Reviews (Occasional Papers, volume 2)

165 Reviews A Variorum Edition of the Works of Geoffrey Chaucer Volume II The Canterbury Tales: The General Prologue. Part One A, edited by Malcolm Andrew, Daniel J.Ransom, Charles Moorman with the assistance of Lynne Hunt Levy. xxvii+298 pp; Part One B Explanatory Notes, by Malcolm Andrew. xxiii+623 pp. (Norman OK and London: University of Oklahoma Press, 1993.) ISBN 0-8061-2552-7. This latest production of the Variorum Edition of Chaucer is the largest to date. The General Prologue [GP] is issued in two volumes, divided as follows. Volume 1 contains, firstly, the critical commentary dealing with sources and analogues, the date of composition, development of scholarly writing, particular approaches and topics, interpretations, and language and style. Secondly, there is the textual commentary, which contains the textual tradition, evidence of the glosses, table of correspondences, and descriptions of the manuscripts and printed editions. Thirdly, the text itself follows and on each page underneath the text there is the apparatus criticus together with comments on manuscript variations and readings as well as editorial comments and proposals. This is followed, finally, by an appendix containing those passages with translations from the texts which have been identified as sources for GP, an extensive bibliographical index which relates to both volumes, and a general index which relates only to this volume. The second volume contains a comprehensive account of all comment on GP together with an index restricted to the commentary in this volume. This arrangement was forced on the editors by the large amount of critical work done on GP and this is why so many contributors have been involved. Some of the sections are those which form part of the normal pattern established for the Variorum. In GP, however, the critical commentary in volume one is separated from the critical commentary found in volume two. Much of the comment on GP has been on individual lines or individual portraits, and this makes it difficult to trace a coherent attitude to criticism of GP itself. Consequently Professor Andrew who is responsible for the commentary in both volumes has divided the main outline in volume one into the various parts outlined in the previous paragraph. As far as possible Andrew keeps his comments to what he has accepted as significant comment for GP as a whole, whereas detailed comments on individual lines or pilgrims are kept for volume two. Naturally this can lead to some disjunction because it is not always easy to separate the one from the other. One must accept that he has made a reasonable decision in the light of the mountain of material he faced. In an analysis of the text the Variorum maintains its tradition of considering only ten manuscripts and Caxton’s first edition for detailed comparison, though unique readings in them are traced in the other witnesses. The apparatus criticus is based on all printed editions up to that of Pratt (1974); more recent editions are not recorded, even though several have been issued since then. These more recent editions have been consulted and from time to time their readings are commented on in the accompanying explanatory notes. THE CANTERBURY TALES PROJECT OCCASIONAL PAPERS II, 1997, 165-179 166 Canterbury Tales Project Occasional Papers II Differences in punctuation are not normally recorded in the apparatus criticus, although occasionally a difference in punctuation is recorded in the notes. This is unfortunate as the punctuation can make a considerable difference to an understanding of what a passage means and, for example, comments on the punctuation used at lines 346 and 367 would have been welcome. No attention is likewise given to punctuation in the commentary on language and style in the introduction, for this section is essentially concerned only with style, and language as such receives little or no mention. The base text is Hengwrt [Hg] as is standard for the Variorum, but its text is given modern punctuation including capitalisation and word-division. The edition sometimes keeps words separate, sometimes introduces hyphens, and sometimes joins words as compounds. Where the edition’s word-division differs from Hg or from other editions, this is not recorded in the notes, even though once again word-division may differ from one editor to the next and occasionally this can cause a difference in interpretation. What the editor thinks of as abbreviations are expanded silently in the text,and occasionally these are commented on in the notes. Thus the cross-stroke through final ll in Averylle (line 1) is interpreted as e; this expansion is referred to, but not justified in, the notes. The claim in the note that ‘all the printed editions read some form of Aprille’ (p.127) is true for the editions used for comparative purposes, but my own 1980 edition does not since it keeps the Hg reading. Since the notes sometimes refer to differences in my own and other recent editions, a reader might be forgiven for assuming that all differences found in such editions was recorded; they are