BMCR 2021.06.35

Response: Thompson on Goldhill on Stray, Clarke and Katz, Liddell and Scott: the history, methodology, and languages of the world’s leading lexicon of Ancient Greek

Response to 2020.11.11

Response by

I would like to respond to Professor Goldhill in relation to my paper in this volume ‘Βάπτω: An Illustration of the State of our Ancient Greek Dictionaries’ and the tricky subject of lexicographic definition. My use of the phrase ‘true lexicographic definitions’ (p. 357 of the book) is quoted by the reviewer as though I was asserting that the use of definitions is the best methodology for a lexicon entry, as opposed to glosses, or indeed any other style. As the reviewer says, the choice is not so simple, and in fact I agree.

In this context, a definition is one designed to give a precise description of the meaning of a headword, consisting usually of a phrase, though occasionally it may be a single word provided that, to all intents and purposes, it is a direct equivalent. A gloss is a word that may be used as a translation in some passages but in its full semantic range is only a partial equivalent of the headword.

I should point out first of all that my paper in the volume is not focussed on any particular methodology. It is about a re-reading of texts with a view to building an ongoing database with a full reassessment of the senses of dictionary headwords, along with increased emphasis on typical contexts, sociolinguistic factors, and the relationship with other words in the language. It need not contain an exhaustive gathering of all attestations, only a reasonable sample taken from existing dictionaries with the addition of some new material, especially from later periods not well covered. In this way, repetition of labour could be avoided in future projects, and it would become easier to concentrate on achieving greater accuracy than in our existing dictionaries, where time-pressed (and usually understaffed) copying from predecessors and patching with new material have reached their limits. Unlike the fresh start made for The Oxford Latin Dictionary (on which full-time work began in 1933), the first edition of Liddell and Scott (1843) was based on a German work (also not entirely original), and was revised through several editions until the ninth (1940, known as LSJ or LSJM), and later with a Supplement that was in its turn also revised (1968 and 1996). Franco Montanari’s Vocabolario della Lingua Greca (1995, with further editions in 2004, 2013) has innovations but remains derivative of LSJ —similarly, the versions of this in other languages: Σύγχρονο Λεξικό της Αρχαίας Ελληνικής Γλώσσας (2013), The Brill Dictionary of Ancient Greek (2015), and Wörterbuch Griechisch-Deutsch (2021). The incomplete but ongoing Diccionario griego-español (1980-), founded by Francisco Rodríguez Adrados, acknowledges a debt to LSJ and retains something of its style, but is an independent re-evaluation of all sources, including post-classical texts omitted from the other work. The intermediate-sized Cambridge Greek Lexicon (2021), where a complete re-reading of the source texts was facilitated by the limited number of authors covered and the timely advent of good digital tools, represents a significant move of a different kind away from the Liddell and Scott tradition.

A general problem, however, is that there has been a tendency amongst those embarking on new projects to plan for increased coverage, improved layout, and modernizing the language of the entries, rather than a thoroughgoing reassessment of the senses of basic vocabulary words, the need for which can be less obvious. It remains the case that ancient Greek vocabulary has yet to be explored fully and systematically. The task may look impossible, but if at the outset only a selection of words were examined, without the pressure of having to complete a whole dictionary, it would be a great advance. This is not a myopic dive down a rabbit hole but a plea for cooperative study of the details in order to create a firmer base for working on new styles of presentation for both learners and advanced scholars and, importantly, what the reviewer refers to as the ‘bigger picture’.

It is claimed by Professor Goldhill that the philologist contributors to this book all insist that lexicographic definition is the right route to follow. It is not quite clear which names are on this list, and it is hard to find in the book any such statement. But even if broadly true, a lot depends on the kind of dictionary being written, and the amount of space available. The Oxford Latin Dictionary (1982, 2nd edition 2012) employs what might be called a ‘mixed method’ of both definitions and glosses, successfully I would argue, along with quotations that also bear some of the load of describing meaning. The overriding consideration in any methodology should be the identification of senses and other relevant material with as much accuracy as possible, expressed in succinct and unambiguous language.

Some misunderstandings between the makers and users of dictionaries, even in philologically-oriented fields, stem from difficulties with terminology. Amongst contemporary biblical Greek lexicographers, the distinction between definition and gloss is generally accepted, with the former being seen as a superior and necessary method, though it is always allowed that sometimes a single word may suffice as a definition: e.g. ὑετός rain. It may be noted that in a monolingual dictionary the situation is different, as in The Oxford English Dictionary where the first section of the entry for ‘rain’ begins with the definitions ‘condensed moisture of the atmosphere falling to the ground visibly in separate drops; the fall of such drops; rainwater’ ( The starting point for widespread use of the definition method in the biblical field was the publication of Johannes Louw and Eugene Nida’s Greek-English Lexicon of the New Testament based on Semantic Domains (1988), the first to apply it in a thoroughgoing way. The landmark significance of this is set out in John Lee’s A History of New Testament Lexicography (2003). Frederick Danker introduced the method into his third edition of Walter Bauer’s A Greek-English Lexicon of the New Testament (2001, known as BDAG), using the term ‘extended definition’ in his Foreword (p. viii). Like LSJ, Bauer’s Lexicon had already run through quite a few editions in both German and English. These definitions in BDAG are highlighted in the text by a different font style. Danker continued the use of definitions in his shorter and more individual work The Concise Greek-English Lexicon of the New Testament (2009). Takamitsu Muraoka explains his own approach to definition method in the Introduction (p. XII) to his work A Greek-English Lexicon of the Septuagint (2009). Both Danker and Muraoka make the point that one of the benefits is the extra help given to readers who are not native speakers of the lexicon’s target language.

