Mind and World in Aristotle’s De Anima

Here is a fact about humans: we use our senses to pick up on things around us and our intellect to understand whatever is out there to be understood. In Mind and World in Aristotle’s De Anima, Kelsey argues that this fact is, in Aristotle’s view, in need of an explanation. He finds one in De Anima 3.8’s suggestion that “intelligence [is] form of forms, and sensibility form of sensibilia” (432a2–3; quoted on p. 2). Roughly, his proposal is that our sensibility and intelligence “enter into the very idea” of their objects; they know them because they help make them what they are (20).This is an admirably adventurous thesis, and Kelsey’s arguments for it are likewise so. A particular strength, in fact, is the way the book brings out what is at stake philosophically in familiar and seemingly obscure doctrines alike. Two highlights, which I discuss below, are its discussions of how Aristotle’s engagement with his predecessors shapes his questions (and then makes it hard to answer them) and of how his account of perceptible qualities helps him meet this challenge. This book is therefore a significant contribution to scholarship on the De Anima (DA), and it will be of great value to scholars working on Aristotle’s philosophy of mind. Part of what makes it valuable, moreover, is how it encourages us to ask better questions about core Aristotelian doctrines: while some of Kelsey’s proposals (especially his account of per se causation, which I discuss below) are provocative, they are always productively so.The introduction sets up Kelsey’s core question. It is: “What about” our sensibility and intelligence “makes” them “subject[s] of” some “attribute” (6)? What must they be they like—in their essence (8)—to know what they do? The next three chapters argue that the DA is concerned to answer this question, and, moreover, to do so in a particular way: to show why sensibility and intelligence know “real beings” as they really are—not as they appear.Kelsey’s argument for this claim is a highlight of the book. It takes off from the observation that DA 2.5 answers two foundational questions in a way that, according to DA 3.3, should be problematic. These are: (A) whether perceivers and perceptibles “are like or unlike,” and (B) “whether perceiving is a matter of ‘being affected’ or ‘altered’” (40). The difficulty is that 2.5 wants to answer that perception is (A∗) like-by-like and (B∗) a case of being altered—where 3.3 suggests that those very commitments got Aristotle’s predecessors into trouble. Those thinkers held that “both understanding and judging are held to be like a kind of perceiving” (427a17–b6), apparently because they thought these are (A∗) like-by-like and (B∗) being altered (43).This “diagnosis,” Kelsey argues, is interesting because it “connects” (A∗) and (B∗) to another question: whether “how things are” just is “how they appear” (43). (A∗), for instance, reflects the view that “our judgments are … the mere projecting of a random and fluctuating piece of ourselves,” so that that our “verdicts” are like us, not things as they are (45). And (B∗) expresses the view (roughly) that in acquiring the state in virtue of which we perceive and judge, we come to judge differently (because we get altered), but not better. This puts the resulting judgments—also alterations—all “on a par,” so that all appearances are true (47–49). The upshot is that what lies behind (A) and (B) is the question of whether knowledge and perception get at things as they really are (43).How, then, can 2.5 safely claim that perceiving is like-by-like and a case of being altered? Kelsey’s answer is that it revises (A∗) and (B∗) to avoid the difficulties 3.3 identifies. The improved version of (A∗) turns out to be an application of the view that “having been affected by something is a matter of having become what the affecting agent is in the business of making things be” (53). This principle then does important work: applied to perception, it entails that it belongs “to colors by nature to be seen.” This means that we are like the things we perceive because it is in their nature to make us be like them—and so we perceive them as they are, not as they appear (54–55). (B∗) gets likewise qualified: one way of being altered—which perceivers experience in perception—is “being busy upon [one’s] appointed work” (58).This is an ingenious argument with significant upshots. One is a richer sense of the importance 2.5 attaches to (A) and (B). Another is a clearer sense of the work that Aristotle’s theory of per se causation can do—and the questions that we need to ask about it. This emerges from Kelsey’s discussion of the principle that “having been affected by something is a matter of having become what the affecting agent is in the business of making things be,” which he glosses as the claim that “(accidents apart) the way things interact is in line with their respective natures” (54).Kelsey’s account of this principle is a welcome and important addition to the literature, and it will undoubtedly spark debate. In Kelsey’s formulation, the principle is quite strong—as, indeed, it needs to be in order to justify the claim that it is in the nature of colors to be seen by us. Kelsey’s idea appears to be that the principle lets us take the full specification of the effect of an agent’s