not. Macrons over and flourishes attached to a final consonant are usually interpreted as an abbreviated e and this leads to such forms as sope (334), tope (590), sleepe (397) and youre (804.) There is no discussion of whether the first two abbreviations might not have been better expanded as soppe and toppe. Where the final flourish is attached to a letter in a group that can be interpreted as ou, it is interpreted as n, as haubergeoun (76), though whether the flourish represents the final n or the penultimate u is not clear. This point is significant because at line 820 this flourish is interpreted as a medial o to give the reading echoon (to rhyme with anoon) rather than echoun. The normal spelling of words like this is simple -on as at lines 747-8 where euerichon rhymes with anon. One can understand why the editors did not want the reading echoun, but it rather throws in question whether this flourish should be interpreted as an abbreviation at all. This reading is not commented on. As it says in the introduction ‘Despite Hg’s number of uncommon readings, the present edition emends Hg only three times’ (p.64.) An uncommon reading is one which is not supported by at least ten manuscripts, though this is a strange figure to choose since the editors appear to accept the Manly/Rickert proposal that for GP we should accept five major lines of descent. This could suggest that what is important is whether there are variations in these five lines of descent rather than across ten manuscripts. In GP there are 45 such readings in Hg as compared with 37 in Ellesmere [El]. It is difficult to evaluate the significance of these figures. There is a difference between the first part of GP up to line 300 and the second part. In the first part Hg has 23 unique readings and El only 7, whereas in the second part the respective numbers are Hg 22 and El 30. In other parts of the poem Hg tends to have a lower number of unique 167 readings than El. The three emendations in Hg made in this edition are, firstly, the change of is to was in Hg’s ‘There as this lord is kepere of the selle’ (172.) Hg’s reading is unique here, but its meaning is satisfactory and hardly merits alteration. The change is not justified, though the notes contain some discussion of the punctuation of the neighbouring lines. The position seems to be roughly as follows. Many scholars assume that line 172 follows directly on to line 171 and in that event the reading was makes better sense. But several editors think that line 172 goes better with line 173, and this is particularly true if the line reads is. It may well be that a change of is to was was made by a scribe or editor who thought that 172 should go with 171 rather than with 173 and consequently felt the past tense was more appropriate. Secondly, at line 324 (wrongly given as 342 on page 64, and not included in the list of Hg’s unique readings on page 65) Hg and one other manuscript do not read the in ‘That from tyme of kyng william weere falle’; but the is included in this edition. Once again the reason for the emendation is not given, but it may be partly a matter of language since determiners were becoming more regular at this time or of metre since the line has otherwise only nine syllables. But there are many instances where tyme has no determiner in the poem as a whole (cf. 1:1637, 1834, 3666) and there are equally many examples of nine-syllable lines. The third emendation is the inclusion of the couplet 637-8 in the description of the Summoner which may have been caused by eyeskip, though the couplet is not found in one other manuscript and its lines are bracketed in another as though they are not part of the original text. A fourth manuscript omits one of these lines and changes the rhyme. So it is clear that something unusual applied to this couplet; it may have been added or marked for deletion. The decision whether to include this couplet is a fine one, but the other two emendations are less justified. This seems particularly so in view of the fact that the edition reads Rusus at line 430. This reading instead of Rufus is found in several manuscripts and this edition assumes with Manly and Rickert that it must, therefore, have been in the copytext used by the scribes. But the reading is erroneous and the commentary clearly accepts that the reference is to Rufus. The preservation of the reading raises the question what text the edition is trying to offer, and this is a point which is not discussed in the volume. Is the edition trying to create this assumed copytext or is it trying to reconstitute Chaucer’s original, assuming that these two are different? Although the edition states that Hg is the base text for the edition, it does not say how it intends to handle it or what precisely the edition represents. Why emend the one unique reading in Hg and two readings supported by other manuscripts, when other readings with little manuscript support are retained and when a reading which makes no sense is not emended? Two further readings in the edition merit mention. At line 32 the edition reads anoon, although it is accepted that the manuscript