The Alfredian Boethius Project
In October 2002 we formally launched an ambitious collaborative research project on the Alfredian Boethius, with Malcolm Godden (Oxford University) as director, Dr Susan Irvine (University College London) as co-director and Dr Rohini Jayatilaka (Oxford University) as project co-ordinator and research associate; we subsequently enlisted Dr Mark Griffith (New College, Oxford) as co-editor. The project was granted funding (£334,000 for a five-year period) by the Arts and Humanities Research Board of Great Britain. I am very grateful to the editor of the OEN for an opportunity to describe the project and invite help and advice from Anglo-Saxon scholars.
The main impetus behind the project was the desperate need for a new and comprehensive edition of this important work (or pair of works), to replace the 1899 edition by Sedgefield.  But we were also very aware of the important issues surrounding the text—the authorship of the two versions, the relation to ninth-century scholarship and culture, the relationship between prose and verse—and the amount of weighty scholarship which had developed in relation to these issues and needed to be brought to bear on them. The main objectives of the project are to:
1. Establish a sound text of each of the two versions of the Old English Boethius, and determine the relationship of the two and their textual history.
2. Research thoroughly the background and possible sources for all the Alfredian material in the Old English texts, and parallels and influences on the Old English author's interpretation of Boethian philosophy, in order to identify both the origins of his reading of Boethius and the ninth- and tenth-century context of Boethian interpretation and response within which he worked.
3. Examine closely the stylistic, linguistic and intellectual relationships between the Old English Boethius and other works of the period, especially the two most closely linked Alfredian texts, the Soliloquies and the Orosius, with a view to establishing a sounder base for views on authorship and on the cultural connections of these texts.
4. Make a thorough study of the language of the Old English Boethius.
The most important outcome of the project will be a new hard-copy edition of the Alfredian Boethius to be published by Oxford University Press. The plans have been approved by the Press and we have undertaken to deliver final copy in 2007. We are envisaging a volume of about 900 pages, containing full texts of each of the two versions (that is, the prose version represented by Oxford, Bodleian Library, MS Bodley 180 and the prosimetrical version represented by London, British Library, Cotton Otho A.vi), a Modern English translation, and a substantial introduction, commentary and glossary. We have also commissioned high-resolution digital images of the two Oxford manuscripts (the twelfth-century MS Bodley 180 and the seventeenth-century transcript by Francis Junius) which will complement the digital images of the burnt Cotton manuscript which are being produced by Kevin Kiernan of the University of Kentucky; ours will be made publicly available, probably in the Electronic Boethius edition which Kiernan plans to publish. The project will also host an annual symposium on issues related to the Alfredian Boethius, and aims to promote research publications, in web form or hard copy, arising out of its work. The topic for the first year was the Latin textual and commentary tradition. Papers from this symposium are appearing on the project website, and subsequent topics are likely to include the metres, textual and editorial strategies, politics, and classical myth and legend.
Editing the Old English Text
The urgent need for a new edition was emphasised by Paul Szarmach in an article published in 2002.  The problem is not just that the current edition is now more than a century old; it is more seriously that the text in that edition is a hybrid with a complicated history behind it, and even the most dedicated of readers finds it difficult to discover quite what its relation is to the manuscript witnesses. The textual complexities are unique. The two surviving medieval manu-scripts of the text represent two different versions, both of which may be authorial. Neither manuscript is satisfactory as a witness to the text. One was copied in the twelfth century, by a scribe who mostly did his best to reproduce the text handed down from two centuries earlier in a language now nearly obsolete, but sometimes struggled and often nodded. The other was copied in the tenth century but was very badly damaged in the Cotton fire and much is now illegible or questionable. In addition to those we have a third unsatisfactory witness, the handwritten version made by Francis Junius in the twelfth century, which provides a modified copy of the twelfth-century manuscript together with variants and additional text copied from the tenth-century manuscript.
