University of Virginia Library

Speght's Chaucer Aand MS. GG.4.27
by
George B. Pace

Although the primary reason for textual criticism is determining texts, there may be other reasons. One may simply want to set the record straight so that erroneous views need no longer be expressed. Or one may hope to discover the actual copy used for an early printed book. Or one may wish to answer a question of provenance or even of human biography. Something of all these "other" reasons underlies the present paper, which employs textual criticism not to establish a text but to throw light upon the books in which the text is contained.

The books are: Cambridge University Library MS. Gg.4.27 (a Canterbury Tales manuscript; "of the highest importance" — Manly and Rickert)[1] and Thomas Speght's 1602 edition of Chaucer's works, the last, except for reprintings, of the Blackletter Chaucers. The point of impingement is Chaucer's poem An A B C, which occurs on the opening folios of the manuscript and appears for the first time in print in Speght's volume.

When Speght published, in 1598, the first edition of his Chaucer, he obviously knew nothing of the A B C, since otherwise he would have printed it. So between 1598 and 1602 Speght came upon a manuscript having the poem. Can this manuscript be identified? I believe it can be: that it is Gg.4.27, which is now known to have been in the possession of Joseph Holland, antiquarian and lover of Chaucer, in 1600.[2] That Holland and Speght were acquainted is a virtual certainty.[3] What more likely place, then, for Speght to obtain the poem than Holland's manuscript where, as the poem begins the volume, it could hardly be missed?

A close relationship between the Gg. 4. 27 and Speght texts has generally been recognized, but they have been looked upon, because of certain differences, as sister texts deriving from a lost common parent.[4] I believe I

can show that the Speght text exhibits a sufficient number of the marked eccentricities of the version in Gg.4.27 to leave no doubt that Holland's volume is Speght's source.

Assuming I am correct, we have a rarity — the actual copy underlying an Elizabethan printed book — [5] and I shall be interested in examining the changes Speght made, the amount and kind of editing to which he subjected his copy. Finally, I shall also wish to consider the significance of the derivation for the provenance of Gg.4.27 and for the question of the authenticity of Speght's well-known assertion that Chaucer wrote the A B C for Blanche, Duchess of Lancaster and wife to John of Gaunt. But first I must establish the derivation.

I

I shall make the following assumptions:

• I. That shared errors indicate close relationship;
• II. That unique readings are probably spurious;
• III. That the simplest derivation (tree) which will account for the variants is to be preferred.

These are usual initial assumptions.[6] They are not laws; evidence may controvert them.

The A B C survives in sixteen manuscripts.[7] The following readings occur only in Gg and Speght:

• 9. myn (for thin)
• 11. omission of him
• 19. for to (for for)
• 38. ben (for be)
• 46. close in with þyn owene grace (for clothe with thi grace)
• 49. Gracyouse (for Glorious)
• 75. with þe (for to yow)
• 83. peyne (for peynes)
• 85. with (for of)

---

227




• 87. omission of oure
• 89. of (for with)
• 103. as litil (for litel)
• 105. þo (for that)
• 124. schal (for may)
• 132. fulle (for rightful)
• 133. ioye (for merci)
• 158. on (for unto)
• 179. out (for ought)

The number of shared variants is large; some are patently errors (e.g., lines 46 and 132). Therefore, according to Assumption I Gg and Speght are closely related.

Chronology prevents the derivation of the fifteenth-century manuscript from the late Elizabethan printed book. Even so, it is useful to be able to show that the derivation of Gg from Speght is implausible on purely textual grounds (otherwise one might justifiably wonder if Speght were not simply a printed facsimile of some lost manuscript, a virtual duplicate like, say, the Chaucer Society transcriptions). Speght has the following unique readings:

• 53. nor (for not)
• 56. sinke (for stynk)
• 77. ye (for that; Gg þt )
• 84. bostaunce (for bobaunce)
• 90. then (for ther)
• 92. can (for gan)
• 121. right (for yit)
• 136. that of pitie will (for that wole of pitee)
• 137. he (for that he)
• 144. royall (for rial)
• 172. a (for as a)
• 174. sured me (for mesured)
• 175. or (for and)
• 179. will (for wel)

Again the number of variants is large; some of the variants are patently spurious (e.g., lines 84 and 174); all of the variants are opposed in Gg by the generally supported readings.[8] Therefore, according to Assumption II Gg is not derived from Speght.

Gg also has unique readings: 156. han (for have); 173. besech (for preye). Moreover, in line 90 Gg has a variant which, although not unique (it is in Gg's sister manuscript, Coventry; see fn. 7 above), is a patent error:

þere brende (for brende). All of these variants are opposed in Speght by the generally supported readings. Here the number of variants is indeed small, but this fact does not seem to have disturbed previous editors. According to Assumption II, then, Speght is not derived from Gg. Since Gg has been shown not to be derived from Speght, the (apparent) final conclusion is that the two derive from a lost common parent.

