles in his life. Son of the archdeacon of Huntingdon, he received a superb Latin education at Lincoln, where he was a member of the glittering episcopal curia of Robert Bloet (d. 1123). In Robert’s entourage, he also frequented Henry I’s court (during c.1120–3) and very likely had aspirations towards being a court poet. However, Robert’s fall from grace, in 1123, seems to have cut this career short and, for the remainder of his days, Henry was to be archdeacon of Huntingdon, a position he had inherited from his father in 1110. Over the years, Henry also married, had children, and managed his family estate.Footnote ⁸⁴

Alongside his duties and other activities, all through his life Henry wrote. He produced love poems, epigrams, a herbal, a lapidary, epistles, and what undoubtedly was his life’s central literary project, the Historia Anglorum (HA), a history of the English people from their first arrival in Britain until Henry’s own time.Footnote ⁸⁵ Whereas some of Henry’s works do not survive at all, and others do so in only a small number of copies, the HA is preserved in around forty medieval manuscripts, of which thirteen can be dated to the twelfth century (including manuscripts dated to the turn of the century).

At a glance, the publishing history of this moderately successful work might look simpler than that of William’s GRA. In all those manuscripts that include the prologue (and these are the great majority), the HA is dedicated to the same person, Alexander, bishop of Lincoln (d. 1148), Henry’s superior in the ecclesiastical hierarchy and successor to his first protector, Robert Bloet. Indeed, it was Alexander, Henry tells us, who had commanded him to compose the history.Footnote ⁸⁶ Besides a single patron, Henry also gives us a single date for the completion of the work, the year 1135, about which he wrote at some length in the epilogue of the HA:

Laying such emphasis on the date of writing is quite extraordinary, as John Gillingham has noted.Footnote ⁸⁸ As if this were not enough, in several of the manuscripts this date also appears as a colophon and as an introductory rubric, both apparently autograph.Footnote ⁸⁹

Henry’s words, however, conceal, quite intentionally, a complicated history of composition and publication. Diana Greenway, the text’s modern editor, has identified no fewer than six different versions which entered circulation at various dates, none of them ending in 1135 and none of them necessarily even published that year.Footnote ⁹⁰ In earlier scholarship, these versions were labelled as ‘editions’,Footnote ⁹¹ a term abandoned by Greenway in preference for the more neutral word ‘version’, since, as she rightly points out, ‘the different forms of Henry’s text are not polished “editions” in the modern sense: rather, they reproduce his work at particular stages in its evolution, when copies were made.’Footnote ⁹² However, while some of the six versions did not result from intentional publication, others certainly did. An examination of the genesis of these different versions demonstrates that the process of publishing was, in Henry’s case, intimately linked to advancing the distribution of the text.

To understand this story, we need to start from the two earliest versions of the HA known to us – versions which show us a still unpublished text. The first thing to observe is indeed that Henry withheld finishing and publishing the HA for a long time. Whereas the chronological narrative of the first version ends at 1129, it contains references to events which took place later, in 1131 and 1133. These have been interpreted by Greenway, plausibly, as later additions made to the margins of a draft whose writing had concluded in 1129 or 1130.Footnote ⁹³ The second version in other respects resembles exactly the first but it contains an added epilogue which explicitly dates the composition to 1135.Footnote ⁹⁴ Each version is preserved in one twelfth-century copy and neither of these manuscripts was copied from a finished exemplar. The first version, found in Edinburgh, National Library of Scotland, MS Advocates’ 33.5.4 lacks rubrics, the so-called laudes of the Roman emperors (which constituted the first book of the HA in its finished state) and it also lacks the epilogue.Footnote ⁹⁵ The laudes and the epilogue, but not the rubrics or the prologue, are found in the one twelfth-century witness of the second version, Aberystwyth, NLW, MS Peniarth 382, copied c.1196. As Diana Greenway has demonstrated, this manuscript descends from an intermediary manuscript copied while Henry was still at work.Footnote ⁹⁶

In other words, what evidence we have for the earliest distribution suggests that Henry, first, sat on an essentially completed work for the period 1129–33/5 and then it was in some way made available as essentially unfinished (versions one and two). The imperfect state of the preserved copies of these versions suggests that this initial circulation was limited. The presence of the epilogue with the date of composition in the second version does not imply otherwise. While it used to be assumed that the epilogue would fix the date of ‘official’ publication to 1135, Greenway’s and Benjamin Pohl’s more recent studies have refuted this idea.Footnote ⁹⁷ Greenway’s examination of the textual history has demonstrated that the Peniarth manuscript (the sole witness of the second version) descends from the same draft as the Advocates manuscript (the sole witness of the first version), simply representing the authorial draft in an updated state. However, the mass of manuscripts with the 1135 epilogue represent completely different, later, and more polished textual states.Footnote ⁹⁸ What is more, the previously held proposition that Robert of Torigni would have had access to a version published in 1135 has been thoroughly discredited by Benjamin Pohl’s reconsideration of Robert’s use of the HA. We can now be certain that Robert only had recourse to one of the later versions.

Admittedly, Henry’s HA was not completely unavailable in c.1135. Geoffrey of Monmouth, who likewise belonged to the literary circle around Henry’s patron Alexander, had access to the HA, most likely sometime between 1135 and 1137.Footnote ⁹⁹ The manuscripts of the first two versions of the HA themselves, moreover, constitute evidence of limited circulation taking place in the 1130s. The most obvious case is that of the Edinburgh manuscript, which, as discussed above, does seem to belong to this period. However, we have also observed that these manuscripts represent an unfinished work and we know that when Henry went to Le Bec in January 1139 he did not give Robert of Torigni a copy, even though he must have known that Robert was a keen collector (and distributor) of historical works. This visit to Le Bec, moreover, occasioned a curious document which perfectly fits the idea that the HA was essentially still a work in progress in early 1139. Robert of Torigni introduced Henry to Geoffrey of Monmouth’s De gestis Britonum, previously unknown to him, and Henry wrote about this discovery to a person called Warin the Briton.Footnote ¹⁰⁰ Much later, Robert included the same letter in his own chronicle, with a preface that underlined his own role in making Geoffrey’s work available to Henry. In this preface, he mentioned how he had inquired of Henry on this occasion about the HA, which shows that he was interested in Henry’s activities and suggests that a copy of the HA was not available at their meeting. The content and purpose of Henry’s letter point to similar conclusions. The letter was addressed to Warin who, it is stated, had commented on the HA, in particular wondering why it had nothing to tell about the early British kings. The fact of his having written a long letter, running to twelve and a half pages in the printed edition, demonstrates that Henry valued Warin’s opinion and indicates close personal contact between the two. This kind of exchange would be well accounted for if Warin had been part of a small circle to whom Henry had made the work available for comment. While reliable identification is not possible, one of Henry’s contemporaries at Lincoln may have been a certain Guarinus the canon.Footnote ¹⁰¹ Certainly, the scholarly community around Lincoln Cathedral, where Henry himself had studied and to which he maintained very close links, would be prime territory for that sort of limited early circulation.

