Bryn Mawr Classical Review

Bryn Mawr Classical Review 2008.02.24

Albert Rijksbaron, The Syntax and Semantics of the Verb in Classical Greek: An Introduction. Third edition (American edition).   Chicago:  The University of Chicago Press, 2006.  Pp. xvi, 214.  ISBN 10: 0-226-71858-1.  ISBN 13: 978-0-226-71858-3.  $20.00.  

Reviewed by Coulter H. George, University of Virginia (
Word count: 2429 words

Table of Contents

When scanning through the table of contents in this book, one immediately thinks of Goodwin's Syntax of the Moods and Tenses of the Greek Verb. After all, the first chapter (pp. 1-8) is an "introduction to the moods and tenses", the second (pp. 9-48) covers "the main uses of the single moods and tenses in independent sentences", the third (pp. 49-94) deals with dependent clauses, the fourth (pp. 95-133) turns to non-finite verbs, and it is only with the fifth (pp. 134-163), on voice, that Rijksbaron discusses material not found in Goodwin. Given this duplication of subject matter, the obvious question is: does this book say anything new? Happily, the answer is a definite "yes".

Rijksbaron's work may look at the same subject matter as Goodwin's, but the perspective is vastly different. Goodwin does have the edge on Rijksbaron in terms of detail and the sheer number of examples presented, but Rijksbaron has the advantages of brevity, clarity of organization, and a consistent theoretical outlook, namely that of Functional Grammar (FG). For the uninitiated, FG is a linguistic framework developed by Simon Dik and associated in particular with Dutch scholars, which has proven an excellent vantage point from which to tackle various linguistic problems that have long troubled classical philologists. It differs from the Chomskyan tradition in being less focused on working out hard-wired syntactic structures and more interested in pragmatics: how does the particular context in which a sentence is uttered or written affect how it is structured?1

Such an approach yields predictably useful results when applied to questions about Greek tense and mood usage. The difference between ὥστε with the indicative and with the infinitive receives clear treatment (pp. 63-66), and the various conditional clauses are explained (pp. 66-74) with close attention both to subtle overlap between types usually classed separately (ex. 188 on p. 71 shows how close εἰ with the optative can come to expressing a counterfactual protasis) and to flagging up distinctions that often go obscured (on p. 74, Rijksbaron calls attention to the fact that, when following the main clause, conditional clauses can indicate purpose and mean roughly "in the hope that"). Temporal clauses with ἕως and ἐν are neatly distinguished (p. 77), as are causal clauses introduced by ὡς and ὅτι (pp. 84-86). Despite such wealth of detail, Rijksbaron keeps the volume accessible to all by consistently employing the user-friendly convention of putting all the basic information that the beginner ought to know in a larger typeface, while consigning to notes at the end of each section (literally, the fine print) those details which will be of interest primarily to more advanced students of Greek.

Of course, any book that tries to cover as much ground as this in such a small amount of space will inevitably suffer from some shortcomings caused by compression of the material. In particular, unfamiliar grammatical terminology, often derived from FG, is not always explained in great enough detail for its use to be justified in a work presumably aimed at readers with relatively little Greek. One example is Rijksbaron's use of the term "'state of affairs', instead of 'action', as a cover term for 'that which is expressed by a predication'" (p. 3, n. 4), in contexts like "the [historic] present marks states of affairs that are of decisive importance for the story" (p. 22). On the one hand, Rijksbaron is absolutely right to note that "action" is a poor choice to express this concept, for, as he says, it is generally restricted to particular types of states of affairs. On the other hand, "state of affairs" is itself problematic in that it has "state" as its head noun, a term which should also be restricted to particular states of affairs, namely those in which no dynamic action takes place and which, in Greek, are associated with the perfect--and not, for example, the decisive turning-points marked by the historic present (see below). While there do not seem to be any terms for this that are completely free from criticism, I suspect that "event" (more accessible) or "eventuality" (more technically correct) would yield a much smoother reading than "state of affairs" in most instances.2

