Bryn Mawr Classical Review 2006.08.58
Coulter H. George, Expressions of Agency in Ancient Greek. Cambridge: Cambridge University Press, 2005. Pp. 288. ISBN 0-521-84789-3. $85.00.
Reviewed by Jonas S. Green, Philosophy, King's College London (email@example.com)
Word count: 1631 words
Table of contents
Dr Coulter H. George's [G.] 'Expressions of Agency in Ancient Greek' ['Expressions'] is a clear, well-written, well-structured, well-argued, and therefore easily accessible exposition and discussion of passive with agent constructions in Ancient Greek. The accessibility of this book will make it available to undergraduates, and even to laypeople with very little knowledge of Greek, since C. has translated all the Greek. He has not, however, translated the inserts of the German and French literature he discusses (see, for example, p. 79), and I think that he perhaps should have paraphrased them into English -- as he has done with the literature he is criticising in chapter four -- since he has taken the trouble to make everything else accessible for readers who only know English.
The book promises to discuss its central question -- 'under what conditions is the agent expressed by a construction other than ὑπό + genitive' -- by 'tracing the development of these expressions from Homer through classical prose and drama, paying attention to the semantic, syntactic and metrical conditions that favoured the use of one preposition over another.' It is mainly chapter four and five which deals with constructions other than ὑπό + genitive, but the book fulfils its promise, although the scope of texts upon which it bases some of its conclusions could have been more extensive. This, however, is mainly true of the first chapters, which are introductory, and not of the central fourth chapter. And narrowing the scope, and not quoting all the references, has allowed C. to keep a good pace throughout the book and avoid belabouring the points he is making. His conclusions of each chapter, which are clearly summarised at the end of the book, are always well grounded.
The bibliography cites German, French and Italian as well as English material, and the book contains three indexes: a general index, an index of Greek words, and a useful index of passages discussed.
'Expressions' is printed as part of the Cambridge Classical Studies series, which, according to the website, 'contains a series of monographs on all aspects of classical learning: literature, philosophy, history and archaeology.' Previous publications in the series include widely read works such as, for example, G.R.F. Ferrari's 'A Study of Plato's Phaedrus,' and Robin Osborne's 'Demos. The discovery of classical Attika,' so 'Expressions' is in good company. It is also, as far as I am aware, the first and only book in the series that solely discusses a grammatical issue, and is, in this respect, unique. Since the series contain a high number of books with a philosophical perspective ('Epicurus and Democritean Ethics' by James Warren, 'The Unity of Plato's Sophist' by Noburu Notomi, 'Method and Politics in Plato's Statesman' by Melissa S. Lane, 'Aristotle on the Sense-Organs' by T. K. Johansen, and 'The Chain of Change' by Robert Wardy, for example) it might be worth pointing out that there is no philosophical discussion of agency, nor any references to this philosophical debate, in G.'s book, which is purely philological.
'Expressions' is divided into six well ordered chapters: 1) Passive verbs and agent constructions; 2) Agent constructions in Homer; 3) Agent constructions with perfect passive verbs; 4) Agent constructions with prepositions other than ὑπό: prose; 5) Agent constructions with prepositions other than ὑπό: tragedy and comedy; and 6) The decline of ὑπό in agent constructions.
In the first chapter G. explains what he means by a passive with agent construction [PAC]. He rightly uses it in a functional sense to denote verb detransitivization accompanied by object promotion regardless of whether the verb is middle or passive (p. 6). After he has given a brief story of the origins of the passive voice in Greek, he turns to the question of why there are PACs at all. His main answer is: to keep the narrative theme as the grammatical subject of the sentence. In English, G. thereafter observes, the agents in PACs are new information critical to the meaning of the sentence and lower on a scale of animacy than the patient. The scale runs from first- to second- to third-person pronouns, to animated nouns, to inanimate nouns. So in English we say 'he was seen by a lizard' rather than 'a lizard was seen by him' (G.'s example, p. 32). But this is not the case with the first twenty PACs with ὑπό in Herodotus, the eighteen PACs with ὑπό in Xenophon's Anabasis 1, and the eleven PACs with ὑπό in Plato's Apology. In some of these cases the agent is not new information and the passive may be used irrespective of the relative placement of the agent and patient in the hierarchy of animacy. On this evidence -- which is not exhaustive but sufficient -- G. concludes that 'maintaining the narrative theme as subject must be more important to these authors than structuring sentences from the standpoint of the most animated participant' (p. 33), and that 'such a desire for textual cohesion is perhaps to be expected of a language that regularly avoids asyndeton' (p. 33-34). This conclusion is carefully restricted to the prose authors since G. finds no evidence for it in comedy-writers such as Aristophanes. The difference between prose and comedy is explained by their relative closeness to the spoken language. The prose writer may manipulate the voice to achieve a stylistic effect to a larger degree than the comedy writer, who has to remain truer to the spoken language.