A crucial part of our scheme is to give full weight to each of the two versions of the Old English Boethius, the all-prose version represented by Bodley 180 and the prose and verse version, or prosimetrical version, represented by Cotton Otho A.vi, without of course prejudging the question of authorship for either. Both seem to have had a history of medieval reading: it was the all prose version that Ælfric read and adapted in the tenth century,  but it seems to have been the prosimetrical version that Nicholas Trevet read and cited in the early fourteenth century.  But the modern textual practice, starting with Junius in the seventeenth century, has always been to conflate the two versions, and we have not really seen either plain. As I argued in 1994,  we have never had a proper edition of the Cotton version. As I work on the Bodley manuscript it becomes increasingly clear to me that we have never had a proper edition of the prose version either. Although Sedgefield's text reflected the structure of the prose version his main focus in his 1899 edition was the readings of the newly recovered Cotton manuscript, and a (generally) faithful record of that manuscript's readings is indeed the great virtue of his edition. But his claim to have given a record of all substantive variants from Bodley seems far from justified by the reality. Much of the time he seems to have been consulting not the Bodley MS itself but earlier printed editions, perhaps that of Samuel Fox,  which he no doubt used as the basis for his collation of Cotton, and also the transcript by Junius.
It is also becoming clear that Junius's 'transcript' (Oxford, Bodleian Library, MS Junius 12) needs to be used with great caution. Recent work by Kees Dekker and Daniel O'Donnell has emphasised that Junius's transcripts of Old English texts were designed as editions not as faithful records of the underlying manuscript,  and he frequently corrected or improved, or silently supplied readings from other manuscripts, according to his own sense of what the text ought to read. His sense was rather a good one, but that makes it all the more difficult to spot that a reading is Junius's and has no basis in an Anglo-Saxon manuscript. In the case of the Alfredian Boethius we have the benefit of being able to compare his version with the manuscript from which it was supposedly copied, MS Bodley 180, and it is easy to discover how frequently his 'transcript' incorporates his own normalisations and improvements. It was evidently not a direct copy of the manuscript but a second or subsequent version embodying a great deal of change. The great problem is that MS Junius 12 also contains his collations of the Cotton manuscript and transcriptions of the verse metres, and is often the only record of the original Cotton readings, before the manuscript was burnt. His texts of the metres do not look like direct copies from the manuscript and presumably incorporate similar editorial improvements and corrections to those evident in his text of the Bodley manuscript. Moreover, his method of reporting Cotton variants in the prose sections is rather variable: sometimes he gives the Cotton reading in the margin of his text of Bodley with the letter 'C' beside it; sometimes he alters his main text to match the Cotton reading, usually by interlinear additions or substitutions, without indicating whether the alteration is derived from Cotton, from Bodley or from his own conjecture; and sometimes he ignores the reading altogether. The differences of method presumably reflect his own view as to whether the Cotton reading was an interesting and possibly valid alternative, a correct reading where Bodley was in error, or itself erroneous or inferior. Subsequent editions in turn treated these different processes differently; while the readings which he placed in the margin and labelled have been recognised as Cotton readings and generally reported as such, his in-text corrections have generally been treated as if they were Bodley readings, and are so treated generally by Sedgefield.
Let me give three brief examples of the issues, one of the failure to reflect the Cotton text, one of the misreporting of the Bodley text and one of the misreporting of both manuscripts.
1. At 147.26-148.3 Sedgefield's text reads as follows:
Wast þu þæt þreo ðing sindon on þis middangearde? An is hwilendlic, ðæt hæfð ægðer ge fruman ge ende; 7 (ic) nat ðeah nanwuht ðæs ðe hwilendlic is, nauðer ne his fruman ne his ende. Oðer ðing is ece, þæt hæfð fruman 7 næfð nænne ende; 7 (ic) wat hwonne hit onginð, 7 wat þæt hit næfre ne geendað; þæt sint englas 7 monna saula. Þridde ðing is ece buton ende 7 buton anginne; þæt is God.
Dost thou know that there are three things on this earth? The first lasts for a time only, and has both beginning and end; yet I know nothing of that which lasts for a time, neither its beginning nor its end. The second thing is eternal, and has beginning but no end; of this I know when it begins, and I know that it never ends; such are angels and men's souls. The third thing is eternal, without end and without beginning, even God. 