Another example, which affects the structure of the book at a basic level, is the fact that Rijksbaron organizes the subchapters on dependent clauses into "Clauses with the function object or subject (obligatory clauses)," "Clauses with the function satellite (optional clauses)," and "Relative clauses." Rijksbaron does explain in a note that these are traditionally called substantive, adverb, and adjective clauses respectively (p. 49), but his preferred terminology might well be off-putting for classicists not trained in linguistics. Nor does it seem a particularly good way of arranging the material anyway: insofar as relative clauses can be either restrictive or non-restrictive--that is, semantically obligatory and optional respectively--it seems that the more practical way of classifying subordinate clauses remains that of noting what part of speech (or, in modern parlance, word class) the clause in question is equivalent to. Note that Rijksbaron himself divides his discussion of these clauses into three chapters corresponding to the parts of speech, not two chapters as would most properly reflect a putative fundamental split between the obligatory and the optional.

This is not the only organizational choice in the book to raise questions. On several occasions, Rijksbaron seems keener to subdivide syntactic phenomena into categories of the same order than to try to work out the big picture by determining which of these categories more closely resemble each other. One good example of this is his description of the various functions of the mediopassive. The mediopassive forms of transitive active verbs are dealt with in four sections, covering respectively (i) those forms that have passive meaning (pp. 138-144), (ii) those that have direct-reflexive meaning (pp. 144-147), (iii) those that have indirect-reflexive meaning (pp. 147-150), (iv) and those that have pseudo-reflexive or pseudo-passive meaning (pp. 151-155). The first category is fairly straightforward and, prototypically, includes mediopassive constructions with ὑπό marking the agent. The direct reflexives are also a familiar category, including primarily verbs of habitual physical treatment, such as λούομαι "wash", γυμνάζομαι "train oneself", and, as Rijksbaron matter-of-factly notes, δέφομαι: not a word one is accustomed to seeing under the rubric "verbs of grooming"! While these direct reflexives can be understood as roughly equivalent to an active verb with a reflexive pronoun in the accusative, the indirect reflexives resemble active verbs with the reflexive pronoun in the dative, e.g. πλοῖα παρασκευάσασθαι "to procure ships for oneself".3 So far, so good. Only with the last category, the pseudo-passives and pseudo-reflexives, do real questions arise. Included in this group are verbs that fall into several different categories in Rutger Allan's recent study of the Greek middle:4 Rijksbaron's pseudo-passives, which have subjects that do not control the action of the verb, include spontaneous process middles like ῥήγνυμαι "break" and mental process middles like φοβέομαι "be frightened"; his pseudo-reflexives, which have subjects that do control the action of the verb, include body motion middles like ἵσταμαι "stand up" and collective motion middles like ἀγείρομαι "assemble".

Now, there is of course nothing wrong with such a classification in principle: the verbs Rijksbaron groups together clearly do belong together. What is problematic about his scheme is how it works in practice in a book generally aimed at a fairly elementary audience: in particular, it is confusing, primarily because the functional approach adhered to by Rijksbaron leads him to discount morphological and syntactic criteria that might well serve as additional aids for the beginner seeking to sort out the sundry uses of the Greek middle into mental pigeonholes. As far as morphology is concerned, it would perhaps have been beneficial to make the role of the aorist formation more prominent. Rijksbaron certainly does not ignore the difference between mediopassives that take an aorist passive in -(θ)η- and those that take a (generally sigmatic) aorist middle, but it could have served as a useful first-order division: the true passives, the pseudo-passives (usually), and the pseudo-reflexives (often) take the former, while the direct and indirect reflexives take the latter. As for syntax, a slightly greater emphasis on the detransitivizing nature of the passive could have reinforced the morphological split in the aorist: with both true passives and pseudo-passives, the subjects of active verbs are demoted to an oblique relation, while the erstwhile objects become subjects in their place; with the reflexives, direct and indirect, the subject of the active verb remains the subject of the mediopassive verb. The pseudo-reflexives then form something of a pivot: morphologically, they often behave more like the passives (note the passive aorist ἐκλί(ν)θην), but they can also have middle aorists instead (κλινάμενος);5 syntactically, they resemble the reflexives in that the subject of the active verb generally remains the controlling agent in the mediopassive, but they show similarity to the passives in that some detransitivization has clearly taken place. It is only the fact that the active verb is causative (and thus has an object that is itself potentially an agent) that allows the detransitivized mediopassive still to retain the agentive subject characteristic of the reflexives. All in all, while it should be stressed that Rijksbaron does not ignore such morphosyntactic criteria, they could have been made more prominent in his presentation. I imagine that, rather than dealing with Rijksbaron's categories, beginners would find it easier to come to terms with a morphological division between mediopassives with middle aorists, those with passive aorists, and those which fluctuate between the two. Put more simply, it seems reasonable to retain the traditional distinction between middle and passive at least as a preliminary guideline.

