Inside Roman Libraries: Book Collections and Their Management in Antiquity by George W. Houston (review)

This informative volume adeptly combines literary and material evidence to provide an engaging overview of Roman book collections in the period between Cicero and Constantine. Thus, “book” in this context refers to the papyrus roll. Libraries that would have contained early Christian codices are outside the scope of Houston’s study, although the collection of Dioscorus of Aphrodito features periodically in the notes as comparative evidence.

The first chapter surveys different methods of assembling collections of books (making copies, buying books, and plundering existing collections). The second chapter turns to the extant lists of books preserved on papyrus, showing that these lists represented collections that varied greatly in size and focus. The next two chapters treat actual book collections that have survived from Roman antiquity. The Villa of the Papyri at Herculaneum receives a chapter of its own, and the fourth chapter analyzes five identifiable collections of books thrown out in the rubbish heaps of Oxyrhynchus in Egypt, excavated first by Bernard Grenfell and Arthur Hunt and later by Evaristo Breccia. The final two chapters answer the question of what we might find in a Roman library aside from the books themselves (types of spaces, storage systems, equipment and artistic decoration), as well as personnel that we might encounter in a library and what sorts of activities they might have carried out. A short conclusion is followed by three appendices: texts, translations, and commentaries on the papyri that contain lists of books; an inventory of books found at the villa at Herculaneum; and a catalogue of the books that make up the largest of the Oxyrhynchus finds.

Scholars will be grateful for the large body of evidence that Houston has gathered and for his generally astute treatment of it. Some of Houston’s more daring conclusions, however, should probably be taken with a grain of salt. The contours of the remains of the collections from Herculaneum and Oxyrhynchus are perhaps not as clear as Houston sometimes implies. While Houston does mention the tentative nature of paleographic dating of manuscripts, a number of his arguments depend quite heavily on the accuracy of palaeographic assignments, especially his claims about the useful life of papyrus rolls both in Herculaneum and Oxyrhynchus. One example can illustrate both of these points. In his analysis of Grenfell and Hunt’s second find from Oxyrhynchus, Houston concludes, “some manuscripts clearly could last for half a millennium” (175). This claim is based on the inclusion in this group of a single outlying papyrus, P.Oxy. 15.1790, which Houston accepts was copied “ca. 130 BCE.” If, as is commonly believed, this group of papyrus rolls was thrown on the rubbish heap in the fifth century CE, then it would seem that P.Oxy. 15.1790 had a life of about 500 years. But questions arise. Unlike Grenfell and Hunt’s first find, which consisted of torn up papyrus rolls said to have been found concentrated together in a single small area and excavated in a 24-hour period, the second find was more widely strewn and apparently excavated over a considerably longer period of time. We do not know exactly where these individual papyri were found in relation to one another. How certain are we that P.Oxy. 15.1790 was in fact discarded with the other pieces in antiquity? And even if this papyrus did form part of this collection, how secure is its date? Palaeography is the only guide, and expert opinions diverge by a century. The date of “ca. 130 BCE” was assigned by Eric Turner, whereas Hunt, in his original edition of the papyrus, preferred a date in the middle or latter half of the first century BCE. This is not to say that Houston’s conclusion is definitely incorrect. Rather, I emphasize that his conclusion could be correct if P.Oxy. 15.1790 was actually part of a collection with the other fragments and if...