As A. K. Cotton acknowledges at the beginning of her monograph Platonic Dialogue and the Education of the Reader, ‘the idea that a reader's relationship with Plato's text is analogous to that of the respondent with the discussion leader’ within the dialogue, and ‘that we engage in a dialogue with the text almost parallel to theirs’, ‘is almost a commonplace of Platonic criticism’ (4). But Cotton has the merit of articulating this commonplace much more clearly and precisely than is often done, and of asking how exactly the dialogue between interlocutors is supposed to affect the dialogue of the reader with the text, and what kind of reader response Plato is inviting. Not surprisingly, her starting point is Plato's notorious (written) concerns about written texts expressed in the Phaedrus: ‘writing cannot contain or convey knowledge’, and will give to the ‘receiver’ the mistaken perception that he or she has learned something – that is, has acquired knowledge – from reading (6–7). She claims that the Phaedrus also suggests, however, that a written text, in the right hands, ‘may have a special role to play in awakening the soul of its receiver towards knowledge’ (17). I have no doubt that Plato thought as much, but Cotton's reference to the language of hupomnēmata at 276d3, and to the way in which sensible images act as hupomnēmata for the recollection of the Forms earlier in the dialogue, fails to support her case: Socrates remarks in that passage that writings can serve only as ‘reminders’ for their authors (16). The book's central thesis is that the way in which writing can awaken the reader's soul ‘towards knowledge’ is not by pointing the reader, however indirectly, implicitly, non-dogmatically, or even ironically, towards the right views, but by developing the reader/learner's ‘ability to engage in a certain way’ in dialectical inquiry (26). The familiar developments between ‘early’, ‘middle’, and ‘late’ dialogues are thus accepted but seen as part of a single coherent educational project towards the reader's/learner's full development of what Cotton calls ‘dialectical virtue’. Plato's reader is invited to treat the characterization of the interlocutors within the dialogues, and the description of their dialectical behaviour, ‘as a commentary on responses appropriate and inappropriate in the reader’ (28). Cotton's programme, clearly sketched in Chapter 1, is ambitious and sophisticated, and is carried out with impressive ingenuity in the following six chapters (the eighth and final chapter, besides summarizing some of the book's conclusions, introduces a notion of ‘civic virtue’ which does not appear to be sufficiently grounded on the analyses in the rest of the book). An especially instructive aspect of her inquiry is the attention paid to the ‘affective’ dimension of the interlocutor's and reader's responses: through the representation of the interlocutors in his written dialogues, and the labours to which he submits us as readers, Plato teaches us that ‘the learner's engagement must be cognitive-affective in character; and it involves a range of specific experiences, including discomfort, frustration, anger, confusion, disbelief, and a desire to flee’ (263). Perhaps because of her belief that what the Platonic dialogues are about is not philosophical views or doctrines but a process of education in ‘dialectical virtue’, Cotton has remarkably little to say concerning the psychological and epistemological underpinnings of the views on, and methods of, education which she attributes to Plato. The Cave allegory in the Republic, which is unsurprisingly adopted as an instructive image of Plato's insights on learning and educational development in Chapter 2, is discussed without any reference to the various cognitive stages which the phases of the ascent in and outside the Cave are meant to represent. Two central features of Plato's conception of learning identified by Cotton – the individual learner's own efforts and participation, and the necessity of some trigger to catalyse the learning process (263) – are not connected, as one might well have expected, to the ‘theory of recollection’ or the related imagery of psychic pregnancy or Socratic midwifery. Even Cotton's laudable stress on the ‘affective’ aspects of the learning process could have been helpfully complemented by some consideration of Platonic moral psychology. Despite these reservations, and the unavoidable limitations and oversimplifications involved in any attempt to characterize Plato's corpus as one single, unified project, I believe that readers with an interest in Platonic writing and method will benefit greatly from Cotton's insightful inquiry.