book history as their subjects. A principal example for English bibliography is Howard-Hill (1969–99); for American bibliography, Tanselle (1971). These bibliographies are readily approached through such general reference guides as Harner (2002).
All of these bibliographical attributes can exist in different combinations in a single bibliography. However, in every instance, the compilation of a list depends on the bibliographical (analytical) examination of copies of books. The longest bibliography starts with the first copy. Not even book historians appreciate the extent to which their work depends on the products of enumerative bibliography: that is, lists of books. Enumerative bibliographies and library catalogues are constructed from descriptions of copies of individual books that are taken to represent, more or less faithfully, individual works that contain distinct texts. Incorporating the products of analytical and descriptive bibliography, it is enumerative bibliography that provides the basic material for the history of books. If books incorporate the collective memory of humankind – that is, preserve what is worth preserving – then without enumerative bibliographies access to the record of civilization would be random: civilization itself would experience a kind of Alzheimer’s disease. Enumerative bibliographers and library cataloguers bind together the elements of civilization and society, providing access that magnifies the power of each element. The increasing sophistication of libraries and the development of bibliographical method exactly parallel the progress of civilization as we know it, not merely as a consequence but as an essential enabling factor. More narrowly, as book historians participate in the extension of knowledge, they build on foundations erected by bibliographers.
I will elaborate more specifically. Usually, bibliographical description for any purpose starts with a single copy of a document. (I will use “bibliographer” for “cataloguer” mostly hereafter.) Identification of the copy to hand is the first concern of the bibliographer. When the cataloguing is “original” (that is, when the bibliographer is not simply matching the copy to hand against a description written by someone else), identification may not be easy, particularly if the work itself was hitherto unknown to bibliographical history. Information sufficient to identify the work or book may be lacking or be false, or the bibliographer may not have the means to make a correct identification. To illustrate this, there are records of twenty-five Hookham and Company Circulating Library catalogues, scattered amongst eleven libraries in my database. For all but three of the catalogues, the dates are conjectural, in some instances pro forma. For instance, the Bodleian Library conjectures “[1829 ]” for a volume (Bodleian Library 2590 e.Lond.186.1) that consists of a catalogue that contains “Addenda 1821” and a separate 1829 supplement with its own pagination, register, and printer. The Bodleian cataloguer apparently dated the book 1829 as the year in which the three parts were issued together, but that obscures the fact that the volume was produced in three different years.
Further, the extent of anonymous and pseudonymous books in the early period is considerable and the bibliographer may have great difficulty in determining what the authority of such a book is (Griffin 1999). Many books lack much of the information that may allow a bibliographer readily to put them into their historical context exactly. Of 10,904 monographs recorded in my database in June 2002, 1,058 (roughly 10 percent) did not identify the author on the title page, 129 were pseudonymous, 1,407 were anonymous, 2,672 did not supply the place of publication, 2,587 did not give the name of the publisher or printer, 2,293 did not give the date of publication, and in 1,087 r e cords the date of publication is doubtful. Identifying such books is