Definitions of bibliography, and in particular of the variety called Analytical

A list of books is a list of books.

By ‘book’ I mean any artefact written, printed or created in whatever fashion and in whatever medium with a purpose of communicating a text, as distinct from an ‘archive record’ which is made and conserved in order to show that some action has been performed. (The other fundamental function of bibliography, the recovery of ‘information’ independently from the text through which it is vehiculed, is more proper to the pure sciences and will not be treated here.) If the books in the said list are united by some common external factor, most usually the circumstance of being stored physically on the same shelves in the same building, in a structure we call a library, the result is a catalogue and does not concern us here; if, on the other hand, the books are widely dispersed but have some internal factor in common, such as the same author, or the same printer/publisher, or the same topic/theme, the list is either a bibliography or has the makings of becoming one. Albeit often with less synthesis, most elementary definitions of the difference between a catalogue and a bibliography run along these lines.

Now a definition is a wholesome thing. Like the opening credits to a film, it sets the scene and tells us what the theme music is going to be like. As far as analytical bibliography as a form of scholarship is concerned, its most penetrating and evocative definitions can be found in the writings of Walter Wilson Greg (1875-1959) and in those of his American counterpart, Fredson Bowers (1905-1991). As far as the former is concerned, here is a vintage and oft quoted definition of [analytical] bibliography: “Bibliography is the study of books as tangible objects. It examines the materials of which they are made and the manner in which those materials are put together. It traces their place and mode of origin, and the subsequent adventures that have befallen them. It is not concerned with their contents in a literary sense, but it is certainly concerned with the signs and symbols they contain (apart from their significance) for the manner in which these marks are written or impressed is a very relevant bibliographical fact. And, starting from this fact, it is concerned with the relation of one book to another: the question of which manuscript was copied from which, which individual copies of printed books are to be grouped together as forming an edition, and what is the relation of edition to edition. Bibliography, in short, deals with books as more or less organic assemblages of sheets of paper, or vellum, or whatever material they consist of, covered with certain conventional but not arbitrary signs, and the relation of the signs in one book to those in another”  (W.W. Greg, ‘The function of bibliography in literary criticism illustrated in a study of the text of King Lear’, Neophilologus, vol. 18 (1933). The very useful edition of his Collected papers, edited by J.C. Maxwell (Oxford, at the Clarendon Press, 1966), where this essay is reprinted at pp. 267-297 [quote at p. 271], is unfortunately rare in libraries outside the English-speaking world, so I follow a policy of including a reference to the original place of publication of the articles cited).

Otherwise Greg’s thoughts on bibliography, its nature, its functions and the methods with which it should be applied to the study of textual transmission are best expressed in the following articles:

– ‘What is Bibliography?’, Transactions of the Bibliographical Society, vol. 12 (1914), pp. 39-53, reprinted in Collected papers, pp. 75-88. Basically an aiming shot that shows his thought taking form.

– ‘The present position of Bibliography’, The Library, vol. 11 (1930), pp. 241-262, reprinted in Collected papers, pp. 207-225. Contains the statement that “bibliography is the study of books as material objects”.

– ‘Bibliography – An Apologia’, The Library, vol. 13 (1932), pp. 113-143, reprinted in Collected papers, pp. 239-266. Easily the most important of the three essays, with some memorable and oft quoted remarks, such as: “Books are the material means by which literature is transmitted; therefore bibliography, the study of books, is essentially the science of the transmission of literary documents” (p. 115; p. 241), and: “I start then with the postulate that what the bibliographer is concerned with is pieces of paper or parchment covered with certain written or printed signs. With these signs he is concerned merely as arbitrary marks; their meaning is no business of his” (pp. 121-122; p. 247). A reading of this essay makes it clear that in Greg’s view no false distinction can be allowed to exist between manuscripts and printed books: his focus is entirely on the making of books, by whatever means, as the essential fact in the transmission of literary documents. Nevertheless his writings, especially for readers coming from different literary cultures, are often applied to very minor English Renaissance texts and require a certain familiarity with the bibliographical debate of the time.

