University of Virginia Library

Variants in the 1479 Oxford Edition of Aristotle's Ethics
by
Dennis E. Rhodes

Not since the days of Falconer Madan and E. Gordon Duff have the results of detailed research on early Oxford printing been published. Now, as the British Museum's cataloguing of English incunabula proceeds, it is possible to give an account of the variants which occur in the second book printed at Oxford—Aristotle's Ethics in the Latin translation of Leonardo Bruni. The book is a small quarto of 174 leaves, the first blank, with signatures a-x8 y6 and with 25 lines to a page. It is not relevant to the present note to argue whether or not Theodoric Rood was the printer of the first three Oxford books: controversy which has raged for years has never settled this point. Meanwhile it does not appear ever to have been noted in print hitherto that the Aristotle of 1479 exists in different states.[1] The variants which I have been able to discover in the British Museum's two copies seemed to be sufficiently numerous and interesting to justify a detailed examination of all known copies. I am very grateful to all the librarians who have answered my questionnaires on this subject.

First, however, it is necessary to re-examine Duff's entry for this book in respect of the number of extant copies. Of the nine copies which he records, only seven can be traced today, with the addition of an eighth in the Broxbourne Library belonging to Mr. and Mrs. Ehrman. There never was a copy in Norwich Public Library, although the Broxbourne copy was formerly in the possession of Norwich Cathedral, which may account for Duff's error. The Earl of Pembroke's copy, sold in 1914, is now in the Chapin Library at Williamstown, Mass. Lord Amherst's copy, wanting four leaves and (says Duff) sold in 1908, I have not traced.[2] Further, the Gesamtkatalog (2373) lists a copy at Cambridge University Library which is in fact a single-leaf fragment; it repeats from Duff

the error of locating a copy at Norwich; and it unaccountably locates one at the Henry E. Huntington Library in California, which again is untrue. Extant copies other than fragments are therefore as follows: (1) British Museum, C.2.a.7; (2) British Museum, Grenville 7930; (3) Bodleian Library, Oxford, S. Seld.e.2; (4) All Souls College, Oxford, L.R.4.e.14; (5) Chetham's Library, Manchester; (6) John Rylands Library, Manchester, no. 15969; (7) Chapin Library, Williamstown, Mass.; (8) Broxbourne Library (imperfect).[3] In the account which follows, these copies are listed thus: BM1, BM2, Bd, AS, C, JR, W, and Bx.

Variants in Reset Forme

Inner sheet a (inner forme)

• 2nd setting: BM2
• 1st setting: BM1, Bd, AS, C, JR, W [this sheet is wanting in Bx]
• a4 verso
• 5 eutrapelia:] eutrapeliɐ
• 5 ſupabūdancia] ſupabundācia
• 18 tum comoditatē] tum comitatē
• 22 copiā lingue] copiaȝ lingue
• 23 Superabundācia inquit] Superabundancia īquit [in original setting the 'u' of 'īquit' is badly broken]
• 24-25 ſcurrilitas . . . et qui illam] ſcurrilitas . . . et qui illam [in second setting the final 's' of 'ſcurrilitas' is a peculiar sort with tall, angular arms, sloping to the right]

a5 recto

• 1 ſcurca: qui autē] ſcurra' qui aute3 [in original the 'u' of 'aute3' is broken]
• 5 abhoret] abhorret [in original the first 'r' of 'abhorret' is gothic, the second roman; in resetting the single 'r' is roman; the spelling is an example of a new error introduced by the compositor into the resetting]
• 7 vidētur . . . in lingua] videntur . . . ī lingua
• 18 cōtine] contine
• 20-21 ſemp la/tine bonum] ſemp lati/ne bonum
• 21 quidē:] quidem.
• 22 tamē] tamen
• 25 omne animal ſubſstancia ſit] ome animal ſubſtancia ſit.

