F. offers a discussion of Greek rhetorical practice, focusing on the enthymeme. Greek rhetoricians understood themselves as masters of a skill, able to impart it to others. However, we have little evidence of their own teachings concerning what enthymeme was or how it worked. For this reason, those attempting to reconstruct this teaching have generally turned to Aristotle's Rhetoric. F.'s book takes a different approach. He argues that Aristotle's special interest in deductive inference led him to present an account of enthymeme that distorts its nature as a rhetorical technique and the uses to which it was put. Instead, he argues, it is better to look at how and when the rhetoricians themselves use the term enthumizesthai. The verb has the sense of ‘keep in mind’ and is employed when speakers are asking their auditors to keep something (whether a particular fact or a generalisation) in mind, in order to make sense of a narrative, by which one can understand how and why events unfolded as they did.

F. first attends to Aristotelian and post-Aristotelian understandings of enthymeme, in order to clear the deck for a new consideration of the matter. Accordingly, Parts 1 and 2 concern the Aristotelian tradition, while Parts 3 and 4 have as their focus surviving Greek oratorical texts.

Aristotle identified enthymeme as a kind of pistis, a manner of discourse intended to lead others to assent or action. Traditionally it has been thought that Aristotle understood an enthymeme as a deduction with one or more premises unstated. In Part 1 F. decisively rejects this understanding of the enthymeme. His main point is that rhetorical speeches are not themselves inferential structures, and that, a fortiori, audiences do not hear them as such. Accordingly, ‘the imagined “filling in” of premises corresponds to no recognizable psychological process on the part of the listener because no one could complete the enthymeme and remain a listener’ (p. 50). F. is here indebted to M.F. Burnyeat's classic ‘Enthymeme: Aristotle on the Logic of Persuasion’ (most easily available in Burnyeat's Explorations in Ancient and Modern Philosophy, Vol. 1 [2012]).

In Part 2 F. turns to enthymeme as Aristotle understood it. Aristotle repeatedly describes enthymeme as a kind of sullogismos. F. accepts scholarly orthodoxy according to which, in the Rhetoric, a sullogismos is not a syllogism, structured in accordance with the formal requirements of the Prior Analytics. It is rather a process of reasoning towards an as yet unknown insight. F. cannot find a distinguishing characteristic that sets enthymeme, rhetorical sullogismoi, off from other sullogismoi, and concludes that for Aristotle, what distinguishes the enthymeme is simply its context: arguing before a mass audience in the hope of that audience making one choice rather than another. F.'s account is best understood in juxtaposition with that of Burnyeat, who argued that for Aristotle, enthymeme is marked by a certain kind of premise, one that that is only likely to be true. As such, Burnyeat takes enthymeme to be rhetorical inferences, the kind of inferences based on ‘common sense’ beliefs, which are sullogismoi only in a qualified sense, as their conclusions do not follow necessarily. On this view, though such quasi-deductions are especially found in the assembly and the courts of law, the context is not what is directly responsible for their being classified as such. F. does not discuss the texts in Rhetoric 2.25 and 26, on which Burnyeat bases his account of Aristotle's discussion of enthymeme.

In support of his own account F. points out that Aristotle's definition of sullogismos at Top. 100a25–6 need not be understood as referring exclusively to deductions. A narrative, too, qualifies as a sullogismos. In Part 3 F. argues that it is this kind of discourse that was employed by Greek orators as an enthymeme. If Aristotle sometimes writes as though rhetorical syllogisms need to conform to the formal requirements of valid inference, he is unduly swayed by the power of his own logical theory.

Part 3 goes back to the Greek orations themselves. Within the lawcourts, both prosecutors and defendants would have begun with an unquestioned incomplete account and would then have asked the auditors to take note of (enthumizesthai) certain facts that would lead the auditors to fill in the gaps in a way that gives rise to a narrative supporting their case. The enthymeme, F. argues, is to be identified with this technique. The logos being developed here is not a logical inference, but a narrative, a story that makes sense. F. supports this argument with reference to the Greek orators, with special attention paid to Lysias 6, Against Andocides, and Andocides 1, On the Mysteries.

Part 4 is devoted to a close analysis of Lysias 1, On the Death of Eratosthenes. F. shows how the carefully constructed speech masterfully presents a series of enthymemes, which leads its auditors to fill in the gaps and create a coherent narrative according to which Lysias was blameless for killing his wife's lover. F. deftly shows how the same gaps could have been the targets of the enthymemes in a speech given by a rhetorically skilful prosecutor.

Part 1 offers a solid account, accessible to non-specialists, of why enthymeme as incomplete syllogism has no role in classical rhetorical theory and is worthy of no such role. F.'s discussion of Aristotle's understanding of enthymeme is the most problematic part of the book. The question of whether Aristotle understood enthymeme as a formally valid deduction is closely related to another much discussed question: is Aristotle's metatheory of scientific explanation strictly deductivist? Aristotle certainly presents his explicit account of explanation in the Posterior Analytics as a matter of formally valid inference, and the syllogistic theory of the Prior Analytics is given as a propadeutic to the theory of demonstration. But Aristotle is surely aware that few, if any, explanatory accounts, least of all his own, conform to the formal requirements that he himself presents. One way of understanding Aristotelian metatheory is to see him as offering an account of the underlying logical, epistemic and ontological structure of successful explanatory theoretical accounts. In principle, a complete theoretical explanation would make all of these relations explicit, though in practice much is left implicit. Likewise, Aristotle could be understood as positing enthymeme as a sullogismos, as a formally valid inference that underlies the sequence of practical reasoning in which the speaker hopes that the auditor will engage. There is no need for the speaker's logos to consist in inferences. Narrative would on this account supply particular premises to which general premises would be applied. The consistency of a narrative would be based on the internal coherence of these particular premises along with their coherence with commonly accepted beliefs (endoxa). To argue that narrative plays a larger role than this in Aristotle's account of practical reasoning would require more textual support than F. provides. Perhaps an innovative rereading of Aristotle's discussion in the Poetics of the unity of plot (muthos, which corresponds to ‘narrative’) could support F.'s hunch that Aristotle's account of the enthymeme need not presuppose the centrality of inferential reasoning in rhetorical discourse.

The volume offers a valuable account of a key element of Greek rhetorical practice and serves as a worthy reminder of the importance of the insights of the Greeks for our own theory of narrative and vice versa. It invites us to revisit difficult and unresolved issues of the relationships that hold among narrative, rhetorical speech and logic, and how they were regarded in antiquity.