Whatever the precise extent and context of the preliminary distribution of the HA in the 1130s, it is clear that, early in the 1140s, the status of the HA changed. The next – the so-called third – version of the HA continued the chronological narrative up to Christmas 1138; its writing finished in 1141 or 1142.Footnote ¹⁰² For this version, the state of the text, its paratexts, and the manuscripts all join forces to give an impression of a work thoroughly polished and duly published. Whereas the two previous versions, each surviving in single twelfth-century manuscripts, were copied directly from Henry’s evolving autograph, all the manuscripts of the third version descend from a common source that was different from the autograph – the presentation copy, one may logically presume.Footnote ¹⁰³ By now, the laudes had been integrated into the text, it had received rubrication, and a colophon had been added. The two earliest manuscripts have an introductory rubric as well.Footnote ¹⁰⁴ The number of manuscripts witnessing only parts of the third version (6) is double the number of manuscripts of the two previous versions put together (3), which also suggests wider circulation.Footnote ¹⁰⁵

Only now, more than ten years after the main narrative had been completed, Henry took the necessary steps to create a finished literary product of the HA and put it into open circulation – all the while keeping up the appearance that the work had been written in 1135. ‘This is the year which contains the writer’, he stated, and ‘[t]his, then is the year from which the writer of the History wished his age to be reckoned by posterity’.Footnote ¹⁰⁶ Why such a delay and why such an emphasis on the year of writing? Henry’s hesitation in publishing and his decision to obscure the date of composition had to do with his handling of his portrait of King Henry I. As Alan Cooper has demonstrated, a series of modifications turned the relatively favourable portrait of the first two versions – written while the king was alive – into a brutally critical one in the third.Footnote ¹⁰⁷ Furthermore, the king’s cruelty and wilfulness were underlined in the De contemptu mundi letter, now integrated into the HA, which also showed the king scheming to bring down Robert Bloet – Henry of Huntingdon’s own link to the court in the golden days of his youth.Footnote ¹⁰⁸ The character assassination was completed by the newly written tenth book of the HA, which included a new, damning portrait and the oft-cited description of the king’s disintegrating corpse.Footnote ¹⁰⁹ Henry of Huntingdon was, of course, aware that he could be accused of hypocrisy in having changed his view of King Henry and he discussed this accusation explicitly in De contemptu mundi, finding various excuses.Footnote ¹¹⁰ The obfuscation of the date of writing was undoubtedly one part of his authorial strategy of defence: it created the appearance that the work had been finished within the king’s lifetime.

While we have no independent evidence on the precise circumstances of the publication of the HA in 1141 or 1142, the most obvious procedure would have been to present a copy of the finished work to Alexander of Lincoln, by whom the HA had originally been commissioned – a very long time ago by this point. This presentation did not, however, lead to any very wide dissemination. The number of manuscripts of the third version (6), while bigger than that of the first (1) or second (2) versions, is not high if compared to the more successful versions of William’s GRA, let alone Geoffrey of Monmouth’s De gestis Britonum, and a number of details renders it still less impressive. While four of the witnesses date from the twelfth century, only two of them, Egerton 3668 and CCCC 280, faithfully reproduce the whole third version. One of the partial witnesses is the Advocates copy of the first version, into which the new chronological coverage provided by the third version was simply added. In other words, we see a copy stemming from the period of limited circulation, and possibly from Henry’s closest circles, being updated. The other is Ushaw College, MS 6, an idiosyncratic Yorkshire amalgam integrating Henry’s work with Geoffrey’s De gestis.Footnote ¹¹¹ Moreover, the earlier of the two faithful copies, one contemporary with Henry – Egerton 3668 – was corrected against the master copy of the next version, i.e. the fourth, which, we know, travelled to France soon after its making.Footnote ¹¹² This indicates that the Egerton manuscript remained close to Henry, perhaps in Lincoln. The manuscript evidence for successful distribution of the third version of the HA as a self-standing literary work thus almost completely disintegrates when interrogated in detail.

The slow take-off of the circulation of the now, nevertheless, undoubtedly published HA may have been one reason why Henry kept on seeking new avenues for its distribution. Over the next few years, he created a fourth version, which saw stylistic improvement throughout the narrative and took the story chronologically to 1146. It ended with King Stephen’s crown-wearing in Lincoln at Christmas in that year, an event at which Henry himself was present. This ceremony symbolised the normalisation of the relationship between King Stephen and Alexander of Lincoln, providing a suitable place for a halt in the narrative. The textual history shows that a complete new master copy (delta) was now produced from Henry’s draft, just as was done when the third version was finished.Footnote ¹¹³ The logical assumption that delta was also a presentation manuscript given to the patron, Alexander of Lincoln, is compatible with what textual history tells us about its movements. First, delta was used to correct the Egerton manuscript (possibly in Lincoln) and that is the only direct textual trace of it in England. Very soon afterwards it travelled to Normandy, where a new recension was produced out of it by Robert of Torigni, then prior of Le Bec.Footnote ¹¹⁴ All surviving descendants of delta belong to this recension, called beta by the editor. The movement of delta parallels the itinerary of Alexander of Lincoln, who travelled to meet the pope in Auxerre in August 1147, just after the fourth version had been finished.Footnote ¹¹⁵ Le Bec would not have been a surprising stopover on such a journey.

As to how precisely Robert of Torigni gained access to the text of the fourth version, two scenarios suggest themselves. Either he had a copy made of the presentation copy delta or delta itself ended up in his hands. The latter proposition receives support from the manuscript evidence. On the Continent, the revised delta quickly gave rise to two new manuscripts. These are its earliest surviving descendants, Paris, Bibliothèque nationale de France, MS lat. 6042 and Cambridge, University Library, MS Gg. 2. 21, both dated to the 1150s.Footnote ¹¹⁶ The Paris manuscript was demonstrably used by Robert of Torigni himself and the Cambridge one belonged to Philip, bishop of Bayeux.Footnote ¹¹⁷ Patricia Stirnemann has noted that the two first initials of the Paris manuscript (C on fol. 3r, B on fol. 4r) show decorative motifs that are of English origin and unknown in continental manuscripts of the period. At the same time, the illuminator’s native decorative idiom was continental and, indeed, he may have been the same draughtsman who decorated two other manuscripts produced close to Robert of Torigni. By this reckoning, Stirnemann has concluded that the source of the designs was most probably an insular exemplar.Footnote ¹¹⁸ This is compatible with the hypothesis that the exemplar was the presentation copy, received from Alexander and revised by Robert. In addition, the structure and the motifs of the insular-influenced initial ‘C’, which begins the prologue in the Paris manuscript, find close parallel in the much less well executed initial ‘C’ found in the earliest of all the manuscripts of the HA, the Advocates manuscript of the first version.Footnote ¹¹⁹ It is a possibility that both manuscripts echo styles current in Lincoln at the time.

At the same time, it is clear that the Paris manuscript, certainly used by Robert, could not have been the first copy in his possession. It is already a fair copy of the text as revised by Robert. In other words, Robert’s changes and interpolations were drafted somewhere else – probably in the margins of the exemplar, which is known to have been his approach with other manuscripts. What is more, both the Paris and the Cambridge manuscripts are densely annotated with distinctive ‘a’ and ‘N’ annotations, very likely to have been copied from the exemplar since their distribution in the two manuscripts is near identical. Similar, and similarly dense, annotations are found in a well-known manuscript that was certainly copied under Robert’s supervision and revised by him – Leiden, Universiteitsbibliotheek, MS BPL 20. In this manuscript, it can be demonstrated that the annotations are, indeed, in Robert’s own hand.Footnote ¹²⁰ The close similarity of the annotations in the Paris and Cambridge manuscripts to those of the Leiden manuscript, certainly made by Robert, indicates that their exemplar was also a manuscript annotated by Robert personally. Set alongside the art-historical evidence, the most economical explanation is that the exemplar was the presentation copy delta itself, which Robert had interpolated and marked up.

