Thursday, March 17, 2016


Stefano Valente, The Antiatticist: Introduction and Critical Edition. Sammlung griechischer und lateinischer Grammatiker (SGLG), 16. Berlin; Boston: De Gruyter, 2015. Pp. xxviii, 292. ISBN 9783110401479. $140.00.

Reviewed by Eleanor Dickey, University of Reading (

Version at BMCR home site


In the second century AD Greek writers made great efforts to use the vocabulary of the fifth and fourth centuries BC, a practice known as Atticism. Since the Greek language had changed immensely during the intervening half millennium, Atticism was difficult to practise, and writers' resulting linguistic insecurity gave rise to numerous lexica of Classical words. The 'Antiatticist' is one of these second-century lexica; its name, which was only attached to the work in modern times, comes from the fact that this work argues for a less strict approach to Atticism than some other surviving lexica.

For some writers, Atticism entailed using the most impressive language available: avoiding words current in ordinary second-century Greek whenever they could be replaced with ones that were no longer current. This often led to obscurity. But the Antiatticist argued that current words were acceptable as long as they were attested somewhere in the Classical canon, a canon that he defined more broadly than many Atticist writers by including authors such as Herodotus, Lycophron, and Menander as well as the standard fifth-century Athenian writers and Homer. The Antiatticist's work is therefore a valuable source of fragments from a wide range of lost works, which are cited to prove attestation of particular words; the lexicon is also interesting for understanding the Atticist movement itself and the diversity of views about language and literature in the second century AD.

Unfortunately the Antiatticist's lexicon has suffered considerably in transmission, being severely abridged both by elimination of entries (as Valente points out, the lexicon gets scantier as the alphabet progresses, with 44 entries for beta but only 10 for tau) and by omission of material from surviving entries. The lost references and quotations, in particular, are a great pity from our perspective, but nevertheless much useful material remains.

Until the publication of Valente's edition, the most recent edition of the Antiatticist was that of Bekker from 1814.1 A new edition after 201 years is very welcome, even though no new manuscripts of the text have surfaced in the intervening period: there is only one surviving manuscript, the tenth-century Coislin 345 in the Bibliothèque Nationale in Paris, and Bekker's transcription of this manuscript was perfectly good. Valente's edition is an improvement chiefly because of the incorporation of two centuries of scholarship into the text and the addition of a wealth of related material: in comparison to the 40 pages the lexicon takes up in Bekker's edition, it occupies 167 pages in Valente's edition. Some of the expansion is due to Valente's helpful addition of modern references to the texts cited by the ancient writer, but most is due to three large apparatuses. The first 'engages with problematic or uncertain identifications of the loci classici, as well as with the context of the glossed expression' (p. 61). The second gives parallels in the indirect tradition (i.e. other texts that seem to preserve extracts from the Antiatticist's original lexicon, which was evidently rather different from the version preserved in the manuscript), and the third is the apparatus criticus proper, dealing with the manuscript readings and emendations to the text of this version of the lexicon. In practice it is not always easy to tell how information has been divided among these apparatuses, and one usually ends up having to look in more than one of them — but that is not difficult, as the typography and layout of the apparatuses makes it easy to find the information one is looking for.

Valente's text is notably different from Bekker's. I investigated as a sample the section for words beginning with the letter epsilon, which has 132 entries in Valente's version and 133 entries in Bekker's; in this sample there are 49 differences in wording between the two editions, as well as 12 differences in word division, brackets, diacritics, or punctuation (not counting differences in punctuation that do not affect the sense). Of those 61 differences, 21 consist of numbers that Valente gives in figures and Bekker spells out (e.g. Πλάτων ε΄ Πολιτείας in Valente versus Πλάτων πέμπτῳ Πολιτείας in Bekker for 'Plato [uses this word] in the fifth book of the Republic' in entry ε 55); here Valente's usage follows the manuscript. A further 10 consist of orthographic variations that have no effect on the sense, such as οὕτως in Valente versus οὕτω in Bekker (ε 22 etc.), Ποσίδιππος in Valente versus Ποσείδιππος in Bekker (ε 36), and κἂν in Valente versus κἀν in Bekker (ε 68). If one sets aside all these, there remain 30 substantive differences in the 132 entries, or an average of one every 4.4 entries.

In most of these differences Valente's text is arguably an improvement over Bekker's: for example, Valente's combination of two entries into one in ε 21, Valente's Ἑλλάς· ὁ ἀνήρ. Σοφοκλῆς Αἴαντι Λοκρῷ 'Hellas: the man's name. Sophocles uses this in the Locrian Ajax' versus Bekker's Ἑλλὰς ὁ ἀνήρ: Σοφοκλῆς Αἴαντι Λοκρῷ in ε 119, Valente's ἐχόμενον· ἀντὶ τοῦ ὕστερον '"held back": instead of "behind"' versus Bekker's ἐχόμενον· ἀντὶ ὑστέρου at ε 72 (here Bekker's text is closer to the manuscript), Valente's correction of erroneous book numbers at ε 6, ε 63, and ε 105 (but erroneous book numbers are left uncorrected when the errors cannot easily be explained palaeographically, e.g. at ε 38 and ε 83; there are so many wrong references that one wonders if the Antiatticist sometimes just made them up). Many of the changes are Valente's own ideas; others were suggested by earlier scholars, who are duly acknowledged in the apparatus.

