and on display. The collected object becomes an icon in a neater, more synthetic history, such as literary history.39 The collector functions in this way as a maker of meaning—an “allegorist,” in Benjamin’s account, who writes and reinforces a new synthetic history in each of the objects he collects.40
The AB Catalog and the Cambridge University Library
Ultimately, Benjamin concedes that the status of books in modern collecting practice is problematic and in need of investigation because the book collector is “the only type of collector who has not completely withdrawn his treasures from their functional context.”41 For this reason—the insuperable functionality of texts—William Pugh’s eighteenth-century catalog survives, permitting a glimpse into what lies behind the synthetic order of a modern collection. During Pugh’s tenure, the books in the AB class were not collectors’ items in the modern sense—they were not, as a rule, clean, individually bound, or perfect—and they would not begin to seem so until some sixty years after Pugh’s dismissal. In 1859, as J. C. T. Oates has shown, the librarian Henry Bradshaw set out to rearrange the early printed books at the University Library according to the somewhat undiplomatic assumption “that there was a right place for every book and that every book should occupy it.” Any volume that seemed to exceed its boundaries “could only be made to conform if it were cut up into its constituent item.” “This,” Oates explains, “Bradshaw did with horrifying ruthlessness,”42 extracting books from their bound contexts with little sense of the information being eliminated in the process. But as the texts in the AB class were being reshaped and placed into magic circles, they could not be fully withdrawn from their archival functionality. Part of an active collection, they had to be available for reading, whether or not they were bound individually and set out in triumphant order on the display shelves. Pugh’s catalog therefore remained in use until late in the nineteenth century, when Bradshaw’s reforms finally made the AB class of books obsolete.
Because the catalog remained in use for so long, it had to be kept up to date during all subsequent modernization campaigns undertaken at the University Library. In this updating, we find the present catalog’s most striking asset: it was annotated by successive Cambridge librarians each time an early printed book in the AB class was reshaped or reorganized.43 Pugh’s original entries, in other words, constitute only the bottom-most layer of bibliographical records, corresponding to the appearance of the books as they originally came into the university’s possession from the library of Bishop Moore. Whenever those books changed shape, the entries in the AB catalog were changed in turn to facilitate access, usually with a set of brackets and shelf marks written in pencil over the old ones (Fig. 5). If, for example, an early volume originally containing multiple printed texts was split into smaller individually bound books during Bradshaw’s tenure, Pugh’s AB catalog would be marked up to reflect the new grouping and the location in the library where the texts were now to be found. Thus, the present document, annotated in layers, allows us to do more than generate a rough reconstruction of a set of early printed books before modern interventions, valuable as that reconstruction is. It also allows us to track the subsequent interventions as they were carried out by a determined institutional collector. Each textual mutation, from each item’s acquisition to its current shelf mark at the University Library, is visible.
The wide-ranging entries in the AB catalog offer a concrete introduction to the ways in which taxonomies of reading and book ownership vary over time and to the curatorial procedures through which traces of this variability have been submerged in modern collections.44 From Pugh’s records, we ascertain that at least half of the books originally placed in the AB class transgressed in some way the bibliographic categories that were institutionalized in modernity, prompting (we assume) Bradshaw’s tenacious desire to reform them.45 From the web of later annotations that now surround Pugh’s records, we find that all but a few of these books were brought into line with modern standards during the reform campaigns undertaken in the nineteenth century. Throughout, a clear pattern emerges in which premodern and early modern books with relatively flexible—at times uncertain—boundaries are transformed into modern, individual, self-enclosed texts. When these selfenclosed texts are reassembled according to the information in the catalog and thus restored to how Pugh first saw them, lines of affiliation open up that help us begin perceiving histories of reading, book storage, and situated interpretation that have been systematically obscured.
Figure 5. A leaf from the AB catalog, showing the hands of many later librarians. Each set of brackets represents a new mutation of the books originally in the AB class. This page shows the original contents of seven books, AB.8.46 to AB.8.52, containing twenty-seven early texts between them. Reproduced by permission of the Syndics of Cambridge University Library.
Many former AB-class volumes appear to have been assembled into early anthologies out of the exigencies of bookselling and bookbinding in the handpress era; their constituent items shared a printer or publisher as their organizing criterion before they were rearranged by Cambridge librarians so that they now stand in single textual units. Shelf mark AB.1.28, for example, once comprised a multitext volume of incunabula—works by Cicero, Martial, and Ovid in folio—that, Pugh notes in his entry, likely issued from the same press.46 When brought together today, the books contain marginalia from the same early reader, including manicules (or pointing fingers) and underlining in the same ink, revealing valuable traces of how they were used early on in intellectual alignment with one another as a collection. According to a binder’s stamp on the inside cover, the titles were separated into independent leather-bound books and given independent shelf marks in or around 1890. A similar example in the vernacular is former shelf mark AB.10.27, which comprised five books printed by Caxton, including Virgil’s Boke of Eneydos and Chaucer’s Boke of Fame.47 Beneath each of the printer’s colophons, a fastidious early reader named “R. Johnson” recorded that he bought the books at the same time in 1510. We might assume that Johnson had them bound together in a single volume to save money or that a later sixteenth- or seventeenth-century collector, who perhaps came into possession of Johnson’s books, did so for similar reasons. The texts are now bound and cataloged separately.
In calling up one of these texts today, one would have no way of knowing that it was once packaged in a composite book or anthology. When we become aware that such configurations were normal in the period—that very few works, particularly literary works, stood alone as they do today—important questions about the interpretation of early printed texts and literatures arise. What readings were enabled or disabled by putting Ovid into material proximity with Cicero, or Chaucer with Virgil, in a single binding? What species of taste or sensibility motivated these assemblages of texts, which we now experience only separately in modern editions? Certainly, many of the volumes in the AB class exhibit thematic connections that allow us to reconstruct their earliest uses and recover information about readers’ desires. In a study unrelated to the AB catalog, Seth Lerer has reassembled thirteen early printed books from the Cambridge University Library that, according to Pugh’s catalog, once made up a single volume (shelf mark AB.5.37) and that can be linked to a particular recusant family from the late sixteenth century. The volume was made of incunabular and nonincunabular texts, literary and nonliterary texts, including a play by John Heywood, works by John Lydate and Stephen Hawes, books on manners and hunting, and a fragment apparently drawn from a larger book and bound in. For the modern collectors who took it apart, this mass of text was disorderly. But as Lerer determines from the continuous threads of marginalia and subject matter that emerge in the reconstruction, the volume functioned for its sixteenthcentury users as a thematic anthology: a “collection of largely instructional works, calibrated to the education of young boys in late medieval and early modern England.”48
A similar example from the opposite side of the political spectrum can be found in AB.5.58, which once comprised eighteen books printed in the mid-sixteenth century, recorded in Pugh’s catalog as a compilation. Each of these is numbered
