Bibliographical Terminology: Excerpt from an Imagined Catalogue

The following is a part of an introductory chapter (‘INTRODUCTORY NOTE’ pp. xi–xix) that I drafted for an imagined catalogue of 1977 copies of The Silmarillion, describing it as a ‘rather lengthy exposition on bibliographical terminology in respect to book-collecting generally, and on this writer's use in this catalogue in particular’. Serious stuff.

This imagined catalogue of my own books currently exists in draft only, and will likely remain so (forever). I share part of it here (slightly modified) to provide some insight into my understanding and use of bibliographical terminology, and possibly as a reminder to myself, too. It is worth stating that what follows is in reference (unless clearly stated) to 1977 UK copies of The Silmarillion only. The title is new to this post.

 

BIBLIOGRAPHICAL TERMINOLOGY

WHEN describing and cataloguing modern books the tricky matter of terminology inevitably arises, unless one chooses to tactfully ignore it. In a world of mass-produced machine-pressed books—as it emphatically was in 1977 when hundreds of thousands of copies of The Silmarillion were to be printed—what terminology to best describe the differences between individual copies and the relationship they have to each other? On this count the views of bibliographers, publishers, booksellers, and book-collectors are at variance.

Twentieth-century bibliographers like Bowers and Gaskell grappled with applying the terms of the past—those applicable to the hand-press period—to the “modern” machine-press age. A close reading of Bowers' Principles of Bibliographical Description (1949) and Gaskell's A New Introduction to Bibliography (1972)—along with Tanselle's more recent reaffirmation, in his Descriptive Bibliography (2020), of many of the bibliographical principles they put forth—will show how difficult, complex, and sometimes contradictory-sounding this task can be. There is not always a tidy one-size-fits-all solution to every given bibliographical scenario. A cursory glance at the multiple copies printed by WILLIAM CLOWES & SONS (printers), to take just one example, should illustrate some of the specific difficulties encountered with describing The Silmarillion in particular.

I will herewith attempt to define the use of edition, impression, issue, and state.

Edition: All 1977 copies of The Silmarillion are first editions; bibliographically they all form part of one original edition. As a consequence, ‘edition’ is not a word I often use to differentiate between copies. It might be argued that BOOK CLUB ASSOCIATES (BCA) published their own distinct edition in 1977; a not uncommon designation. However, the likely simultaneous (or near-simultaneous) printing of their sheets alongside sheets for use by the primary publisher GEORGE ALLEN & UNWIN (GA&U) suggests a closer relationship than edition describes. In fact, this clearly falls under Gaskell's description of ‘one act of composition’ i.e. they both belong to one original edition. BCA would go on to produce their own distinct edition, in 1978, which clearly does call for edition to be applied.    

From a publishing perspective, aside from the various paperback and book club editions published throughout the 1980s—all of which do, almost certainly, constitute separate editions (although some of these may fall under Bowers' ‘subsidiary edition’ definition)—all copies of The Silmarillion published through to 1999 are textual first editions too. At this time, errors—that remained largely uncorrected in hardback releases (or at least inconsistently corrected)—were finally corrected and new introductory material added to the prelims. These changes were deemed sufficient, in the publisher's view, to merit the designation of ‘Second Edition’. In this new edition the original 1977 Foreword is now followed by a ‘Preface to the Second Edition’; this consists of a very brief introductory note by Christopher Tolkien followed by a 1951 letter by Tolkien to Milton Waldman. The copyright page remains unaltered with no mention of new edition.

Whether any of this is bibliographically significant is another matter entirely. Bowers makes the comment, that while copyright law—taking in account his 1940s American perspective (where copyright laws were still at variance from Europe)—allows for a ‘new edition with fresh copyright’ when only small changes have taken place, this is of ‘no bibliographical consequence’; pointedly reminding the reader that:

in this matter bibliographical requirements for clarity and uniformity in terminology must take precedence even over legal.

Publishers, in his view, are both unreliable and inconsistent.

Until the HARPERCOLLINS years, there does not seem to have been any further hardback copies of The Silmarillion even printed. GA&U (and later UNWIN HYMAN) appear to have distributed (mainly second impression) copies from existing stock; they did this from 1977, throughout the eighties, and into the early nineties. Only in 1992 was stock of hardback copies, printed in 1977, finally exhausted (assuming further copies weren't merely pulped). HARPERCOLLINS still had bound GA&U second impression copies to hand; these they distributed in John Howe dustjackets along with their own newly printed edition.

So while hardback copies (excluding book club copies) up until 1992 are certainly part of the original first edition—having all been printed at around the same time—it could be argued that copies even up to today are, in the broadest sense, also still part of the same original edition, as the published text has remained substantially unchanged to this day.