A few of the changes to the text, however, are not improvements, including several that seem to be typographical errors: Valente's ungrammatical τὴν ἀριστερὰ χεῖρα versus Bekker's τὴν ἀριστερὰν χεῖρα 'the left hand [acc.]' in ε 116, Valente's meaningless φιάλη τι ὅμοιον versus Bekker's φιάλῃ τι ὅμοιον 'a thing like a bowl' in ε 130, and Valente's οὐξηγητής versus Bekker's οὑξηγητής (= ὁ ἐξηγητής 'the exegete') in ε 105 (in the last example Valente's apparatus also mis-reports Bekker's reading). Such errors are easy to make, of course, and one should not reject a useful edition like this one merely because it contains mistakes. Nevertheless typographical errors detract significantly from the value of an edition, and to judge from my sample they are considerably more frequent in Valente's edition than in Bekker's. The result is that Bekker's edition is not completely superseded by Valente's work: readers will need to check Bekker to verify the accuracy of Valente's readings.

Valente (unlike Bekker) numbers entries consecutively within each letter of the alphabet, a very useful feature as alphabetization is by first letter only and some sections are long. Bekker's page numbers are given in the margin to enable users to follow up references using those pages; again this feature is useful, as is the concordance of Bekker numbers at the end of the book.

The edition proper is preceded by 108 pages of preliminary material. Most of this consists of a wide-ranging introduction divided into the following sections: 1 'Status quaestionis', a brief overview of modern work on the Antiatticist; 2 'The textual tradition', a lengthy explanation of the single manuscript and the complex indirect tradition; 3 'Sources', a discussion of the relationship between the Antiatticist and other lexica; 4 'Lexicographic typology', an extremely interesting overview of the different type of entry found in the Antiatticist (e.g. X ἀντὶ τοῦ Y, indicating that word X is found with the allegedly post-Classical meaning Y in the Classical author named in the entry; οὔ φασι δεῖν λέγειν X, ἀλλά, indicating that word X, despite having been condemned as post-Classical by some Atticists, is attested in the Classical author named in the entry; ἐκβάλλουσι X, indicating that some Atticists have condemned the word X as post-Classical and providing evidence of its Classical attestation); 5 'The Antiatticist and the Atticist lexicography', an investigation of the Antiatticist's relationship to Phrynichus, Philetaerus, Pollux, and others; 6 'Authorship, dating, and nature'; and 7 'Principles of the present edition'. These are followed by an 'Appendix' reproducing Ruhnken's notes on the Antiatticist, made in 1755; the notes are of significant historical interest, and their presence here is an excellent bonus.

Basically the introduction is very good; the discussion of the indirect tradition will be of crucial importance to future work on the Atticist lexica and their transmission, and the explanation of lexicographic typology is a vital tool for understanding the Antiatticist. But for many readers these advantages will be obscured by the fact that the introduction is distinctly less user-friendly than it might have been: many people will be unable to follow much of the discussion, and in a few places I could not follow it myself. The main problem is the level of background assumed: this work was obviously not designed to be used by people not already familiar with the Antiatticist and the tradition in which he operated, and that restriction of audience is sad given how few Classicists have the necessary background. A second problem is occasional errors (e.g. on p. 18 Valente discusses a scribe's writing ΛΛΑ for ΛΛ, when what he means is ΑΛΛ for ΛΛ), which make it difficult for the reader to decode the text. Non-native-speaker English also causes comprehension difficulties, for example 'Furthermore, it is impossible to attribute every single entry of an ancient Greek lexicon to a well-determined source, other than it happens in the editions of Byzantine lexicographical works' (p. 31). Of course, Valente is not a native speaker of English, and it was kind of him to write in English at all: the blame here must lie with the publisher, since a press of De Gruyter's standing, charging the astronomical prices they do, ought really to exercise better quality control over their books.2

Another factor that will cause comprehension difficulties is the abbreviations. The work begins with a list of abbreviations and sigla that is 20 pages long and nevertheless omits many signs and abbreviations used in the introduction, text, and apparatuses.3 The omissions are unkind to the reader, and the amount of abbreviation used (particularly in the introduction, which being narrative prose should have been largely free of abbreviations) is high enough that even if every abbreviation used had been included in this list, many readers would still have found the abbreviations an obstacle to comprehension.

Overall, this publication is very welcome, not only for the edition, which despite its occasional errors will no doubt replace Bekker as the standard text of the Antiatticist, but also for the introduction. Nevertheless an opportunity to make this very interesting text more widely accessible has been missed because of carelessness and insufficient attention to the needs of readers.


1.   I. Bekker, Anecdota Graeca I, Berlin 1814, pp. 75-116; this text is available on the TLG.
2.   The situation is not improved by misspellings, such as 'Hernnius' for 'Herennius' (p. 39) and 'outern' for 'outer' (p. 25).
3.   Among the omissions are curly brackets { }, which are used to mark letters present in the manuscript that Valente would like to delete (e.g. in ε 15), and Σ*, which seems to stand for the Συναγωγὴ λέξεων χρησίμων. For the symbols Σ΄, Σ΄΄, Σ΄΄΄, etc. the reader is referred to p. 14, which does not explain (or even use) them; these are expanded versions of the Συναγωγή. A beautifully clear explanation of the different versions of the Συναγωγή can be found in I. C. Cunningham, Synagoge, Berlin 2003. In fact, readers who cannot understand Valente's introduction should read Cunningham's introduction and then try again, as much of the background Valente assumes is provided in comprehensible form by Cunningham.

No comments:

Post a Comment