In addition to the Paris and Cambridge manuscripts, two other twelfth-century copies descend from Robert’s recension. It (or a lost intermediary) was used as the exemplar in the copying of a historical compilation, Rouen, Bibliothèque municipale, MS U.74 (1177), for the library of Jumièges, at the very end of the twelfth century.Footnote ¹²¹ Perhaps more surprisingly, this circulated back to England, as is evinced by London, Lambeth Palace, MS 327, (fols. 1r–176r, s. xii^2/2). This manuscript has a fifteenth- or sixteenth-century ex libris of the Augustinian priory of Bourne in Lincolnshire and its decoration, script, and the placement of a list of the archbishops of Canterbury as the first item in the codex would all indicate an English origin.Footnote ¹²² In terms of advancing the distribution, the inclusion of Robert of Torigni in Henry’s publishing circle proved in this way to be very beneficial. It is most likely that Henry had agreed with Alexander that he make the text available to Robert, in the hope that Robert would show the HA to others, just as he had shown Geoffrey’s De gestis Britonum to Henry.

Having run this errand for Henry, Alexander of Lincoln died in February 1148, of a disease he had acquired on his continental journey.Footnote ¹²³ Henry’s publishing activities were, however, far from over. He immediately went back to working on the HA, creating the so-called fifth version, which extended the narrative to the ascent of Alexander’s successor, Robert de Chesney (6 January 1149). Although this version retained the earlier dedication to Alexander, and still purported 1135 to be the year of writing, it is highly likely that a presentation copy of it was given to bishop Robert, as Greenway has suggested.Footnote ¹²⁴ The fifth version gained popularity over time. There exist three copies from the turn of the twelfth and thirteenth centuries and three others from the first half of the thirteenth, all of which are almost certainly of insular origin.Footnote ¹²⁵ Given that it provided more extensive chronological coverage, it was likely to be preferred over the earlier versions as time passed. However, there is less manuscript evidence on its immediate distribution during Henry’s lifetime than is the case with the previous versions, of which we have at least some contemporary copies. The impression is that, despite the relaunch, the work was still not as effectively distributed as those of William of Malmesbury or, as we shall see, Geoffrey of Monmouth. Henry, in any case, kept on trying to find new outlets. The sixth and final version – the only one which refers explicitly to the giving of a presentation copy – attempted an ambitious extension of the publishing circle. This time, the text of the HA was brought down to the coronation of Henry II, to December 1154, and a poem praising his virtues was added to its end.Footnote ¹²⁶ After the poem, Henry of Huntingdon added the line ‘Et iam regi novo nouus liber donandus est’, indicating at least an intended presentation of a copy of the sixth version to the king himself. Not surprisingly, Henry again revised the portrait of King Henry I – the new king’s grandfather – this time removing some of the vices he had added to the third version.Footnote ¹²⁷ Like the fifth version, the sixth also became popular over time and, in some manuscripts, its text was used to provide the ending for an earlier version.Footnote ¹²⁸ However, the manuscript evidence on early distribution is again scanty, with only one more or less complete copy from the twelfth century, datable to its second half, very likely after Henry of Huntingdon’s own death.Footnote ¹²⁹

To sum up, the publishing history of the HA falls into two distinct phases. First, a long period of authorial hesitation and limited circulation, witnessed by the so-called first and second versions which, from Henry’s perspective, were probably not ‘versions’ at all but just two copies of the draft at different stages. The start of the second phase is signalled by the so-called third version, made available in 1141/2. With this version, which could even be called the first ‘edition’ of the HA, we start to find clear traces of active authorial promotion of the text. Henry kept on revising and continuing the HA over a long period of time and he released it during that period again and again. The circle of people whom Henry sought to involve in the distribution of the finished versions is an impressive one: Alexander of Lincoln, Robert of Torigni, Robert de Chesney, and King Henry II.

In the long run, Henry’s obstinacy in publishing paid off and the HA became widely available, remaining to the present day a standard authority for English history. However, while the overall count of copies datable to or antedating c.1200 is substantial (13), one also notes that no single version achieved great popularity quickly and that some of the versions appear to have remained available primarily in Lincoln and its environs. The only version that certainly circulated further immediately after its making was the fourth one, distributed via Robert of Torigni. Here – but probably only here – Henry’s publishing circle was functioning effectively.

Indeed, perceived failure to achieve satisfactory dissemination may well have been one reason why so many different versions were launched in the first place. One explanation for the relatively limited early distribution could be that, so far as we know, Henry’s text was not quickly available in any important insular monastic libraries, unlike William of Malmesbury’s Gesta regum and Gesta pontificum, for instance. This may have been because Henry, as a secular clerk, did not have similarly natural connections to such centres, which were still at that point crucial textual supply points. The politically sensitive nature of his work might also have been an issue, ensuring that not only monastic but also aristocratic circles were wary of the HA.

4 Geoffrey of Monmouth

Geoffrey of Monmouth’s (d. 1154/5) career was connected to the development of the schools of Oxford, where he was almost certainly a canon of St George’s, a rich and short-lived college of secular canons located at the castle. Geoffrey is called magister in some of the documents he witnessed, which suggests that he had some involvement in teaching. Late in life he was elected bishop of St Asaph, a see he was, however, unable to enter because of the unstable political situation in northern Wales. During his years in Oxford, he composed three literary works, interlinked to one another, and it is to them that he owes his fame. First, there was the Prophetia Merlini (henceforth PM), a booklet released before late 1135, containing the Latin translation of a prophetic vision that Merlin had purportedly received in the sixth century. This was followed by De gestis Britonum (henceforth DGB), better known by its later title Historia regum Britanniae, another alleged translation of an old Brythonic text, which incorporated also the previously published PM (Book VII in DGB). Finally, in or soon after 1148, Geoffrey published a long hexameter epic, Vita Merlini, dedicated to the newly elected bishop of Lincoln, Robert de Chesney, his erstwhile colleague from St George’s College.

DGB, by far the most popular of these works, survives in more than 220 medieval copies – more than five times as many as William’s and Henry’s histories.Footnote ¹³⁰ Moreover, almost eighty of these copies can be dated to the twelfth century (including manuscripts dated to the turn of the century). There is little doubt that this contemporary popularity was in large measure due to the novelty and attraction of the work’s content. DGB provided a previously unknown past for pre–Anglo-Saxon Britain, starting with the island’s settlement by Trojan refugees. It also contained the first comprehensive narrative of King Arthur, who was at this time emerging as a popular figure of chivalrous literature. However, as I seek to demonstrate, Geoffrey was also exceptionally successful in creating a publishing circle that secured immediate distribution for his work.

The process leading to the publication of the complete DGB began with the release of a prelude, the PM, which was presented to Alexander of Lincoln sometime before late 1135, that is, well before the rest of DGB was finished.Footnote ¹³¹ According to Geoffrey’s own testimony, the PM was sent to Alexander with a dedicatory letter.Footnote ¹³² Alexander spent much of 1134 and 1135 in Normandy.Footnote ¹³³ We also know that Orderic Vitalis had read the PM in St Evroult (Normandy) well before December 1135 and that six out of nine twelfth-century copies of the independently published PM are continental.Footnote ¹³⁴ It appears likely that Alexander was responsible for initiating the continental dissemination, in a manner that is similar to his role in the distribution of the fourth version of the HA. Either the PM was presented to him while he was in England, and he took it to Normandy with him, or Geoffrey somehow had it delivered to him across the Channel.

