This study, as noted in the preface, is not primarily an examination of the breviary identified in the title or of the liturgy it contains but rather a vehicle for examining methods of cataloguing early printed books. The preface and prefatory lists (xi-xxii) constitute the first two chapters, explaining the choice of the 1493 printing of the York breviary as the subject and providing a chronological list of the catalogues and other reference works published since 1715 which describe this liturgical book. Chapter 3 sets out various types of discrepancies in descriptions of early printed books, illustrating the confusion that cataloguing errors may generate: "One simple misstatement or misreading...can cause ghost volumes to appear authentic" (4). Chapter 4 provides a physical description of the York breviary printed by Johannes Hamman in Venice in 1493 and an account of the apparent discrepancies in the attempts to catalogue the book, which survives in a complete copy formerly in St. Helen's Church, Ashby de la Zouch, and now in the Bate Collection in the library of Loughborough University (13), a nearly complete copy in the Bodleian Library, Oxford, and four sets of fragments out of an unknown number of copies originally printed. Chapter 5 presents the liturgical context, offering a discussion of the York Use, late medieval additions to the liturgical year at York that might have affected the contents of books produced for that Use, and certain chants and readings that distinguish the office of Thomas Becket in York books from that in Sarum books. On the basis of features found only in particular York manuscripts and printed books, an attempt is made to identify the manuscript breviary that the printer may have used as his model for the 1493 printing. Chapter 6 presents an overview of York manuscripts and printed books, and the last two chapters examine problems in modern reproductions of the 1493 breviary and offer recommendations for the cataloguing of printed liturgical books. Three extensive appendixes provide: (a) a detailed inventory of the breviary's contents as it appears in the extant copies and microfilm, microfiche, and digital reproductions; (b) descriptions of the manuscripts and printed Office books of the York Use; and (c) a list of resources for early printed books.
The volume is a valuable source of information for the study of York Office books and a useful discussion of the potential for confusion in the cataloguing of early printed books. It shows how problems resulting from mistakes in the printing process can be compounded by inadequate cataloguing, leading to further confusion as a description of an early printed book in one catalogue becomes the source on which subsequent catalogues depend for their descriptions of the same book. While the general points made in this discussion are valid, some of the particular discrepancies used to illustrate these points are themselves excellent examples of how ghost volumes can appear as a result of misread catalogues.
One instance of this is in the discussion of the folio numbers in the printed breviary. In the two extant copies of the 1493 printing, the section ending on folio 284v is followed immediately by the Sanctorale on folios numbered 301-478. Noting that the 1917 description by E. Gordon Duff in Fifteenth Century English Books uses the numbers 285-462 for the folios numbered 301-478 in the known copies of the breviary, and noting that the 1932 description in vol. 5 of the Gesamtkatalog der Wiegendrucke [GW] uses the numbers 1-178 for these folios, Hughes writes, "We have to postulate two other copies, from two different printing runs," suggesting that the printer might have changed the folio numbers of the Sanctorale twice, originally starting this section at folio 285 but finding that "beginning the Sanctorale with f.301, as in the two known copies, or with f.1, as in the book described by GW, allows for additions" before the Sanctorale or allows it to be used in revisions of the breviary in which the foliation of the preceding sections may differ (43). Hughes concludes, "If there were indeed two or three differently numbered Sanctorales, we probably do not need to propose two or three different prints, least of all in different years. But we must contemplate the possibility of separate print runs in 1493" (66). There is, however, no need to imagine multiple print runs or to postulate copies that had different folio numbers, since any apparent discrepancies in the foliation used by Duff, GW, and the known copies can be explained: Duff, in writing 285-462, and GW, in writing 1-178, were not using the folio numbers printed in the volume they were describing; instead they were counting the actual folios of the Sanctorale in the volume as a whole or as a separate section. The entries in GW in fact distinguish between folios actually counted ("Blätter," abbreviated "Bl.") and folios as numbered ("gezählte Blätter," abbreviated "gez."), and accordingly the description of the 1493 breviary has the Sanctorale listed not only as "Bl." 1-178 but also as "gez." 301-478. Duff and GW both indicate which copies or fragments of the 1493 breviary were known to them, and there is no basis for the suggestion that copies were available to those cataloguers which are no longer extant or that there ever existed copies with folio numbers different from those in the extant copies.
