Children provide a test case for psychological and epistemological theories. Many theories fail: They do not suffer the little children to come to them. That's part of the point of Vasudevi Reddy's fascinating How Infants Know Minds. The title contains the thesis: Infants know other minds. They don't act by mere instinct, or mere reflex. And they aren't just responding to stimuli or bodies. Infants know other minds. Reddy began to formulate her thesis when she had children of her own: "The most striking thing my babies were telling me was that they could understand me and others as persons. They were teasing, joking, playing with our expectations and attitudes and interests, being shy, and showing offer long before they were able to speak. Understanding other minds didn't seem to be a problem to them" (1). Consider the fact that "neonates can imitate tongue protrusions" and the "considerable evidence that they can imitate mouth opening, and some evidence that they can imitate finger movements and eye blinking and even one vocal sound - an elongated 'aaaa'" (47). This should arouse more wonder than it does. How does an infant even know he or she has a tongue, much less that it's somewhere near the part of the face as the adult's tongue? The adult intends to be imitated; somehow the infant knows that too, discerning intentions and not mere bodily movements. Piaget thought infant imitation impossible, for two reasons: "(1) because imitation requires quite a complex ability to rasp the similarity between self and the person or animal to be imitated, an ability presumed to be impossible at birth without considerable experience of self and others; and (2) because imitation of specific acts was presumed to require specific learning about the similarity between self and other in the parts of the body producing those acts" (45). Infants haven't studied themselves in mirrors: How can they know these thing. Yet they do imitate. Which means something is amiss in post-Piaget child pscyhology. Reddy provides evidence that tongue protrusion isn't a mere reflex. One study found that infants sometimes delay imitation, suggesting that the protrusion isn't merely imitative but a provocation to interaction (53). She also probes the notion that "mere" imitation doesn't indicate any contact of mind with mind. Savages, Darwin and other explorers have said, are keen mimics; it's a low and primitive form of interaction. Reddy doesn't believe it: "imitation can be a means of making contact in the absence of any common language. . . . It seems to establish something shared, some common ground on which both interactants can stand. . . . It seems to be a psychological door through which one is immediately led into world of intentional relations with another person" (45). The sorts of other-directed activities don't start with birth: "Even in utero, and as early as 15 weeks gestational age, ultrasound scanning studies have revealed at least fifteen different well-coordinated non-reflexive movement patterns, including independent finger movements, rapid and slow mouth opening, hand movements, repetitive contacting of mouth with fingers." She thinks it "indisputable that human neonates are capable not only of intentional actions, but also of intentional actions motivated by curiosity and interest rather than merely those driven be physiological needs" (49). Why has this been missed? Reddy faults the dualistic assumptions that are embedded in psychology, whether cognitivist or behaviorist. When she began talking about infants knowing other minds, she was warned to "mind the gap" - the gap between mind and body, the gap between mind and mind. If there is such a gap, there needs to be a bridge to cross it. What if there's no gap? What if minds aren't concealed inside bodies but communicated through bodies? What if minds were never disconnected in the first place, thus undermining the ned for bridge-building? What if minds are constituted by their interaction with one another? What if, in short, psychology is best pursued not as a first-person nor as a third-person discourse, but as a second-person science (cue Buber here, though, sadly, not Rosenstock!). Psychologists reject Cartesian dualism; but the warnings to "mind the gap" show that Descartes shadows the discipline. In one clever riposte to the denial of intersubjectivity between infants and others, Reddy writes: "The reason for such distinctions is often attributed to the scientifically respectable desire for parsimony. Parsimony, however, exists only in a theoretical context: if, and only if, we accept a theory which says that perceiving the physical is simpler than perceiving the mental, is it more parsimonious to suggest that infants perceive only physical qualities. But if we reject the dualism separating mind from body  . . . it may well be more parsimonious as well as more coherent to reject the dualism of separating the perception of the body from the perception of the mind" (23). Reddy recognizes the radical implications of her suggestion. If psychology is second-person, then the best approach to observation may not be detached experiment but engagement (note the incisive discussion of Piaget on 37-38). Maybe a child psychologist will learn things from playing with children that he could not learn by watching them play. Perhaps "Suffer the children to come to me" is the most scientific stance available.