In a number of different studies, several primate species (most of the research is on chimpanzees) demonstrate an ability to assess and react not just to another’s overt behavior, but rather to the goal they are attempting to achieve—and in some cases the behavioral plan (intention) of the other as well.
There are two very different types of studies, from a variety of different experimental paradigms, requiring at least five different behavioral responses, most of which have also been used with human infants, that support this contention. First are studies in which the subject reacts not toward the behavior of the other but toward his goals. For example, chimpanzees react differently to very similar behaviors when a human is refraining from giving food because he is unwilling versus unable to do so (Call, Hare, Carpenter, & Tomasello, 2004). They (and other apes) also react differently when a human does something on purpose versus by accident (Call & Tomasello, 1998). They also discern that another’s goal is to reach an out-of-reach object and then they either help him to reach it (Warneken, Chen, & Tomasello, 2006; Warneken, Hare, Melis, Hanus, & Tomasello, 2007) or grab it first if it suggests the location of food (Hare & Tomasello, 2004). And having reliably anticipated an experimenter’s movement to a feeding station when he stands and turns in a certain direction by going there to meet him, they hesitate when novel contextual cues change the goal he is probably pursuing (D. Buttelmann, S. Schütte, M. Carpenter, J. Call, & M. Tomasello, unpublished data). Recent research on capuchin monkeys has found that they too help a human experimenter to gain an out-of-reach reward and distinguish between unwilling and unable acts (Barnes, Hill, Langer, Martinez, & Santos, 2008; Phillips, Barnes, Mahajan, Yamaguchi, & Santos, 2009). It is important to note that in some of these studies, the behavior at the time the subject must react is identical in experimental and control conditions, with the only difference being in the immediately preceding context, which is the only cue that can be used to infer the different goals involved. Given the range of behaviors and contexts investigated, predicting outcomes based on rules abstracted from past experience which do not involve attributing goals to other individuals seems unlikely.
A second set of studies show converging evidence that chimpanzees infer goals from surface action: Chimpanzees act out what they understand the other to be attempting to do, not what he is actually doing. In these three studies (all modeled on similar studies with human infants), chimpanzees imitated what the human was trying to do, not what he actually did (Tomasello & Carpenter, 2005); they imitated his purposeful rather than his accidental actions (Tomasello & Carpenter, 2005); and they selectively imitated his actions based on an understanding of why he chose this particular behavioral means toward his goal (Buttelmann, Carpenter, Call, & Tomasello, 2007)—which may be construed as understanding his intention, in the sense of the action plan chosen for pursuing a goal. These actions and contexts were novel, as is usual in studies of social learning, and so rules based on likely behavior in given contexts do not apply.
The only reasonable conclusion to be drawn from these studies, in our opinion, is that chimpanzees, and perhaps other nonhuman primates, understand the actions of others not just in terms of surface behaviors but also in terms of the underlying goals, and possibly intentions, involved.