Do 5-month-old infants see humans as material objects?.

Infants expect objects to be solid and cohesive, and to move on continuous paths through space. In this study, we examine whether infants understand that human beings are material objects, subject to these same principles. We report that 5-month-old infants apply the constraint of continuous motion to inanimate blocks, but not to people. This suggests that young infants have two separate modes of construal: one for inanimate objects and another for humans.

Infants have appropriate expectations regarding the nature of objects. As early as 4 months of age, infants are sensitive to violations of the constraints of object physics such as cohesion (objects move as bounded wholes), solidity (two objects do not occupy the same space at the same time), and continuity (objects move on connected paths). It is also the case that infants distinguish between inanimate objects and animates, namely humans, in important ways. For example, they recognize that humans are self-propelled while inanimate objects move only after contact with another object.

However, do infants recognize that, in some cases, the principles they readily apply to inanimate objects also apply to humans? In other words, do they understand that people are objects that are subject to physical laws such as cohesion, solidity, and continuity? If they do, it would imply an understanding of the duality of humans: animate objects engaging in self-initiated action, yet fundamentally material objects. Here, we examine this question by using the test case of the continuity principle.

In study, infants were familiarized to an object moving across a stage that contained two occluding screens. Infants who saw the object slide behind the first screen, then "reappear" from behind the second screen without having traversed the intermediary space, looked longer at a subsequent test display of one object than at a display of two identical objects, implying that the former was more novel or surprising than the latter. That is, the infants recognized the discontinuity of motion in the original presentation and, since a single object cannot move in such a manner, inferred the presence of two objects. Consistent with this interpretation is the fact that infants in the same study who were familiarized to an object moving in a continuous manner (behind one screen, through the intermediary space, then behind the second screen and out again) showed preferential looking to the test display of two objects.

General discussion

The present results suggest that while 5-month-old infants apply the principle of continuous motion to inanimate objects, they do not readily apply it to humans. There is evidence from prior studies that infants differentiate between animate and inanimate objects in appropriate ways. However, the present study represents a situation in which they mistakenly differentiate between the two, suggesting that at 5 months, infants do not readily view humans as material objects.

The precise nature of the distinction here is an open question. It might be best characterized as humans versus everything else, but there is also good reason to think that the category of "humans" might be too narrow. After all, infants treat a wide variety of animals, including humans, as similar to each other but different from inanimate objects. In fact, they will even regard non-animals such as geometrical figures as goal-directed agents, akin to humans, so long as they are imbued with multiple animate characteristics (e.g. behavioral adjustment and contingent reactivity). On the other hand, recent research has also suggested that, in some cases, infants treat humans as separate not only from inanimate objects, but other animate entities. It thus remains an open question whether the results of the present study are due to a distinction between animates versus inanimates, intentional agents versus non-intentional objects, or humans versus other entities.

Regardless of whether the current results apply specifically to humans or to all animate/animated entities, the present study still suggests that, early in development, infants are mistaken about the physical constraints that apply to humans. Of course, one must be cautious when interpreting a null result. Yet, the combined data are consistent with the current interpretation: the expectations that infants have of inanimate objects are not readily applied to humans in analogous situations in which the application is appropriate.

Infants do have certain expectations about humans—and these expectations are often (correctly) not applied to inanimate objects. For example, infants interpret humans—but not inanimates—as social, goal-directed entities. To illustrate, after witnessing a human arm repeatedly reach and grasp one of two toys, 6-month-old infants expect the arm to continue to reach for that particular goal toy, even if its location has been moved. In contrast, infants do not have such expectations of an inanimate rod engaging in the same motion patterns, implying that the rod's motion was not interpreted as goal-directed. Additionally, while infants will imitate the intentional actions of a human, they will not imitate the motions of an inanimate object.

It is possible that the dissociation between objects and humans found in the present study forms the complement to the distinction between humans and inanimate objects in terms of social, goal-directed behavior. Together, this double dissociation suggests that young infants may have different modes of construal for humans versus inanimate objects: humans are construed in terms of social and intentional actions, while inanimate objects are interpreted via a system sensitive to object physics. The existence of this human/inanimate distinction, and the differential application of principles to each, may help infants to define these areas of knowledge early in development. The appreciation that these construals overlap—that in certain regards, people are just objects—may be a developmental accomplishment.