In Britain, too, the fact that two of the three twelfth-century copies have a connection to Lincoln may also be a reflection of Alexander’s connections. One manuscript, of unknown origin and provenance, now belongs to Lincoln Cathedral Library (MS 214). The medieval provenance, and probably the origin, of the other is the Gilbertine house of Sempringham in Lincolnshire, Alexander’s foundation.Footnote ¹³⁵ A Durham origin has been suggested for the third insular manuscript, Liège 369 C, a connection not incompatible with a Lincoln origin.Footnote ¹³⁶ While the PM apparently did not become a great bestseller on its own, the circumstantial evidence suggests that Alexander was involved in its early distribution. Notwithstanding, Geoffrey later complained that this dedication to Alexander had not been richly rewarded.Footnote ¹³⁷ This sounds more like a complaint about a lack of career advancement than of literary success, reminding us of Felicity Riddy’s point that what authors sought with patronage may not chiefly have been publicity and literary fame.Footnote ¹³⁸

Once Geoffrey had completed his main work, DGB, he in any case turned to a different dedicatee, one who had already demonstrated interest in William of Malmesbury’s historical writing: Robert of Gloucester. While the dedicatory history of DGB, discussed further in what follows, has its famous complications, recent textual work has demonstrated beyond doubt that its first version was dedicated to Robert alone and he, consequently, must have received a presentation copy soon after work was finished.Footnote ¹³⁹ The terminus post quem non for the public availability of DGB and, thus, for this presentation, is January 1139, when Robert of Torigni surprised Henry of Huntingdon with his copy of Geoffrey’s work.Footnote ¹⁴⁰ At the same time, the presentation probably did not occur very much before that date, since it is unlikely that Henry could have remained unaware of DGB for long, especially if it had entered into circulation in England.

Examining Robert of Gloucester’s itinerary in this light makes it possible to define more closely the possible times and places for the original presentation of the work. Robert came from Normandy to England for King Stephen’s Easter court of 1136 and returned to Normandy on Easter Sunday, 1137. His next visit to England was in September 1139.Footnote ¹⁴¹ This means that Robert was presented with DGB either in England between the Easters of 1136 and 1137 or in Normandy between Easter 1137 and the end of 1138 (to allow time for the making of Robert of Torigni’s copy, finished by January 1139). Either way, the earliest evidence for its dissemination comes from the Continent and the textual history also indicates that it was there that the work first became generally available, via the presentation copy. If Geoffrey had given the book to Robert in England, it would probably have been only shortly before Easter 1137 and his departure.

Robert’s role both in initiating the circulation of DGB on the Continent and, later, in contributing to its availability in Britain turns out to have been substantial. Much may be uncovered by fitting together evidence from the textual history and the manuscripts with available information about his movements and activities. The starting point is the observation that all early continental manuscripts descend from a single (now lost) exemplar, called phi by Michael Reeve, editor of DGB.Footnote ¹⁴² Phi had the earliest of the three authorially released versions of the text that have been identified and it bore the dedication to Robert. Manuscripts descending from phi display a more structured division into books than members of other textual families and have more systematic rubrication. In other words, the layout of their exemplar was carefully executed. On this account, Reeve has suggested that phi was the presentation copy, as such a copy would have been likely to have received complete chapter markings, rubrics, and initials.Footnote ¹⁴³ The fact that all the early continental copies of DGB descend from phi, without any interference from the other versions (whose influence is seen in the insular transmission from its beginning), and that they more faithfully than the early insular copies echo its division into eleven books, supports the idea that this dissemination originated with a single book – the presentation copy that Robert of Gloucester received from Geoffrey, either immediately before his arrival in Normandy (Easter 1137) or else once in Normandy (probably Caen), 1137–8.Footnote ¹⁴⁴

Certainly, the continental copying of DGB was initiated with great speed. Robert of Torigni had a copy, apparently a descendant of phi, in his possession by January 1139.Footnote ¹⁴⁵ Furthermore, all continental descendants of phi descend through a curious bottleneck that came about during the very first steps of the continental transmission, probably because of accelerated copying. For the first half of the text (§§1–108), all early continental copies directly follow Paris, Bibliothèque Sainte-Geneviève, MS 2113 (henceforth, G). We know this for certain since they share unique transpositions that occurred during the copying of G. However, for the second half, they all (except one) descend directly from the exemplar of G.Footnote ¹⁴⁶ This textual situation can only be explained by the existence of a manuscript copied partly from G and partly from its exemplar (which did not, of course, have the transpositions present in G), a copy we may call post-G, which then came to father the rest of the tradition. This break in the structure of post-G coincides with a section break in G itself. The first section, copied rather carefully by a single scribe, ends on the last folio of the fourth quire, six lines from the bottom of its verso. On this line, another scribe took over, continuing on the next quire and copying the rest of the work. The logical explanation for the structure of post-G would seem to be this: G was being copied from a copy of phi and, once the quires constituting the first half of G were finished, these quires were put to use as an exemplar for post-G, while the copying of the rest of the G was still going on using the original exemplar. Once §§1–108 of post-G had been copied from the first quires of G, G itself was completed and its exemplar became available and was used for the rest of post-G. This would have been an obvious choice, particularly since the second half of G is the result of visibly hurried scribal work, with many of the book divisions and initials indicating chapter division being omitted or sloppily executed.

The very first stages of copying appear to have been frantic. Not only were G and post-G produced with the emphasis on speed of production but they had already been preceded by a copy of phi, a lost intermediary which served as their immediate exemplar.Footnote ¹⁴⁷ Robert of Gloucester’s involvement in this early continental copying is suggested by an important piece of textual evidence. Very soon after G and post-G had been copied, the continental transmission bifurcated. Two-thirds of the continental copies datable to the twelfth century descend directly from post-G but one-third (12 out of c.35), while in other respects completely similar, omit the name of the dedicatee in the dedicatory passage.Footnote ¹⁴⁸ Neil Wright and Mary Garrison have suggested that the omission came about in a manuscript in which the place for the name of the dedicatee was left blank for later completion in another colour.Footnote ¹⁴⁹ This was sometimes done: Robert of Gloucester’s half-sister, Empress Matilda, owned a copy of Ekkehard of Aura’s chronicle in which her name was inscribed in gold letters, obviously done at a stage in the manufacturing process later than the transcription of the text.Footnote ¹⁵⁰ While most ‘nameless-dedication’ copies do not directly point out any omission in the exemplar but rather try to amend the situation, one of them does draw attention to an incomplete exemplar. This is Paris, Bibliothèque nationale de France, MS lat. 12943, a manuscript whose making can be localised to Paris and dated to the late 1160s.Footnote ¹⁵¹ In this book, no attempt has been taken to smooth out the sentence made unintelligible by the missing words of the dedication. Instead, where the dedication ought to be, we read only ‘deest’ (‘is missing’). In other words, the copyist noted the absence of the dedication in the exemplar, probably because there was an empty space.Footnote ¹⁵²

If, as seems likely, the nameless dedication went back to an unfinished luxury manuscript (the nameless master copy), its most probable destinataire would have been Robert of Gloucester himself. The changing of the single line which is missing would not have offered sufficient means to rededicate the work to another, since the dedicatory passage is much longer and contains, for example, a reference in another sentence to Robert’s status as son of Henry I. It is furthermore very likely that the nameless master copy was one of the very first copies made on the Continent, since it is difficult to explain how a manuscript with such a glaring shortcoming at the very beginning of the text would come to have been used so extensively as an exemplar if the text had already been widely available. The probable use of the nameless master copy as an exemplar for a book made in Paris in the late 1160s hints at its presence there at that time, an interpretation supported by what we know about the medieval origins and provenances of other members of the nameless-dedication family. These manuscripts, in as much as they can be localised, hail from the regions of Paris, Burgundy, Champagne, and Anjou, whereas the direct descendants of post-G cluster in Normandy, Flanders, and Picardy. Could it have been that the nameless master copy, once its text had been transcribed from post-G in Normandy, travelled south, perhaps to Paris, to receive fashionable illumination at the hands of the best artists of the time? On the Continent, the availability of bookmaking expertise was at any rate greater than in Geoffrey’s Oxford ambit, where the original presentation copy had probably been produced.

