University of Virginia Library

In writing about the phenomenon of 'signing by the page' (Studies in Bibliography 48 (1995), 259- 268, esp. pp. 266-267) I referred to the 1743 Oxford quarto Bible, which I was able to record existing in two states: (i) totally innocent of cancellation, represented by one exemplar only, Bodleian Bib.Eng. 1743.d.1; (ii) containing at least 63 (vere 62) cancellantia, represented by three exemplars, British Library 3050.ee.8(2), National Library of Wales BS185.d43(4to), and Durham University Bamburgh Castle L.iii.7-10.[1] This edition is the first Bible to be printed by Thomas and Robert Baskett in succession to their father John—the imprint reads 'OXFORD: Printed by THOMAS BASKETT and ROBERT BASKETT, Printers to the UNIVERSITY. M DCC XLIII.' The volume is gathered in eights, as is usual for English Bibles in quarto, and the gatherings are signed, in conventional fashion, $1-4—i.e. leaves $5-8 are unsigned. In state (ii) 61 of the 62 obvious cancellantia occur in the $5-8 sequence (the other is $3 in a four-leaf gathering) and are stigmatized by being signed by the page—i.e. according to the page within the gathering occupied by their recto; thus cancellans $7 is signed '$13' in such a system, $7r occupying the thirteenth page in the gathering. The system of signing by the page can therefore be taken to be a guide to replacing unsigned leaves for agents apparently not accustomed to inferring foliation (and it might be noted that, as here, Bibles are customarily not paginated, so that that form of location was not available either). At the time of writing I opined that 'there is no reason to suppose that the cancellation in this volume is confined to the second half of gatherings. Indeed, one would expect that there would be about the same number of cancellantia in the first halves of gatherings, signed in the conventional manner—i.e. as their corresponding cancellanda' (p. 266). In fact, disturbances to the patterns of watermarks, chainlines and tranchefiles in the BL exemplar suggested that there were at least a further 51 cancellantia in the $1-4 sequence, though since that exemplar has been rebacked some leaves may have been repositioned, thus erroneously suggesting cancellation; additionally some cancellantia may in that exemplar agree in the three criteria with their corresponding cancellanda

and consequently may have escaped identification. Nonetheless the likelihood was that the $1-4 sequence too would contain 60-odd cancellantia.

What at that stage seemed required was the capacity to bring together the Bodleian exemplar and one of the exemplars in state (ii) in order to carry out a textual comparison. The opportunity for such a comparison was later afforded by the presence—not initially realised—of a second exemplar in the National Library of Wales, BX5145 A4 d43(4to), which is in state (i).[2] NLW BS185 (the exemplar in state (ii)) is defective, lacking nine leaves (3R2 3R8 3S1 4A8 4E7 4E8 4H8 4I1 4I2), none of which, however, are cancellantia in the other exemplars of state (ii); discounting those leaves but taking account of B1 and B2, which in BS185 are from state (i) but in other exemplars of state (ii) are cancellantia, the comparison reveals that, as anticipated, there are 70 identifiable cancellantia in the $1-4 sequence. Hence the total number of cancellantia in the 1743 Bible is at least 132.

In the process of cancellation the opportunity was taken in a few instances to simplify the binder's task by making changes in conjugacy, though on the evidence of surviving exemplars by no means all opportunities for such simplification were taken (it remains possible, however, that binders have separated cancellantia which were printed as conjugate pairs).

Exemplars in state (i) collate: πA2 A-3P8 3Q4 3R2; (NT:) χ3Q4 χ3R8 3S-4H8 4I2 State (ii) in its 'ideal' form is probably best represented by the following formula: πA2 A8(±1,2; -5,6 +5.6; ±7) B8(±1,2; -3,4 +3.4; ±5,7,8) C8(±1,3) D8(±6,8) E8(±3,4,6; -7,8 +7.8) F8 G8(-4.5,6,7 +4.7, 5.6; ±8) H8(±2,4,7) I8(±1; ±4.5) K8(±1.8; ±2; -3,4 +3.4) L8(±1,3) M8(±2; -6,8 +6.8; ±7) N8(±2,6) O8(±5) P8(±1) Q8(±1) R8(±7) S8(±5) T8(-3,4 +3.4; ±5,6) U8 X8(±6,7,8) Y8(±1,2,5,7) Z8(-1,2 +1.2) 2A8(±1,2,4,6,8) 2B8 2C8(±4,5,8) 2D8(±7) 2E-2F8 2G8(±1.8; ±2,4) 2H8(±1.8) 2I8(±5,7) 2K8(±3,4,8) 2L8(±2,3.6,4) 2M8(±4,5,6) 2N8(±3,8) 2O8(±3.6) 2P8(±8) 2Q8(±1.8; ±4.5) 2R8(±5) 2S8(±1) 2T8 2U8(±4,7) 2X8 2Y8(-2,3 +2.3; ±5,8) 2Z8(±1) 3A8(±6) 3B8(±2,6) 3C8(±1,4) 3D8 3E8(±4) 3F8(±4) 3G8 3H8(±4,6) 3I-3N8 3O8(±3) 3P8(±1) 3Q4 3R2; (NT:) χ3Q4(±3) χ3R8 3S-4A8 4B8(±1.8; ±6) 4C8(±2,4,7) 4D8(±1,3) 4E-4F8 4G8(±2) 4H8(±5) 4I2 (Note that the cancels at I8 and R1 have been excluded from the formula; both were corrected at press and so are found with state (ii) readings either as integral leaves or as cancellantia. Similar instances may be awaiting discovery in exemplars as yet unexamined.)

