Bryn Mawr Classical Review 2005.07.70
Michel Buijs, Clause Combining in Ancient Greek Narrative Discourse. The Distribution of Subclauses and Participial Clauses in Xenophon's Hellenica and Anabasis. Mnemosyne Supplementa 260. Leiden: Brill, 2005. Pp. 277. ISBN 90-04-14250-9. €85.00.
Reviewed by Steven D. Smith, Hofstra University (firstname.lastname@example.org)
Word count: 1483 words
This book is an extensive study of the use of subordinate clauses and participial clauses in Xenophon's prose. Buijs (hereafter B.) successfully demonstrates that the choice between a subclause or a participial clause in the management of textual information may best be understood through discourse analysis, a method of examining linguistic phenomena in context, moving beyond the limitations of a sentence-by-sentence approach. B. intends his audience to be not just specialists in Greek linguistics and discourse analysis but anyone interested in Xenophon and Greek historiography generally. All Greek in the book is translated. Translations are either taken directly or are adapted from the Loeb editions of Xenophon.
The book begins with a lengthy introduction, in which B. reveals as insufficient the explanations which traditional grammars1 have given for the seemingly arbitrary choice between participial clauses and subclauses. For his discourse analytic approach, B. announces that he will not define "rules," but "regularities" of linguistic choice, drawing primarily from Xenophon's Hellenica and Anabasis, though a good deal of evidence is also drawn from the Agesilaus and the Cyropaedia. As his study treats narrative phenomena, B. regularly distinguishes between the "level of the Real World construction" and the "level of text articulation" (13-16).
Chapter two is a close reading of six near minimal pairs (NMP's), parallel passages from the works of Xenophon which differ in their choice of subclauses and participial clauses for relaying similar narrative information. B.'s comparative analysis of Hellenica 3.4.15-16 and Agesilaus 1.19-25 is representative of his methodology as a whole. Both passages depict the same real world event -- the raising of a cavalry force by Agesilaus -- and much of the wording is similar in both passages. There are, however, differences in the deployment of finite verbs, participles, and subordinate clauses. When Agesilaus is represented raising a cavalry in the passage from the Hellenica, it is as part of a larger narrative episode, whereas the same narrative material in the Agesilaus is given more elaboration. Grammatical differences, argues B., are due to the texts' different discursive aims: in the historiographic narrative of the Hellenica, Agesilaus' action is "hierarchically downgraded" by the conjunct participle προειπών, while in the encomiastic prose of the Agesilaus, the actions of the laudandus, depicted through a series of finite main verbs (μέν ... κατέλεξε ... προεῖπε δέ ... ἔταξε δέ), are shown to be the organizing principle of the text. In the conclusion to this chapter, B. warns against formulating hypotheses from the evidence that he has provided: understanding an author's choice of, for example, a participle or a subordinate ἐπεί clause, depends ultimately not on a series of grammatical rules, but on the context in which the author uses them.
In chapter three, B. examines real world relationships conveyed by the use of subordinate clauses and participial clauses. The chapter is divided into four sections. First, B. analyzes causal and temporal relations conveyed by subclauses which begin with the semantically specific relators διότι, ὅτι, ὅτε, ἥνικα or ἐν ᾧ. B.'s conclusion in this section is sound, though not surprising: subclauses headed by the aforementioned relators, whether the subclauses are preposed or postposed in the sentence, while indicating causal or temporal relationships within the sentence itself, may also "evoke other parts of the text, and therewith add to the comprehension of the content of the matrix clause within the discourse as a whole" (111). Next, B. analyzes the real world relations conveyed by subclauses beginning with the semantically non-specific relators ἐπεί, ἐπειδή, and ὡς. Whereas traditional approaches have attempted to interpret such subclauses as temporally or causally connected to their matrix clauses, B. asserts that the point of such semantically non-specific relators is to indicate not causal or temporal relations (though this is of course possible) but to indicate situations which the reader needs to comprehend in order to understand what follows in the text. A present tense verb in the subclause means that the text which follows should be read "in view of the Real World situation" (136) articulated within the subordinate clause. An aorist tense verb in the subclause means that the text which follows should be read as "following on the occurrence" of the event articulated in the subordinate clause. In the last part of this chapter, B. addresses participial clauses which are headed by semantically specific relators. Generally speaking, participial clauses are not headed by relators, but when the author chooses to express a participial clause's relationship to the matrix clause, it is to prevent the audience's misinterpretation of real world events.