According to Gaimar’s Estoire des Engleis, Robert loaned a copy of DGB to Walter Espec, a Yorkshire magnate. This took place soon after his return to England in September 1139 (early 1140s at the latest).Footnote ¹⁵³ Textual history and codicological evidence support this testimony and indicate that the book loan had a real effect on the distribution. Essentially, the influence of Robert’s presentation copy is visible as an anomaly in the work’s textual landscape in Britain. The great majority of the early British copies of DGB belong to a different branch of transmission than do the early continental phi manuscripts. They descend from so-called delta, which was apparently produced from Geoffrey’s draft (of which phi was an earlier fair copy) after a series of authorial revisions. However, the predominantly continental phi tradition has a group of relatives in England, some of which are particularly important witnesses to the phi text and were used by Reeve in its editorial reconstruction.

The first group clusters around the twelfth-century manuscript Cambridge, Gonville & Caius College, MS 406/627, called Y by the editor. Y was kept, by c.1300, at Bridlington priory (OSA, Yorks ER), which belonged to a Yorkshire network of Augustinian houses in which Walter Espec took an interest.Footnote ¹⁵⁴ Its twelfth-century marginalia suggests that the manuscript was already in Yorkshire by then.Footnote ¹⁵⁵ The manuscript, furthermore, has unique and uniquely representational decoration. The initial B on the first page of the book (beginning the second chapter of DGB) contains two portraits, one in each lobe of the letter. The upper one is a figure with crown-like headgear. The one below is a grumpy-looking clerk. The images are not of the highest artistic standard and their iconography is by no means clear. In fact, the figure above, with flowing hair, might well be female. Yet one wonders whether these images, found directly below the dedicatory passage, in fact depict the author and the patron and derive from the decoration of the exemplar of Y.

Altogether nine British manuscripts are related to Y, six of these datable to the twelfth century.Footnote ¹⁵⁶ Most of these have no early provenance but the manuscript textually closest to Y, Oxford, Bodleian Library, MS Fairfax 28, was in Nun Appleton, just outside of York, in the seventeenth century, as one part of the small and locally sourced Fairfax family collection.Footnote ¹⁵⁷ This manuscript has an opening rubric reading ‘Domino Rodberto Comiti’. A similar address to Robert, similar in diplomatic form to epistolary address clauses, is included in the rubrics of some of the continental descendants of phi as well, including G and post-G. It is not found in other branches of the textual tradition and the logical assumption is that it reflects the rubrication of the presentation copy.Footnote ¹⁵⁸ Moreover, a member of the Y group was used by Alfred of Beverley in Yorkshire, c.1143, as he created his epitome of DGB.Footnote ¹⁵⁹ We know that Walter’s copy of DGB was kept at Helmsley Castle, twenty-four miles north of York, and it appears highly likely that these manuscripts descend from it. Apparently, the introduction of the text to Yorkshire by Robert’s and Walter’s cooperation provided a significant boost to its distribution.

We have already encountered the other important insular witness to the mainly continental phi tradition: London, British Library, MS Royal 13 D. II, which was, by the thirteenth century (at the latest), at Margam Abbey (OCist, Glam), one of Robert of Gloucester’s few monastic foundations. This manuscript contains excellent textual witnesses of William of Malmesbury’s GRA, the ‘Robertian recension’ of his Historia novella, and DGB.Footnote ¹⁶⁰ According to Reeve, the Margam manuscript is textually the best representative of the phi family of DGB – and also, arguably, the best witness to the text of DGB overall. In Michael Reeve’s words, ‘a transcript of M [= Margam Abbey MS] would be a tolerable substitute for an edition’.Footnote ¹⁶¹ Given the ensemble of texts and their quality and the early home of this manuscript, it is difficult to escape the conclusion that the exemplar used in the copying of DGB was the original presentation copy given to Robert in 1137 or 1138.

By all accounts, Robert of Gloucester was the key figure in Geoffrey’s publishing circle. He certainly contributed significantly to the rapid dissemination of the text which bore his name as dedicatee and in fact his role may have been even more far-reaching. DGB is also found with two double dedications, both objects of much previous debate. The first, and by far the more common in the manuscripts, dedicates the work jointly to Robert and Waleran, count of Meulan. Robert and Waleran were not friends or allies. During the reign of Henry I, Waleran, who had been brought up at the royal court, had plotted against the king, Robert’s father and protector, and been imprisoned for five years. Stephen’s seizure of the crown opened new avenues for Waleran’s ambition. He quickly became the new king’s trusted counsellor and had great power vested in his hands. This put him in direct conflict with Robert of Gloucester, whose commitment to Stephen was never anything but superficial and who eventually challenged Stephen’s right to the crown, backing his sister Matilda’s claim.

It used to be argued that the double dedication to Robert and Waleran was the original one and that a book (or, more likely, books) bearing it had been presented to the magnates at King Stephen’s Easter Court of 1136. This presentation, so it was thought, carried a plea to the two rivals to maintain peace. Such a chronology, once obstinately championed by the discoverer of the double dedication, Acton Griscom, and echoed by some relatively recent scholarship as well,Footnote ¹⁶² has been made untenable by the recent uncovering of the textual history of the work.Footnote ¹⁶³ It is now clear that the double dedication post-dated the dedication to Robert alone and this makes its suggested early date unlikely in the extreme. Publication of DGB before Robert and Waleran’s falling-out would require that two versions – one dedicated to Robert alone and another dedicated to Robert and Waleran – were put into circulation in England in 1136 and such widely publicised early availability is impossible to reconcile with the fact that Henry of Huntingdon was unaware of the work until January 1139.Footnote ¹⁶⁴

A much more plausible date for the creation of the double dedication is provided by the circumstances in which Robert and Waleran made their peace. King Stephen’s defeat and capture at the Battle of Lincoln, February 1141, made his position utterly hopeless. In the following summer, Waleran abandoned Stephen’s cause and was reconciled with Robert and Matilda after two years of open and destructive warfare. At this point, the two counts met, witnessing charters by which Bordesley Abbey passed from Waleran to Matilda, as part of the peace-making process.Footnote ¹⁶⁵ The Bordesley charters were written at Devizes but, throughout the summer of 1141, Matilda’s court lodged at Oxford, Geoffrey of Monmouth’s probable place of residence. I suggest that this was the context for the making of the first joint dedication and that it was produced with Robert’s participation.