reads anon with a gap between ano and n. But there is no trace of a further o and no indication that one was ever written. It does not seem satisfactory to read what might have been there, when no trace of this assumed reading can be detected. At line 60 the edition accepts armee rather than ariuee as Hg’s reading, though it is impossible to achieve certainty as to what the scribe actually meant as the notes make clear. Reviews 168 Canterbury Tales Project Occasional Papers II The second volume is a monument to patient reading and hard graft. Professor Andrew has waded through a mass of material but one may question whether the expenditure of all that energy was worthwhile. Since the coverage is comprehensive, one could argue that the compiler has spent too much time on scholarship that in many cases hardly deserves that name. The volume is arranged in what might described as a descending order of length. A portrait as a whole, then individual groups of lines within the portrait, then individual lines within a group and finally individual phrases or words may elicit comments. This arrangement naturally leads to some overlap since what an individual scholar says may appear in the discussions of the portrait as a whole, of groups of lines and of individual words. Thus the association of various pilgrims with signs of the zodiac as proposed by Spencer figures with reference to the Monk in comments on lines 198-206, on lines 198-9 and on lines 201-2. There are many cross-references, but they are not always helpful. The cut-off date for comment is 1985, which is different from the date used for editions, but references to works written after that date are found. Scholars quoted have to be looked up in the first volume, because the bibliography is not repeated in the second. Professor Andrew has spent a considerable amount of time in reading older scholarship and what he has found is sometimes interesting in predating what was re-discovered in later scholarship. Whether the results justify this length is uncertain. It is to my mind a weakness that authors are sometimes quoted through a single phrase or clause as though this quotation provides the kernel of what they thought. Thus on p. 127 it is said that Hales understood line 119 to mean that ‘the Prioress never swore at all’; that Skeat claims there is support here for her ‘love of gold and corals,’ and that Kittredge claimed ‘that in those days everybody swore.’ Inevitably these quotations are plucked out of the work in question and do not necessarily give a balanced account of the whole work. Because the arrangement is largely chronological, Andrew cannot pick out the major themes and subsume the various accounts under a more general approach. He is very restrained in his account of the various opinions he refers to; he rarely says that an opinion is wrong or stupid. He confines himself to coded criticism by using such adverbials as ‘curiously’ and ‘surprisingly.’ The result is a blow-by-blow report of what has been written on each part of the Prologue which is at times somewhat indigestible. This is not Andrew’s fault since he has carried out his brief remarkably well. What it does call into question is whether this is the most sensible use of a senior scholar’s time and energy. The second volume will remain a veritable repository of information on what has been written on GP and future editors will no doubt use it to help them in their work. Other scholars are likely to dip into those parts which might seem of value to what they are doing. Few will actually try to read it from cover to cover. The production of this book is excellent. The editors are to be complimented on the accuracy with which they have accomplished their task, for misprints and mistakes are few. The result is somewhat lengthier and organised less helpfully than most scholars might wish for, but then GP is different from the rest of Chaucer’s output. Whether it should have been pruned is a question that constantly arises as one reads through the two 169 volumes. But we should be grateful for the material crammed between these four covers even if few will exploit it to the full. N.F. Blake University of Sheffield The Canterbury Tales: Fifteenth-Century Continuations and Additions. Edited by John M. Bowers. A Teams Middle English Test Series Book. Kalamazoo, Michigan: Western Michigan University, 1992. ISBN 1-879288-23-0. vii + 200 pp. Price $8.95. This volume consists of a brief introduction (pp.1-5), select bibliography (pp.5-10), editions of five texts or groups of texts (pp.11-196), and a short glossary (pp.197-200.) The five edited works are the prologue to Lydgate’s Siege of Thebes, The Ploughman’s Tale from the Christ Church manuscript, The Cook’s Tale with the additions found in Bodley 686, two short sets of spurious links (one from Lansdowne 851 and the other from Royal 18 C ii), and The Canterbury Interlude and The Merchant’s Tale of Beryn. Each of these edited works consists of a short introduction, the text itself with glosses of difficult words in the margin, and notes at the end. The purpose of the glossary at the end of the