This conclusion, however, ignores a large body of sublexical evidence which has never before been presented.

"In spelling," Manly and Rickert remark (I, 177), "Gg [the entire volume] has long been recognized as unique among Chaucer MSS." The peculiarities of the Gg volume have been extensively studied by Caldwell, and both he and Manly conclude that the scribe was probably not English but a Dutchman or a Fleming. Among the forms they especially cite in evidence, quoting Canterbury Tales A 1527 and A 1579, is bosch for bush (MDu bosch).[9] This form appears twice in the Gg copy of the A B C (lines 89 and 92):

The corresponding lines in Speght are: All other manuscripts read bush or something closely resembling bush. It is hard not to believe that Speght got its strange bosh from Gg's bosch.

Manly and Rickert (I, 177) comment upon another peculiarity of the Gg scribe: "He divides and joins words wrongly and makes nonsense"; they give myn che kys as an example. In line 174 Gg reads:

The corresponding line in Speght is:

The other manuscripts all read mesured. It is hard not to believe that Speght's unique variant resulted from a misinterpretation of the Gg reading with its unnatural space.

A striking peculiarity of the Gg scribe is described by Manly and Rickert (I, 177) thus: "The n of possessive adjectives [is] regularly retained before consonants: myn self, myn lyf myn lust." This feature, foreign to Chaucer's English as well as to Speght's,[10] appears over and over in the Gg

copy of the A B C and is often paralleled in Speght's; the Speght reading is given second:

• 7. myne perlious : mine perillous
• 16. myn schip : mine ship
• 18. myn synne, myn confusioun : mine sinne, mine confusioun
• 19. þyn presense : thin presence
• 23. myn dampnacioun : mine damnatioun
• 40. myn werk : mine werke
• 41. þyn tente : thine tent
• 48. myn deth : mine death
• 52. myn fadir : mine fader
• 55. myn socour : mine soccour
• 56. myn gost : mine ghost
• 68. þyn pete : thine pitie
• 74. þyn name : thine name
• 79. myn fo : mine fo
• 81. þyn sorwe : thine sorow
• 88. þyne petous eyne : thine pitous eyen
• 111. þyn goodnesse : thine goodnes
• 123. myn soule : mine soule
• 125. þyn sone, myn mene : thine sonne, mine meane
• 126. þyn self : thine selfe

One could continue extending this listing (further examples may be found in lines 134, 143, 147, 159, 161, 167, 180). Such spellings occur in no other manuscript of the A B C.

A few additional spellings could be cited but enough evidence seems to have been given.

Assumption III asserts that the simplest derivation (tree) which will account for the variants is to be preferred. The bifid tree, with Gg and Speght deriving from a lost common parent, explains the lexical variants but leaves the strange spellings unaccounted for. The derivation of Speght directly from Gg accounts for the spellings (and for most of the lexical evidence) but leaves the variants in lines 90, 156, and 173 unexplained. Thus either of the possible trees leaves something unresolved. In this circumstance Assumption III applies. The simpler derivation is, of course, the second. Therefore the true conclusion is that Speght is derived from Gg.

How are we then to explain the three lines where Gg is wrong and Speght right? I take the Speght readings to be conscious emendations. Speght was, after all, an editor, though an early one. Three emendations to the right reading in a poem of 184 lines is not beyond the bounds of a priori probability.

Line 90 is rendered thus in Gg: Brennynge of which þere neuere a stikke þere brende. The repetition is so obviously in error that almost anyone would be inclined to strike the second þere out.

The trivial variation in line 156, haue in Speght for Gg's han, is viewable as a mere modernization, han having become archaic by Speght's day.[11]

The third variation, besech for preye in line 173, becomes much less striking when viewed in context:

Obviously, as the rhyme scheme of the poem is ababbcbc, line 173 should rhyme. Apparently preye is the only possible rhyme.[12] Anyone who tried to improve the defective rhyme scheme would produce the right reading.[13]

To summarize: Between 1598 and 1602 Speght came upon a copy of the A B C. In 1600 Gg, with the poem on its opening folios, was in the possession of Joseph Holland. The manuscript was thus in London at the right time for Speght to use it, and in the hands of a man whom he almost certainly knew. Comparison of the Speght text with the text in Gg shows the two versions to be decidedly similar, the similarity including eighteen readings found in no other manuscript and more than thirty unusual spellings, likewise in no other manuscript, of a kind peculiarly characteristic of the Gg scribe and explainable in Speght's copy only on the assumption that it derives from Gg. Textual theory also requires this derivation. Surely one may write Q. E. D. Gg.4.27 was Speght's actual source.

