O’Connell’s Rhetoric of Seeing joins a growing list of titles interested in restoring performance and visuality to our understanding of ancient Greek culture and, especially, political and legal culture. This work distinguishes itself through its limited focus on the rhetorical function of seeing and visuality in extant forensic speeches. Each chapter addresses a different kind of seeing, often beginning with an overview of the relevant secondary literature, then considering other ancient genres or fields—Plato and Aristotle, poetry or history, medical or rhetorical treatises, and finally examining two or three important or representative examples from legal speeches. O’Connell divides the work into three “kinds” of seeing.First, he looks at what the audience can literally see. Part 1, “Physical Sight,” considers examples of visual bias concerning the physical appearance of litigants or others. This includes familiar arguments from probability (eikos) based on appearance: one need only look at Antigenes to know that he could not have overpowered Pantainetos (Demosthenes 37, Against Pantainetos); a glance at the pensioner’s disability and we can see that the charge of hubris is ridiculous (Lysias 31, For the Disabled Man). It is surprising here that O’Connell does not do more with the visual aspect of eikos arguments, which are said to have begun with Hermes’ infantile appeal to his own youth in his defense speech against Apollo: “Do I look like a cattle driver to you, a burly fellow?” (Hymn to Hermes 265). This is a central and well-trod aspect of ancient rhetorical theory that seems to call out for inclusion and that could have been given a new layer of interpretation through O’Connell’s visual approach. Counter-probability is rare in legal arguments but equally important in the development of rhetorical theory and with similar implications for visual rhetoric. The strong (or young) man who asserts that he would not have assaulted the weak (or old) man because he would be the first suspected depends in part upon similarly visible features of his person (Antiphon 2.2.3; cf. Aristotle 2.24.10–11).The final chapter of Part 1 takes up issues of movement and gesture, with references to gesture in Plato and Aristotle, a brief review of physiognomy, and then a discussion of Aeschines’ widely studied Against Timarkhos. There is brief mention of the rhetorical cannon of delivery or hupokrisis and the recommendations of Cicero, Quintilian, and Dionysius and more recent work on categories of gesture by Karsouris and Hughes, but O’Connell does not take up the rhetorical canon of delivery (hupokrisis) in depth. His discussion of delivery faces the same problems that most scholars face: there simply is no good way to talk about it as a general category. Either atomize the body to talk about hands, then faces, then movement, or settle for vague gestural and expressive categories and recommendations: modest and appropriate or excessive and inappropriate. Attending to specific cases and speeches is often more successful. O’Connell’s discussion of Aeschines’ speech Against Timarkhos goes further toward demonstrating his overall thesis than do his general comments.Second, we can observe the language of visuality in the speech itself, when the speaker asks the audience to look at something literally and directly as visual evidence, or figuratively or indirectly through terms of demonstration, display, and witnessing. Part 2, “The Language of Demonstration and Visibility,” looks at terms of seeing in the orations: deiknumi (demonstrate or display) and its variants (apodeixis, epideixis, endeixis, etc.), phaneros and phainomai (visible) and their variants (kataphanēs, apophainō, etc.), and martus (witness) and its variants. Chapter3 considers the language of display and witnessing, where speakers seek to prove their case by describing what has been shown and seen by witnesses, or where they demand witnesses to prove what has been asserted. “How else,” says Antiphon in On the Chorus Boy, “can I make true things trustworthy” except through the consistent affirmation of witnesses who were present? (Antiphon 6.29). This section is valuable for bringing into focus the centrality of visibility and sight to notions of truth, a factor that can easily be lost in translations. Thus, the speaker of On the Chorus Boy emphasizes not only that he was appointed a counselor and entered the council-house as such, but that he was seen (horōntes) and was visible (phaneros) doing so. O’Connell does not claim, but he enables one to conclude, that the infamous dichotomy between truth and probability in rhetorical theory typically devolves into these two kinds of seeing: what has been witnessed (and is therefore true) and what the situation “looks like” to the audience (and must be probable).Included here is a section on medical and philosophical interest in the visible as an epistemological link to the invisible. O’Connell quotes Anaxoagoras’ maxim, “Visible things are the face of things which are unclear” (101). This could lead to a discussion of the complex and rhetorically important doctrine of signs as tools of rhetorical argument. Instead, O’Connell moves on in chapter four to discuss how speakers use the language of visibility and demonstration to describe arguments. This, argues O’Connell, places jurors into the position of virtual witnesses themselves of something publicly known, as it was known that some grain dealers had been changing their prices over the course of a day (Lysias 22, Against the Grain Dealers). Or they are witnesses of arguments as demonstrations (epideixō). Speakers contrast what the opponent simply says (legei) with what the speaker will “demonstrate in an evident manner” (110). The language of display is thus used to differentiate mere telling from showing. This reference to visual metaphors for the persuasive effects of argument suggests a larger connection with rhetorical argument generally and the role of vision therein.Third, we can attend to imagination as internal sight, or what O’Connell calls “shared spectatorship,” when speakers “try to make the jurors visualize their version of events and accept it as true” (123). This includes a discussion of techniques of vivid description like enargeia, hyptyposis, or ekphrasis via detailed description. O’Connell looks specifically at described scenes of civic suffering, as when Lycurgus describes the panic after the Athenian defeat at Chaeronea. Shared spectatorship can also occur through the construction of “internal audiences—characters in a narrative who witness what is being described and whose reaction can function as a prompt and model for the jury, as when, in the speech Against Diogeiton (Lysias 32), the speaker recounts Diogeiton’s daughter speaking to the family about her father’s embezzlement and lying (150). Visualization can also be heightened through deictic pointing to the persons in court whose actions or suffering is being described, fusing what is physically seen (demonstratio ad oculo) with what is imagined (deixis ad phantasma): “this man here they seized and tied to the pillar” (Lysias fr. 279, 155). This takes us back to the beginning, which addressed seeing in performance space itself. This last section was for me the most interesting and informative, and it seemed the most widely applicable to forensic, and indeed all genres of oratory. Here too, I saw connections to a basic category of rhetorical discourse: narrative and narrative theory, to notions of realism and verisimilitude, to the conjuring of story worlds and the work of narrative inference.Certainly, anyone interested in visual and spatial rhetorics, bodily rhetoric, performance, and related topics will want to be familiar with O’Connell’s work. I found much to admire in every chapter, and more so as the book advanced to later sections and chapters. At the same time, in each section I found myself thinking about some clear and relevant connections to fundamentals of rhetorical theory—theories of probability and signs, of argument and narrative—that the work brushed up against but did not explore. Of course, O’Connell writes as a classicist, not a rhetorician, and we cannot expect any work to follow up every thread that it pulls on, particularly those outside the author’s bailiwick. So, we might rather say that this work promises to amply repay the attention of scholars of rhetorical history and theory for its insights into the operation of sight and seeing—physical, lexical, and imaginary—in Attic forensic speeches.