University of Virginia Library

The Little Gest of Robin Hood: A Note on the Pynson and Lettersnijder Editions
by
J. C. T. Oates

Pynson's edition of The Little Gest of Robin Hood (STC 13688), a quarto with 31 lines to the page, survives only in the following fragments: (1) a leaf signed cii, in the University Library, Cambridge (Oates, Catalogue of the Fifteenth-Century Printed Books, &c., no. 4206); (2) 1½ leaves formerly in the possession of Mr. Boies Penrose (A Selection of Rare Books . . . from the Library of Boies Penrose, &c., p. 29) but now by his gift in the Folger Library (The Folger Library: A Decade of Growth, 1950-1960, p. 25); (3) a leaf in the Bodleian Library, Douce fragments f. 51 (3). I shall call these leaves ULC, BPF1, BPF2, and BLO.[1]

ULC has been recovered from a binding. It was acquired in September 1888 by Francis Jenkinson (Cambridge University Librarian, 1889-1923) and was presented to the University Library by him on 20th August 1917. It is the fragment registered by Duff, Fifteenth Century English Books, No. 362 ('Private Library') and STC. BPF1,2 have also been taken from a binding. They are doubtless the fragments offered for sale in 1927 by Myers & Co. of London, catalogue 257, item 270. BPF2 is the upper half of the leaf only, containing the first fourteen lines. BLO is very defective. This leaf was known to F. J. Child, The English and Scottish Popular Ballads, No. 117. He cites it as e but does not attempt to give it a printer or date.

Since there are 31 lines to the page, the deficiencies of BPF2 and BLO are easily supplied from Child's text, if allowance is made for a cross-heading, perhaps with a white line above and below it, at the beginning of each 'Fytte'. I give the first and last lines of each page and identify them in Child, using his system of reference (stanza-number

with a superior number denoting the line within the stanza). The 'Fyttes' run thus: I, 1-81; II, 82-143; III, 144-204; IV, 205-280; V, 281-316; VI, 317-353; VII, 354-417; VIII, 418-456.                                

ULC	recto	And euill thyrfte on thy hede . . .}	2201-2273
		No force therof saide robyn}
	verso	Agaynst the lawes and ryght}	3194-3272
		And went hym on his way}
BPF1	recto	And Robyn hode wente to grene wode}	3273-3351
		Whan she cam in the forest}
	verso	The truthe than tell thou me}	2432-2504
		A moche better borowe}
BPF2	recto	Under the grenewode tree}	3124-3193
		[Thou kepest here the kynges enemys]}
	verso	For curtesy can he none}	2274-2352
		[Of all this longe day]}
BLO	recto	[With hym] all for to gone}	4354-4432
		[It may no better be]}
	verso	[Seuen nyght I gyve] the leve}	4433-4511
		[Yet he was begyled, i-wys]

It is apparent that BLO is the only leaf on which the verso text follows on correctly from the recto. The text printed on the other leaves extends from 2201 to 3351 — a total of 461 lines, to which must be added the headings &c. at the beginning of Fyttes V and VI, making material for 15 pages in all. The contents of these pages may with reasonable certainty be reconstructed thus:

• 1. 2201-2273ULC recto
• 2. 2274-2352BPF2 verso
• 3. 2353-2431
• 4. 2432-2504BPF1 verso
• 5. 2511-2583
• 6. 2584-2662
• 7. 2663-2741
• 8. 2742-2813
• 9. 2814-2892
• 10. 2893-2971
• 11. 2972-3044
• 12. 3051-3123
• 13. 3124-3193BPF2 recto
• 14. 3194-3272ULC verso
• 15. 3273-3351BPF1 recto

Since ULC recto is signed c2, it follows that BPF2 verso is c2 verso, and that BPF1 verso is c3 verso. If the book were a quarto in fours,

BPF1 recto would be e1, or d1 if it were a quarto in eights, but in fact BPF1 is unsigned. It follows that the book must be a quarto in sixes. Thus BPF2 recto (of which the lower half is missing) is d2 recto, ULC verso is d2 verso, and BPF1 recto is d3 recto: or, to state the facts another way:

• ULC contains c2 recto backed with d2 verso
• BPF2 " d2 recto " " c2 verso
• BPF1 " d3 recto " " c3 verso

At this point readers of Studies in Bibliography will not find it necessary (as I did) to fiddle about with folded pieces of paper in order to conclude that, all the formes of c and d being in type at the same time, c (i) has been perfected with d(o) both in the outer sheet of the quire (leaves 1, 2, 5, 6) and in the inner half-sheet (leaves 3, 4).