The extent of the cancellation in the 1743 Bible is in itself a matter of surprise: 132 cancellantia among the 622 leaves—i.e. 21.2% of the leaves in state (ii) may be cancellantia. I am not aware of any sizeable publication approaching this level of cancellation, whatever the period or country, the closest that I can come being the quarto issue of Baskerville's edition of

Ariosto's Orlando Furioso, 1773, which contains what Philip Gaskell regards as 'the amazing total' of 66 cancellantia—this in a total of 901 leaves, a rate of 7.3%, or not much more than a third of that in the 1743 Bible.[3] But the Baskett brothers and John Baskerville were quite different in their customary approach to their texts, even if Baskerville did not manage to secure the accuracy that he strove for; and while Baskerville was producing, with obvious concern for appearance and textual accuracy, a foreign-language text which might be expected to be available for sale for a number of years, the Baskett brothers were producing, with no obvious concern for appearance, what might be considered a run-of-the- mill text in English in which accuracy was not by any means a customary concern and which in the normal course of events would be exhausted within a couple of years. The cancellation in the 1743 Bible should also, therefore, be seen in the context of a particular tradition of textual transmission and in the context of Bible-publishing at Oxford.

Bibles as a genre show scant regard for textual accuracy, and in the eighteenth century probably only the 1762 Cambridge folio/quarto (DMH1142/1143)[4] and the 1769 Oxford folio/quarto (DMH1194/1196) were actually edited, as opposed to being set from whatever exemplar was at hand, regardless of the state of its text and apparently without any thought of reproducing it literatim. It can confidently be asserted that the textual condition of the Authorised Version (first published in 1611—DMH309) is one of virtually unrelieved progressive deteriora- tion. The deplorable textual condition of the AV was a matter of concern as early as 1659, when William Kilburne—in Dangerous errors in several late printed Bibles to the great scandal, and corruption of sound and true religion—could claim to have 'discovered . . . many thousands'. Similar complaints continued to be voiced well into the nineteenth century—for example, Thomas Curtis in 1833 calculated that there were 'upwards of Eleven Thousand' intentional departures from the 1611 text, a figure 'not at all including the general alterations of the orthography or minute punctuation'.[5]

One difficulty in assessing the various claims is knowing what standard

to judge a particular edition against. The 1611 text has obvious general authority, being closest to the translators' manuscript, which is presumed to have served as the immediate copy, but in some eyes it has total authority, even in accidentals—this despite the fact that as a text it is a composite work translated from sources themselves translations of indeterminable authority and that as a publication it is the product of an early-seventeenth-century London printing house. In other words specific substantives may be subject to challenge and the accidentals have no particular warrant.

An exemplar of 1611 cannot have been the source of the corrections made in 1743, since some of those corrections are in the marginal references (which had become more numerous by 1743) and in the marginal dates (which were not included in an English Bible until 1679, at Oxford (DMH 744/745)). In both states, by and large, 1743 represents a modernisation of 1611's spellings (of the breake/break, hee/he kind), though, in resetting, state (ii) tends to take that process further; otherwise most of the changes in state (ii) are either corrections of obvious errors in (i) or errors introduced into (ii)—indeed, state (i) is more often in agreement with 1611 than is state (ii). The changes in the marginalia suggest a printed source, while the changes in the text could just as well have been made independently of any such source by a careful reader in the printing house. In other words I have not been able to establish that the corrections made in 1743 were effected in order to bring the text into strict conformity with an identifiable printed source. We must wonder, therefore, what the motive for making the corrections was.