In chapter four, B. explores the impact of preposed subclauses and participial clauses on text segmentation, basing his understanding of a text's thematic units on continuity between the four "coherence strands" (referential coherence, temporal coherence, locational coherence, and action-event coherence) delineated by Givón.2 To exemplify the different levels of text segmentation possible, B. provides a close reading of the story of Mania (Hellenica 3.1.10-16). The level of text segmentation indicated by a preposed subclause or participial clause may be reflected in its relative desententialization, or grammatical dependence upon the matrix clause to which it is joined.3 Each of the four types of subordinate, embedded predications fall somewhere along a continuum of dependence upon the matrix clause. They are, from least to most dependent: subordinate clauses, infinitival clauses (which B. does not treat), genitives absolute, and conjunct participles (which agree in case, number, and gender with a noun component in the matrix clause). Consequently, preposed subclauses, having a low dependence on the matrix clause, signal a comparatively high level of text segmentation, which in turn reflects a proportional discontinuity in the representation of real world events. Genitives absolute and conjunct participles, on the other hand, while indicating text segmentation, do not reflect significant discontinuity in the representation of real world situations. Preposed participial clauses, in other words, indicate textual boundaries within the larger boundaries of a thematic unit.
Chapter five is divided into two sections. In the first section, B. examines the distribution of (a) postposed "causal" clauses headed by ἐπεί and ἐπειδή and (b) postposed participial clauses. B. finds that postposed causal clauses prompt the reader to reconsider information which has already been processed from within the matrix clause. They are, in other words, "motivating clauses" which provide justification for utterances in the preceding text or, in pure narrative contexts, the reason(s) for a narrative agent's actions. Postposed conjunct participles, however, which have a lower degree of linguistic coding than postposed subordinate clauses, relate information of a motivating nature that is more integrated into the matrix clause to which it is linked. Furthermore, postposed genitives absolute, while less dependent upon the matrix clause than conjunct participles, relate information that is not essential for understanding the preceding utterance. In the second section, B. looks more closely at genitive absolute constructions as alternatives to conjunct participles. B.'s focus here are those instances when a genitive absolute is preferred by the author, even when the author could have used a conjunct participle to refer to a noun component in the matrix clause. The less dependent quality of the genitive absolute provides the author with a means of conveying information without "interfering with the construction of the sentence as a whole" (253).
Chapter six serves as a summary and general conclusion. B. reminds readers that in any good discourse analysis there is an inescapable interdependence between the text as a discursive whole and its syntactic parts. It is doubtful that B.'s approach will supplant that of the traditional grammars in the short term, especially for beginners still grappling with the complexities of the language. But this study provides an important corrective to the ambiguities and blind spots which have stymied traditional approaches to the language, in particular as regards clause combining and the use of subclauses and participial clauses. It is the nature of the discursive approach to language "to describe regularities which will have to be reconsidered in each individual context" (260).
The eight-page bibliography is followed by an index of linguistic terms (which will be helpful for those readers unacquainted with the terminology), a general index, and an index locorum.
The book is clearly intended for specialists in Greek linguistics and discourse analysis. I am less certain, though, that B.'s technical study will attract those who are interested in Xenophon and Greek historiography in a general sense. Nevertheless, this book's insights will reward many readers, and anyone interested in Greek prose narrative should, after reading this book, return to the primary texts with a renewed sensitivity towards the language.
The book is unfortunately flawed by a number of typographical errors and by some confusing and sometimes ungrammatical sentences4 which seem to have eluded the eye of the editor. Surprising in a study of such a technical nature are a few misprints in the Greek.5 These are superficial criticisms, though, and do not ultimately detract from the value of the book as a whole.
1. Kühner, R. & Gerth, B., Ausführliche Grammatik der griechischen Sprache. Zweiter teil: Satzlehre. Hannover: 1898-1904; Schwyzer, E. & Debrunner, A., Griechische Grammatik. II: Syntax und syntaktische Stilistik. München, 1950; Goodwin, W. W., Syntax of the Moods and Tenses of the Greek Verb. London: 1889; Smyth, H. W., Greek Grammar. Cambridge, Mass.: 1920 (rev. ed. 1956); Bornemann, E. & Risch, E., Griechische Grammatik. Frankfurt am Main: 1973; Gildersleeve, B. L., Syntax of Classical Greek from Homer to Demosthenes. Groningen: 1980 (reprint); Humbert, J., Syntaxe Grecque. Paris: 1960.
2. Givón, T., Syntax. A Functional-Typological Introduction. Volume II. Amsterdam and Philadelphia: 1990. 827 n. 3. See also Givón, T., English Grammar. A Function-Based Introduction. Volume II. Amsterdam and Philadelphia: 1993. 286-287.
3. For the notion of desententialization, B. follows Lehmann, C., "Towards a Typology of Clause Linkage." Clause Combining in Grammar and Discourse. Typological Studies in Language 18. J. Haiman and S. A. Thompson, eds. Amsterdam and Philadelphia: 1988. 181-225.
4. "Although is self evident" (2); "verbal contituent" (39); "the discourse structure that is abuilding" (94); "Rather, since ... that adds to text comprehension" (135); "when the subclause contains an aorist verbal constituent, will have to be interpreted" (136); and "The subclause thus ...to the communicative mode" (146).
5. For example, there is a peculiar recurrence of ῲ where a reader expects a comma and a blank space dividing Greek words on pp. 26 and 28 (e.g., οὖνῲκαὶ for οὖν καὶ). Less egregious, there is κτἑ for what I presume to be κτλ on p. 105.