A simple interchange of the heaps before the sheets were perfected accounts for the mistake in the outer sheet. As regards the inner halfsheet the problem is a little more complicated, there being two different ways in which the mistake might have been made:

• (1) The paper may have been cut before printing began and the half-sheets c3.4 and d3.4 each worked from two two-pages formes. The mistake would then have been caused, as with the outer sheet, by an interchange of the heaps before perfecting.
• (2) The half-sheets may have been printed by half-sheet imposition ('work and turn'). The mistake would then have arisen by an interchange of the heaps after the sheets had been turned end-to-end before perfecting.

Of these two methods of working the half-sheet the first, since it doubles the labour involved, cannot be regarded as anything more than a theoretical possibility. The second is therefore the more likely.

Since we know that sigs. c and d each had six leaves, and since c2 recto begins at 2201 while d3 recto ends at 3351, we can now attempt to reconstruct the collation of the whole book. There are 402 lines of text and two cross-headings (for Fyttes VII and VIII) between the beginning of d3 verso (3352) and the beginning of BLO recto (4354). This is material for the thirteen pages d3 verso-e3 verso. BLO therefore becomes e4. The last line of its verso is 4511, and only 23 lines of text remain before the poem ends at 4564. The last quire therefore seems to have been e6, with text ending on e5 recto; we may presume a device on e5 verso, and that e6 was blank.

In reconstructing the first two quires it will be necessary to take into account certain features of the other early editions which are extant. There are two editions by De Worde, both in quarto, of which

the earlier (STC 13687, Child's text c), dated [1500] by Duff, op. cit., No. 361, survives in two leaves at the Bodleian Library. His later edition (STC 13689, Child's text b), which Dr. F. S. Ferguson dates [1506?], survives in the unique copy at the University Library, Cambridge. Its collation is A-D6 E8. It has a short title and a wood-cut, composed of three factotums, on A1 recto and De Worde's device McKerrow No. 19 on E8 verso. The text begins on A1 verso beneath a drop-title which, with the white lines above and below it, occupies the equivalent of nine lines of text. The third extant edition (not in STC, Child's text a) survives in the unique but imperfect copy in the National Library of Scotland, being the eleventh and last item in the celebrated volume which contains nine unique tracts by Scotland's first printers Walter Chepman and Androw Myllar. Robert Proctor (Jan van Doesborgh, Bibliographical Society, 1894, p. 24) attributed it to the press of Jan van Doesborch at Antwerp and dated it [1510-1515?]. Nijhoff-Kronenberg (Nederlandsche Bibliographie, &c., No. 3080) admits the probability of this attribution but points out that there can be no certainty since the Lettersnijder type in which it is printed was used by many Dutch printers. This edition (which I shall call the Lettersnijder edition) is most easily available in Dr. William Beattie's facsimile edition (The Chepman and Myllar Prints, &c., Edinburgh Bibliographical Society, 1950). I shall try to show at the end of this article that it is a reprint of Pynson's text. For the moment it is only necessary to note that the text begins on the first recto beneath a two-line short title and a reduced copy of the woodcut of the Yeoman which Pynson used in his Canterbury Tales of [1492?] and 1526 (Hodnett No. 1643).

We can now return to Pynson's edition. The text preceding c2 recto (stanzas 1-219) with cross-headings for Fyttes II, III, and IV must have occupied about 880 lines, of which 62 are required for c1. The remaining 818 lines are too much for 26 pages (806 lines), unless we suppose that some early pages had more than 31 lines. It seems more likely that they occupied 27 pages (837 lines) and that the gap of some 19 lines was filled either by a lengthy drop-title ornamentally set out, as in De Worde's later edition, or, more probably, by the woodcut of the Yeoman, as in the Lettersnijder edition. These hypothetical calculations cannot, of course, claim to be more than approximately accurate: but 19 lines of Pynson's text measures 90 mm., and the Yeoman cut is 85 mm. in height. In either event we have fourteen leaves (a8 b6, or, less probably, a6 b8), with the text beginning on the first verso, as in De Worde's later edition; and we may visualize the first recto as carrying a short title and either Pynson's device or the Yeoman woodcut, whether the latter appeared on the verso or not.

The complete collation, as here reconstructed, is therefore: a8 b-e6, 32 leaves; a1 recto, title with device or woodcut; a1 verso, text begins beneath woodcut; e5 recto, text ends; e5 verso, device; e6, blank.

