Thursday, July 28, 2011


Christopher Stray (ed.), Classical Dictionaries: Past, Present and Future. London: Duckworth, 2010. Pp. viii, 229. ISBN 9780715639160. $80.00.

Reviewed by Tom Keeline, Harvard University (

Version at BMCR home site

[Table of Contents is listed at the end of the review.]

When once pressed at a party about what he really did for a living, D.R. Shackleton Bailey is said to have acerbically replied, "I just look things up all day." This remark, however ironic, carries more than a grain of truth: classicists do in fact devote vast portions of their lives to looking things up, especially in dictionaries of Greek and Latin. It is thus salutary to reflect on the nature of the tools we all spend so much time using. Classical Dictionaries, an edited collection of papers delivered at an Oxford conference in June 2009, does just that, considering the stories of both familiar and lesser known lexica. The book is subtitled "Past, present and future," but it is mostly devoted to the history of scholarship, and in that field it scores an unqualified success: it is excellent both in treating dictionaries past and in evaluating the present lexical offerings as products of that past. When it comes to discussing the future of dictionaries, the book is occasionally on less sure footing, but nevertheless opens up important fields for discussion and debate.

First, readers should be clear on what this book is not. There is very little discussion of lexicographic theory or of the thorny philosophical issues underpinning any attempt at compiling a lexicon. While two of the chapters are written by practicing lexicographers, the other seven are not, and the target audience is not writers of classical dictionaries but curious users of such works. Coverage of different dictionaries is uneven, with several chapters reacting to LSJ, but only one treating the OLD. The most noticeable omission is the TLL, which receives no chapter of its own, although its learned pages are sighted from afar in several of the contributions. Readers will also search in vain for names like Stephanus and Forcellini; indeed, the whole collection is decidedly Anglo-centric. However, the book does not purport to be systematic, and within its chosen remit it offers up a variety of interesting and informative essays.

After a paragraph of preface and a brief introduction, the book plunges into the past with Eleanor Dickey's discussion of Byzantine lexica. She describes the format and content of a series of entries in different Byzantine dictionaries, comparing them with the information in LSJ. Dickey is a reliable guide through treacherous terrain, and the chapter provides a clear and understandable survey of what the Byzantine lexica had and have to offer. It concludes with the provocative thought that ancient lexica invariably omit the most common words and focus on the rare and unusual, while modern dictionaries do the opposite—the more unusual the word, the larger the dictionary we must consult. She asks whether the modern system is actually useful, inasmuch as it is precisely the rare words that we look up in the dictionary.

Joshua Katz's second chapter treats etymological dictionaries of Greek and Latin. Setting out to investigate the value of having multiple competing etymological dictionaries, the essay quickly turns into an extended musing on the value of historical linguistics and etymology more generally. While sometimes overly discursive—one suspects that neither the audience of this book nor the attendees at the conference needed evangelization on the origins and value of comparative historical linguistics, complete with Sir William Jones address to the Asiatick Society on "the Sanscrit"—the chapter makes an important theoretical point and several useful practical observations. Theoretically, Katz rightly sees etymology as part of intellectual history, and thus recognizes that it can be (for example) just as useful to be aware of a folk etymology that ancient speakers believed in as to know the "true" origin of a word. More practically, he concludes with a candid appraisal of the virtues and vices of the current etymological dictionaries of Greek and Latin.1

In Graham Whitaker's third chapter on lexica that cover a single author, we meet with one of the outstanding strengths of this book, thorough archival research. Whitaker covers a huge amount of ground succinctly and with interest, focusing largely on description and eschewing any generalized typology. He consistently tells fascinating stories that illuminate the background of the lexica he treats: to single out just one of many examples, he studies the slips that J. Enoch Powell used to compile his Lexicon to Herodotus, thus letting us into the lexicographer's workshop and allowing us to see him ply his trade.

The fourth chapter, David Butterfield on the history of that sine qua non of schoolboy versification, the Gradus ad Parnassum, couples bibliographical industry with a keen eye for revealing detail. For readers unfamiliar with the genre, a Gradus is a dictionary that helps in verse-making, giving the prosody of a word and some verses plucked from classical authors demonstrating its scansion and use, often complete with synonyms, epithets, and other helpful hints for the budding versifier. Butterfield traces the development of such works throughout Europe across three centuries, well illustrating both their progress and their tralatitious nature. The chapter concludes with a detailed appendix that lists the major editions of the Gradus from 1652 to 1967.2

The book's central chapter, by Christopher Stray, provides sensitive and nuanced insight into the world of 19th-century English classical scholarship. In another example of first-class history of scholarship and archival research, Stray discusses the history of LS(J), interweaving the process of its composition and revision with the lives and personalities of the people involved. The piece's most valuable contribution is an understanding of the constraints governing the origin and revision of the lexicon. These constraints were intimately bound up in the press's desire to market a product and make a profit: as an example, to save money and simplify revision, the type was sometimes left standing or electrotyped for subsequent editions. This did indeed facilitate revision—but at the cost of allowing for only minor changes. Such a fundamentally conservative process has left a lot of venerable absurdities in its wake, and it goes a long way toward explaining LSJ's current state.