II

Since we have no other instance in which the manuscript used for one of the Blackletter Chaucers has been identified, I shall examine in some detail Speght's handling of Gg, even when the results may seem predictable. I believe it may be of interest to have definite knowledge as to what one early editor of Chaucer did when he was faced with an actual manuscript.[14]

The A B C has 184 lines or approximately 1400 words. Speght differs from Gg in 22 lines or 23 words. Not all of the differences are misreadings. As has been observed, Speght corrected Gg in five instances, the most notable being the besech-preye variant. These readings, successful emendations, must be subtracted if a meaningful assessment of Speght's accuracy is to be made. Moreover, the three readings in which both Gg and Speght are unique (fn. 13) must also be subtracted, the Speght variants being attempts,

although unsuccessful, to correct a defective text. Finally, one of the Speght variants is regardable as simply a modernization (royall for rial, line 144); it too should perhaps be subtracted. If nine of the differences, then, are disregarded, Speght averaged one error every 13 lines, or 14 errors out of 1400 words.

How is this revised figure to be assessed? We have nothing exactly comparable, for none of the other manuscripts of the A B C appears to be a copy of an extant text. However, by collating Robinson's basic manuscript (Cambridge University Library Ff.5.30) with his reconstructed text one may arrive at some notion of a reasonable expectation. The manuscript exhibits ten differences. To be sure, this figure is imperfect. Even so, it bears out one's intuitive feeling, that Speght's copy is no better than what one might expect of a competent medieval scribe.

The preceding paragraph deals with lexical differences only. The medieval scribe was notoriously unconcerned about spelling; Speght's attitude seems to have been similar (conceivably some of the spellings were due to the printer). The spelling of 630 words, or 45 per cent of the total, is preserved. These are mostly words like of, to, he, good, and name, where Gg and the usual Elizabethan spelling are in agreement. For the remaining words the spelling is sometimes modified only slightly (e.g., þe becomes the, occasionally ye; ȝow, you; Virgyne, Virgine), but often the change is considerable: herte becomes heart; reles, release; myn, mine; refeut, refute; pete, pitie; Bounte, Bountie; faderis, faders; iuge, iudge; sorwe, sorow; sauacioun, salvatioun, etc. The underlying principle may appear to be modernization but is more likely, perhaps, simply conformity with Speht's own usage; nor, of course, is the apparent modernization carried out consistently (e.g., modir now remains modir, now becomes moder). The attitude resembles the medieval scribe's: fidelity to the spelling of the exemplar counts for little.

Gg is without punctuation. Speght, like most modern editors, supplies punctuation. He uses only comma, period, and colon, and he often employs a comma where modern usage would prefer a semicolon. If we disregard these differences, Speght's punctuation is good and agrees with Robinson's in 121 out of 184 lines, or 65 per cent of the time. The figure is not an absolute indication; in some 32 lines more than one punctuation is possible (e.g., line 28, in Speght thus: For certis, Christis blisfull modir dere; Robinson puts a comma after For also). Speght and Robinson punctuate one stanza identically (with the allowance mentioned).

So much for general observations. A few of Speght's misreadings are interesting in themselves (and further proof, if such is needed, that Gg was Speght's exemplar). In line 56 Gg reads: To stynk eterne he wele myn gost exile; Speght changes stynk to sinke, one guesses from a wish to ameliorate the diction (cf. the probable amelioration in St. John's College G.21: To lastande Paine). In line 77 Speght reads ye where the correct reading is that; one would suspect origin from þt — and þt is the Gg variant. In line

84 Gg reads bobaunce ("boast"); apparently Speght did not know the word and so, taking his cue from the context, coined the pseudo-archaism bostaunce. In line 144 Gg reads in so rial wise; Speght removed this archaism, substituting royall for rial. In line 150 Speght reads sore for the correct yore; one would suspect origin from ȝore, and such is the Gg reading. Speght's striking misreading of Gg's me seured as sured me has already been discussed.

For one change, the title, Speght may have gone to another source. Today the first four folios of Gg are missing; the text of the A B C begins the fifth folio. The opening four folios may have been missing in Speght's day (Holland wrote his name on what is actually the fifth folio; one would assume that he would inscribe the first page of the manuscript). Someone, not the Gg scribe but probably Holland, wrote a title, in blue ink, at the top of the fifth folio: CHAUCERS A. B. C. [15] Otherwise the poem is untitled in the manuscript. Speght has the following title: Chaucers A. B. C., called La Priere de Nostre Dame: made, as some say, at the Request of Blanch, Duchesse of Lancaster, as a praier for her priuat use, being a woman in her religion very deuout. This is really two titles and a statement. The first title, Chaucers A. B. C., appears in this form in Fairfax 16 as well as in Gg, but there is no need to suppose that Speght went beyond Gg for it.[16] The second title, La Priere de Nostre Dame, occurs only in Pepys 2006. If Speght consulted this manuscript, he did so purely for the title; no influence is to be seen in his text. The statement about Blanche appears nowhere else. It is possible that Speght obtained it from some lost source. It is also possible that he found all of his heading in Gg: the English title on folio 5, where it still is; the remainder on the missing fourth folio, which may have been merely so badly deteriorated in 1600 that Holland preferred to write his name on the next page.