R. Dickson and J. P. Edmond (Annals of Scottish Printing, p. 68) thought that the Lettersnijder edition had been printed in Scotland. 'The text of the piece', they wrote, 'is, with trivial variations, the same as that of the 'Lytell Geste of Robyn Hode', printed by Wynkyn de Worde, without date.' They then draw attention to certain features, to which I shall return, of the Lettersnijder edition, but do not comment upon them. That the Lettersnijder woodcut was copied from Pynson's Yeoman was first noticed by Frank Isaac (English & Scottish Printing Types, 1501-35, 1508-41, in his discussion of his figs. 92-3). Isaac knew that an edition of the Gest of Robin Hood printed by Pynson survived in the leaf recorded by Duff. 'We must infer,' he continued, 'that whoever printed the present edition had before him this fifteenth-century one or some other from Pynson's press in which the woodcut of the yeoman . . . was made to serve for Robin Hood, while the text, as it easily might, agreed fairly closely with De Worde's.'

The lines of text common to both the Lettersnijder edition and Pynson's as they survive, the one imperfect and the other in fragments only, are 3142-3161 and 3194-3351. A collation of the two editions within these 70 lines shows only three variants of any significance: 3194, L lawe / P (and De Worde) lawes; 3273, L (and De Worde) to / P wente to; 3311, L the / P this (De Worde that). On the other hand L agrees with P on 22 occasions where P differs from De Worde, the most important of them (I give the LP readings first) being: 3153, xl.dayes / twelue dayes; 3161, clothes were spredde / clothes spred; 3234, were so (o) noble and gode / noble were and good; 3261, Go nowe home shyref sayde our kynge / Go home thou proud sheryf; 3323, led hym to / lad hym home to; 3324, Bounde bothe fote and hande / I bonde both honde and fote; 3332, rode / a tre; 3333, C.li/ hondred pounde; 3341, harde (herde) / lady. Similarly, L differs from P in spelling and capitalisation much less frequently than L and P agree against De Worde. Isaac's inference is thus completely supported by an examination of the two texts.

A printer reprinting another man's edition of a popular text may be expected to do so as economically as possible. We have seen that Pynson's edition probably contained 64 pages, including a title-page, a page with his device, and a blank leaf at the end. The collation of the Lettersnijder edition has been reconstructed with, I think, complete certainty (though Dr. Beattie cautiously queries the number of leaves in the last signature) as A6 b6 c4 d6 e4, 26 leaves, of which the surviving copy lacks leaves 6, 7, part of 8, 13-18 and all after 20. It has no titlepage,

and the last page must have been fairly full. Its only concession to commercial embellishment is the woodcut which precedes the beginning of the text. It has no white lines above or below the cross-headings for the Fyttes, and the cross-heading for Fytte VI is printed on the same line as the last line of Fytte V. From A5 recto onwards the surviving leaves have 33 lines of type to the page, but by printing two lines of text, with a paragraph-mark between them, in one line of type the printer has got two or three additional lines of text into every page except b4 recto, into which he contrived to get no fewer than eleven additional lines. All these peculiarities are characteristic of a cheap reprint in which economy of paper is more important than appearance.

The book's most curious feature, however, is that stanzas 1-12 (with 1½ lines omitted from stanza 2) are printed as prose, beginning beneath the woodcut on the first recto and ending half-way down the first verso. The desire to save paper might reduce a compositor to so desperate a device as he approached the end of his copy, but hardly at its beginning; nor is it likely that at stanza 13 the Lettersnijder compositor received a sudden revelation that what he was setting was verse, not prose. The explanation of this curiosity must surely rather be that in setting the early pages the compositor failed to leave space for the initial woodcut, whether negligently or ignorantly or because the decision to insert it was not taken until after work had begun. The opening stanzas were therefore reset as prose in order to create the necessary space; and there was some consequent adjustment of the succeeding pages. A2 recto has 33 lines of type and 35 lines of text; but A2 verso is unnaturally short, having only 28 lines of type (and 28 lines of text), while the pages of A3, 4 show an unusual regularity in that each of them has 32 lines of type and 32 lines of text. It seems probable that the compositor by his resetting made a little more space than he actually needed; but he could not adjust his type-pages beyond A2 verso since A3, 4 (the inner half-sheet of the quire), being ready for the press before the outer sheet, had already been partly or wholly printed off.[2]

Notes