Wisdom (the speaker) is saying that he as Wisdom has no knowledge of things beneath the level of man. That has a sort of plausibility, but seems at odds with Wisdom's general familiarity with, and interest in, the lower orders of existence (e.g. the proof that all things seek their continued existence).
If we happen to notice that the pronoun 'ic' in lines 147.28 and 148.1 is in round brackets and italics in Sedgefield, we find that there is no explanatory note, but if we track back to his introduction to find what that signifies the answer should be that the words are not in the Cotton manuscript, and never were, but are supplied from either the Bodley MS or from Junius's transcript. If we now look at the Bodley manuscript, we find that the word 'ic' is indeed present in both cases. If we look at the transcript, we find that Junius records 'ic' in each case from Bodley but says nothing about the omission in Cotton, although he normally does cite such differences. But Kevin Kiernan has confirmed for me that the text of Cotton is clear at this point and the word 'ic' is clearly absent in both lines; and Fox's edition of 1864 does in fact say this. Junius presumably disregarded the Cotton reading because he thought it was a scribal error, and subsequent editors have all followed his lead. But the Cotton reading clearly makes perfectly possible sense; the subject of knowing is not wisdom himself but creation. The meaning would be:
Do you know that there are three things in this world? One is transient, and that has both beginning and end; and yet nothing that is transient knows either its beginning or its end. The second thing is everlasting, and that has beginning and has no end; and it knows when it begins and knows that it will never end; that is angels and the souls of men. The third thing is eternal without end and without beginning; that is God.
The implication is that animals and plants and stones have no knowledge of their beginning or their end, whereas angels and human souls know when they begin and know that they will never end. There is a somewhat similar formulation much earlier in the Old English Boethius: 'þam neatum is gecynde þæt hi nyton hwæt hie send; ac þæt is þara monna unðeaw þæt hi niton hwæt hie sen'.  And on men and angels, one might compare Alfred's Soliloquies:
Þonne wene ic þæt hyt wille þe andweardan, gif hyt gesceadwis is, and cwæðan þæt hit forði wilnige þæt to witanne þæt ær us wæs, forði hit simle wære syððan god þone forman man gesceape hafde; ... and ic wene æac þæt hyt wille cweðan þæð hyt forði wilnige to witanne þætte æfter urum dagum geweorðan sceal, forðam hyt wat þæt hyt a beon sceal. 
The difference between the two manuscripts here is probably not a matter of copying errors: there is no reason for the Cotton scribe to omit ic accidentally twice in succession. That ic is the original reading and that it was deliberately deleted in Cotton to change the subject from Wisdom to animals, souls and angels seems to me possible but unlikely, because it seems such an intellectually profound change to make (that is, if the original author did it, it indicates a radical change of mind about the argument he is making, and if another reviser did it, it indicates a radical challenge to the argument). More likely is that Cotton has the original reading and that a reader in the tradition drawn on by Bodley read the text too carelessly, thought the subject was missing or unexpressed (because in the first sentence the subject in fact comes after the verb and can be mistaken for an object) and tried to tidy it up by adding ic.
Tantalisingly there is one other early testimony for this passage but its evidence is inconclusive. Ælfric rewrote this passage in his Lives of Saints 1 as follows:
Ðreo þing synd on middanearde . An is hwilwendlic . þe hæfð ægðer ge ordfrumman ge ende . þæt synd nytenu . and ealle sawul-lease þing þe ongunnan þa þa hi god gesceop . and æft geændiað and to nahte gewurðaþ. Oðer þing is ece . swa þæt hit hæfð ordfruman . and næfð nenne ende . þæt synd ænglas and manna saula . þe ongunnen ða þa hi god gesceop . ac hi ne geendiað næfre. Ðridde þing is ece . swa þæt hit næfð naðor ne ordfruman ne ende . þæt is se ana ælmihtiga god. 
But unfortunately Ælfric omitted Alfred's reference to knowledge altogether. One might hypothesise that he did so because he had a corrupt text like the Bodley version and the question of Wisdom's knowledge would have been irrelevant and confusing in the context (since the speaker in his version is not Wisdom but Ælfric himself). But although it is certain that the version which Ælfric knew was the all-prose version like Bodley, it does not follow that his copy contained all the corruptions and changes that appear in the twelfth-century manuscript. His copy may have had something that read like the Cotton text at this point and he simply may have been less interested in the question of the self-knowledge of animals, men and angels than Alfred was.
