Sunday, May 22, 2016


Serena Ammirati, Sul libro latino antico: ricerche bibliologiche e paleografiche (premessa di Mario Capasso, prefazione di Guglielmo Cavallo). Biblioteca degli "Studi di Egittologia e di Papirologia," 12. Pisa; Roma: Fabrizio Serra editore, 2015. Pp. 195. ISBN 97862278140. €120.00 (pb).

Reviewed by George W. Houston, University of North Carolina at Chapel Hill (

Version at BMCR home site

Table of Contents

There are far more Greek than Latin literary papyri — perhaps fifty times as many — and scholars of the ancient book have understandably tended to concentrate on the Greek side of things. Yet there is a good deal of evidence on Latin books as well. In this volume, an updated version of her 2010 dissertation, Serena Ammirati assesses that evidence through a description and discussion of two hundred and eighty-three papyri, eighty-five parchment codices preserved in libraries, and a handful of inscriptions, ostraca, and wooden tablets.1

Given this mass of material, the essential problem is organization. How can the bits and pieces of parchment and papyrus (seldom more than a page, and often much less) be organized and presented in a meaningful and informative way? Ammirati's solution is to divide her material into five broad categories, each presented in one chapter. Within each chapter, she identifies and discusses a number of different book types — distinguished from one another by some combination of content, script, size, format, layout, and other characteristics. She then illustrates the type by describing in detail the specific examples of that type that she has found. Throughout, Ammirati attempts to place each type and each fragment within a cultural or historical context. The result is not only a series of observations on the history of the book in general, but a number of mini-histories of certain types of book.

Chapter One treats what remains of books created in the period extending from the first century BC to the beginning of the third century AD. Among the groups into which Ammirati sorts her material, we may note: writing exercises (25–28, 36–37); bilingual glossaries (39–42); and wax and wooden tablets (33–34). The latter are important because they sometimes took the form of a polyptych, which Ammirati sees as a precursor of the codex. Latin books in this early period had various characteristics that distinguished them from Greek books: interpuncts (dots between words) were regularly used in Latin books but not in Greek ones, and the columns in Latin prose texts were much broader than those in Greek texts.

Among the groups of books discussed by Ammirati in her second chapter — on Latin books found in the East and dating from the third to the seventh century — are word lists for students of Latin, Greek fables translated into Latin, and glossaries to the works of specific authors (48–53, 65–72). All seem to have been intended for Greek-speaking students of Latin. There are also more advanced textbooks, with wide margins for comments, Greek glosses, and diacritical signs to aid in pronunciation. Ammirati believes these were special texts for student use (53–57, esp. 54 on "codici di studio"). Yet one wonders if at least some were not simply good readable texts. An example might be P.Oxy. 24.2401, a compact text of Terence's Andria (54). Ammirati isolates a number of other book types too: codices in which a variety of pre-existing materials was assembled and bound together (57–61), items found at Nessana (63–65), and bilingual glossaries that she subdivides into volumes with four, and then two, columns per page (66–72).

In these first two chapters, Ammirati traces several changes in the Latin book. The use of interpuncts became sporadic in the late second century and eventually disappeared. Columns gradually became more slender, like those in Greek books. Letter forms changed, and some Latin and Greek letters — the letter A and alpha, for example — were sometimes written in identical fashion in bilingual texts. In short, Latin books gradually became more and more like Greek ones. Ammirati proposes a theory to explain this phenomenon: in this period, more and more Greek-speaking natives of the eastern provinces sought positions in the imperial administration. Seeking to learn Latin, they wanted books that were easy to read, such as old fables, and they appreciated books in familiar formats (73). In addition to changes in layout and script, this period — the second and third centuries — is when rolls gradually disappeared and codices became the norm. By the second half of the third century, almost all Latin books were codices (45).

Chapter Three, on late Latin books found in Egypt but produced in the West, is organized not by content or form but by script: capitals first (75–79), subdivided by content (Sallust, Vergil, Christian, and grammatical texts), and then uncials (79–81). Ammirati makes no attempt to be exhaustive, opting instead to provide examples. The texts in uncials gathered here — all parchment codices — share one or more of a set of characteristics. There may be small prick marks to be used in drawing guidelines, red ink may be used for the first line of a new section, fascicles may be numbered, and sections may begin with a line projecting to the left of the column (79; Ammirati lists further items).

Interesting points large and small emerge in the fourth chapter, an earlier version of which was published in the Journal of Juristic Papyrology for 2010.2 The chapter is in effect a monograph: a comprehensive study of Latin manuscripts with juridical content, beginning in the first century AD (the earliest surviving such texts) and continuing to the eighth century.The organization of the evidence is complex. Ammirati divides the texts into two main groups by geography: codices produced in the eastern provinces (85–100), and those produced in the West (100–104). The former are divided into two groups chronologically, those dating before the great codifications (86–91), and then manuscripts containing the Theodosian and Justinian codes (91–100). The latter are subdivided into manuscripts in uncials and those in slanting letters. The chapter's further subgroupings need not be enumerated here.