exercising some power on a patient and read off it the correct specification of that power. In Kelsey’s example, “pavement does not just happen to afford automobiles a smooth ride, that is what it is for” (54). Pavement’s effect is a smooth ride—because it has in its nature a power for this. (Thus, the agent’s “business” is what it has “a power of making things be”—its “defining work” [53].)If this is right, there is pressure to ensure we specify the effect from which we read off the agent’s power in just the right way. At a minimum, we must find the per se effect. After all, while mudflats do afford cars a smooth ride, their per se effect on cars is something else—for they, unlike pavement, surely do not have a power for affording smooth rides. The lesson is that our specification of the effect should omit things it is not plausible to think the agent has a power for. This sharpening, however, may make us wonder about the idea that in being perceived, a perceptible object exercises a power for “revealing itself” to sentient creatures (55)—a claim that is an important piece of groundwork for Kelsey’s ultimate view that sensibility is part of the form of perceptible objects.Another question concerns how best to characterize the dialectic between Aristotle and his predecessors about (A∗). In Kelsey’s discussion of 3.3, the relata of the likeness are “the qualities of the persons passing judgment” and the “verdicts” they issue (“what judges ‘see’”) (45). In his discussion of 2.5, they are “sentient creatures” and the “objects they perceive”—which “appear as they do thanks in part to something of them” (54–55). These two formulations raise an interesting question about the extent to which Aristotle is maintaining his predecessors’ picture of the explanandum: if the objects’ “appear[ing]” this way is the “verdict” we issue, he will have kept his predecessors’ formulation of the relata and revised his explanation of their likeness—but if it is their effect on sentient creatures while they are perceiving, he will have modified their conception of the relata too.With this groundwork in place, the next two chapters introduce two ingredients in Kelsey’s answer to his original question. The first builds on his account of like-by-like causation; this, he argues, requires “likeness in form” (69). The second expands on those forms, appealing to the notion of a “measure.” Measures, Kelsey argues, are not just like but also “conceptually prior to the objects known by them” (85). Purple, for instance, lies on a spectrum characterized by a “mean” or “middle”; moreover, it is “in the nature of” purple to lie “on one side or the other” of that middle (92–95). But, the next chapter argues, this middle is itself defined with reference to sensibility—which is its measure (see also 434a9; quoted on p. 100). When DA 2.12 defines sensibility as “a kind of ratio” (424a27–28)—itself a middle (424a4–5)—its point is that sensibility is a form that is the measure of perceptible forms (103–17). The upshot is that sensibility knows purple, because sensibility “enter[s] into” purple’s nature (90–91; see also 112–17). Something similar holds in the case of intelligence. The last two chapters therefore argue, roughly, that intelligence is a measure of its objects, insofar as “the clarity and distinctness which characterize its activity are (as it were) the very form” of those objects—what makes them intelligible (154).This section again sheds helpful new light on familiar doctrines. In Kelsey’s hands, Aristotle’s view that sensibility is a “middle” is not just an arcane detail or a convenient explanation of our ability to discriminate a variety of qualities but a reflection of their very nature—one that is well equipped to explain why we perceive them as they are. Kelsey’s book is to be commended not only for sketching a promising new position but for compelling us to get clear on what it would be to answer the question it addresses.Kelsey’s reasoning here, of course, is not uncontroversial. Consider his claim that where qualities lie on a spectrum, “each particular quality” “will lie” either “in the middle” or “on one side of the spectrum,” “by its very nature” (95). Why “by its very nature”? Kelsey’s answer is that it is “composed of” contraries according to a particular “ratio” (94)—one that makes it lie in such a position “in its very own nature” (94). But if this is the evidence that perceptible qualities are defined with reference to the middle, it may be challenging to reject the alternative view that they are “defined by the ‘ratios’” of contraries (95n18).1 For the reason they are defined with reference to the middle is that their ratio gives them the features that situate them with respect to it—and it is hard to see how the ratio can do this if it is not in their nature.Kelsey’s position, as the book’s conclusion emphasizes, is a surprising one: it assigns “a kind of priority—the priority of measure to measured” to sensibility and intelligence to what they perceive and understand (159). This, I would emphasize, is a welcome surprise: by clearly identifying the considerations that might push Aristotle to such a position, Kelsey’s excellent book opens up new questions and makes sure we think hard about what they are asking.Many thanks to Reier Helle for helpful comments on a draft of this review.