Such a possibility had actually been considered by Acton Griscom and his rejection of the idea buried this obvious solution to the dedicatory puzzle for decades.Footnote ¹⁶⁶ Griscom’s thinking was, however, conditioned by the incorrect assumption that the joint dedication would have been the first and original one. He also made questionable interpretations about its making and purpose. First, Griscom supposed that the making of such a double dedication, placing Robert and Waleran on an equal footing, would have been Geoffrey’s decision alone. But would Geoffrey not have needed to take into account Robert’s reaction to the joint presentation of the work to a new patron with whom Robert had a particularly fraught relationship? Second, Griscom assumed that the wording of the double dedication would have necessitated a thorough and sincere reconciliation between the two magnates, primarily because it addressed Waleran as ‘the other pillar of our kingdom’ (altera regni nostri columna).Footnote ¹⁶⁷ But does a courtesy of this sort warrant conclusions about what Robert and Waleran really thought of one another? If we suspect that Robert may not have felt very warmly towards Waleran, who had ravaged his lands and with whom he shared a long history of mistrust extending back into the reign of Henry I, the text of the double dedication leaves plenty of room for ironic interpretations as well. The dedication lauds Waleran’s courage: ‘you surpassed your comrades in boldness, and you learned, like your father before you, to become a terror to your enemies and to be a protector of your own.’Footnote ¹⁶⁸ But they probably both remembered that, in the Battle of Lincoln, Robert’s triumph, which had brought the pair to the negotiation table, Waleran had conspicuously fled the battlefield on horseback while his lord, King Stephen, whom he had vowed to protect and who had engaged in the battle on foot, kept on fighting until captured.Footnote ¹⁶⁹

However we interpret its message, the dedicatory text makes it clear that a copy of the work was given to Waleran, since Geoffrey asked him to take under his protection both himself and ‘the book given out for your delight’ (codicemque ad oblectamentum tui editum).Footnote ¹⁷⁰ Whether Waleran contributed much to the further circulation of the work is another matter. He travelled to Normandy soon after the summer of 1141 but, on the Continent, the double dedication is found only in two manuscripts. Both these manuscripts, furthermore, represent textually the continental mainstream of the transmission – the text descending from phi presented to Robert alone – with the dedication to Waleran simply added from another textual source.

In England, at the same time, a total of eight copies survive bearing the joint dedication from the twelfth century – not an insignificant number.Footnote ¹⁷¹ All of these copies descend from a single manuscript, Paris BnF, MS lat. 6040, of English origin. We can be sure of their dependence on this particular book since one of its quires was copied from a different textual source than the rest, by a different scribe. This combination of textual traditions, resulting from contemporary repair of physical damage that took place as the manuscript was being bound, is repeated in all the manuscripts with the double dedication (except the continental manuscripts which conflate it with a phi text).Footnote ¹⁷² The main source of Paris 6040 was delta, the authorial version post-dating phi from which most of the early British copies descend. However, the odd quire follows phi, i.e. the presentation copy given to Robert. Intriguingly, the phi part of Paris 6040 descends from phi through a similar intermediary copy as the Yorkshire group, the probable result of Robert’s book loan to Walter Espec. Paris 6040 is an important textual witness to both phi and delta and was used by Reeve in establishing the text of DGB for both these textual families. It is thus textually close to manuscripts that we can associate with Geoffrey and with Robert and, in all likelihood, it stems from the context in which the presentation copy to Waleran was originally produced, although its poor quality makes it unlikely to have been the presentation copy itself. Whereas Waleran’s role in the distribution must have been limited, the involvement of Robert in the creation of the joint dedication would fit logically with his attested active role in the circulation of DGB.

Circumstantial evidence, furthermore, points to the possibility that the other joint dedication, to Robert and to King Stephen, was also launched with Robert’s (or Matilda’s) participation. This famous dedication is found in a single manuscript, Bern, Burgerbibliothek, MS 569, datable to the second half of the twelfth century. It was executed with minimal effort, by changing thirteen words in the earlier joint dedication so that King Stephen was put in Robert’s place as the first dedicatee and Robert was moved over to occupy Waleran’s position. Given the lazy nature of this reworking and the textual corruption of the manuscript in which it is found, it is not possible to account straightforwardly for the origin of this dedication and it has been suggested that someone other than Geoffrey may have written it.Footnote ¹⁷³ But it is difficult to imagine a person entirely disconnected from the author imposing such modifications to a copy so as to present it to the king. In what sense could this kind of faked dedication – for it is always Geoffrey who appears as the composer in the prologue – have pleased Stephen, especially since the text was already in circulation with a dedication to his bitter enemy?

The more likely alternative seems, again, joint agency on the part of the author and the original dedicatee(s). A probable context for the giving of the book to Stephen emerges from the curious circumstances of September to October 1141. Robert had by then also been taken prisoner and the negotiations for exchanging him with Stephen (who remained in Matilda’s custody) were taking place. We know that the text was fashionable at this time. Writing very soon after, in 1143, Alfred of Beverley stated that ‘it was a sign of rusticity’ not to be familiar with the stories told in DGB and, indeed, that he had often blushed in conversations because of his ignorance of the work.Footnote ¹⁷⁴ Robert and Matilda could well have played for diplomatic effect on Robert’s role as the primary dedicatee of such an uncommonly sought-after work. It would not have hurt them, either, that the text also argued the case for the viability of female rulership in showing several capable historical queens who had governed Britain or that it portrayed Robert’s crucial allies, the Welsh, in a more positive light than was typical for the Anglo-Norman literary imagination.Footnote ¹⁷⁵

As with the copy presented to Waleran, the book given to King Stephen had very little impact on the further dissemination of DGB. The only manuscript to contain the dedication was copied in Fécamp, probably in the last quarter of the twelfth century.Footnote ¹⁷⁶ In this manuscript, DGB is followed by Aelred of Rievaulx’s Vita Sancti Edwardi, composed sometime between 1161 and 1163 and dedicated to Henry II. The exemplar used for this text was of English origin. In the absence of earlier continental witnesses, one suspects that the same was true of the exemplar that the Fécamp scribes used for DGB.Footnote ¹⁷⁷ Given Robert’s active role in the distribution, and the probably widely known association of the text with him, it is not surprising that we see much less trace of the secondary dedicatees promoting the dissemination of DGB.

While Robert was undoubtedly a key person within Geoffrey’s publishing circle, he was not the sole distribution channel for the text. Evidently, institutional collections were important for the transmission of DGB as well. We know that Robert of Torigni had a copy by January 1139 and that, c.1150, he arranged for the production of a compilation containing it for the library of the abbey of Le Bec.Footnote ¹⁷⁸ Over the course of the twelfth century, copies of DGB stemming from the Le Bec book or its exemplar (Robert of Torigni’s copy?) were produced at least for the libraries of Anchin (Flanders),Footnote ¹⁷⁹ Marchiennes (Flanders),Footnote ¹⁸⁰ Chaalis (Picardy),Footnote ¹⁸¹ and Jumièges (Normandy).Footnote ¹⁸² In Britain, there is less firm evidence for the ownership of copies of DGB by institutional libraries in the twelfth century. However, a list of books copied for Glastonbury Abbey at the instigation of Henry of Blois (abbot, 1126–71) includes Geoffrey’s work, showing that, at an early date, DGB had entered the Glastonbury collection, which had been a significant factor in the success of the GRA.Footnote ¹⁸³ While these are clear instances of transmission relying on monastic networks, both the Le Bec and the Glastonbury cases also make it evident that the copying happened at the initiative of an individual bibliophile. It is highly likely that other similarly scholarly monks, invisible to us now, were of equal significance in the general monastic transmission.

