Reviewed by: Locke's Science of Knowledge by Matthew Priselac Martha Brandt Bolton Matthew Priselac. Locke's Science of Knowledge. New York: Routledge, 2017. Pp. 240. Cloth, $140.00. This interesting and challenging book addresses the apparent gap between the empiricist account of the origin of ideas and the theory of knowledge in the Essay concerning Human Understanding. Matthew Priselac makes an impressive argument that they are complementary parts of a coherent program. It consists of a naturalistic interpretation on which the Essay's main aim is to provide the kind of understanding of the mind, knowledge, and probability afforded by modern methods of natural scientific inquiry. On this view, the Essay advances a hypothesis intended to explain the epistemic phenomena of everyday life, that is, what we take ourselves to know rather than hold as probable. It describes a process by which the mind generates experience from nothing but simple ideas received by sense or reflection, operating on them to form various kinds of compound ideas, and culminating in sensory and reflective experience. Completed by Locke's definitions of knowledge and probability, the hypothesis is confirmed by its explanatory adequacy: its ability to show that we know instances of the four main types of propositions commonly counted as exemplars of knowledge (propositions that express identity, a relation among things, co-existence of things in the same subject, and real existence), and to explain why we regard them as such. The project is explanatory, not justificatory. It is purely epistemic in that claims of speculative metaphysics are excluded. The hypothesis of the origin and taxonomy of ideas is said to be grounded on introspection, observation, and, importantly, as the author has it, Locke's analysis of the sorts of ideas that contribute to the mind's activity of constructing experience that can be intelligibly understood. Completed by the definition of knowledge, its severest test is posed by Locke's account of sensory cognition of the existence of things outside us. Understanding how that fits the definition of knowledge is a well-known interpretative crux. To my knowledge, no previously published interpretation is fully consistent with the structure and strictly internal theory of warrant it implies, but Priselac urges a credible account that fully comports with the definition. The most central points of this interpretation are these. The notion of "genetic structure" (5–6) adds the following to the familiar doctrine that compound ideas of modes, substances, and relations are formed by the mind's uniting two or more simpler ideas by means of certain relations. Because the mind is always conscious of its current operations, the relations are [End Page 405] known by reflection. Therefore, each compound idea exhibits its simple constituents and the relations by which, and order in which, its more complex constituents have been formed. The simple idea of power is important because it or one of its simple modes enters the genetic structure of all ideas of qualities. The structure of ideas of substances is not taken from the chapter on ideas of substances or the account of their archetypes (Essay, 2.23.1–6; 30.5; 12.6). Rather, it is based on what, as Priselac has it, Locke takes our thoughts about substances to be, that is, things that are existentially self-subsistent, unified, have various qualities, and interact with ourselves and each other—nothing more. Accordingly, ideas of substances are "confused," "obscure," and "inadequate," for reasons having to do with nothing but their genesis within the mind. According to Locke, knowledge is perception of the agreement of two ideas that are joined by the mind; probability is the presumed, but not perceived, agreement between ideas so joined. Priselac contends that, for Locke, two ideas agree just in case one contains the other. Three points are offered in support of this claim: it makes sense of the project—because ideas are either simple or have genetic structure, the idea-containment model applies to all four types of knowledge recognized by Locke; genetic structure is explicitly mentioned in connection with "trifling propositions" (i.e. propositions that do not increase our knowledge [Essay, 4.8.4]); and the containment model comports with the foundational role given...