Interesting as the point may be that "each early printed book must be treated like a manuscript; that is, as a unique copy" (3), there does not seem to be any compelling evidence that there were multiple print runs of the 1493 breviary or differences among the individual copies resulting from stop-press changes. The obvious differences between the extant copies have to do with the binding process or with the insertion or removal of pages, ownership marks, and other changes made after the printing of a book. One difference that might appear to result from different print runs is the colophon printed in red ink, which can be read clearly on folio 478v of the Bate copy, reproduced by Hughes as plates 4.2 and 4.3. The assertion that "the Bate copy has a textual colophon not present in the Bodleian copy" and the suggestion that "the printer must have changed the last pages after pulling a copy from the press" (27) seem to imply that the Bodleian copy had been printed before the colophon was set. But there is a more likely explanation for the apparent difference between the two extant copies of folio 478v: the colophon simply does not show up in the Bodleian copy or at least in the digital reproduction of that copy because it is the part of folio 478v printed in red. The apparent difference between the two copies may be due to the inking process, the fading of red ink on the last page of the Bodleian copy, or the quality of the reproduction rather than the actual typesetting of the page.
One other apparent difference between the extant copies of the breviary which might be an argument for separate print runs is indicated by the comparative inventory in Appendix 1: according to this inventory (105), the text ending on folio 198r of the Bodleian copy ends on folio 198v in the Bate copy. But there is no further discussion of this reported difference between the two copies. If the inventory is correct in indicating that the text ending on folio 198r in one extant copy ends on the following verso in the other, this difference between the two extant copies needs to be emphasized and discussed in detail, since it would be the only difference that pertains to the actual typesetting of the breviary. Of the cataloguing discrepancies which are discussed in detail in this study, some may be the result of differences between the extant copies as they have come down to us, but none seems to be the result of differences due to multiple print runs or stop-press changes.
Cataloguing Discrepancies is a practical guide that offers researchers in liturgy and other fields helpful illustrations of the types of problems they may find in working with early printed books and with reproductions or descriptions of these books. In producing a study of this sort whose specific purpose is to clear up confusion that errors in the printing or cataloguing of the 1493 breviary might cause, particular care ought to be taken in order to avoid introducing new errors that could cause further confusion. Hughes acknowledges this in the preface by writing, "We hope that our work will clarify rather than blame, and will not result in too many additional difficulties" (xiii). Some oversights that could create confusion are corrected below: xvii: for "Catalogue of Books Printed in the XVIth Century" read "Catalogue of Books Printed in the XVth Century"; 17: for "Although its location in Ashby de la Zouch is not stated, the earliest certain reference to the Bate copy is in ELB.1924" read "The earliest certain reference to the Bate (Ashby de la Zouch) copy is in Duff (1917)"; 33: for "80ff." read "180ff."; 35: for "273-281 277 282 278 284" read "273-280 277 282 279 284"; 36: for "277-279 were retained" read "277 and 279 were retained"; 41: "(which is in fact 479 & 480, and not 301 + 179, which results in 480 & 481)" should be deleted since the catalogues have 300 + 179, not 301 + 179; 51: Figure 4.5, showing the imposition of one side of the sheet containing signature 3, should be redone, since the shown sequence of pages (8 - 9 - 16 - 1 on the top and 5 - 12 - 13 - 4 on the bottom) does not reflect the arrangement seen, e.g., in Figure 4.3 in the diagram for the corresponding side of signature 35, and the pages indicated for four of the headings are incorrect: "Dominica quarta," "Quarta adventus," "In festo nativitas xpi" and "In vigilia" appear respectively on pp. 4, 13, 9, and 8 of signature 3, not pp. 8, 9, 13, and 4, and the pages with the headings transposed by the printer are 1 and 13, not 1 and 9; accordingly, on p. 50, "pages 1 and 9 of signature 3" should read "pages 1 and 13 of signature 3"; 102-7: there should be a corresponding number of entries in columns 3- 7 of the inventory table except in those instances where there are gaps in the reproductions or the individual copies; as it stands the table suggests that there are more differences between the individual copies than there are; 105: (and 111) for "De dominicis ordinacionis" read "De dominicis ordinandis"; 106: for "(o.257)(o.257)" in the Bate column, read "(o.257)(o.262)"; 108: (in n. 16) for "o. 313" read "o. 412"