In Britain, besides aristocratic and monastic circles, the text was also available directly from the producer. As has been said, not all manuscripts descended from the presentation copy given to Robert of Gloucester. Geoffrey kept on revising the text and produced the so-called sigma and delta versions of the text (sigma covering only §§118–208). Delta was probably the main source for the book made for Waleran but this version additionally circulated with the dedication to Robert. Four of the delta manuscripts, all British ones, are considered by Reeve to have independent textual authority and thus descend from Geoffrey’s delta autograph by independent routes.Footnote ¹⁸⁴ All of these manuscripts are early ones, datable to around the middle of the twelfth century or soon after, and they suggest frequent early copying of Geoffrey’s own working manuscript.Footnote ¹⁸⁵

One of these delta manuscripts, furthermore, conveys distinguishable signs of further authorial revision. This is Oxford, Bodleian Library, MS Rawlinson C. 152, from the middle of the twelfth century, having a peculiar tall and narrow format (300 × 165 mm).Footnote ¹⁸⁶ In this manuscript, the prologue to the PM is idiosyncratic. Whereas manuscripts regularly include the full dedicatory letter to Alexander of Lincoln, in this copy the letter is replaced by a newly formulated, apparently autobiographical passage to the same effect, in which Geoffrey refers to himself as pudibundus brito (‘bashful Briton’).Footnote ¹⁸⁷ Furthermore, the text of DGB has a unique and curious opening rubric, ‘Gaufridi Monemutensis de gestis Britonum secundum Caratonum Editio’.Footnote ¹⁸⁸ It has been plausibly argued that the work of Caratonus, given here as Geoffrey’s source, should be equated with the book that Geoffrey mentions as his source in the prologue and the epilogue – the very old book in the Welsh language, which he had from Walter, archdeacon of Oxford.Footnote ¹⁸⁹

Besides the textual history, one contemporary reference hints at the distribution of exemplars directly from Oxford, something which can perhaps be associated with the version of the text presented by Rawlinson C. 152. In his epilogue to the Estoire des Engleis, Gaimar mentions not only the book that Walter Espec was given by Robert of Gloucester but also another book of Welsh history, which he had ‘purchasé – u fust a dreit u fust a tort – le bon livere de Oxeford, ki fust Walter l’arcediaen’. This book, Gaimar writes, he used as a supplementary source as he was himself penning a copy of the book he had received from Walter.Footnote ¹⁹⁰ It seems highly unlikely that Gaimar would have had access to Geoffrey’s purported Welsh source, which very likely did not exist and which Gaimar even more likely would not have been able to read had it existed at all. This reference in effect suggests that Gaimar had in fact two copies of the De gestis at his disposal, one that he had procured from Oxford and another that came via Walter Espec. The Oxford copy may have been related to Rawlinson C. 152, with its rubric referring to Caratonus, i.e. the purported source. The origin of C. 152 is unclear and it has no early surviving descendants. However, while two thirteenth-century copies descending from it were made at St Albans,Footnote ¹⁹¹ the most numerous of its late medieval descendants have associations with the north of England, i.e. in the landscape familiar to Gaimar and Walter Espec.Footnote ¹⁹²

One cannot but conclude that Geoffrey played the game of publishing deliberately and successfully. He succeeded in creating an effective publishing circle, making use of aristocratic, monastic, and learned networks of the time. The narrative sources combined with information from textual history and manuscripts suggest that individuals with high status and resources, like Robert of Gloucester and Henry of Blois, or with special influence over the transmission of texts, like Robert of Torigni, advertised the work, commissioned new copies, and provided exemplars for others seeking access to the text. Very probably, Robert was also involved in arranging new publishing events, such as presentations of tinkered copies to Waleran and King Stephen. That Geoffrey’s work ultimately became so much more popular than Henry of Huntingdon’s and William of Malmesbury’s must have owed a great deal to its novel and attractive content; but one cannot fail to notice that his publishing circle also left us with many more direct and indirect traces of its activity. Geoffrey’s success was won over within his own lifetime, as is evinced both by the large numbers of surviving early copies and by contemporary accounts. In his Vita Merlini (c.1148) Geoffrey stated that DGB was celebrata per orbem,Footnote ¹⁹³ and, despite the self-congratulatory tone, he provides here, for once, an accurate historical testimony.

5 Conclusions

The processes by which William of Malmesbury, Henry of Huntingdon, and Geoffrey of Monmouth made their works available to their audiences unite the two main components of our definition of publishing – the releasing of intellectual content and the making of its materiality publicly available. For these authors, publishing involved releasing, and sometimes rereleasing, texts, as well as building a circle of patrons and bibliophiles to advance their distribution. This should obviously not lead us to think that successful publishing in a manuscript context would always need to be in equal parts authorial and promotional. In reaching an audience, many twelfth-century authors, such as Hugh and Richard of the abbey of Saint-Victor in Paris or Alexander Nequam, prior of Cirencester Abbey, continued to depend on a community that would promote their work and help in producing copies of it. It is also true that achieving bestseller status depended in significant measure on factors inherent to the literary product. In our case, the fact that William, Henry, and Geoffrey are still among the most widely read of medieval historians is indeed a lasting testimony to their skill as writers. Furthermore, William and Henry wrote the first real histories of England since Bede, and Geoffrey’s history, largely of his own invention, covered an even greater patch of uncharted territory. There was, in other words, a demand for the content they produced.

Nevertheless, the speed at which all these works managed to enter wide circulation, taken together with the quantity of publishing activity that can be documented across the authors’ lifetimes, would indicate that the role of authorial effort in these cases was significant. To contextualise the matter of publishing and the success of our three authors further, it is useful to take a brief look at other writers active in the same genre and in the same period. Such a comparative exercise reminds us, first, that many historical works were published in a very small way, with hardly any authorial aspirations concerning distribution, and that many such works never entered wide circulation. A famous contemporary example is the Historia ecclesiastica of Orderic Vitalis (1075–c.1152). Orderic’s work was accessible in his own monastic library of St Evroult to those who had an interest, as is demonstrated by several known extracts from it. However, the only existing manuscript of the work as such is Orderic’s own autograph.Footnote ¹⁹⁴ To judge by the massive volume of his writing and its largely very local content, Orderic never seriously sought wider literary fame, his work being principally intended for himself and his family, that is, the present and future brethren of St Evroult. Gervase of Canterbury (b. c.1145, d. in or after 1210) formulated this sentiment, probably widely shared by monastic writers of historical texts, when he stated that his work was not intended for ‘the public library’ (biblioteca publica) but only for his own community.Footnote ¹⁹⁵ The intention to contribute to the biblioteca publica, that is, to make one’s work generally available in the public sphere, certainly set William, Henry, and Geoffrey apart from Orderic, Gervase, and many other monastic historians.