My second example is much slighter. In the Bodley version of chapter 29 Wisdom says caustically, in the course of his attack on kingship:
Hwæt wille we nu secgan be þam þegnum buton þæt þæt ðærof gebyreð . þæt hi weorðað bereafode ælcre are ge furðum þæs feores fram heora leofan cyninge . 
When first transcribing the Bodley copy I took this last phrase as a fine piece of sarcasm, only to see that Sedgefield, following the Cotton manuscript, printed not leofan but leasan 'false'.  The Cotton reading may be right, but it is a pity that this rather attractive Bodley reading is not even recorded in the apparatus.
My final example is a question of possible ghost words, but it illustrates well the perils of the Junius transcript. At 115.7, in a discussion of men who are like swine, Sedgefield prints: 'hy næfre nellað aspy[ligan on] hluttru[m] wætrum' (they will never wash themselves in clean water). His mixture of roman, italics and brackets serves to indicate that in the Cotton manuscript the letters aspy- can be deciphered and the -ligan is supplied from Bodley; his note says 'aspylian J', meaning apparently that Junius's transcript records the reading aspylian from Cotton. The word does not occur anywhere else in Old English but is duly recorded on the basis of these two readings in Bosworth-Toller and the new DOE, in the sense 'to wash'.  In fact however neither manuscript has this reading. The Bodley manuscript quite clearly reads aswyligan, with a wynn, and Susan Irvine tells me that wynn, rather than p, is equally clear in the Cotton manuscript. If we turn to the Junius transcript, the source of the problem, we find that he does indeed have aspyligan in his text and reports aspylian as the Cotton reading in the margin. But a harder look at the transcript suggests that in both cases the p is heavily inked over and rather different in shape from his usual p. It seems very likely that Junius originally recorded aswyligan/aswylian as the reading of both manuscripts but then decided this was an error and silently corrected it. Both the erroneous readings, with asp-, are recorded in Fox's edition and passed down to Sedgefield, and neither editor seems to have looked closely at the manuscript at this point. But aswylian is not only the reading of both manuscripts but almost certainly the correct form, matching the simplex swilian which is well attested. 
Our first aim then is to provide a reliable text of both versions. Going back to the manuscripts and not relying on the past editorial tradition will certainly help, but there is no denying that the task will be difficult. While we can all agree that the two versions ought to be treated as distinct works, it has to be acknowledged that both manuscripts do contain scribal errors and each needs to be corrected at times in the light of the other. There will also be places where Cotton can no longer be read and Junius cannot fully be trusted.
The Sources of the Old English Text
While some of our colleagues undoubtedly think that fussing about sources is a modern obsession closely allied to navel-gazing, in this particular case the obsession starts with Alfred himself. It was he who claimed that books had virtually vanished from England in his time and that there were no resources available for an adapter of Boethius except his own limited knowledge of Latin and a bit of help from his learned foreign friends. On the Continent, however, commenting on the Consolation of Philosophy had become a heavy industry, and it is hard to believe there was no connection. When Sedgefield published his edition in 1899 he was able to take account of the latest discovery about the composition of the text, Georg Schepss's demonstration in 1895 that "many illustrations and amplifications found in the Old English version … are undoubtedly derived from Old Latin commentaries" on Boethius.  Much has been written on the subject since then, but it is perhaps even more of a minefield now than it has ever been. Whether Alfred (or whoever actually composed the Alfredian Boethius) drew on such commentary, or on other kinds of sources, or on what he learned orally from the scholars gathered around him, or on his own general education, is still in dispute. To my eye, some of the parallels adduced by Otten remain very persuasive.  But the whole problem is bedeviled by the uncertainty as to what commentaries might have been available to the Alfredian circle. The two commentaries referred to by Sedgefield were subsequently identified as the work of Remigius of Auxerre, but Courcelle's redating of the Remigius commentary to the beginning of the tenth century seemed to remove that commentary from contention.  