Bryn Mawr Classical Review

[For an expanded Table of Contents, see the end of this review.]

The simple title on the cover The Roman Book is deceptive; the lengthy subtitle is more revealing. This book is as much about oral performance as about written books, arguing that a major step in publication for some (most?) types of Roman literary works was the recitatio, the reading out of all or part of a composition by the author and/or a chosen proxy, often a trained slave. The period chosen for analysis is roughly the first two centuries of the Common Era (p. 101 specifies 80 BCE to 170 CE), and the starting point is a strong critique of modern theories about publication and the book trade during Roman times. The author attempts to recreate the socio-literary world from which "classical" Latin "literary" works emerged--especially, but not exclusively, poetry. He argues that oral performance played the major role in creating interest and demand for these works.

The author is trained in the classics (this book began as his Ph.D. thesis at London University) and has pursued a successful career in journalism, which makes for a very informed and readable presentation. Ancient sources are cited in abundance, almost always accompanied by standard translations or even new and lively English renderings by the author, along with a plethora of modern works (see the book's valuable bibliography, pp. 223-229; internet resources are also sometimes mentioned). The text is also amply annotated, although the notes are inconveniently located as a separate section at the end of the work (pp. 181-222; fortunately those page headers specify the pages to which the notes apply) rather than as footnotes. A further frustration is that the index (pp. 231-236) is to the main text only, and does not cover the notes, which frequently add important information to the discussion. Nor is there an index of the ancient passages cited. Two brief appendices deal with the interesting side issues of the origins of Latin tachygraphy (perhaps Tiro was significantly involved at some point), and the claim made by some ancient authors that their works were "known throughout the world" (judged unlikely). See the Table of Contents at the end of this review, to which the numerous sub-headings in each chapter could be added for more detail. It is not difficult to thumb through the book and follow the argument.

With regard to evidence often cited for "publication" and "book trade" (based on modern analogies popularized by Theodor Birt and similar authors since the late 19th century), Winsbury is a minimalist. If no clear claim exists in the ancient references, no inferences are accepted as evidence. "There was no developed marketplace, no substantial book trade, no technology, no high rate of literacy through which written texts could have been 'commodified'" (p. 170). On the other hand, for the author's theories about the significance of performance in the issuance of classical works, various inferences are allowed that go far beyond the "firm" evidence. While Winsbury would not limit ancient Roman publication to what was performed orally, "orality" is clearly emphasized as the essential element leading to dissemination for much Roman literature: "The recitatio was therefore the pivotal event in the life of a new work with literary pretensions (though not necessarily of all works that have come down to us), the event to which all the efforts and concerns of its author were directed. The recitatio followed by editio if the new work warranted it, was in effect publication for most genres, even if libraries may have been an alternative route for some others" (p. 110). What he does not discuss at any length or in any depth are the admitted exceptions to such a generalization (e.g. lengthy histories, letter collections, reference works, medical texts, commentaries--see pp. 104-106, 132f.).

Winsbury is also highly selective in his choice of evidence. He attempts to focus on "Roman" materials, with Latin at center stage, but also recognizes the close relationship of "Greek" developments as both prior to and parallel with the Latin. Sometimes he appeals to the Greek materials (e.g. Galen and Lucian on books), but often he does not, even when his presentation could be strengthened.(e.g. Homeric textual reconstruction by Alexandrian critics is not even mentioned in his section on recognizing errors in transmitted copies, pp. 130f.). Indeed, on Winsbury's reading it is partly in conscious imitation of the Greeks that Romans abandoned word division and punctuation, "from about the time of Hadrian" (p. 46), "the Graecophile emperor" (p.139): "It is this prestige status that largely explains why the Romans maintained for so long the format of the scroll...in preference to the ultimately successful codex format, and why they discarded spaces between word[s] and retained texts as river-of-letters without any system of punctuation, in the Greek manner. It was all part of keeping the book as a specialist upper-class possession, to be mastered only by the few" (pp.164f). This sort of socio-cultural explanation of Roman intentionality runs throughout the book, but without any systematic exploration of the Greek side of the same issues.