Press Variants

Inner sheet n (outer forme)

• Corrected: BM2, Bd, C, W, Bx
• Uncorrected: BM1, AS, JR
• n3 recto
• 25 ſed vt ſanitas ſic ſapiēcia felicitatem. cū] ſed vt ſanitas ſapiencia felicitatem. cum

n6 verso

• 4 borum et vituperabiliū.] bo rm et vituperalium.
• 9 propter racionem] racionem

Outer sheet q (inner forme)

• Corrected: BM2, Bd, AS, C, JR, W
• Uncorrected: BM1, Bx
• q2 verso
• 1 ſupins] ſupins [(u) has dot over the 'n']

• q7 recto
• 6 no fāiliaritate neq3 mictua &c.nv;uerſacōe] nō miliaritate neq3 mitua cōuerſacōe
• 22 ingenuitas] ingenntas [(u) has dot over second 'n']

Observations

(1) Two pages of the inner forme of inner sheet a turn out to have been completely reset.

(2) All the remaining variants occur by press correction in the second half of the book, and in each case both pages of a forme are affected. However, neither of the formes here involved has been reset: indeed, in the press correction the setting of as few letters as possible in each of the lines affected has been changed.

(3) BM1 is the only copy which shows the uncorrected state of n and q combined with the original setting of a; and BM2 the only copy showing the resetting of a (combined with the corrected states of n and q): this is perhaps purely fortuitous but the copy may have been made up from late-printed sheets. Three copies show a mixture of the original setting of a and the corrected states of n and q. Two copies show a mixture of original quire a, uncorrected n, and corrected q. Bx (defective in quire a) is the only copy which shows (u) in q and (c) in n.

Conclusions

(1) It is clear from the arrangement of the variants that the book was printed in formes of only two pages. It now occurs to one to ask whether the sheets were cut in half before printing or whether each full sheet was printed on the one half with a two-page forme and then put through the press a second time to print the other half with its appropriate two pages, the process being repeated in the perfecting. In three copies examined to check this interesting point (BM1,2 and Broxbourne, the latter missing five leaves in quire a), in some one sheet of each quire in the series a-f there is an impossible combination of watermarks or of their absence; and this evidence shows that quires a-f were printed by cut sheets. On the other hand, in these three copies all the remaining quires exhibit normal watermarking, this indicating that printing by full sheets (each sheet going through the press four times) was the method in quires g-y.

(2) It is not likely that the inner forme of a4 was reset merely to correct its two small misprints (a4v,l.18 and a5r,l.1). Indeed, the fact that this, the only forme to be reset, is also the forme which would normally be the first to be printed, strongly suggests a decision taken during its printing to increase the size of the edition. Other explanations are possible: for instance, the forme might have been hopelessly upset during printing, and have had to be reset; or

insufficient copies of it might have been printed, and this error have been noticed before or during the printing of the outer forme. But then one would have to ask: why did the accident happen to this of all formes? The balance of probability points to a decision to increase the size of the edition, because no other interruption involving resetting would be specially associated with the printing of a first forme. It would be unsafe to assume offhand that the state of this forme in which the misprints are corrected is the later. The compositor makes so many misprints in the remainder of the book, that he might well here have introduced new errors in resetting a forme which he had previously set correctly. But the fact that one state survives in six copies and the other in only one may perhaps indicate that the single-copy state is the later; and this is also the one in which the misprints are corrected. Lastly, a sudden discovery by the printer that the possible market for his book was larger than he had hitherto foreseen (or an unexpected order for further copies from his patron or agents, etc.) is particularly liable to occur with a book in considerable demand. Such, certainly, was Aristotle's Ethics in fifteenth-century Oxford. Other similar cases will readily be called to mind, as for instance that of Fust and Schoeffer's Mainz Psalter of 1457.

(3) No variants have been discovered in the remainder of the book, although similar misprints occur throughout, even in the non-variant formes of quires n and q. It would be useless to speculate what accident led to the correction of misprints on only two formes of n and q. It is noticeable however that each page of these formes contains a corrected misprint. Evidently in each case whoever noticed the first error (it need not necessarily have been the careless compositor) was thereby stimulated to find the second; but he did not notice or understand the misprint 'mitua' for 'mutua' on q7r,l.6, and this was only made worse by the substitution of 'mictua'.

Notes