The other great definer of bibliographical purposes, ways and means in our time has been Fredson Bowers, professor of English at the University of Charlottesville in Virginia and founder of the journal Studies in Bibliography (1948). Any and every judgement on Bowers’ sometimes controversial style has to take account of the fact that, not unlike F.R. Leavis in the parallel field of English literary criticism, he represented a militant, crusading form of bibliography that spared no quarter. His target was the dilettante, amateurish fashion in which evidence relating to the physical transmission of the text was marshalled and interpreted by scholars, most of them employed in universities, in the years after the Second World War (his favourite Aunt Sally, already singled out as a target by Greg, was the Shakespeare critic John Dover Wilson). This militancy led to insistence on analysis as a “bibliographical state of mind”, to be practised by properly trained adepts, as he most famously states in the brief pamphlet The Bibliographical Way, Lawrence, University of Kansas, 1959, reprinted in his Essays on Bibliography, Text and Editing (Charlottesville, published for the Bibliographical Society of Virginia by the University Press of Virgina, 1975, pp. 54-74).

Previous critics had already sought a terminology to define and describe the different sorts of bibliography, above all in terms of the different activities performed by bibliographers (Greg, for instance, regularly speaks of “critical bibliography”). Bowers, however, is the first to offer a cogent and workable organisation of the same in the essay entitled ‘Bibliography, Pure Bibliography, and Literary Studies’, Papers of the Bibliographical Society of America, vol. 46 (1952), pp. 186-208, reprinted in his Essays (1975), pp. 37-53. In his view the following five-part structure covers all the processes involved:

1) Enumerative (or Compilative) Bibliography is the simple construction of lists of books, articles and other writings on a given theme or subject, either as an ancillary document or as a hand list;

2) Historical Bibliography is research conducted into the context in which the document is produced, involving archive work, biographical studies and so on, where the evidence is external to the book itself (“... we should include here all biographical and historical studies of printers, papermakers, binders, typefounders, engravers, publishers, booksellers, and anyone else in any way concerned with the materials and the production of the book and its subsequent dissemination. Under the history of such I should also include studies of costs and prices, methods of sale and distribution; studies of the meaning of imprints, colophons, copyright entries, and of advertisements; all aesthetic studies of printing and its materials as an art; all studies of sizes of editions from the collateral evidence of publishers’ records or other external material; all investigation into the circumstances of literary composition which have any relation to the physical form of the literary work, the transmission of literary documents, and the relation of authors to the commercial process of publciation. It is difficult to limit this grouping narrowly, but let us say very much in general that it concerns itself chiefly with the discovery and interpretation of external evidence” ( p. 40);

3) Analytical Bibliography is the “technical investigation of the printing of specific books, or of general printing practise, based exclusively on the books themselves, not ignoring, however, what helpful correlation may be available with collateral evidence” (p. 191; p. 41), wherein “what is important is that the impressed symbols which are letters and words are treated in a physical and not in a literary way” (p. 192; p. 42);

which leads to two further, self-explanatory branches:

4) Descriptive Bibliography is the transformation of the analysis into a report in which the bibliographer describes what has been found (i.e. “There is some need to emphasize that descriptive bibliography which is not based on analytical is pratically useless, for it is as important to explain the reason for the peculiarities as it is to give the external facts about them; and, truly, the external facts often cannot be properly described until analytical bibliography has provided the reasons for their existence”, p. 194; p. 43);

5) Textual (or Critical) Bibliography is “the application of the evidence of analytical bibliography, or at least of its pertinent methods, to textual problems where meaning of some sort is involved and where it does make a difference whether a book is printed in English or in Sanskrit” (pp. 194-195; p. 44).

Much the same statement, albeit written for a non-specialist reader, can be found in his entry Bibliography in the 1960 edition of the Encyclopaedia Britannica, vol. 3, pp. 539-543. Since the term ‘Textual Bibliography’ has frequently been misunderstood, especially when later scholars have sought to apply the Greg-Bowers canon to other modern languages, it should be noted that in the latter’s definition it refers only to the application of knowledge acquired through analysis of the making of the book to textual decisions and choices (the common mistake is to confuse it with the activity more properly denominated ‘Analytical bibliography’). In 1971 Bowers provided a different, in many ways inferior, four-part definition of the varieties of bibliographical activity, in which ‘Historical bibliography’ is omitted and the order of the remaining four is given as “enumerative, descriptive, analytical, and textual”; see ‘Four Faces of Bibliography’, Papers of the Bibliographical Society of Canada, vol. 10 (1971), pp. 33-45, reprinted in his Essays (1975), pp. 94-108. The same article contains a significant novelty in that the various kinds of bibliography are now viewed in terms of a pyramidal or ascending structure, with ‘Textual bibliography’ at the summit, confirming Bowers’ view that the ultimate purpose of bibliography is the determination of how texts are transmitted. What should be noted is Bowers’ evident inability to find a satisfactory definition of what he had previously called ‘Historical bibliography’, or all the relevant information garnered from sources external to the artefact itself, since, in the exquisitely scientific terminology of Swift’s Houynmhms, from a purely bibliographical viewpoint it is “the thing that is not”. His sense of a conceptual incoherence therein brought his whole structure to totter and led to the weak solution of omitting the troubling item.