Though the differences between the two states can be isolated it is not a straightforward matter to distinguish needed corrections identified in the process of proof-reading from indifferent variants introduced in the process of resetting. A few random textual comparisons (confined to the early gatherings) show that only very occasionally—when the differences are few—can the reason for the cancellation be established. Thus in the marginal notes at the head of column 2 on A5r state (i) reads '2281. | 2247. | d 1 Chr. | l. 19', state (ii) '2281. | d 1 Chr. | l. 19 | 2247.'; since the only apparent change on A5v is the introduction of a space in the marginal notes of no significance the reason for the cancellation must have been to effect the reordering on A5r. Elsewhere, where the differences are more numerous, it is difficult to be categorical in distinguishing the intended from the incidental, as a comparison of the two states of B5, comprehending Genesis 27- 29, reveals (asterisks are used to isolate the element of variation):

Changes in punctuation (1, 3, 7, 10, 13, 14, 16) improve the rhetorical effect of the text and bring it into general conformity with the text as a whole (e.g. by having a comma after the frequent 'behold')—but the resulting text is by no means consistent in this respect. The replacement of 'lift' by 'lifted' (2) is part of the process of modernisation, which sees 'borne' replaced (erroneously according to modern usage) by 'born', as in 'I have borne/born him'. The supplying of an apostrophe (5, 6) should be set against an omission (8) and against the fact that on the two pages there are about 20 other possessives—including one proper name (Abrahams)—in which the apostro-

RECTO

                               

Chapter:Verse	State (i)	State (ii)
1. 27:37	and what shall I do now unto *thee˄* my son?	thee,
2. 27:38	And Esau *lift* up his voice	lifted
3. 27:43	Now therefore, my *son˄* obey	son,
4. 28:Chapter head	*Padan-aram,*	Padan- aram.
5. 28:Chapter head	*Jacobs* ladder	Jacob's
6. 28:Chapter head	*Jacobs* vow	Jacob's
7. 28:4	give thee the blessing of *Abraham˄* to thee, and to thy seed	Abraham,
8. 28:5	*Jacob's* and Esau's mother	Jacobs
9. 28:9	Mahalath the *daughter* of Ishmael	daughters
10. 28:12	and *behold˄* a ladder set upon the earth	behold,
11. 28:13	And behold, the *LORD God* stood above it,	LORD
12. 28:13	and said, I am the *LORD* of Abraham	LORD God
13. 28:14	and in *thee˄* and in thy *seed˄* shall	thee, / seed,
14. 28:15	And *behold˄* I am with thee	behold,
15. 28:15	I have spoken to *thee*	the

VERSO

     

Chapter:Verse	State (i)	State (ii)
16. 29:10	And it came to *pass˄* when Jacob saw Rachel	pass,
17. 29:24	*handmaid*	hand-maid

phe is not supplied in resetting. The hyphenation of 'handmaid' (17) should be contrasted with its retention as one word in the resetting of B6r. The resetting corrects a final mark of punctuation (4), but it also introduces two errors of its own (9, 15). The remaining changes (11, 12), though in isolation not obviously needed corrections, do in fact bring the text into conformity with the tradition of 1611 and may well be the occasion for the cancellation. Nonetheless, given the general textual state of the Bible in the middle of the eighteenth century, these do not seem to be the kinds of error that would cause even a moment's pause.

The cancels cluster in the early gatherings (63—virtually half—are in the first alphabet), but, as the variations in B5 demonstrate, it is difficult to know why—i.e. it would be impossible to determine with any confidence whether the decline in the scale of cancellation as the volume progresses results from a greater tolerance of error or from improved standards of setting or initial proof-reading, since to modern eyes most of the variations are indifferent in the absence of a known yardstick against which the 1743 text may have been judged.

Another context within which to view the 1743 edition is that of Oxford quarto Bibles. Assuming that no edition has disappeared completely, there were five Oxford quartos in the preceding decade (1733, 1736, 1738, 1739, 1740) and seven in the following (1744, 1746, 1747, 1749, 1752 (two editions), 1753). Thus the 1743 appeared in the middle of a 21-year period in which

there are 13 surviving Oxford quarto Bibles, an average of one every 19 months or so. In other words the Baskett brothers can have anticipated printing another edition in the same format within a period not exceeding two or three years. (In the event a new edition in quarto must have been set in train very soon after the completion of the 1743.) In that light the extensive correction in 1743 appears all the more puzzling: Why go to the trouble of inserting so many cancellantia in a volume which they could anticipate disposing of practically straightaway and therefore be starting to set a new edition of almost immediately? In view of the general textual state of the Bible in the eighteenth century and of the nature of the corrections themselves I cannot better my earlier suggestion: that in producing their first Bible Thomas and Robert Baskett were anxious to improve on their father John's standards. John was possibly no worse than Bible-printers of the preceding century, but his 1717 Oxford folio—the so-called 'Vinegar Bible'—had earned the additional sobriquet 'a Baskett-ful of Errors'. If such was their intention, Thomas and Robert were unsuccessful. Or could it be that the brothers were reacting to the resumption of Bible printing at Cambridge? In 1743, after a lapse of sixty years, the Cambridge University Press published a Bible, a duodecimo on two qualities of paper (DMH1063); unlikely as it may seem, could Thomas and Robert have been attempting to demonstrate that their product was textually superior to their competitors'?

