Bryn Mawr Classical Review

Bryn Mawr Classical Review 2004.06.04

Andreas Willi, The Languages of Aristophanes. Aspects of Linguistic Variation in Classical Attic Greek. Oxford Classical Monographs.   Oxford:  Oxford University Press, 2003.  Pp. xiv, 361.  ISBN 0-19-926264-0.  $74.00.  

Reviewed by Jerker Blomqvist, Lund University (
Word count: 2647 words

The language of Greek comedy was the theme of a recent collection of essays under the editorship of Andreas Willi, the author of the book under review here.1 Willi's new book is in one respect more limited in scope than the previous collection, since it concentrates on Aristophanes and says little about the other comedians. On the other hand, it takes up one of the themes of the previous book and develops it with the amplification it deserves, viz., Aristophanic comedy as a reflection of the sociolinguistic situation in contemporary Athens.2 Consequently, as the author himself points out (pp. 1-2), the work has some affinities with Victor Ehrenberg's classic The People of Aristophanes of 1962; it is a book, not in the first place about Aristophanes, but about Athens as mirrored in the more or less distorting looking-glass of Aristophanic comedy. Ehrenberg's target was social and economic life in general, Willi's is the linguistic situation, i.e., what different varieties of Attic Greek existed in the last decades of the fifth century BC, what (groups of) people used them, in what socio-cultural contexts they were used, the existence of specialized terminologies in certain lines of trade, and the characteristic features of women's and foreigners' talk.

Ehrenberg's and Willi's choice of Aristophanes as source material for their investigations is both natural and necessary. As is well known, sociological studies of classical Athens, whether focussed on economy, religion, politics, language, or culture in general, are hampered by the dearth of relevant source material. Archaeology provides little that allows an unambiguous interpretation; inscriptions are not very illuminating either; and other literary texts are precisely literary. Only the comedies, though literary as well, permit more than occasional glimpses of the ordinary Athenian citizen's everyday life and attitudes, ambitions and practices, prevailing in population groups outside the political and intellectual élite.

But even if its focus is not so much on Aristophanes per se but on his background and environment, the book also makes important contributions to the interpretation of the individual comedies, or of particular passages in them, and to the study of the genre of comedy. The second chapter of the book ends with a summary that epitomizes its conclusions under three different headings, "(1) literary, (2) methodological, and (3) linguistic" (p. 47). Summaries with comparable content end the other chapters as well. Not only the varieties of the language and the often complex methodological issues involved in their study but also Aristophanic comedy as an example of Athenian and Greek literature fall within the scope of the book; to quote its author (p. 7): "... the positive facts assembled in this book ... show what a complex structure comic discourse has, how literary styles are composed and configured, and where the potentials and limitations of our analytic methods lie."

Let me say it at once: this is an important book. It opens up a new field of studies of classical Attic, rarely cultivated before, its author shows an excellent understanding both of the empirical data provided by the texts and of the theoretical and methodological obstacles continually surfacing in studies of this sort, and his conclusions are well founded and well presented, their uncertainties pointed out whenever motivated. The book is provided with an ample bibliography (pp. 271-304), likewise ample indexes of passages (305-336), Greek words and phrases (337-350), and general matters (351-361). Its usefulness is further increased by an appendix entitled "Aristophanes' Attic: A Grammatical Sketch" (pp. 232-269), intended for relatively advanced classical scholars who know Attic grammar but have not familiarized themselves with Aristophanes' individual choices between existing phonetic, morphological and syntactical alternatives (say, between ἤν and ἐάν or between infinitive and personal mode in ὥστε clauses). Those relatively advanced Hellenists also seem to be the intended readership of Willi's book.

The bulk of the book consists of eight chapters, including "Introduction" (ch. 1) and "Conclusion" (ch. 8). Chapters 2-7 treat "Religious Registers", "Technical Languages", "Scientific Discourse", "Sophistic Innovations", "Female Speech", and "Foreigner Talk". I summarize them one by one.