Twenty years ago I think I should have signed, in the most unquestioning of manners, the dotted line at the bottom of Bowers’ 1952 statement and expounded it as gospel truth. Today I believe very little of it or, rather, I believe it to be extrinsic to the nature of bibliography itself. It is certainly true that the uncertainties of Bowers are those of the discipline as a whole, not only because nobody has found a convincing way of reconciling bibliography to history of the book, but also because true analytical bibliography has remained an art form practised by very few scholars and this fact has led others to conclude that it is a waste of time. The five-part structure nevertheless confirms Bowers’ personality as a formidable organiser of bibliographical research and I believe it provides an excellent guide as to how bibliographers ought to go about their chosen business. They begin by gathering together all the information relative to their chosen theme (enumerative bibliography); they explore the historical, social and economic context in which the work or works that interest them were produced (historical bibliography); they study the physical books in order to obtain information about their making (analytical bibliography); if the project warrants, they draw up a report on those same physical features (descriptive bibliography); and, again if the project warrants, they analyse the significance of their discoveries for the transmission of the text and the constitution of a critical version (textual bibliography). My personal conviction, however, is that all bibliography is simultaneously enumerative, historical and analytical, though I agree that descriptive and textual, being applicative rather than pure, need to be considered as being somehow different.

Let me try and explain better. In the history of any and every ‘book’, three essential phases can be distinguished. The first is as a “work”, when it is sold by a publisher, through bookshops or directly, to a public of readers or of libraries, and is perused for its intellectual, informative, narrative or creative contents. From this point of view the ease with which the book can be obtained and the familiarity of the reader with its nature as a contemporary artefact mean that little consideration is paid to the object itself. In bibliographical terms its immaterial, abstract form dominates with respect to its physical nature. This phase ends, more or less, when the said ‘book’ is no longer available through a conventional commercial circuit (excepting the antiquarian or used book trade), either because the print-run, in the case of a typographical document, has been sold out and perhaps replaced by later editions, or because there is no longer a consistent interest for the title on the part of the book-purchasing public. The second phase, which can considerably overlap with the first, occurs when the book is utilised as a “text” for documentary purposes, either because somebody wants to understand the form of expression it represents, or because it conserves a particular version of a work. This same phase is mainly library based, since the “book” can only be obtained through the second-hand market with difficulty and expense; on the other hand, except for special cases such as first editions of important titles, the libraries do not impose onerous restrictions on the use of the book in this particular fashion. It can be said therefore that its immaterial and material aspects are essentially in equilibrium. The third and final phase is when the book is considered as an “artefact”. Such artefacts, when they survive, are usually rare and sometimes in poor condition. Where we have accurate information about original print-runs, we usually discover that only a very small percentage of the copies fabricated at the time are still extant (for most incunabula, for instance, the survival rate is less than 1%). The library function is absolutely dominant, since these artefacts, on account of their age and rarity are considered precious, while relatively few scholars have the means to purchase them, even if they were available on the market. In practical terms these books are not ‘read’, since the work concerned is usually available in a more recent edition (in terms of my definition I consider that collation of the text of the original artefact by a critic does not constitute ‘reading’).