The faults of LSJ are too well known to need rehearsing here,3 and John A.L. Lee's sixth chapter rightly claims it needs serious and substantive revision. The first part of the chapter is a perceptive analysis of the entry ἀγαπητός through successive editions of the lexicon, explaining how it got to be the (problematic and misleading) way it is today. He catalogs LSJ's failings in some detail; I might simply say that it is uninformed by modern lexicographic method and that its formatting is a disaster. In any event, all would agree that the next stage of revision must enter the electronic world, and Lee devotes the last part of his chapter to sketching out a vision of what such a digital lexicon might look like. Unfortunately his prescriptions are both unrealistic and not universally helpful. Although he doesn't phrase it this way, in essence he proposes that the lexicon should contain the same information as a TLL entry with the addition of translations of all passages, and this may be a logical, if lofty, goal. He further suggests, however, that the lexicon contain every single extant instance of each word, all appropriately categorized—an undertaking far beyond the ambit of even the TLL. The full collection and classification of the Greek evidence, even with electronic tools, would take forever, and one cannot imagine finding sufficient money, manpower, and time for such an enterprise in today's world. Perhaps more importantly, it ultimately would not prove especially enlightening: once a word's meaning is securely established, we are primarily interested in later instances only insofar as they deviate from or innovate on that meaning. This review is not the place to put forward a program for revisions to LSJ, but I might suggest that the key issues are those of lexicographic philosophy and principles. Formatting is of secondary importance; so long as the data is all appropriately encoded, its actual display should be infinitely fungible.

In John Henderson's chapter 7 we return to the history of a dictionary, this time that of the OLD. Henderson discusses the project's genesis and usefully explains the origin of certain fateful decisions, like the notorious chronological limit,4 telling the story through the correspondence of the main players, sequences of early specimen entries with comments, plans, etc. Of particular interest is the close relationship between the OED and the OLD in everything from shared lexicographic principles to shared lexicographic workspace to shared lexicographers. Two personalities dominate the OLD's early years, those of Alexander Souter and James M. Wyllie, who taken in tandem were responsible for many of the basic decisions on the layout and arrangement of the dictionary. Both had remarkable rises and falls, which Henderson chronicles in detail.5 After two decades of difficult gestation, Peter Glare took the helm in the mid-1950s and smoothly guided the publication of fascicles to a triumphant and (mirabile dictu!) on-schedule finish in 1982. This all makes for fascinating reading and greatly fleshes out the skeletal "Publisher's Note" found in the OLD itself.

The final two chapters take us to the world of contemporary lexicography, examining two dictionaries currently in preparation. For reasons of space I cannot discuss in detail Richard Ashdowne's ninth chapter on the Dictionary of Medieval Latin from British Sources, which usefully describes the background, scope, history, methods, and future of the lexicon from the perspective of one of its current editors. I will note in passing that the section on present editorial practice contains an excellent description of a lexicographer's daily work.

Chapter 8, by Patrick James, treats the Cambridge Greek Lexicon, a dictionary targeted at intermediate learners. The chapter provides case studies of three words, showing how the CGL's treatment differs from that of LSJ. The criticisms of LSJ are by and large just, and the great virtue of the CGL is its simplicity: it appears well organized and straightforward to use. It offers both definitions and translations, which are typographically delineated and easy to understand. The lexicon is based on a fresh examination of a corpus of canonical Greek authors,6 and it will be disseminated both digitally (in integration with the Perseus project) and in print. My only reservation concerns one puzzling decision that the project has made: in its articles the CGL does not provide references to passages and only rarely gives quotations, preferring English paraphrase. If this dictionary were an abridgement of an existing lexicon and designed for print, perhaps such a decision would be understandable. As it is, however, the lexicographers have done the work of examining the passages afresh and drawing up their articles based on that examination; it seems perverse to discard this useful information. Providing references to passages allows lexicographers to justify their work; it also allows users to check it or arrange the material differently—to say nothing of the fact that illustrative quotations clearly give a deeper sense of the meaning of a word. James defends the decision primarily on the grounds of concision and clarity, but in an online world these considerations must be viewed differently. The lexicon already notes in which authors a given meaning occurs; it would be trivial to make the author's name a clickable link that would expand into the specific passages underlying the definition. In this way learners could have the best of both worlds: a clear and simple presentation that can unfold into more detailed information if they so desire. Furthermore, with some tagging of the electronic data the press could easily decide to issue the dictionary in multiple print versions with no additional effort: the most basic (and cheapest) version containing no citations or quotations, an intermediate version containing references deemed important, and a full version containing all available information.