The Introduction is devoted mainly to the method employed in Willi's investigation. Among his predecessors he points out especially Stephen Colvin and Antonio López Eire. Whereas those two scholars have treated language varieties that are easily definable, viz., non-Attic dialects and colloquial language, Willi raises the question whether a more sophisticated categorization of the linguistic ingredients in comedy could be possible. The book, Willi emphasizes, is not a full treatment of this theme. When selecting the linguistic varieties to be studied, Willi was guided partly by his own observations of "irregularities" in the Aristophanic text which could be suspected to be features of an identifiable variety (or "register"; cf. below) and partly by sociolinguistic literature on modern languages, which may exhibit varieties likely to have existed in ancient Greek as well. After selecting the registers, the next step was to identify the features characteristic of each of them. For this, three techniques were used: (i) stylistic comparisons, by which items which deviate stylistically from their immediate context were identified; (ii) statistical comparisons, by which a tendency of stylistically unmarked features to aggregate in certain contexts could be ascertained; and (iii) cross-linguistic comparisons which take into account relevant phenomena existing in comparable contexts in other languages.

The chapter on Religious Registers starts with a definition of the sociolinguistic term "register": it denotes a variety of a language that is defined by being used in a certain contextual situation. Registers are to be distinguished from dialects, which are geographically defined, and sociolects, defined by social grouping. Register should of course also be kept apart from the less precise concept of style, which is not defined situationally. The chapter discusses two such registers, both belonging to the field of religion, viz., prayer and hymn, of which there are rather numerous examples in Aristophanes. A detailed study of forms of invocation, epithets, speech-act verbs, other vocabulary features, syntax, argument structure, etc. makes it possible to draw fairly reliable conclusions, and in the summary of the chapter Willi concludes that the method chosen by him is effective and likely to be suitable for similar studies of ancient and modern languages. In this particular case Willi's analysis indicates that prayer and hymn were two separate registers in classical Attic; unlike e.g. English, Attic did not have a single category of "religious language". This, in turn, may indicate that prayer and hymn were also culturally distinct, i.e., "religion was not perceived as one unified cultural domain", to use Willi's words. That last assertion perhaps goes a little too far, but the chapter clearly demonstrates the potential of Willi's methodological approach.

Technical Languages are not to be regarded as registers. They are not defined by situation but by topic, i.e., when you discuss technical matters you should preferably use a technical vocabulary, but technical matters could be the topic discussed in a number of different situations. The prime distinctive feature of a technical language is its vocabulary, and Willi's third chapter is mainly devoted to tracing technical terminology in the comedies. This proves a complex operation, for technical words do not always clearly demarcate themselves. Partly, this is due to the fact that ancient Greek, unlike most modern European languages, could not use easily recognizable loan-words for creating a technical terminology. But, as Willi demonstrates, there is also a need for distinguishing between different degrees of technicality of words, based on their comprehensibility among non-specialist speakers of the language and the situations in which the words are used. Willi analyses these methodological issues thoroughly -- the chapter is the longest in the book -- but must content himself with a meager result, in quantitative terms at least: there is little technical vocabulary in Aristophanes. However, precisely that meager result leads to interesting conclusions regarding the linguistic situation in late fifth-century Athenian society: technical vocabulary in most areas was only imperfectly developed and had not reached the independent status it has in modern societies (illustrated by the fact that "scholarly" figures like Socrates and the geometer Meton use very little technical vocabulary). In another field, alleged to have been of particular interest to the average Athenian citizen, viz. the dispensation of law, the technical vocabulary seems to have been so commonly known that its technical character did not make itself felt. In the highly specialized field of medicine there existed a technical vocabulary, but it has left remarkably few traces in Aristophanes. According to Willi, the reason is that doctors used their technical vocabulary only among themselves and used ordinary language when offering their services to laymen. Aristophanes avoided medical terminology since it was not readily comprehensible to his audience -- even if it was so to himself.