Consider too Rijksbaron's discussion of the historic present, one of the sections revised in the third edition (pp. 22-25).6 In his introduction, Rijksbaron appears to set up a two-way split between those historic presents in which 'the narrator plays the role of eyewitness', and those in which this is not the case. He then goes on, however, to divide his examples of the historic present into two categories, those in which it marks a decisive state of affairs, and those in which it punctuates a narrative. In the first category he includes the messenger speech from the Medea, which contains three historic presents in very quick succession (lines 1156-1169), a single historic present in the Candaules and Gyges episode (Hdt. 1.10), which is followed by two others in chapters 11 and 12, and the eight historic presents, again very densely packed, at the beginning of the Anabasis (1.1.1-4). In the second category, he gives as an example only the repeated use of ἐξελαύνει from Anabasis 1.2.5 and passim. My impression is that his examples could have been more effectively subdivided by the criterion he hints at in his introduction, that of the eyewitness, than by twofold division into decisive states of affairs and states of affairs that punctuate a narrative. As it stands, it is unclear what the difference is between the historic presents in Herodotus, which are said to mark decisive states of affairs, and the use of ἐξελαύνει in Xenophon, which is said to punctuate the narrative. But had Rijksbaron continued in the direction he was going in the introduction, a much clearer dichotomy would have presented itself: the historic present of the messenger speech, in which the historic presents are clustered closely together in a passage filled with the vivid description of a single event, could be called, say, an eyewitness present, whereas those of the other passages, all in history, and all, in contrast to Euripides, used at turning-points in the story to signal a new stage in the narrative, could be called a punctuating present or, for that matter, the historic present proper. One suspects that his reason for choosing the headings he finally opted for lay in an understandable desire to avoid the oft-misused label of the "vivid" present. This is clearly not a description one wishes to attach to the humdrum repetition of ἐξελαύνει in Xenophon--but is it really such a misleading characterization of the historic present of a messenger speech?7

While the book has relatively few typographical errors, a handful of minor infelicities should be pointed out:

(p. 64) The use of archaizing translations based on or borrowed from the Loeb series (e.g. Hdt. 7.13.2 "... my youthful spirit did for the nonce take fire, whereby there brake from me a more unseemly answer ... than was fit") occasionally sits ill at ease with the modernizing grammatical terms used in the book: the translation just quoted comes from the chapter on "Clauses with the function Satellite".

(p. 77) In example 206, I would translate "until" rather than "before", because the action of the main verb appears to last continuously until that of the subordinate clause (rather than being a punctual event: note the imperfect ἔπασχον).

The influence of a Dutch substrate is revealed in a couple of typographical slips: (p. 77) for "en", read "and"; (p. 163) for "het", read "the". Throughout, "jussive" is spelled "iussive".

(p. 121) The constructions of φαίνομαι with the participle and the infinitive are glossed as "appear" and "seem" respectively, verbs which are in fact pretty close to synonymous in everyday English usage. Perhaps "be clear that, be clearly" would be a better gloss for the participial construction.

(p. 135) "John decided to find the letter" and "John decided to wake up" are both said to be ungrammatical because find and wake up are non-agentive verbs in which the agent does not control the action. While I agree that these verbs in their basic sense are non-agentive (and thus incompatible with the degree of control entailed by "decided"), I also find it easy to reinterpret the semantics of these verbs after such a verb in such a way as to make the sentence grammatical, e.g. as roughly equivalent to "look for (doggedly)" and "get out of bed" respectively.

In closing, I would like to emphasize that these criticisms of the book represent, above all, the splitting of linguistic hairs--as I hope is already obvious from the nature of said criticisms. Rijksbaron's sensitivity to the subtle nuances of Greek is of the highest order, and this reviewer would like nothing more than to see him do for Denniston what he's already done for Goodwin. The time seems ripe for a new introductory account of the Greek particles, of about the same length as the volume reviewed here, and fully incorporating recent research elucidating the pragmatic differences between particles of apparently synonymous semantics. That must surely be a desideratum.