In chapter two G. examines PACs in Homer, where the most common agent marker is dative. But G. argues that this is due to the fact that verbs frequently used by Homer take dative so that the dative of agent cannot be taken to be 'a default agent marker' (p. 265). ὑπό, the standard agent marker in classical Greek, is already used in Homer, although it might often take dative and not genitive depending on the verb.
In chapter three G. turns to agent constructions with perfect passive verbs. Perfect originally denotes a state of affairs -- the door was open -- rather than an action -- the door was opened -- and takes a dative of interest rather than a dative of agent. But the perfect loses its stative nature gradually and the dative of interest becomes in some cases dative of agent. Furthermore, in some instances, agents in PACs with perfect passive are expressed by ὑπό + genitive. G. shows that this occurs in Herodotus and argues that ὑπό + genitive is used in instances where both the agent and the patient are animate. In cases where the agent is animate and the patient is inanimate a relatively ambiguous agent marker like the dative of agent is used (p. 87). G. concludes that 'the decisive factor in the decline of the dative of agent seems to have been more a reduction in PACs with the perfect passive altogether than an overwhelming shift to the use of ὑπό as an agent marker with these verbs' (p. 99).
The fourth and the fifth chapters deal with PACs with prepositions other than ὑπό in prose, tragedy and comedy. These two chapters form the core of G.'s project and jointly fill half of the book, but it is the fourth chapter, which is by far the longest, and the most interesting. It is in the fourth chapter that G. mainly engages critically with the literature. G. criticises both Kühner-Gerth's' 'Ausfurhliche Grammatik der griechischen Sprache' from 1898, and Schwyzer's 'Zum personlichen Agens beim Passiv, besonderes im Griechischen' from 1943, for being too cursory (p. 103-107). G. has taken over and worked within parts of the framework of a number of Luraghi's articles in which she pays attention to how animacy affects the choice of preposition -- for example 'On the distribution of instrumental and agentive markers for human and non-human agents of passive verbs in some Indo-European languages' from 1986, 'Cause and instrument expression in Classical Greek' from 1989, and 'Animated nouns in cause expressions' from 1994 -- then he criticises her for not treating the material in great detail, which he sets out to rectify by going through all of the examples of PACs with prepositions other than ὑπό in Herodotus, Thucydides, Xenophon, Lysias, Plato and Demosthenes. On this evidence G. concludes that the choice of preposition in PACs depends mainly on the semantics of the verb. IF the verb denotes motion of the patient away from the agent, such as for example 'sending' or 'giving,' then the agent is marked with an ablative preposition. In Herodotus it is usually ἐκ and in Attic prose it is usually παρά + genitive. If the Verb does not depict the patient as moving away from the agent, such as for example 'thinking' or 'believing,' then the agent is marked with a locative preposition. In Herodotus it is usually πρός + genitive and in Attic prose it is usually παρά + dative.
The fifth chapter is brief. G. points out that ὑπό + genitive is not the most common agent marker in Attic tragedy but that both ἐκ and πρός + genitive are used much more widely because of their 'metrical utility' since monosyllables fit more easily into the iambic trimeter than ὑπό (p. 269).
Finally, the last chapter rejects the claim that ὑπό merely was replaced by ἀπό as the agent marker in post-classical Greek. G. shows how this view relies too much on the evidence of the Judeo-Christian writings. In the New Testament ἀπό primarily occurs when it marks inanimate agents or is found with verbs that have the prefix ἀπό.
G. has managed to write an extremely clear, well-argued and engaging book about passive with agent constructions in Ancient Greek, and I would recommend it to anyone who has an interest in this area.