book is not clear since glosses are given in the margin of individual texts, but it may be intended to provide glosses for common words which are not glossed in the margin. The purpose of the book, as its title indicates, is to provide editions of fifteenth-century continuations and additions. No information is provided as to what a continuation or addition might be, for there is no discussion of what might constitute Chaucer’s text. It appears to be accepted that what is in the Riverside Chaucer is indeed the definitive text of the poem left by Chaucer and so anything not found in that edition constitutes an addition. However, there are many pieces which are not in that edition which are not included here. Thus, the Tale of Gamleyn is not edited in this volume, though the reason for its omission is not clear. Bowers writes ‘while Gamelyn survives in a large number of manuscripts as a replacement for the unfinished Cook’s Tale, the fourteenthcentury romance was not written in response to Chaucer but was, quite possibly, found among the author’s papers as a possible source for a substitute tale that was never composed’ (p.vii.) It would appear it was because this tale was written in the fourteenth century that it is not included here. But he includes the Ploughman’s Tale from the Christ Church manuscript, though not that from Thynne’s edition of 1542. Neither was written ‘in response to Chaucer,’ even if both were written in the fifteenth rather than the fourteenth century, and yet only one is included. Naturally, there are many additions in a large number of the manuscripts of The Canterbury Tales, but the additions included here are somewhat restricted. No reason is given for the arbitrary decision of what is included or not. It is true that the editor claims that his book Reviews 170 Canterbury Tales Project Occasional Papers II ‘is a practical edition aspiring to no great scholarly rigor’ (p.vii), but the premisses on which the edition is based do need to be made clear. Bowers isolates four types of addition or supplementation of the poem. These are: to allow for a return trip to Southwark; to provide links for tales in the poem; to provide endings for unfinished tales; and to give pilgrims without a tale a narrative as his or her contribution. He neglects additions made within the tales, though he edits The Cook’s Tale from Bodley 686 and highlights not only its new ending, but also the additions made within the body of the tale itself. Indeed, many tales in different manuscripts have similar additions. This edition, therefore, contains some additions and continuations, but it is by no means comprehensive. Bowers is of the opinion that many of the additions are politically motivated. He sees in them the hand of Thomas Chaucer, Geoffrey’s son, and his attempts to provide support for the new Lancastrian dynasty. This position is not argued in any depth and seems to be in conflict with the other view he expresses, namely that editors of the poem in the fifteenth century were distressed by the unfinished nature of the poem and tried manfully to create the impression of a complete work. The select bibliography is very patchy and does not contain many of the books or articles which have been devoted to the textual development of the poem over the last fifteen to twenty years. No one looking at this bibliography would have any inkling that there had been so much controversy over the manuscript tradition recently. There are, on the other hand, references to more sophisticated discussions of how to edit Middle English works, which appear to sit rather uncomfortably with the expressed view that this edition aspires to ‘no great scholarly rigor.’ References are provided throughout the book, as in the introductions and notes to each edited text; but it may be difficult for the reader to find out full bibliographical information about such books. In the opening note to the Siege of Thebes a reference is given to ‘The Middle English encyclopedia On the Properties of Things’ (p.19) and page references to various discussions in this work are added. But a student might find it hard to track down what edition was intended and how to find the specific passages referred to through the page numbers. The texts are edited with modern paragraphing, punctuation, capitalisation, and word-division, and the letter forms have been modernised. Otherwise, the original spellings have been retained. Generally speaking the texts have been edited conservatively, and any changes are recorded in the notes. This decision is quite sensible and Bowers has largely taken his lead from previous editors who have edited these texts. The glosses in the margin are not always very helpful, and sometimes they are wrong. Consider some examples from Lydgate’s Siege of Thebes. On p.15 l.96 the expression wel broke ye youre name has the gloss ‘broached,’ presumably as a translation for broke, though how the readers would understand the gloss is difficult to imagine. However, on p.62 l.66 the expression now broke we thy name is glossed in the margin ‘do credit to.’ Not only is this gloss more correct, but there is no cross-reference to the previous example. Similarly the gloss for I dar my hede assure (p.15 l.103) as ‘guarantee’ hardly strikes the right note. An annotation for this whil (p.14 l.66) might have been welcome by students reading this text. Throughout all the texts there are 171 numerous words which might not be known to students in their Middle English form and which could reasonably have been glossed and others which are glossed in a somewhat unhelpful way. Common Middle English words which might have been glossed include skill, hoppe, by any way, had up and religiouste. Words which are not helpfully glossed include sook ‘suckled’ (p.26 l.21), though the subject is the baby Jesus and not his mother as the gloss ‘suckled’ implies, and excitacioun ‘spiritual exercise’ (p.27 l.46) where the meaning is probably ‘awakening’ as suggested in the Middle English Dictionary. The problem with this edition is that it is not quite clear for whom it is intended. It appears in a series which is devoted to student texts and so one must assume it is meant for undergraduates and new postgraduates. For them a complete collection of the additions in the Chaucer apocrypha might have been welcome. That some texts have been edited elsewhere seems irrelevant, since students might not have access to such editions. The impression from the title of the volume is that the whole of the apocrypha is included and it might have been more helpful to students if that implied promise had been honoured. Students might also need a little more discussion about how one decides what is added to the text, which means some discussion about the text as left by Chaucer. A little more care as to what should be glossed or commented upon would also have been helpful. Students are not likely to have much interest in editorial minutiae and to include these with general notes might confuse them as to what was important in these texts. Nevertheless, it is difficult to get hold of some of the texts edited in this volume and Professor Bowers has done us a service by making them more readily available at a reasonable price. N.F.Blake University of Sheffield Geoffrey Chaucer. The Legend of Good Women. Edited by Janet Cowen and George Kane. East Lansing, Michigan: Colleagues Press, 1995. xii + 344 pages. For the past three decades—indeed, since George Kane published his edition of the A version of Langland’s Piers Plowman in 1960—debate about editing of Middle English texts has centred around the problems of the Piers Plowman tradition. By contrast, there has been relatively little attention given to the editing of Chaucer. Editions of Chaucer there have certainly been, prominent among them the Variorum Chaucer and the new Riverside, but the less-expert observer might be forgiven for thinking that Chaucerian editing is a peaceful business, compared to the warfare surrounding Langland’s text. There have been signs that this consensus (if that is what it is) is more apparent than real. Blake’s articles, and his edition of The Canterbury Tales, have pointed to fundamental unresolved issues in the textual history of this work; other studies by (among others) Moorman, Owen, Ramsey, and Hanna have focussed on some of these issues; the very different approaches taken by the Reviews 172 Canterbury Tales Project Occasional Papers II Variorum and Riverside editors imply disagreement, without necessarily articulating it. Above all, there has been Kane’s own article on Manly and Rickert’s work (printed in Paul Ruggiers’ Editing Chaucer: The Great Tradition) on The Canterbury Tales, and Kane’s judgement in this piece that no part of Manly and Rickert’s work is be trusted, and it must be entirely redone. In a long footnote to that article Kane outlines the procedures he thinks such an edition should follow: Step 1. All the variants at every point, in every witness, should be identified; Step 2. These variants should be scrutinized to establish, by editorial judgement, the original reading at each point; Step 3. The agreements in unoriginal readings among witnesses at each point should be gathered, and the witnesses grouped into families on the basis of shared agreements. A text may then be issued, giving the text of the putative original as established by the editor in (2) above, with all the variants from this original (as in 3 above) contained in an apparatus. This is exactly the model Kane established in his 1960 edition of the A version of Piers Plowman, and deployed in his 1975 edition of the B version with Talbot Donaldson. This, too, is the model for this edition of The Legend of Good Women: thus the editors’ declaration in the first sentence of their Preface (p. vii.) As the first application of these methods to a text outside Langland, and especially to a major Chaucerian text, this edition is accordingly of the deepest interest. The components of this edition will be familiar to readers of the Piers Plowman editions: a lengthy introduction, in this case consuming half the volume and containing intricately-detailed accounts of the manuscript variants and much else; an original-spelling text; a full corpus