Bookrolls and Scribes in Oxyrhynchus (review)

The aim of Johnson's impressive book is to find information about the production and use of classical literature in its original context. For the study of the bookroll, i.e., the papyrus roll that contains literary texts, he coins the term "voluminology." The papyrus roll as an artifact has attracted little attention in the past and, when it has, the scope for examination has been very limited (see 5 for a bibliography of past work on this subject). Johnson has collected and closely analyzed a sample of over four hundred bookrolls preserving extant literary texts, all published in The Oxyrhynchus Papyri (vols. 1–61). Of these, 317 come from the Oxyrhynchus site (appendix 1A) and 95, Ptolemaic and Roman, from elsewhere (appendix 1B); he calls them "the comparison sample," as he intends to check the relationship of the Oxyrhynchus papyri to other papyri of the same period (Roman) but of different provenance and also to earlier bookroll traditions (Ptolemaic). He announces that his examination has led to the discovery of essential differences in the rules that govern the production, aesthetics, and usage of codices and papyri. One nevertheless misses an illustrative example among the magnificent plates at the back of the book when, e.g., Johnson refers to the "rakish lean" as a deliberate design feature in papyri. Since the glossary serves the nonspecialists, the phrase could also be defined there.

For all the book's intention to target the layman, it remains of interest mainly to specialists. Johnson's work consists primarily in measuring the physical details of the papyri: these were neither regularly reported nor systematically measured by the first editors. He describes his way of measurement, explains the terminology, and proceeds to the supplementation and correction of previous readings,¹ some of which prove the (neglected) consistency of a given scribe. After this technical groundwork, Johnson goes on to trace the relationship between dimensions, genre, and script, and how these changed over time.

Since "the detailed history of the bookroll is an urgent desideratum," Johnson proceeds to four case studies which investigate the conventional scribal habits and features of the bookroll and distinguishes them from idiosyncratic tendencies or from what could be part of the paradosis. He does this first through examining papyri written by the same scribe (identified by the unusual script). The sample is not large, but he observes that scribes were remarkably consistent in the column width, both in the same roll and also when copying texts of different prose genres, making it likely that scribes must have used a tool of measurement. He then uses some examples of scribal error to discover how scribes copied from their exemplar. Although line-by-line copying can be deduced in some examples (41), he thinks that the evidence favors neither line-by-line nor column-to-column copying, without, though, telling us what the usual pattern was.

Chapter 3 unpacks the formal characteristics of the papyrus roll. Johnson examines what appears to have been the conventional scribal behavior: when writing a roll, the scribe did not pay attention to the kollemata (joins of papyrus sheets), although the quality and the placements of the joins revealed the quality of the papyrus. Johnson tries to establish that the vertical tilt (Maas' law) is a deliberate aesthetic effect by tracing dots which mark the slant of the margin (pl. 9). Photographs are not provided for all the examples, and, when they are, the occurrences are disputable; sometimes dots mark neither the leading of the lines (distance from base line to base line) nor the column slant. Maas' law is nevertheless persistent in the Ptolemaic and early Roman era, but the infrequent occurrence of ruling dots, combined with the mutilation of the papyri, obscures their function.

The rest of the chapter examines the dimensions of the column. Johnson notes that of all the prose genres, philosophical texts (and not oratorical ones, as is generally assumed) seem to be written in...