There is a long-standing, oft-uttered and widely-held credence that analytical bibliography only deals with printed artefacts of the hand-press era. This credence is false and is tantamount to claiming that literary criticism only deals with the writings of an author when the said author is dead. Another equally useless and artificial distinction is the parallel belief that the principles underlying the study of medieval manuscripts, incunabula and successive products of the hand-press period, the publications of mechanised typesetting and printing, or the CD-Roms and DVDs of our own age, somehow differ in the necessity for analysis and description. But it is all one discipline. It is all bibliography, though it is reasonable to say that the knowledge and experience necessary to understand complex artefacts, widely dispersed in modern collections, often endowed with an extensive secondary literature, are acquired only over the space of a lifetime and therefore it is too much to ask a specialist in medieval codicology to interpret the intricacies of quadrichrome printing and electronic lettersetting in our own day and age. Specialisation is acceptable therefore, as long as it is recognised that, bibliographically speaking, the age, provenance and method of fabrication of an object have no determinant significance.

If we return to the tripartite distinction between “works”, “texts” and “artefacts” posited here, the inherent weakness, or perhaps underlying prejudice, of Bowers’ definition becomes somewhat clearer. It is obvious that the artefacts are primarily the older books, those which are no longer commonly considered as works or texts. The watershed previous to which most books nowadays are considered artefacts falls nowadays towards the end of the Eighteenth century: in this piece I certainly cite some editions of the same century as “texts” (see Fertel in § 2 and the numerous references to the Encyclopédie), but accompany the same with a complaint about the lack of a proper modern critical version. There is therefore a minimal element of practical truth in the observation that analytical bibliography has so far directed most of its energies and attention to, and achieved its most significant results with, artefacts produced on the handpress. But this general perception, however true, has to be modified by the deeper understanding that in due course all books, or at least those few that survive, will come to be viewed and conserved as artefacts. A teasing, contradictory and salutary example, however, of a contemporary publication, the three ‘issues’, termed ‘editions’ by the publisher, of Rino Pensato’s Corso di bibliografia (1987, 1989, 1995), studied as a physical artefact, in which much of the irony stems from the fact that it is a ‘work’ that seeks to define bibliography, is provided by Carlo Maria Simonetti, ‘Cataloghi storici: note e osservazioni bibliografiche’, Il bibliotecario, vol. 15, n. 2 (1998), pp. 29-40.

From our point of view, the principal weakness of Bowers’ definition of three (or five) sorts of bibliography becomes his failure to see that information about “works” is also a form of the same. What we might call ‘Current’ or ‘Ongoing’ bibliography is available through publishers’ and booksellers’ catalogues, including publicity, through reviews in newspapers and journals, and sometimes even through library catalogues. Periodicals that dedicate part of their space to reviews and overviews of critical production in their sphere are, among other things, bibliographies. In the pure and applied sciences very sophisticated forms of bibliographical analysis, which usually go under the name of documentation, acquire and make available with immediacy information extracted from periodical literature, since in these fields research has a short shelf-life and within a matter of years, sometimes months, belongs to yesterday’s science. Lists in which the presentation of writings or the contents of writings on a determined subject have a more ordered form and, especially in the humanities, reach backwards in time obviously form part of enumerative bibliography. Though there is necessarily an overlap between “works” on the one hand and “artefacts” on the other, from this point of view the said writings are listed primarily as “texts”, with emphasis not only on their contents but also on how to find them (i.e. the bibliographical details identifying the book or article are usually cited in considerable detail with the assumption that they will be sought through libraries).

But when all is said and done, the differences are merely nuances, since the decision to consider a book as a “work”, a “text” or an “artefact” lies exclusively in the eye of the beholder. In other words, and this is the point that interests us, all written, printed or otherwise created artefacts, whenever and wherever made, can become objects of bibliographical analysis. Returning now to Bowers’ difficulty in defining ‘Historical bibliography’, like Solomon, we propose cutting the baby in half, but only in theory, in order to understand the distinction between information extrinsic to the artefact obtained through historical research and that intrinsic to the object acquired through analysis. But in practical terms any attempt to exclude extrinsic information would prove messy, if not murderous, since in proper research all relevant information is called into play. It is however necessary to add that in 99% of instances the only reliable source of evidence about how the artefact was made is the artefact itself, while in the 1% of cases in which relevant external information about the same is discovered, for example in the form of printing or publishing archives, the documentary source all too often proves incomplete, ambiguous, or misleading.

Analytical bibliography therefore is the science of how to obtain blood from a stone, or how to extract from an physical artefact every gram of information about the way it was made.