'Scientific Discourse', which is put within single quotes by Willi, is intimately linked to technical language. If technical language, as shown in the preceding chapter, was sparingly used, there existed perhaps an alternative to it. This is what Willi wants to show (p. 96): "I will argue that Aristophanes compensated for the absence of high-profile technical language by substituting linguistic material taken from, or linked to, the tradition of Presocratic scientific poetry." He takes most of his material from Clouds, where Socrates and his fraternity in a Pythagorean setting are heard divulging linguistic theories resembling those of Protagoras and a natural philosophy reminiscent of Diogenes of Apollonia. But the language they use is, according to Willi, tinged with the vocabulary of Empedocles' poetry. Aristophanes' purpose here is parody, and presumably neither Protagoras nor Diogenes lent themselves easily to parodying, whereas Empedocles was judged differently by Aristophanes, perhaps because Empedoclean vocabulary was actually used for philosophical teaching at the time. Willi sees a possible parallel in the Derveni papyrus, where Empedoclean vocabulary is also noticeable, although the philosophical content seems to be inspired by Diogenes.

Under Sophistic Innovations Willi treats word-formations that are regarded as typical of the intellectual milieu of contemporary Athens. In the last decades of the fifth century the sophistic movement is supposed to have had an impact on the Attic language. Cross-linguistic comparison with similar processes in other languages makes it probable that the influence of the sophists on language will manifest itself mainly in two strategies, viz., nominalization and typicalization. By nominalization is understood the formation of nouns -- especially abstract ones -- from verb roots but also a preference for adjectives and participles to (other) verb forms. Thus, nominalization refers to linguistic form. Typicalization, on the other hand, is a semantic phenomenon. It refers to the tendency to treat individual phenomena as instances of something more general. On the linguistic level it manifests itself precisely by nominalization (swimming is more general than Socrates swam) and by a preference for non-finite verb forms which are stripped of markers for such individualizing categories as person, time and mood. Under these two headings Willi discusses six linguistic phenomena, verbal compounds in -έω and -άω, active perfects with transitive function, verbal nouns in -σις, abstract nouns in -μα and -ότης, adjectives in -ικός,3 and verbal adjectives in -τέος, plus a number of other nominalization processes. The analysis reveals a considerable amount of significant material, and Willi ventures the conclusion that it is possible to trace an ongoing process of linguistic change in the Aristophanic texts.

In the chapter Female Speech Willi has also been able to use modern studies on genderlects in a number of present-day societies. Earlier studies on women's talk in the classical languages have in many cases been too much guided by the researcher's preconceived ideas of what are the characteristics of female speech. A further difficulty lies in the fact that Aristophanes, too, may have had preconceived ideas of what constitutes female speech. Willi underlines that only statistical analyses of a relevant empirical material can clarify what are genuinely feminine peculiarities, whether cross-linguistic or limited to a certain language. The catalogue of female features in the Aristophanic texts includes syntactic, pragmatic and lexical phenomena, whereas Willi has found no evidence for phonetic or phonological peculiarities.4 Most of the identifiable features can be classified as either signs of politeness (e.g., 'cooperative' βούλει;, avoidance of obscenities, parenthetical ὁρᾶις; eliciting agreement, the confidence-inspiring particle ἀμέλει), colloquialisms (e.g., the final conjunctions ὅπως / ὡς (ἄν) (more subjective than ἵνα, according to Willi), the affective ethic dative, redundant expressions), or innovations foreshadowing developments of Attic in the fourth century or later (e.g., μου and σου for ἐμός / σός, λαλεῖν as a neutral verb for "speak", ὑπάρχω "to be", the neologism καθάπερ).

Foreigner Talk concentrates on the sole section in Aristophanes where a non-Greek speaker plays a prominent part, viz., the appearance of the Scythian archer in Thesmophoriazusae 1001-1225. The Scythian's Greek has been studied previously by Friedrich and Brixhe.5 They both concluded that the characteristic features of the Scythian's idiolect belonged to a low variety of contemporary Attic, appearing only in spoken language at the time but later becoming constituents of Hellenistic Greek. Willi rejects that explanation. Basing himself on recent studies of 'simplified registers', i.e., varieties of languages used by non-native speakers or by native speakers when conversing with non-natives with limited command of the language, he can show that the characteristics of the Scythian's Attic are paralleled by features known from literary representations of foreigner talk in other languages; coincidences with what we may surmise to be peculiarities of low-class Attic are accidental. Evidently, the Scythian has difficulties mainly with reproducing Attic phonetics and morphology; his understanding of ordinary Attic is not defective, for his native Athenian interlocutors do not have to adapt a more simplified version of Greek when speaking to him; there is no evidence for that sort of simplified register in the scene. Willi's analysis also contributes to the general theory on simplified registers. The simplifying devices of the Scythian mostly have parallels in modern-language foreigner talk. But the reverse is not true, for some features that, in studies on modern languages, have usually been supposed to be cross-linguistic characteristics of foreigner talk are not put to use by the Scythian, e.g., omission of the article or other function words and analytic periphrases of verb forms. Modern theory should obviously be corrected on this point.