On technical aspects of book production, Winsbury is caught between some older and newer views and approaches that are sometimes somewhat problematic. He argues that cultural factors were more influential than practical considerations in the long survival of papyrus (not leather) and scroll (not codex) in the period under examination, despite the knowledge that both leather and codices were available choices in that period (pp.16-25, citing Martial among others), He then proposes that "the victory of the codex, and the victory of Christianity, became one and the same. ... The codex 'piggy-backed' to success on Christianity, and its success arose from the need to 'mobilise God' in a particular way using particular texts. Thus the final choice between papyrus and parchment, scroll and codex, was largely a religious, i.e. cultural, one rather than a technical choice" (p. 25; see also p. 134). And yet further: "Once a precise set of texts or body of literature gathered sacred Christian canonical status, then these texts had to be contained within one sacred book...in a format that could be repeated with exactitude from edition to edition and place to place. The codex was, one might say, an inspired choice for this purpose" (pp. 25f.). This despite the fact that few Christian codices of the entire Bible (pandects) have survived, and even those are far from identical, and also that codices of particular genres of "pagan" literature are increasingly well attested prior to or apart from the "victory" of Christianity.

I have a few other more pedantic quibbles with Winsbury's treatment of the technical matters pertaining to book production. At a couple of places, the picture captions identify as a "pen" what is almost certainly a stylus used to write on wax tablets (pp. 28 and 34--in general, the role of the wax tablet notebook as a proto-codex is underplayed, despite the pictures and the brief discussion on p. 103). In describing how a copyist would have written, he fails to invert mentally the picture based on Egyptian models (where writing went right to left) and thus states that for the Roman copyist, "the left hand held the unwritten scroll and the right hand wrote"--which would indeed be an "awkward position" for right to left writing (p. 37). In discussing the relationship between scroll format and textual "structuring" (pp. 47f.), Winsbury fails to explore questions of "book" lengths (in terms of content, not physical measurements) in relation to genres and to multiple "volumes" attributed to a work (that is, the extent to which "structuring" of a work into multiple "books" is a reflection of how many scrolls were needed for the task). This oversight is reflected in the attempt to compare the number of "books" in scroll collections with those in codex collections (e.g. p. 70 and its n. 29, with the claim that "Roman imperial libraries were...huge by medieval standards--the Cambridge library of 1424 had a mere 122 volumes"; but how many scrolls would that represent? If Trajan's imperial library had 20,000 scrolls, and if a robust codex could hold the equivalent of 50-100 scrolls of various sizes, the comparative capacities of the libraries might not have been so disparate).

In general, the book is well written and edited, with only an occasional lapse (see the list at the end of this review). And it is an enjoyable read, with dashes of humor (including clever sarcasm) interjected here and there. Winsbury raises interesting issues about socio-cultural control over literary production that seem to me to be problematic, or at best one-sided. To his credit, he is well aware that matters are more complex than his treatment allows, but that does not stop him from tending to simplify the picture of elite Roman "literature" (with all the ambiguities of that anachronistic term--see p. 183 n. 26) and literary production, or of the historical developments and processes involved (e.g. Greek parallels, or the later role of Christianity). Still, he provides an interesting and instructive romp through a wide range of "Roman" materials that are related in one way or another to the production of a certain type of "Roman book"--and to some of the associated processes (e.g. theatrical venues, oral performance, role of slaves, distribution and control). I am pleased to have read and learned from it.

Table of Contents, with narrative summary statements from p. 11 interwoven:

1. Myths and anachronisms: the need for a new look at Roman publishing.

What was the Roman book?

"to establish and discuss the manufacture, format and aesthetics of a Roman 'book', and their merits and demerits"

Deconstructing the Roman book trade

"the myths and realities around 'publishers', bookshops and libraries at Rome"

What the Latin tells us

"a close look at the terms used by the Romans themselves to describe their activities, to see how this terminology paints a truer picture of what they did, including their reliance on slaves"

Texts in an oral/aural society

"to place the Roman 'book' within a context of a primarily oral society"

The perils of publishing

"explores the many hazards that Roman 'books' and authors had to endure and survive (or not)"

Gluing it all together

"the social and cultural landscape into which Roman 'books' fitted, and which is itself illuminated by a better understanding of Roman 'publishing.' The objective is to construct a new picture of both the practicalities and the sociology of Roman publishing that is as rounded as the limited evidence allows. This will help to explain why and what the Romans did when they did what we today would call 'publishing,' in accordance with their own, very different, scale of literary, social and political values and traditions--and in accordance with the raw materials, people skills and means of distribution available to them."

Errata: I noted the following typographical errors and/or slips in the editing process:

pp. 21 and 25 "only...only" where only one is needed; p. 44 "reference" for preference, p. 62, "who" for whom; p. 138 and 164, "an" for a; p. 164 superfluous "it" in "what it is still a common presumption"; p. 165 "word" for words; p. 193 n. 31 "CED" for CE; p. 214 n. 24 "spend" for spent; p. 217 n. 68, Josephus War 19.94 (War does not have 19 books).