Courcelle suggested instead that Alfred had used a commentary prepared for him by Asser on the basis of the commentary ascribed to the Anonymous of St Gall, and that Remigius had in turn worked from the St Gall commentary supplemented by Asser's version and his own reading. But current thinking suggests that the St Gall commentary may also have been compiled in the tenth century,  and in any case nothing suggests that it was known in England. Meanwhile Fabio Troncarelli has argued that a third commentary, the one preserved in Rome, Vatican City, BAV lat. 3363, is a summary version of a commentary which Asser prepared for Alfred, and which in turn used the commentary hitherto ascribed to Remigius, which must accordingly be dated earlier in the ninth century.  The whole question of distinct and ascribable commentaries in the period may itself be untenable. There are some sixty manuscripts from the Anglo-Saxon period (that is, up to about 1100) containing some commentary on the Consolatio, and they vary enormously amongst themselves, often showing the contributions of different scribes drawing on different traditions. Recent work has begun to chip away at this mass. Diane Bolton demonstrated the ramifications of the Anglo-Saxon manuscripts of the so-called Remigian commentary in an article in 1977.  Joseph Wittig has identified several distinct strands and manuscript groups among Remigian manuscripts on the basis of detailed collations for III m. 9 and 12.  Fabio Troncarelli has himself demonstrated that the commentary in MS Vatican 3363 is the work of at least four hands, ranging from the Continent in the early ninth century through a Welsh commentator of the late ninth to St Dunstan and a collaborator at Glastonbury in the mid tenth century.  But it is clear that nothing very decisive can be said about the development and transmission of early Boethian commentary until full collations are made and analysed. Dr Jayatilaka has begun work on this for the current project, and although it is an enormous and demanding task we hope to be able to give some account of the overall pattern. Even if, in the end, we cannot demonstrate convincingly that Alfred had access to a commentary and used it, it is important to document how Boethius was being read and interpreted in Alfred's time. And having some idea of the sources from which he was working, to explicate references to ancient history and legend and philosophy and the natural world, will be an important tool for dealing with the still knotty question whether the same person, Alfred or another, wrote both the prose version and the metres, since it might be supposed that if the original translator subsequently turned the prose into verse he would retain some understanding of the material from which the prose version was made.
Those involved in the project are very conscious of the magnitude of the task. The project website, at http://www.english.ox.ac.uk/boethius/, includes a bibliography of relevant works on Boethius and the Old English versions, but we would be very grateful for notification of important items we have missed, or information on work in progress, or indeed any help and advice that readers of OEN can offer.
 King Alfred's Old English Version of Boethius 'De Consolatione Philosophiae', ed. W. J. Sedgefield (Oxford, 1899). Unless specified otherwise, all translations are my own.
 Paul E. Szarmach, "Editions of Alfred: the Wages of Un-influence," in Early medieval English texts and interpretations: studies presented to Donald G. Scragg, ed. Elaine Treharne and Susan Rosser (Tempe, AZ, 2002), 135–49.
 See M. R. Godden, "Anglo-Saxons on the Mind," in Learning and Literature in Anglo-Saxon England, ed. Michael Lapidge and Helmut Gneuss (Cambridge, 1985), 271–98, esp. 296–8, and "Editing Old English and the Problem of Alfred's Boethius," in The Editing of Old English: Papers from the 1990 Manchester Conference, ed. Donald G. Scragg and Paul E. Szarmach (Woodbridge, 1994), 163–76, at p. 168. That it was the prose version that Ælfric knew is evident most clearly from the phrase and þider wilniað oððe þæs þe him lyst oððe þæs þe hi beþurfon, which agrees much more closely with the prose Boethius and þider willniað, oððe þæs þe hi lyst oððe þæs þe hi beþurfon than with the verse wilnað to eorðan, sume nedþearfe, sume neodfræce; for Ælfric, see Ælfric's Lives of Saints, ed. W. W. Skeat, Early English Text Society os 76, 82, 94, 114 (London, 1881-1900), 1.56–57; for the prose and verse Boethius, see Sedgefield, Old English Version, 147.8 and 204.14–15.
