Bryn Mawr Classical Review

BMCR 2018.11.52 on the BMCR blog

Bryn Mawr Classical Review 2018.11.52

Carlo Vessella, Sophisticated Speakers: Atticistic Pronunciation in the Atticist Lexica. Trends in Classics – Supplementary Volumes, 55.   Berlin; Boston:  De Gruyter, 2018.  Pp. xvi, 320.  ISBN 9783110440409.  €99,95.  

Reviewed by Eleanor Dickey, University of Reading (


It is well known that in the second century AD many Greeks attempted to erase centuries of language change and write in the language of fifth-century BC Athens. This Atticism was primarily a feature of writing; it could also affect speech in matters such as lexical choice, but linguistic features not reflected in writing, such as pronunciation, are normally thought to have remained untouched by Atticism.

Challenging this traditional view, Vessella’s book provides considerable evidence that Atticism sometimes extended to pronunciation. Its core is a set of passages from the Atticist lexica that can be argued to prescribe Atticistic pronunciation. Of course, most prescriptive statements that might refer to pronunciation could also refer to spelling, so the focus of this work is on identifying statements that must originally have concerned pronunciation as opposed to spelling. Many of Vessella’s passages are dubious in this respect, but some (such as the statement that the alpha of πελαργός should be pronounced short, p. 238) appear close to unassailable, and overall there are enough convincing passages to demonstrate the author’s thesis: Atticism probably did extend to pronunciation.

But how could Greek speakers of the second century AD know how words had been pronounced in Classical Attic? Often, of course, they did not really know. Sometimes they used the pronunciation found in the Attic of their own day as a guide, and sometimes they guessed or extrapolated from the Classical Attic features they knew, producing phonetic ‘hyperatticisms’.

The work begins with an overview of the evidence for Atticising pronunciation. This evidence is not confined to the Atticist lexica, and indeed some of the most interesting passages occur in other texts. Vessella has done a good job of gathering these and discussing them in the first chapter, but he gives them a subordinate role: the book’s focus is on evidence in the Atticist lexica, and therefore only the passages from the lexica appear in the central passage-by-passage discussion.

The first chapter also provides an introduction to relevant ancient scholarship, particularly the Atticist lexica that make up the corpus on which the book is based: Antiatticista, Aelius Dionysius and Pausanias, Phrynichus, Moeris, Philemon, Philetaerus, Herennius Philo/Ammonius, Pollux, and lexica found on papyrus.

Atticistic pronunciation, according to Vessella, concerned the timbre (quality) of vowels, the quantities of vowels, and gemination (whether consonants were single or double); apart from the question of gemination, there does not seem to be any evidence for Atticizing pronunciation of consonants. Vessella also considers the features marked by accents and breathings to belong to pronunciation rather than spelling, since accents and breathings were rarely written in the second century AD. The features Vessella identifies as ones for which Atticistic pronunciation was possible all probably underwent changes between the fifth century BC and the second century AD, and some understanding of those changes is needed in order to evaluate ancient prescriptions that might pertain to pronunciation. Accordingly, chapter 2 is devoted to vowel timbre, chapter 3 to vowel quantity, chapter 4 to prosodies (accents, breathings, and other diacritic signs), and chapter 5 to degemination; the discussion proceeds sound by sound, examining the evidence for pronunciation changes and comparing it to the pronouncements of Atticist lexicographers. This discussion is detailed (totalling more than 80 pages) and makes use of the most important modern literature on the vexed question of the chronology of Greek vowel changes. It will be of interest to specialists but is best avoided by those without prior background, both because it is not very clear and because it sometimes makes assumptions not shared by many English-speaking classicists, without flagging them as contested.

For example, on p. 68 it is assumed that a stress accent is simply incompatible with the existence of distinctive vowel quantity: when a language has one, it cannot have the other. Many American and British scholars will find this puzzling, since they are likely to consider classical Latin to be an obvious counterexample, that is, a language with both distinctive vowel quantity and a stress accent. In fact the assumption is not as odd as it looks in that context, for many continental scholars believe that classical Latin had a pitch accent — for them, the incompatibility of stress accents and vowel quantities makes more sense. But Vessella does not explain that he is operating in that framework, nor does he defend his argument against the objections of those who do not share his framework. Likewise, on pp. 96–9 Vessella claims that the Atticists prescribed the position of accents more often than their types (i.e. acute or circumflex) — but he does not adequately make clear that this claim is only valid if one believes that prescriptions specifying both position and type were really intended to refer only to position. The words used by ancient scholars for labelling accents automatically specified both position and type (e.g. προπαροξυτόνως ‘with an acute on the penultimate syllable’, περισπωμένως ‘with a circumflex on the final syllable), so in fact most of the words listed in this section as having only their positions mentioned are in fact described in Atticist lexica with terms that specify the accent’s type as well.