What, then, did Speght do with Gg? He copied the poem relatively faithfully, emending occasionally (not always rightly) and making an average of one real error every hundred words. He made the spelling largely conform with his own although he did retain such peculiarities as the n on possessive pronouns before consonants. He punctuated the poem. He provided an elaborate heading, possibly simply combining elements in Gg, possibly going elsewhere for part of it. Finally, a circumstance not remarked upon above, he featured his discovery on his title page, ending his list of the changes from the first edition with the statement, "Chaucers A. B. C. called La Priere de nostre Dame, at this Impression added."

There are suggestions here of the modern editor, but even more of the

medieval scribe. If it is in the latter tradition that Speght is to be viewed, he comes off fairly well.

III

That Speght made use of Gg has now, I feel, been conclusively demonstrated and is in consequence regardable as a fact in the manuscript's provenance.[17] The significance of the fact may be seen in the following quotation from Manly and Rickert (I, 178) concerning additions to Gg thought to have been done, or commissioned, by Holland:

Thus a curiously reciprocal relationship obtains: Not only was Gg used to augment Speght; Speght was earlier used to supplement Gg!

IV

I shall conclude by returning to the possibility that Speght found all of his heading in Gg. This possibility is of great interest because of the statement that Chaucer wrote the poem for Blanche, Duchess of Lancaster, which constitutes, if true, an item in the poet's biography. Nothing, of course, is inherently improbable in Chaucer's having written the A B C for the wife of his long-time associate (and patron) John of Gaunt. Blanche died in 1369; the A B C is agreed to be an early poem. That Chaucer wrote the Book of the Duchess to lament Blanche's death is, although not a certainty, the common belief. Nevertheless Chaucerians have either given very qualified acceptance to Speght's statement or have rejected it — principally, it would seem, because of its late date.[19] If Speght's heading could be shown to have been in so venerable a manuscript as Gg the argument for authenticity would obviously be stronger. And not only because of the earlier date of Gg (1420-40). In the opinion of Manly and Rickert Gg was prepared for "a wealthy patron of literature . . . who knew owners of special texts" (I, 180). Presumably "special texts" may be interpreted to include special traditions.

The fact that the statement about Blanche, and the accompanying French title, must be presumed to have been written at the end of folio 4 (rather than on folio 5 immediately above the text, as would now seem normal) presents only an apparent difficulty. Headings in manuscripts do occasionally appear just so. For example, the Leyden manuscript of Chaucer's Truth (Vossius Germ. Gall.Q.9) has the heading for the poem at the bottom of folio 95b while the poem itself is on 96a. The corresponding folio in Gg is missing. Are there compelling reasons for believing that it had on it the statement about Blanche?

There are, at any rate, reasons.

To begin with, the simplest explanation of Speght's heading is to assume that all of it was in Gg, and if a fact of biography were not involved this reason would doubtless seem sufficient in itself.

Next, the French title is characteristic of Gg. Three of the four short poems which follow the A B C have foreign titles, and one of these is French (Balade de bone conseyl [Truth]). Also, Speght's French title is found in Pepys 2006 (and only found there). Heath's tree (fn. 4 above) shows Gg and Pepys as deriving from a common source, now lost. Making the adjustment made necessary by this paper, showing Speght as derived from Gg, we have the following as the relevant portion of the tree:

A strong textual reason thus exists for believing that Gg once had the French title at least.[20]

Two other arguments apply strictly to the statement about Blanche. The phrasing — "made, as some say" — seems to point to an earlier day. The A B C must have been virtually unknown in the sixteenth century, as Speght

was the first to print it. Finally, a special reason exists for believing that Gg, more than any other manuscript, might well have preserved a tradition about Blanche. Manly and Rickert's study of the early provenance of Gg leads them to conclude (I, 180) that the original owner was probably Humphrey, Duke of Gloucester. As they observe, Humphrey was Chaucer's great-nephew (half blood). Even more pertinent to our discussion, Humphrey's grandmother was Blanche, Duchess of Lancaster.[21]

This paper has shown that the version of Chaucer's A B C in the 1602 Speght, the first printed edition of the poem, is essentially a copy of an extant text.[22] The paper has also considered the editorial treatment which Speght gave to his source, and the relevance of the identification of the source for the provenance of the Canterbury Tales manuscript Gg.4.27 and for the authenticity of Speght's assertion that Chaucer wrote the A B C for Blanche, Duchess of Lancaster, which it rather strongly supports.

Notes