The second important observation that a comparative examination yields is that the social and cultural context in which the publishing took place – the publishing framework – had great significance in determining what authorial aspiration could hope to achieve. Between the late tenth and early twelfth centuries, many German writers, for instance, composed ambitious secular histories, comparable in style, substance, and, probably, their intended audience to the works of Henry, William, and Geoffrey. These engaging narratives, such as Widukind of Corvey’s Res gestae Saxonicae, Thietmar of Merseburg’s Chronicon, Wipo’s Gesta Chuonradi, and the anonymous Vita Heinrici IV, have been much studied by modern scholars.Footnote ¹⁹⁶ The manuscript record of their early circulation is, however, singularly unimpressive. Of Res gestae Saxonicae (967–73), there is one copy predating 1100, one from the twelfth century, and three later ones.Footnote ¹⁹⁷ Thietmar of Merseburg’s Chronicon (1018), a less polished work in literary terms, survives in one manuscript predating 1100 and in one other, later copy.Footnote ¹⁹⁸ The anonymous Vita Heinrici IV similarly exists in one copy from the early twelfth century,Footnote ¹⁹⁹ while the earliest manuscript of Wipo’s Gesta Chuonradi (1040–6) dates only from the end of the sixteenth century.Footnote ²⁰⁰ While the German case is probably the most extreme one, a similar picture emerges if we look at the distribution of newly composed historical works in France and Normandy in the era preceding that of our three authors. Rodulf Glaber’s (c.980–c.1046) Historiarum libri quinque survives in one eleventh-century manuscript, which is in part autograph.Footnote ²⁰¹ Aimoin of Fleury’s Historia Francorum (c.998) exists in five copies, of which one is dated to the turn of the tenth century, one to the twelfth, and the remainder to later times.Footnote ²⁰² Adémar of Chabannes’s (989–1034) Chronicon has a slightly more impressive record of thirteen manuscripts, with three from the eleventh century and six from the twelfth.Footnote ²⁰³ In Normandy, the fact that there is one surviving medieval manuscript of Carmen de Hastingae praelio,Footnote ²⁰⁴ and none at all of Gesta Guillelmi, suggests strikingly limited circulation for those histories occasioned by the Norman Conquest.Footnote ²⁰⁵

At the turn of the eleventh century, the circulation of historical texts appears, however, to have gained momentum. This can be seen in the intensified circulation of earlier works, such as Dudo of Saint-Quentin’s Historia Normannorum (996–1015), which survives in only two eleventh-century manuscripts but in ten from the twelfth century or the beginning of the thirteenth.Footnote ²⁰⁶ Dudo’s continuation, William of Jumièges’s Gesta Normannorum Ducum (c.1050–66), similarly exists in one eleventh-century but twenty twelfth-century copies.Footnote ²⁰⁷ Even more notably, the numbers of manuscripts of the newly published crusading histories of Guibert of Nogent (Gesta Dei per Francos, c.1112, thirty-four twelfth-century copies),Footnote ²⁰⁸ Baldric of Bourgueil (Historia Hierosolimitana, c.1107–8, at least eleven twelfth-century copies),Footnote ²⁰⁹ and Robert the Monk (Historia Iherosolimitana, after 1107, seven twelfth-century copies)Footnote ²¹⁰ indicates that they found an audience more quickly and more effectively than had been the case for historical writing in the previous era. While the reliability of the numbers of manuscripts as indicators of popularity can of course be challenged, the oddities of survival should not in this case distort the overall picture, as it is highly unlikely that the survival rate of eleventh-century historical manuscripts would have been drastically worse than that of manuscripts of the same genre from the twelfth century.Footnote ²¹¹

This admittedly superficial examination of the levels of success of various historical works suggests that, around the year 1100, the patterns of textual transmission and book production were going through a transition. It is not difficult to connect this change with well-known developments in society: rapid economic growth, increasing demand for specialists in the written word, and an explosion in the availability of education, captured by Guibert of Nogent’s famous remark as to how it had been next to impossible to find a decent grammarian in his youth (1060s), whereas wandering clerks skilled in Latin were now (c.1115) widely available.Footnote ²¹² I propose that it was this development of the infrastructure of written culture, the publishing framework, that opened the door for effective authorial publishing.

The method that William, Henry, and Geoffrey used to distribute their texts in this increasingly dynamic literary environment was to involve several people, who held nodal positions in this framework, in the process of publishing – that is, to construct a publishing circle. By way of conclusion, it will be appropriate to recapitulate the key elements of these circles: their personnel and the role these persons had in advancing textual distribution.

Many of the key contacts were forged at the presentation of the work or immediately thereafter and among them the dedicatees form a distinct group. The evidence of their impact on the texts’ distribution is strong, even if rarely completely unambiguous. Some of the patrons we have encountered certainly loaned exemplars for others to read and copy and they must also have provided advertisement by word of mouth. Very probably they also commissioned new copies and participated in presenting copies to other notables. This is not to say that all patrons always did participate in the distribution. Empress Matilda does not seem to have advanced the circulation of the GRA; and, while Alexander of Lincoln quite possibly made the HA available at Le Bec, his role as a literary patron remains ambiguous. Moreover, as Geoffrey’s criticism of Alexander suggests, securing help in the distribution was not necessarily the chief aim of a dedication, as the authors also had other sorts of gains, such as career advancement, on their minds. Nevertheless, an interested patron could give a great boost to dissemination and it would appear natural to suppose that such a boost was one thing – if rarely the only thing – that authors wished for as they constructed the dedication to a literary work. Dedicatees were a very common, one might say near-integral, element of a successful publishing circle in this twelfth-century context.

However, especially when it came to making exemplars available, other kinds of collaborators, such as monastic librarians, were also important partners. Since so little narrative or documentary evidence discusses the day-to-day business of circulating texts, most of these medieval publishing agents will remain unknown to us by name. In our evidence, the one such person, encountered in several roles, is Robert of Torigni. It appears probable that he and the book collections he managed provided the main continental distribution platform for the early circulation of both DGB and the HA. For William of Malmesbury, the libraries of Glastonbury Abbey and Winchester Cathedral seem to have functioned in somewhat similar fashion. Success in getting a good exemplar into an active institutional library was paramount for the circulation of the works of the authors discussed here, just as it was for those authors whose own institutions provided the textual base. It is often difficult to say whether, and to what degree, the institutional librarians (or bibliophiles, if ‘librarian’ is too problematic a word) were intentionally recruited by the author and, as a consequence, how integral their role can be understood to have been to the authorial publishing circle. Nevertheless, authors must have been aware of the significance of institutional librarians and, given that they were seeking distribution, it would be very surprising if they had not taken measures to try to attain presence and visibility in the book collections of, say, Le Bec, or Glastonbury, to which they knew interested people would come for texts.

Finally, while this study has been about writers who can, in their twelfth-century context, be described as literary, it should be borne in mind that the actual constituents of a publishing circle would have been similar for most sorts of authors, whether or not they were members of religious institutions or whether or not they presented their works to dedicatees. The study of publishing in the manuscript context is still a new field and I propose that one of its main objectives should be the reconstruction of such publishing circles for individual works or groups of works.Footnote ²¹³ At the same time, the publishing circles discussed in this Element call for due caution to be exercised in seeking to categorise the publication of texts in manuscript context by (more or less) fixed typologies, such as publishing by patronage, publishing ‘officially’, or publishing ‘organically’. Such categorisations, when they are not closely based on empirical studies, risk imposing artificial interpretative categories on historical reality. They also have the tendency to obscure human agency, on which all these forms of textual circulation depended. Official sanction of a text by a religious order, or the author’s membership of a religious house with strong networks and its own lettered tradition, may have helped in creating an effective publishing circle. But the circle was always made up of a particular constellation of individuals. It was these individuals on whom a work’s entry into the contemporary literary canon depended.