Next comes a passage-by-passage discussion of nearly 200 passages from the Atticist lexica. These discussions vary in quality, but many are good and interesting, at least for specialists. Even if readers are likely to remain unconvinced about many of the passages, Vessella’s discussions usually represent the most thorough investigation that these passages have ever received and are therefore useful for anyone with a serious interest in the lexica. The work ends with an impressive set of concordances and indices.

This book was written for specialists, and many non-specialists will find it difficult or impossible to read. The crucial passages from the lexica, for example, are not translated, and some of them are difficult to understand without a translation; each passage is discussed in detail, but these discussions are hard to understand or evaluate if one cannot read the passage itself. (Other passages of Greek are scrupulously translated, so the non-translation of the passages under discussion must have been a deliberate attempt to avoid prejudicing the argument.) Tables sometimes present data without adequate labelling or explanation, leading to confusion.1 Potentially unfamiliar symbols are frequently used, sometimes unnecessarily and often without any explanation. (Note in particular that the different types of brackets are explained on p. 39, the symbol → indicates a cross-reference to another part of the book, and the two little triangles arranged like a colon and placed after a vowel indicate that it is long, as a macron would; this symbol and others not mentioned here come from the International Phonetic Alphabet, to which a key can be found at Similar issues arise with terminology; for example ‘isochrony’ (meaning the loss of distinctive vowel quantity) is defined on p. 67 but used from p. 48. Arguments are sometimes presented in a more difficult and convoluted fashion than is necessary, and/or without the information needed to follow them easily. The English is often poor, with misused words and ungrammatical sentences; usually it is possible to work out what was intended, but reading takes more effort than some readers will be prepared to invest.2 (A hint for those inclined to try: in this book ‘Hellenism’ means ‘the Hellenistic period’, ‘to base (on)’ means ‘to be based on’, ‘lecture’ means ‘reading’, ‘rising’ means ‘raising’, ‘vital’ means ‘productive’, and ‘redoublement’ means ‘doubling’, not ‘reduplication’.)

The presence of enough English mistakes to diminish readability is disappointing in a book with such a high price tag: could De Gruyter really not afford to employ a copyeditor with a decent command of English? The author, who is not a native speaker of English (and whose English is far better than my Italian), may have thought that he had protected himself against this outcome by seeking out a publisher that on its website promises authors the ‘highest publication standards’.

For those with the background and the patience to read and understand it, however, this book makes a real contribution to our understanding of the Atticist movement and its conception of classical Attic.


1.   For example, the table on p. 37 is taken from another work and cannot be understood without looking in that other work, while the table on p. 89 has headings that only make sense in the context of a note that does not appear on the same page.
2.   For example, ‘Moeris for instance includes Homer and Herodotus, but not tragedy, which together with Old Comedy distinguishes Phrynichus’ canon form other lexica basing mostly authors of Attic prose.’ [p. 35]; ‘These glosses either discuss the accentuation of difficult words, sometimes apparently basing on accented manuscripts or identify Attic forms: in particular various instances of non-Koine accentuations due to Vendryes’ Law, an accent shift exclusive to Attic.’ [p. 120]; ‘The entry in Ammonius is not prescriptive, as it is normally the case with entries in synonymical lexica.’ [p. 125; does ‘normally the case’ apply to ‘prescriptive’ or ‘not prescriptive’?]; ‘The ancient explanation basing on ἄδην has been abandouned (note also that Phrynichus does not explain the change in breathing, a compound of ἄδην should have a rough one). ... The iota can serve the only purpose of signalling the long vowel, therefore implying that the ‘long diphthong’ <ᾳ>/<αι> had already merged with [a:].’ [p. 130]. Mistakes have also been introduced into quotations from other scholars; for example ‘The diction that is appropriate for Solemnity consists of broad sounds’ has been turned into ‘The diction that is apporpriate for Solemnity consists on broad sounds’ [p. 65].

Read comments on this review or add a comment on the BMCR blog

Read Latest
About BMCR
Review for BMCR
Support BMCR

BMCR, Bryn Mawr College, 101 N. Merion Ave., Bryn Mawr, PA 19010