Periphrastic forms feature only marginally in ancient Greek verbal paradigms as we usually learn them: as perfect subjunctives and optatives (λελυκὼς ὦ, λελυμένος ὦ, λελυκὼς εἴην, λελυμένος εἴην, but in the active also λελύκω, λελύκοιμι), and as third person plural perfect and pluperfect middle/passive indicatives for consonant-stem verbs (πεπεισμένοι εἰσί, πεπεισμένοι ἦσαν). Yet we do not have to read much ancient Greek before we come across other apparently periphrastic verb forms. At Herodotus 9.27.5, for example, the Tegeans and Athenians are arguing over who should command the left wing of the army. Adducing their achievements at the Battle of Marathon, the Athenians say ἡμῖν δὲ εἰ μηδὲν ἄλλο ἐστὶ ἀποδεδεγμένον … ἀλλὰ καὶ ἀπὸ τοῦ ἐν Μαραθῶνι ἔργου ἄξιοί εἰμεν τοῦτο τὸ γέρας ἔχειν ‘But even if no other deed has been displayed by us … we deserve to have this honour on the basis of our act at Marathon’. In Sophocles’ Philoctetes (435), Philoctetes has asked where Patroclus was when Odysseus claimed Achilles’ armour, and Neoptolemus replies χοὖτος τεθνηκ ὼς ἦν ‘he too was dead’. In the Antigone (192), after outlining the principles on which he governs the city, Creon explains that he has issued an edict consistent with these principles: καὶ νῦν ἀδελϕὰ τῶνδε κηρ ύξας ἔχω ‘And now I have announced things related to these’. In the Oedipus Rex, when Oedipus starts to grasp his situation, he says δεινῶς ἀθυμῶ μὴ βλ έπων ὁ μάντις ᾖ (OT 747) ‘I fear terribly that the prophet is seeing’. Such usages raise a whole series of questions. Why do Herodotus’ Athenians use the periphrasis ἐστὶ ἀποδεδεγμένον, rather than the usual (Ionic) perfect ἀποδέδεκται, and why do Sophocles’ characters not say ἐτεθνήκει, κεκήρυχα and βλέπῃ? Does μὴ βλέπων ὁ μάντις ᾖ really deserve to be discussed in a book on verbal periphrasis, or is βλέπων in this context effectively an adjective meaning ‘endowed with sight; not blind’? How many different syntactic varieties of verbal periphrasis do we find, and what meanings do they convey? How did all these usages come about in the first place, and what happened to them after the Classical period? Does Latin or Hebrew influence become relevant in some way?

In this volume, B. studies Greek periphrastic verb forms (‘verbal periphrases’) involving the verbs εἰμί and ἔχω, from the eighth century bc to the eighth century ad. The study is based on an extensive (c. 10-million-word) corpus providing 8,094 instances of the constructions under investigation. This corpus has been carefully put together so as to include both higher-register and lower-register texts spanning the whole period considered, and the post-Classical portions include documentary papyri. B. gathered his examples with the help of the Thesaurus Linguae Graecae and the Duke Databank of Documentary Papyri, and then personally inspected each example to see what kind of meaning it expressed.

After a brief introduction, Chapter 1 outlines the theoretical background in some detail. Among other things, this chapter introduces the reader to the major categories of verbal aspect that B. works with (treating these as categories of function, not morphological form), and distinguishes carefully between different kinds of meaning subsumed under each major category. For example, when Creon says κηρύξας ἔχω, he is conveying a past event with current relevance. With τεθνηκὼς ἦν, Neoptolemus is conveying a (past) situation resulting from a past event; the situation is more important here than the event. In B.’s terminology the first of these functions as an ‘anterior perfect’ while the second functions as a past ‘resultative perfect’. This chapter makes essential reading: the distinctions are crucial to the book as a whole and likely to be unfamiliar to most Classicists. Moreover, certain terms have been used in more than one way in literature on the subject (see p. 38 n. 127 on ‘resultative perfect’), and this chapter makes clear how they are used in this book.

In Chapter 2, B. takes on the question of distinguishing verbal periphrasis from adjectival periphrasis: how do we know when a participle is really a participle rather than an adjective? B. shows how attempts to draw a sharp distinction inevitably run into difficulties. Rather than adding to these attempts he argues that there is a continuum between more and less prototypical instances of verbal periphrasis, and that the whole continuum is relevant if we want to understand how periphrastic constructions develop over time.

The real meat of the book consists of Chapters 3 and 4, on verbal periphrases conveying species of perfect and imperfective aspect respectively. In both chapters the discussion is organised first and foremost by chronological period, with a division into Archaic Greek (eighth to sixth centuries bc), Classical Greek (fifth to fourth centuries bc), early post-Classical Greek (third to first centuries bc), middle post-Classical Greek (first to third centuries ad), and late post-Classical and early Byzantine Greek (fourth to eighth centuries ad). The much shorter Chapter 5 deals with the marginal phenomenon of periphrastic forms conveying perfective aspect, which are effectively limited to some uses of εἰμί with the aorist participle in the Classical period. A succinct conclusion recapitulates the book's main results.

A real strength of the book lies in its fine combination of qualitative and quantitative work. A big picture emerges from tables showing how many times a particular construction occurs at a particular period, and how these occurrences are distributed between genres and authors. Against this backdrop, careful discussion of individual passages shows how the picture looks when we zoom in.

To return to Sophocles’ κηρύξας ἔχω, part of the story here is that κεκήρυχα could have expressed the same meaning, except that no non-periphrastic form such as κεκήρυχα was available in Sophocles’ day (p. 76). But we learn that this is not the whole story, because Sophocles and (to a lesser extent) his contemporaries were keen on ἔχω plus aorist active participle even when a non-periphrastic form would have been available. All this is part of a larger picture in which the use of ἔχω plus aorist active participle sees a spectacular increase in frequency in the Classical period, compared to the Archaic period (p. 113). B. argues that this increase is connected to an expansion in the use of non-periphrastic as well as periphrastic perfect forms to convey an anterior perfect sense. Perfect forms were now in demand for many verbs that had hardly needed them before. In Sophocles’ day the use of ἔχω plus aorist active participle helped to meet this demand (p. 125). At the same time, however, non-periphrastic perfects were being created for an increasing number of verbs. By the fourth century bc a non-periphrastic perfect was available for most verbs, and the use of ἔχω plus aorist active participle declined (p. 117). The construction reappears, however, in later authors for whom its Classical associations helped to create a high style; it can even be found in papyri as a mark of formality, and more puzzlingly in a fourth- or fifth-century papyrus document that does not seem to aim at a formal style (see pp. 166, 179–80, 198–9).

The book includes a helpful glossary of technical terms, and an appendix listing the literary works in B.’s corpus. A spreadsheet cataloguing every example in the corpus, and the category of meaning to which B. assigns it, is available via OUP's website (http://global.oup.com/booksites/content/9780198747093/). This makes it possible to check all the evidence and judgements that underpin B.’s quantitative claims, and offers a resource for those wishing to conduct further investigations into Greek verbal periphrasis.

This volume is an excellent study based on a large quantity of well-handled evidence. B. is often able to correct or nuance assertions that have appeared in earlier literature, and there are many new insights. This book puts the subject of Greek verbal periphrasis on a new footing.