1.   For Dik's own work, see e.g. The Theory of Functional Grammar part 1, 2nd edn., Berlin-New York 1997, and the website Two areas of linguistic research that have particularly profited from a more pragmatic approach have been word order (see e.g. Helma Dik's Word Order in Ancient Greek: A Pragmatic Account of Word Order Variation in Herodotus, Amsterdam 1995, and now also Word Order in Greek Tragic Dialogue, Oxford 2007) and particle use (see e.g. Caroline Kroon's Discourse Particles in Latin: A Study of nam, enim, autem, vero, and at, Amsterdam 1995, and many of the articles in the Rijksbaron-edited volume New Approaches to Greek Particles, Amsterdam 1997.
2.   For an introduction to the difference between such terms as states, events, and action, see C. S. Smith, The Parameter of Aspect, 2nd edn., Dordrecht 1997.
3.   A purely syntactic account of the middle does not seem to work in the end, but for a good look at how far it can get you, see E. J. W. Barber (1975) "Voice--Beyond the passive," Proceedings of the Annual Meeting of the Berkeley Linguistics Society 1: 16-24. Rijksbaron appears to include in the direct-reflexive category constructions that have an accusative of respect specifying the part of the body affected, e.g. ἀλειψαμένη τὸ σῶμ' ὅλον, but not those mediopassives that govern an accusative that is external, e.g. κόπτομαι ξύλον "I cut wood (in my own interest)" (p. 146, nn. 2, 4). That mediopassives of two different categories should have such similar semantic and syntactic manifestations suggests that these two categories are not as different from each other as they are from the passive mediopassives. That said, I certainly don't dispute the need to consider the two types separately, as is particularly nicely demonstrated by Allan (see following note), p. 94.
4.   R. J. Allan, The Middle Voice in Ancient Greek, Amsterdam 2003. (Incidentally, this work is listed as "forthcoming" on p. 163, but given its date in the bibliography on p. 167.)
5.   This example is taken from p. 150 in Allan. He notes that the body motion, collective motion (pseudo-reflexive), and mental process (pseudo-passive) middles are the three classes most likely to exhibit fluctuation between sigmatic and passive aorists (see the useful chart on his p. 147 as well as the following discussion).
6.   Other sections revised in the third edition include those on the future indicative, causal clauses, the aorist of performative verbs, and the oblique optative. While these revised sections include twenty or so new examples (some quite nice, e.g. ex. 133a on p. 53 contrasting the indicative and oblique optative in a single indirect statement), they by no means amount to a wholesale re-writing of the book. Still, considering the extremely modest price, libraries and linguists will want to acquire the latest edition. Non-specialists who consult the work primarily for e.g. teaching undergraduates will probably be safe in sticking to the second edition.
7.   On questions of tense and aspect generally, Rijksbaron plays up tense and plays down aspect to a greater extent than most others would nowadays (p. 2, n. 1). On the one hand, it's convenient to speak in tense-based terms of anteriority and simultaneity when discussing the relationship between a participle and the main verb (p. 117); on the other hand, this causes problems for Rijksbaron when he looks at the gnomic aorist (p. 32). Such aorists are much easier to explain if the aorist indicative fundamentally marks perfective aspect rather than past tense. (The anteriority and simultaneity expressed by aorist and present participles respectively can still be easily explained in aspect-based accounts as epiphenomenal.) Significantly, the gnomic aorist which causes him such difficulty comes from Homer, whereas the vast majority of his examples come from Herodotus and Classical Attic (the index locorum contains only two citations from Homer). Based on the late development of the historical present (which, quite the opposite of the gnomic aorist, is easier to explain in a tense-based model like Rijksbaron's than in an aspect-based one), it seems likely, as I argued in a talk at the 2008 APA, that tense gradually became more important relative to aspect over the years between Homer and the classical period. Rijksbaron's focus on the fifth and fourth centuries would then explain his bias in favor of tense.

Read Latest
Index for 2008
Change Greek Display
Books Available for Review

HTML generated at 13:32:38, Friday, 03 April 2009