Impression: All four printers, that were employed to print The Silmarillion in 1977, printed their own impressions: at different sites, from separate plates—and possibly at close, but significantly different, times too. It should be noted that the plates used by each printer may have derived from a single set of type or a ‘mother’ plate; plates may have been made by photographic techniques; and these plates (or type) may have undergone separate independent correction and amendment. That these impressions are only nominally numbered (intentionally or otherwise) on copyright pages does not invalidate the use of the term where applicable.

State & Issue: The definitions of both state and issue are still debated in academia today, with no universal consensus yet reached amongst bibliographers. However, irrespective of exact meaning, both terms are closely linked; the definition of one helps define and inform the other. A good starting point for discussion of these bibliographical terms—and in the avoidance of lengthy Bowersesque analysis—is to be found in the succinct definitions set out by Carter (under ‘ISSUES AND STATES’) in his ABC for Book-Collectors:

 

When alterations, corrections, additions or excisions are effected in a book during the process of manufacture, so that copies exhibiting variations go on sale on publication day indiscriminately, these copies are conveniently classified as belonging to different states of the edition.

. . .

When similar variation can be clearly shown to have originated in some action taken after the book was published, two (or more) issues are distinguishable.

. . .

Yet in fact the onus of proof that an observed variation derives from a deliberate action taken after publication lies, or should lie, like an iron weight on the conscience of anyone who begins to write the word issue.

Put simply, by Carter's definition, issue classifies changes that originated after publication; state, changes that occurred during manufacturing i.e. pre-publication.

Bowers (in particular), Gaskell, and Tanselle all treat this subject at great length; and while small differences exist between their respective definitions, all three are in large part in agreement with Carter. Tanselle—the most authoritative present-day bibliographer—in particular, has further refined his understanding of these terms taking account of modern publishing and printing practises:

an issue is a group of copies of an impression that give evidence of forming a consciously planned publishing unit, whereas state refers to the kind of correction that does not call attention to itself as a marketing effort and that can only categorize an individual part of a book (such as a page or a binding), not a book as a whole (since any given book may contain mixed states).

Carter's defining element of timing (publication) is noticeably absent from Tanselle's definitions; which, nonetheless, remain largely unchanged from those he himself set out in 1975.

Since, with The Silmarillion, all four printers were known to be employed in printing various impressions pre-publication—and no evidence shows any printing or binding variants occurred after publication or that any of these variants were intended to be planned publishing units (with one major exception)—the use of state seems the most applicable bibliographical term to employ in describing the majority of variants encountered. As Tanselle would remark of Bowers: ‘the conservative position of claiming no more than state unless issue can be convincingly demonstrated’.

Where the use of issue is possibly called for is in respect to book club copies and binding variants. One definition of ‘separate issue’ by Gaskell is ‘the alteration of title-pages to suit the issue of a book simultaneously in two or more different forms’; Bowers describes this as ‘the very rare category of simultaneous issue’. Tanselle's more nuanced definition of issue focusing on the planned publishing unit—the publisher's intention rather than the minutiae of publishing timing—is pertinent here. Tanselle also acknowledges the role binding variants play in determining issue; again, if adjudged to be distinct, planned publishing units. As a result, issue should certainly be considered in respect to BOOK CLUB ASSOCIATES (1977) variants, is most certainly applicable to the distinct publishing venture that was the 1982 Collector's Edition [discussed here], but is otherwise not a term I generally use.  

However, while issue is reasonably clear, the problem of when to apply—or perhaps more accurately how far to apply—state still remains. Tanselle's comment that state ‘can only categorize an individual part of a book... not a book as a whole’ is particularly problematic if one wants to stick strictly to accurate bibliographical terminology. Years of book-collecting discussion—with its jargon and loose phraseology—has severely prejudiced this writer's ability to embrace the bibliographical precision here required.

 

WORKS CITED

 

BOWERS, Fredson

Principles of Bibliographical Description

Princeton University Press, 1949

 

CARTER, John

ABC for Book-Collectors

Rupert Hart-Davis, 1952

Oak Knoll Press / The British Library [with Nicolas Barker], 2004

 

GASKELL, Philip

A New Introduction to Bibliography

Oxford University Press, 1972

 

TANSELLE, G. Thomas

Descriptive Bibliography

The Bibliographical Society of the University of Virginia, 2020

 

 

POSTSCRIPT: It strikes me that many booksellers and book-collectors use these terms quite loosely. Personally, as well as tiptoeing around issue, I tend to limit the use of state to describing copies that exhibit textual variations only. I believe there to be three identified (textual) states of the 1977 first edition; priority seems reasonably clear. There are multiple binding states too; but, here, priority is less certain. Of course, the decision of whether to define a copy as an individual bibliographical state (or not) can also be conveniently avoided altogether by simply referring to them as variants, and giving them appropriate variant naming. Clearly this all requires much more thought.