Commonly eighteenth-century English Bibles are bound with a Book of Common Prayer, a metrical psalter (almost invariably the version of Sternhold and Hopkins) and perhaps—particularly with folios and quartos—an Index to the Holy Bible or—particularly with octavos—various devotional works. Surviving exemplars of the 1743 Oxford quarto are therefore unexceptionable in that two (NLW BX5145 and BL) are bound with an Oxford quarto Book of Common Prayer with the same imprint and all five with a quarto Index, inferentially from Oxford (like most, if not all, eighteenth-century indexes which are bibliographically distinct volumes, this edition has only a caption title and no colophon). There is no corresponding Sternhold and Hopkins: at this period when Oxford Bibles have one bound in it is an edition printed by one or other of the London stationers. In both the Book of Common Prayer and the Index there are also two states, corresponding with those for the Bible; in state (ii) both have cancellantia signed by the page. The binder's volumes are 'pure': state (i) of both is found only with state (i) of the Bible, state (ii) only with state (ii)—thus NLW BX5145 contains all three in state (i), BL all three in state (ii).

The Book of Common Prayer collates in state (i) a8 A-F8 G4; in state (ii) there are cancellantia at least at A6 (signed 'A11'), F5 (signed 'F9') and F6 (signed 'F11') and probably at D1, E3 and F3, which have press figures in state (i) but not in state (ii), thus giving the provisional collation a8 A8(±6) B-C8 D8(±1) E8(±3) F8(±3,5,6). The Index collates in state (i) a-d4; in state (ii) there are cancellantia at least at a3 (signed 'A5') and a4 (signed 'A7') but no other patent indications of cancellation, so that state (ii) may well be represented as a4(±3,4) b-d4. I have not attempted at this stage to establish the

full extent of the cancellation in the two works but have confined myself to a comparison of the five leaves identified as cancellantia by their signatures, a comparison which was sufficient to establish the nature of the variations. In the Book of Common Prayer the changes are not of the kind that might have been expected and that are regularly found in the first half of the eighteenth century—i.e. those occasioned by deaths, births and marriages within the royal family—but almost exclusively variations in punctuation and capitalisation, with the occasional correction of an erroneous reference. In the Index the changes are of a similar charac- ter: variations in accidentals, with the odd correction of an apparent error (like 'save' for 'spare'). As with the Bible, in both works it may be that the correction of the odd error was the occasion for the cancellation and that the variations in accidentals are 'indifferent', being a product of the resetting.

Bibliographically speaking, the major point of interest in the 1743 Oxford quarto Bible, Book of Common Prayer and Index is the method of signing cancellantia in the $5-8 sequence—i.e. 'by the page'. As I have already suggested, such a method does appear to be designed to facilitate the process of effecting cancellation by people unaccustomed to inferring the signature of an unsigned leaf (note the more customary—though by no means universal—practice of signing cancellantia by the leaf, according to the position of the corresponding cancellanda, whether signed or not, so that cancellans $7 in the 1743 Bible would be signed '$7'). Admittedly Bibles are normally not paginated, thereby removing an alternative form of reference for those unfamiliar with signatures; on the other hand, with the practice of dividing the Bible into verses, begun in English in 1560 with the publication of the first edition of the Geneva Bible (DMH107), it contains its own reference system—one need only look at the headline to know where a particular leaf belongs. Hence it is all the more puzzling that signing by the page should be adopted for a Bible. It has been tempting to believe that the method was designed to aid owners of volumes to effect the cancellation themselves; such a supposition was at least attractive in the case of the Book of Common Prayer. But here the nature of the corrections renders such a supposition untenable: owners of Bibles can have been no more than indifferent to the changes made on most leaves. Moreover, the cancellation of conjugate pairs and cancellation in which conjugate pairs replace disjunct leaves mean that the process can have taken place only before the sheets were bound—i.e. the practice of signing by the page now appears to have been employed for the benefit of members of the trade, either in the printing house or in the binding shop.

The textual changes in the 1743 Oxford quarto Bible, Book of Common Prayer and Index are themselves of little consequence: most are such as to cause no more than a momentary stumble in reading (silently or aloud). But their very inconsequentiality confers on them an interest and gives rise to a number of questions which have been raised in the body of this essay: Why bother to make the changes in the first place, particularly given the general textual state of the Bible in the middle of the eighteenth century and the

speed with which editions of quarto Bibles published from Oxford were exhausted? What is the temporal relationship between the two issues (noting that they are virtually 'pure', on the evidence of the small number of surviving exemplars examined)? Why were cancellantia in the $5-8 sequence signed by the page? For whose benefit was the system employed, especially given that the privileged printers used the system for bringing the Book of Common Prayer up to date? I would be pleased to be enlightened on any of these points.