The brief Chapter Five deals with papyrus codices produced in the West. Drawing from both Chapter Four and Chapter Five, I note a sampling of useful details. One roll (P.Oxy. 17.2103, Gaius) has a number, XVIII, at the top of one column, a rare instance, perhaps, of the numbering of columns (84). With the late Paolo Radiciotti, Ammirati takes the codex as a format of authority: codices were the standard form for both juridical and Christian texts, i.e. the laws of man and the laws of God (85). Discussing the wide variety of layouts and formats in these texts, Ammirati cites Giovanna Nicolaj: the texts we have were not intended to be normative or to establish standards, but rather were for the use of individuals, and those individuals chose whatever format answered to their needs (90).

One group of codices was given unusually wide margins and extensive annotations. Kathleen McNamee suggested, and Ammirati is inclined to agree, that the notes were added before the volume was bound, and they thus represent a development in book design: annotations included (usually within strict boundaries) as an integral part of the text (97–98). On the question of papyrus versus parchment: when we are dealing with literature, including legal texts, there are more parchment fragments surviving than papyrus. This is not due, Ammirati argues, simply to the fragility of papyrus. For when we look at documents there are many on papyrus, and papyrus continued to be used for documents right into medieval times. Thus the use of parchment for literature was a choice, and the surviving numbers reflect that choice, not simply the chances of survival (105, with the important note 1).

The text is followed by an index of the materials cited, references to standard databases such as the Leuven Database of the Ancient Book, and information on reproductions of the manuscripts. Ammirati provides an extensive bibliography and seventy-seven mostly clear and legible plates. In the bibliography, one notes the surprising omission of Robert Cavenaile, Corpus Papyrorum Latinarum, and of William Johnson, Bookrolls and Scribes in Oxyrhynchus. The latter could have provided details on the normative dimensions of Greek books. They would have been useful, since Ammirati argues that they were different from those of Latin books.

What emerges from the materials that Ammirati has gathered and discussed? It will not be cause for surprise if we note the seemingly endless variety of types, shapes, and sizes of books, of script and content, and the corresponding lack of standardization in these books. There is no single Latin book; there is no single set of standards, but rather many types and various characteristics, often dictated by the demands of the widely varying content. Ammirati's discussion of the range of possibilities will serve as a useful guide through this maze of materials. For specialists, she provides up-to-date bibliography (as recent as May 2015), new dates, and, occasionally, new readings. A good example of this last is on p. 89, with n. 1: two fragments from Berlin, where Ammirati's new reading led her to a previously unnoticed join. She provides a wealth of comparative materials that both specialists and non-specialists will be able to use when assessing a book, or a book type, previously unknown to them.

The book has some shortcomings. The section on the Villa of the Papyri at Herculaneum (23–25) is disappointing. Ammirati here does not describe any of the papyri in detail. Instead, she argues against the theses of Knut Kleve concerning the identification of certain papyri and against Richard Janko's use of the Oxford disegni. I agree with her on the first point, not the second; but more importantly I would like to know what she herself makes of the surviving fragments from the Villa, tiny and unpromising as they are. It is clear that she has a keen eye; turning it to the Herculaneum fragments would be a contribution.

Ammirati's presentation of her material is unnecessarily difficult to follow. The description above of the various groups and subgroups in Chapter Four will give an idea of the complex organization of the book, but only an idea, for there are additional subgroups, defined in various ways and distinguished from one another on various grounds. Granted there are many variables—date, language, script, content, provenance, layout, and so on—but still one suspects that a simpler and clearer organization could be devised.

In her detailed descriptions of papyri, Ammirati has opted for a continuous narrative, even when presenting dimensions of, say, columns, or the number of lines per column. I longed for a few tables setting out these details and facilitating quick and comprehensive comparisons. One might compare, for example, the tables in Eric Turner's The Typology of the Early Codex: they may be messy, but they make it relatively easy to compare one manuscript to another or one group of manuscripts to another.

As she describes the manuscripts, Ammirati regularly provides at least some measurements such as breadth and height of columns or width of intercolumnar spaces. Seldom, however, does she comment on the ratios or proportions of those elements, and in particular she almost never comments on the ratio of written space to blank space on a page (or virtual page). This is surprising. Such proportions might be similar in the books of a given group, and they might help to define the group. Or, put the other way round, are the books of a given group designed in a similar way? Are they all elegant? Imposing? Pleasant to look at and hold? As Johnson showed, many Greek volumes from Oxyrhynchus seem to have been designed to be impressive visually and physically. Is that true of Latin books as well? Maas's Law is frequently evident in Greek papyri. Is it also common in Latin books? A systematic discussion of these and related questions—the aesthetics of the book—would be useful and deserves a place in a study of Latin books.

There are some mistakes. Most are typos and easy to spot, but note the following.

Page 39, right column, 16 from bottom: verso and recto are reversed.
Page 66, left column, line 22: for 21238, read 21138.
Ibid., line 24: for seconda actio, read oratio prima.
Page 98, n. 1: for p. 627, read 672–73.
Plate 22 is mislabeled. For PMich VII 429, read PLondLit 184, and see p. 39.
Plate 54 seems to be upside down. This affects the text on pp. 76–77.


1.   The dissertation is available online at Archivio Aperto di Ateneo (log-in required).
2.   Ammirati, Serena. 2010. "Per una storia del libro latino antico. Osservazioni paleografiche,bibliologiche e codicologiche sui manoscritti latini di argomento legale dalle origini alla tarda antichit√†." JJP 40: 55-110.

1 comment:

  1. I am sorry that the reviewer did not really engage with Ammirati's account of the palaeography of these manuscripts, which strikes me as more original and more important than her treatment of their layout.