of variants in the apparatus. Far the longest section of the introduction, at 69 pages, is the ‘Analysis of variants in the Legend’ section. This is the heart of this edition. Here, as in the similar sections in the Langland editions, the editors gather every variant, in every witness, and group them into categories according to the scribal variation exemplified (in the editors’ opinion) in each variant: nearly eight pages for variants resulting from ‘inducement of preceding copy,’ five pages for ‘inducement of following copy,’ three pages for ‘grammatical attraction,’ and so on. Other parts of the introduction treat the variants in the Gg 4.27 copy, and in the Prologue. These sections of the introduction are vital because it is through these lists that the editors reveal their identification of the original reading at each point of the text, and their justification of this identification: Kane’s Step 2, above. There are some difficulties with this procedure. Firstly, the editors assert, with characteristic rigour, that because only agreement in unoriginal readings is significant of shared descent from exemplars below the archetype, therefore no genetic analysis of the manuscript relations is possible until the originality of each reading is determined. This is not correct. Cladistic methods, and their capacity to create ‘unrooted trees,’ have shown that it is not necessary to identify the original reading at each point before proceeding to stemmatic analysis. O’Hara and I discuss this at some length in our article on ‘Computer- 173 assisted Stemmatic Analysis’ in the first Occasional Papers volume. Consider a tradition of fifteen witnesses, all descending from an archetype ‘O.’ We discover that the witnesses divide neatly into three groups A, B, and C, each of five witnesses, with the witnesses of group A sharing the A set of variants, those of group B sharing the B set of variants, those of group C sharing the C set of variants. We can then examine those three sets of variants. It is likely that two of these sets (say, the A and B sets) will show evidence of being non-authorial and scribal, and thus likely to have each been introduced into the tradition by a single exemplar below the archetype. However, it is possible that the third set of readings (the C set) shows no evidence of being non-authorial and scribal. In this case, the five witnesses of this group do not share any exemplar below the archetype. Thus, they are not a genetic group in the sense that the witnesses of groups A and B are. Rather, the witnesses of C are related only through common descent from O, and the readings in which C witnesses agree against other groupings have come by descent from O itself, and not descent from a hyparchetype below O. This model (which is the method used by the Canterbury Tales Project) inverts the procedures followed by Kane and Cowen. Where they first identify originality, and then identify the groupings, in this model one first identifies the groupings, and then examines the variants in these groupings for evidence of originality or otherwise. There is a simple practical advantage in this model: instead of having to assess the originality of every reading, in advance of stemmatic analysis, one has only to assess the originality of those readings which are apparently characteristic of witness family groupings. This will save considerable editorial effort. However, there is reason to think that the method adopted by the Kane and Cowen may have had other and more serious consequences, beyond the expenditure of their labour. The editors’ method rests entirely on their confidence that through their long experience of scribal variation, and through their special feeling for Chaucer’s poetry, they are able to deduce from the range of variants at every point just what the original reading was, and how this came to be corrupted into the extant variants. The skill, sensitivity, and experience of the editors are evident on every page of this edition. Yet, there appear to me to be profound theoretical difficulties with this model. Chief among these is: exactly what is meant by ‘the original reading’? Is this the reading of Chaucer’s own autograph; of a scribal copy prepared under Chaucer’s supervision but perhaps containing errors; or is it the reading of the most recent exclusive common ancestor of all extant witnesses, which might actually be separated by several stages of copying from Chaucer’s own copy? It is an axiom of stemmatics that it can only reveal the readings most likely to have been present in the most recent exclusive common ancestor. But at various points, it is clear that the editors believe that what they are identifying is Chaucer’s own reading: thus on p. 41 the editors list five readings where they believe all witnesses are in error, and in each case give what they think is the ‘correct’ reading—that is, Chaucer’s own reading. In cases such as this, the editors feel themselves able to go beyond the actual manuscript evidence. This reader does not always feel able to go with them, as I will show in my discussion of some of these readings. Reviews