The short Conclusion first points out the value of identifying Aristophanes' use of discontinuous linguistic characterization for a better understanding of what means he uses for depicting his characters. But the conclusion above all makes it clear that Willi's primary aim was to show how Aristophanes can be used by the linguist of today in order identify the various registers of Attic Greek that were current in late fifth-century Athens. Willi describes his results as "basic but promising". That is a very modest claim, for after Willi's study it is no longer possible to regard classical Attic as the monolithic monument of clarity, beauty and correct usage that both school grammars and much scholarly research makes of it. Rather, we may be tempted to describe the language situation in terms of diglossia. Greek diglossia, it is true, is generally thought to have started with Atticism in the first century AD.6 However, the linguistic norm for Attic literary prose was being established already in early fourth century BC and soon spread throughout the Greek-speaking world; it is not an original creation of the Atticist movement. When Willi points out the existence of linguistic registers existing parallel with this normative literary language, we have reason to ask if one of those parallel registers was so firmly established and characterized by such features that it could be described, in Fresno's terminology, as the Low Variety standing in binary opposition to the High Variety constituted by literary prose. If so, Greek diglossia starts c. 400 BC at the latest. If the relationship between the languages of different literary genres (epic, choral lyric, tragic dialogue, etc) can be described in diglossic terms (as I personally think they could), then it starts much earlier.


1.   Andreas Willi (ed.), The Language of Greek Comedy. Oxford: Oxford University Press, 2002. Pp. vi, 343. ISBN 0-19-924547-9. Reviewed by Stephen Colvin, BMCR 2003.09.08.
2.   In the previous book it is primarily Willi's own substantial contribution that deals with sociolinguistic issues ("Languages on Stage: Aristophanic Language, Cultural History, and Athenian Identity", pp. 111-149), but also Simon Slings' "Figures of Speech in Aristophanes" (pp. 99-109), with its attempt to distinguish between "literary" and "oral" elements, concerns related topics, in particular methodology. Cf. also the editor's introduction, pp. 15-25.
3.   A most impressive instance of this sort of "sophistic" influence occurs at Knights 1378-1381 (quoted by Willi, p. 139), where seven new, more or less parodical, formations in -ικός have been crammed into four trimeters. It could be remarked that another class of presumably "sophistic" adjectives, those in -ώδης are not particularly noticeable in Aristophanes, except for a cluster in the late Plutus (560-562 and 793-797). On this formation in fifth- and fourth-century Greek see D. op de Hipt, Adjektive auf -ώδης im Corpus Hippocraticum, Hamburg 1972.
4.   Although it is not attested in Aristophanes, Willi also discusses the iotacistic pronunciation of eta. This is clearly an innovative feature, but when discussed by Plato in Cratylus 418b-d, ἱμέρα is declared to be the original pronunciation of ἡμέρα, preserved in Socrates' life-time only by the women of Athens. Willi finds an interesting parallel in a present-day Spanish village, where a sociolinguistic study has shown that the women are more prone than the men to adopt innovations of standard Spanish, although, like their Athenian sisters, they live a more secluded life.
5.   Johannes Friedrich, 'Das Attische im Munde von Ausländern bei Aristophanes', Philologus 75, 1918, 274-303; Claude Brixhe, 'La langue de l'étranger non grec chez Aristophane', in: Raoul Lonis (ed.), L'étranger dans le monde grec, Nancy 1988, 113-138.
6.   Cf. Geoffrey Horrocks, Greek. A History of the Language and its Speakers, London & New York 1997, 81: "The resulting dichotomy between an unchanging Attic ideal and the Koine in all its heterogeneity quickly established a formal state of diglossia."

Read Latest
Index for 2004
Change Greek Display
Books Available for Review

HTML generated at 13:29:19, Friday, 03 April 2009