In her closing remarks B. claims that a reading as unorthodox as hers must ‘earn any consideration, let alone its keep, entirely from scratch without prior presumption of a sympathetic welcome’ (p. 206). This volume's interpretation is certainly unorthodox, yet with its dedication to the text and constancy of argument it has without doubt earned its welcome, if not its keep.

The central thesis is that the form of the good is fundamentally interrogative. It stands atop the dialectical enterprise, not as a super-premiss, but as the question: ‘Is — good?’ (B. dubs this ‘the G-question’) (2.8). B. further claims that ‘good’ in the G-question has the common-sense meaning of the predicate ‘good’ in the open sentence ‘good(—)’. Because the good has many varied uses (e.g. Is — a good book? A good friend? A good resolution?), it has no fixed definition (2.12) and is entirely sensitive to the circumstances in which the G-question is asked (2.15). Put simply, the interrogative good just is ‘what is fitting in whatever sphere’ (p. 127). By B.'s own admission this interpretation brings the good's role in dialectic closer to ‘common-sense practical reasoning’ than a mysterious or unattainable part of some ‘perfectionist fantasy’ (2.16). Plato's dialectical method, B. surmises, serves to ‘delineate an idealized version of something not only real but totally familiar, namely ethical reasoning in light of changeable circumstances’ (p. 107).

Perhaps unsurprisingly, long stretches of the volume are devoted to reading counter-textually or at least against the textual tradition. The chief (alleged) misunderstanding that B. is at pains to correct is that the good is a premise or paradigm, with a rich, non-obvious definition (like those definitions that Plato rejects, i.e., it is pleasure or wisdom [505b5–d1]), from which the philosopher deduces practical conclusions. As part of this corrective campaign, B. dismantles a number of popular views, including that the good is a paradigm (2.10), with a discrete, definable nature (2.13), known via generalisation (2.13) or by acquaintance (2.14). She also rejects that dialectic ‘ascends’ to the form of the good (2.11), without empirical content (2.15), produces infallible results (2.16), operates (like mathematics) deductively (2.17), is available only to the rulers (2.18) or appeals either to a ‘mega rule-book [of forms] in Platonic heaven’ (2.19) or a monistic vision of the good from which all other virtue questions derive their answers (2.20).

Perhaps the hardest of these views to let go is that the good has a discrete definition. The text clearly states that the dialectician possesses an account of what each thing is (534b1) and that this is ‘likewise’ (ὡσαύτως) true for the good (b6). However, B. thinks that this passage as a whole suggests not that the dialectician possesses a fixed definition of the good, but that he does not confuse it with anything else (pp. 74–5). Whether or not this is plausible, it is nevertheless notable that B.'s concept of an interrogative and therefore indefinable good is not as vacuous as it first appears. One of the volume's more ingenious moves is to propose that the power of dialectic does not rest with an obscured definition of goodness, a secret for which so many have gone searching; instead, its power derives from ‘reasoners who are intellectually and morally flawless’ (p. 107), whose ‘moral, intellectual, and practical training have made them … as accurate detectors of the truth as any human being could be’ (p. 117). The thought is that ideal reasoners, equipped purely with the G-question and a common-sense predicative notion of the good, might achieve the results the Republic imagines. This explains the scant descriptions of dialectic in the dialogue, which, to the credit of B.'s reading, spends far more time on the education and character of the ideal reasoner. This ingenious proposal is emblematic of something about the entire project: if not always the most natural interpretation of individual passages, B.'s reading stays absolutely true to the spirit of the Republic overall – as a document about the state of the soul best suited to political rule.

Part 3 addresses the claim that the form of the good causes, in addition to the knowability of intelligibles, their being and reality (509b1–9). B. argues that being and reality ‘are added or accrue to’ intelligibles because of the good (p. 140) and proposes two interpretations: (1) intelligibles, especially the virtue forms, become plans or standards when employed in G-question reasoning (3.2) or (2) a neutral thing, i.e. a course of action, becomes virtuous when understood to engage with the good via the G-question (3.4). B. also rejects three popular interpretations: the diffusion account, whereby the good generates the other forms in ‘a timeless sort of pluralizing diffusion of itself’ (3.4); the perfectionist account, whereby the forms, like all good things, are good because they are perfect instances of themselves (3.5); and system accounts, whereby the good just is the appropriate arrangement of the virtue forms and other goods (3.6).

B.'s rejection of perfectionism – the view that the good is essentially the ‘property of being a perfect instance as such, or of perfection or ideality as such’ (p. 156) – is puzzling, especially since Part 2 claims that the good of the interrogative account is, effectively, whatever is fitting in a given context (p. 127). It is unclear how this appeal to fittingness, while necessary for the highly contextual interrogative good to retain any unified sense across all situations in which the G-question is asked, is supposed to escape perfectionism. What, after all, grounds the judgement that something is fitting in a certain context? In ordinary language, we say something is ‘fitting’ when it coheres with whatever standards are relevant. Similarly, we say something is ‘more fitting’ when it coheres better or more perfectly. But if the good enables us to make judgements about fittingness and if fittingness just is a comparison to an ideal, then it is unclear what makes B.'s interrogative good non-perfectionistic.

Part 4 tackles some remaining issues, including ambiguities in Plato's use of ‘the good’ (4.1–2), the importance of mathematical training (4.3) and cosmological concerns (4.4). An important consideration comes to light in 4.2: Proclus thought Plato held two distinct forms of the good, the sun-like good and the participand good. The good of the Sun cannot be the participand good or the good of the one-over-many relation. This is because the sun's relation to ordinary visible objects is not participation; visible objects are not ‘junior suns’ (p. 169), not representations or shares of the sun. B. concludes that the form of the good must have both a sun-like interrogative mode, the focus thus far, and a declarative mode, or the mode of the form participated in by the many good things. Once the declarative mode peeks out from behind the clouds, however, it is tempting to ask whether (and why not) some of the passages B. is at pains to read consistently with the sun-like mode in Parts 2 and 3 might not be better explained by the possibility that Plato, in this part of the Republic, passes from one mode to the other without much warning or that he addresses both modes simultaneously.

Further challenges are sure to be brought against B.'s interpretation. But this is hardly a shortcoming of the project; rather, it is a testament to its brilliance. Rarely does a piece of scholarship offer such a thorough and ingenious revision of such a familiar text. Even if aspects of B.'s revision turn out to be untenable, its comprehensive and unique vision is sure to inspire rich discussions for years to come.