The book concludes with a general index, including Greek and Latin words discussed, which is not complete but is generally useful and occasionally humorous.7 The book is attractively produced and contains numerous well chosen pictures. Typographical errors are relatively few and almost never such as to cause difficulties.8


Christopher Stray: Introduction 1-4
Joshua T. Katz: Nonne lexica etymologica multiplicanda sunt? 5-24
Graham Whitaker: Ploughing a lone furrow? The single-author lexicon 49-70
David Butterfield: Gradus ad Parnassum 71-93
Christopher Stray: Liddell and Scott: myths and markets 94-118
John A.L. Lee: Releasing Liddell-Scott-Jones from its past 119-138
John Henderson: A1-ZYTHUM: DOMIMINA NUSTIO ILLUMEA, or out with the OLD 139-176
Patrick James: Learners' lexica: the approach of the Cambridge Greek Lexicon 177-194
Richard Ashdowne: Ut Latine minus vulgariter magis loquamur: the making of the Dictionary of Medieval Latin from British Sources 195-222
Index 223-229


1.   In brief: the German ones (Frisk, Walde-Hofmann) are in some sense more thorough, but the French (Chantraine, Ernout-Meillet) are better on a word's changes through time (les histoires des mots); the very recent Dutch offerings (Beekes, de Vaan; both written in English) do not always supersede the earlier works and are influenced by the controversial "Leiden school" of Indo-European linguistics.
2.   A minor addition: Butterfield dates the first vernacular-Latin Gradus to 1890 (Ainger and Wintle); at least by Koch's 1879 revision of Sintenis a basic German-Latin appendix is to be found.
3.   The interested reader can start with the introduction to John Chadwick's Lexicographica Graeca (Oxford 1996) as well as id., "The Case for Replacing Liddell and Scott," BICS 39 (1994) 1-11.
4.   This limit (the end of the 2nd century) was present in some of the earliest letters about the dictionary, but already much lamented in committee meetings by the early 1950s. For one of its most scathing indictments, see F.R.D. Goodyear, "The Oxford Latin Dictionary," Proceedings of the African Classical Associations 17 (1983) 124-36 = K.M. Coleman, J. Diggle, J.B. Hall, and H.D. Jocelyn (eds.), F.R.D. Goodyear. Papers on Latin Literature (London 1992) 281-7.
5.   One might have expected more on how Wyllie came a cropper; his spectacular meltdown is only alluded to.
6.   "The major authors now studied in schools and universities from Homer to Xenophon … and Aristotle's major works, Theophrastus' Characters, the better preserved plays of Menander, the major Hellenistic poets (Callimachus, Apollonius of Rhodes, and Theocritus), Polybius, Plutarch's Lives, and the New Testament gospel books and Acts of the Apostles" (191 n. 6).
7.   The Greek index of Words Discussed, for example, lacks the case studies of θωρήσσω and θεραπεύω from James's chapter. For humor see e.g. "Callimachus, J.T. Katz no."
8.   The most serious at p. 116 n. 67 1899-2008 for 1899-1905 (?). There are occasional slips in Greek and Latin: p. 13 ἀσχοί for ἀσκοί, p. 34 κώμος for κῶμος, p. 74 parsum for sparsum, p. 108 μη for μή, p. 111 λοιπὸν for λοιπόν, p. 142 ἀσφαλές for ἀσφαλὲς, p. 183 somnum for somnus, along with assorted other trivialities.

1 comment:

  1. I am at a loss for words at John A. Lee's suggestion for the future of the Great Scott: "Although he doesn't phrase it this way, in essence he proposes that the lexicon should contain the same information as a TLL entry with the addition of translations of all passages, and this may be a logical, if lofty, goal."---TRANSLATIONS OF ALL PASSAGES . . . really??? When I got my PhD from Berkeley, the main rational of our very challenging Greek and Latin "at sight" exams was that we be able to translate for ourselves--even better, UNDERSTAND WITHOUT TRANSLATING basically any small piece of Greek or Latin set before us. THAT was a lofty goal. Any one who needs to have all the passages translated for them has NO BUSINESS with their face such a work at all. Or is it desirable to add yet one more electronic tool for the hands of those who would feign scholarly excellence in classical studies when the truth of the matter is that they can barely translate themselves out of a paper bag! Shame, shame, shame. Where is the pride? Where is the excellence? It has taken ME thirty years of Greek study to get where I am, I the world be damned if I myself at least will approve of a tool that allows . . . (I leave the reader to continue the rant!). On the positive side, I DO look forward to reading Graham Whitaker's account of Powell's creation of the Herodotus Lexicon since I just spent a large part of last year reading the whole of the Histories (minus Bk II) with Powell's wonderful work.