The distinction between definition and gloss however is not well established for classical Greek lexicography, where traditionally ‘definition’ is the term used by writers and readers for any rendering of the the headword in an entry. In practice, the gloss method prevails in Liddell and Scott, as well as in the Diccionario griego-español, and the Montanari dictionaries. The Cambridge Greek Lexicon sporadically adopts what are termed ‘definitional phrases’ (see p. xi), but this refers rather to an element of an entry that is in a specific position and font style. This may mislead some into thinking that the intention is the same as with Danker’s system in BDAG, but the phrases are actually of a mixed type. Some are descriptions of the precise meaning of headwords, but most are devices to illustrate word formation or to clarify distinctions between senses. For example, under βραβεύς, ‘one who ensures the rules of a competition are enforced’ is a definition of the lexical sense in the first section. Under ῥόδινος, ‘made or derived from roses’ is a definition, though here it also mirrors the formation of the headword, giving proper force to the suffix. ‘of roses’ is more usual as a translation. Under γραφή, ‘drawing of lines or application of colour’ and then ‘that which is drawn or painted’ are phrases in somewhat artificial language chosen for the purpose of distinguishing the activity from the result of the activity, where the translation words ‘drawing or painting’ can serve for both but do not signal the difference. Focus on this one element of the sense distinction, albeit for very practical purposes, does however mean the phrases are a step away from precise definitions. Under φιλοκίνδυνος, ‘fond of danger’ is a literal rendering of the elements of the compound, translated in its good sense as ‘bold, intrepid’and ‘adventurous’, but the phrase is not precise enough to be considered a definition.

In trying to avoid confusion in the terminology, I used the expression in my article ‘true lexicographic definition’ as opposed to ‘gloss’, but did not intend to say more than that a phrase has the power in some cases to reduce the imprecision of a gloss. I also had in mind a distinction from some of the types of ‘definitional phrases’ in The Cambridge Greek Lexicon. The case in point was dip, the principal gloss in LSJ for the verb βάπτω, which has more than one sense in English, giving rise to a danger of ambiguity. If the reader already knows the meaning of the Greek word but is seeking further elaboration, or if the reader has the competence to understand the untranslated Greek quotations that follow, then this will probably in an individual instance not matter much, especially when usually there will be the context of a particular passage in front of the reader. Definitions can however sharpen the appreciation of meaning, and over a whole text this can add up substantially. There might otherwise be a series of slight misses when words are looked up. Definitions can also be useful in calling attention to differences between subsenses and between words with a similar meaning, important because these analyses have rarely been carried out as well as they might be for ancient Greek.

However, even if such a method is adopted, there is nothing straightforward about it. Editors differ considerably in their approaches and theory is often not put into practice successfully. Errors and infelicities creep in. I stressed in my paper at several points the risk of inadequacies. To put it in an extreme way, there can be as many wordings as there are definers, even different wordings by the same definers if they attempt a definition of the same word on different days. How many of us would agree that The Oxford English Dictionary’s entry for ‘rain’ above is perfectly worded? The choice of the most accurate glosses to represent a headword can be problematic as well. Traditionally, the art of defining is where the training of lexicographers by those who already have long experience comes in, although now many are looking to formulate a more scientific and published procedure to replace, at least in part, what has hitherto tended to be somewhat intuitive and handed down by word of mouth.

Neither definition nor gloss method excludes the desirability of other kinds of information in an entry, such as quotations, grammar, syntax, dialect, register, context, sociolinguistic comment, frequency. The suggestion by the reviewer that I am at odds with David Goldstein’s contribution in this book on the particle γε is unfounded. Indeed, no lexicographer would argue that such a headword can be defined solely in terms of lexical meaning without reference to function, unless for a simple glossary with word-for-word equivalents. Since I was writing about a verb as a single illustration of problems with our dictionary entries as they stand, I did not think to mention that different parts of speech require different approaches.

The reviewer divides the book’s contributors along the lines of philologist, linguistics expert, literary critic and sociologist, but the very relevant category of lexicographer is barely mentioned. This is not unexpected in that the lexicographer is not always acknowledged as a specialist who requires years of training and experience, the work more often being seen as something on which any advanced scholar is equipped to embark. A good comprehensive lexicon requires linguistic, literary and lexicographic expertise —all three—, and there can be no good result if any one of these is set against the others. Expertise from other fields, such as history, philosophy, medicine, law, botany, and so forth, must also come into it. Lexicography is about cooperative teamwork and is not suited to competitive individual scholarship.

There is no need to assume that in the future we shall want to compile for ancient Greek only one kind of dictionary or lexical reference work, as is amply illustrated by the variety of ideas presented in this book. Electronic publications have the potential anyway to present material in multiple ways. On the other hand, dissension over what needs to be done is not necessarily the rule. There is evidence for some agreement amongst the contributors, such as in favouring a more context-driven approach than in LSJ in terms of both the language and the real world of the speakers.

The reviewer’s suggestion that the construction and politics of LSJ should be discussed by teachers and students is most welcome. We badly need a greater understanding of lexicography, a subject not normally a component of the Classics curriculum. In time this would lead to the development of better lexical reference works. No one would disagree with Martin West that LSJ is a lexicon of ‘remarkably high quality’, but much of its inspiration is very old, from before the time of modern linguistics, indeed all modern classical scholarship. If we keep clinging to LSJ we are denying ourselves to an unimaginable extent a better understanding of ancient Greek.