 Brian S. Donaghey, "Nicholas Trevet's Use of King Alfred's Translation of Boethius, and the Dating of his Commentary," in The Medieval Boethius: Studies in the Vernacular Translations of De Consolatione Philosophiae, ed. A. J. Minnis (Cambridge, 1987), 1–31.
 Godden, "Editing Old English."
 King Alfred's Anglo-Saxon Version of Boethius 'De Consolatione Philosophiae', with a literal English translation, notes and glossary, ed. Rev. Samuel Fox (London, 1864). Fox records of the Cotton manuscript that "for many years it was utterly useless" because of the damage by fire and water, but that after rebinding and re-arrangement in 1844 it "is now rendered so perfect that most of it can be read with the greatest ease!" (p. iii). The manuscript has evidently deteriorated massively over the last century and a half.
 Kees Dekker, "Francis Junius (1591-1677): copyist or editor?," Anglo-Saxon England 29 (2000): 279–96; Daniel Paul O'Donnell, "Junius's knowledge of the Old English poem Durham," Anglo-Saxon England 30 (2001): 231–45.
 King Alfred's Version of the Consolations of Boethius Done into Modern English, with an Introduction, trans. W. J. Sedgefield (Oxford, 1900), p. 174.
 Sedgefield, Old English Version, 32.21–3.
 King Alfred's Version of St Augustine's Soliloquies, ed. T. A. Carnicelli (Cambridge, Mass., 1969), 90.23–91.8; "I know that it [the mind], if it is rational, will answer you and say that it desires to know what was before our time because it always existed since God had created the first man; … and I think also that it will say that it desires to know what will happen after our days because it knows that it will exist for ever."
 Skeat, Ælfric's Lives, 1.25–32; "There are three things on this earth: one is transitory, which has both beginning and end, that is animals and all soul-less things which began when God made them and afterwards end and come to nothing. The second thing is everlasting, so that it has beginning and has no end, that is angels and human souls, which began when God made them but they never end. The third thing is eternal, so that it has neither beginning nor end, that is the one almighty God."
 MS Bodley 180, f. 40v; "What shall we say now about the king's thegns but what happens to them: that they are robbed of all their wealth and even their lives by their beloved king."
 Sedgefield, Old English Version, 66.22.
 J. Bosworth and T. N. Toller, An Anglo-Saxon Dictionary (London, 1898), p. 54 (citing Fox's edition); Dictionary of Old English A-F, Dictionary of Old English Project, CD-ROM version 1.0 (Toronto, 2003).
 Toller had already hinted at a doubt about aspylian in his entry under swilian in the 1898 dictionary. Toni Healey, the chief editor of the Toronto DOE, agrees with my reading of the text and tells me that she will be deleting the entry for aspylian in subsequent versions.
 Sedgefield, Old English Version, p. xxiv. The reference is to Georg Schepss, "Zu König Alfreds Boethius," Archiv für das Studium der neueren Sprachen und Literaturen 94 (1895): 149–60.
 Kurt Otten, König Alfreds Boethius, Studien zur englischen Philologie, neue Folge no. 3 (Tübingen, 1964).
 Pierre Courcelle, "étude critique sur les commentaires de la Consolation de Boèce (IXe-XVe siècles)," Archives d'histoire doctrinale et littéraire du Moyen Âge 14 (1939): 5–140.
 See especially Petrus Tax, "Das Längezeichen e im Fränkischen und Alemannischen schon um 1000? Eine neue Hypothese," Sprachwissenschaft 27, no. 2 (2002): 129–42; I am very grateful to Professor Tax for further advice on this subject.
 Fabio Troncarelli, Tradizioni perdute. La Consolatio Philosophiae nell' alto medioevo (Padua, 1981), esp. pp. 135–49.
 Diane K. Bolton, "The Study of the Consolation of Philosophy in Anglo-Saxon England," Archives d'histoire doctrinale et littéraire du Moyen Âge 44 (1977): 33–78.
 Professor Wittig presented the material and his arguments at a symposium on the project held in Oxford in July 2003.
 Troncarelli, Tradizioni, pp. 135–49.