1. Premack & Woodruff:

This is one of the very first studies to investigate whether chimpanzees could infer others' intention. Chimpanzees were shown videos of an actor doing something and then encountered a problem. Then, pictures are presented to the chimpanzees with several ways to solve the problem. Results showed that in general, chimpanzees would choose the correct picture that solved the problem. Based on these results, the authors suggest that chimpanzees have theory of mind. The authors raise alternatives and handle them with control studies. For instance, one alternative interpretation is that chimpanzees are merely associating elements on the videos with those on the pictures when they are choosing. An additional study was conducted to refute this alternative interpretation. In this study, the chimpanzees were shown an actor being locked in a room. A picture of a normal and broken key was presented to the chimpanzees. Needless to say, the chimpanzees chose the normal key as the solution.

Another alternative interpretation is that the chimpanzees are putting themselves in the others' situation rather than inferring and attributing others' mental states. To control for this, chimpanzees were shown videos of two actors (one actor was a trainer they like and the other actor was a trainer they do not like), and gave them a choice of a good or a bad outcome for the story. The results showed that chimpanzees were more likely to choose the good outcome for the trainer that they like and bad outcome for the trainer they do not like. It was noted, however, that this may not be the best control; one should assume that even the disliked trainer would like to solve the problem.

2. Povinelli et al:

In this study, the actors investigated whether chimpanzees understand intentions by showing them videos of intentional and accidental actions. For the accidental action, the chimpanzees saw one actor giving another actor some juice but accidentally dropped the juice on the floor. There were two intentional actions. For the passive intentional action, one actor gives some juice to the other actor, pulls the juice away and then pours it on the floor. For the other intentional action, rather than pouring the juice on the floor, the actor threw the juice to the floor. The test was to see which actor (the one doing the accidental or intentional actions), the chimpanzees would want to get their food from. The results showed that the chimpanzees have no preferences; suggesting that they didn't appreciate accidental versus intentional actions.

3. Call & Tomasello:

In this study the authors looked at the understanding of intentions in young children, orangutan and chimpanzees. In this study, subjects were trained to expect that a reward would be under a location that an experimenter marked. In experimental trials, one location was marked intentionally, while another was marked unintentionally (by dropping the marker). Non-human primates did poorly, but so also did human children, who we generally think of as appreciating intentions.

Discussion Issues:

1. What can we conclude by the primates' poor performance in these tasks? Can we conclude that primates do not have intentions or that the tasks are not convincing enough?

-- A number of design difficulties were discussed. It's difficult to make a situation ecologically valid from a chimpanzee's point of view while maintaining experimental control. A second difficulty is that many of these designs employ repeated measures, and after the same accidental action keeps happening in the same exact way, it sort of starts to look intentional. This could contribute to poor performances in the later trials (see in particular Call & Tomasello).

-- It was suggested that one good study might be to test understanding of intentions via prediction of anothers' future behavior. Difficulties were mentioned in that there are likely other cognitive mechanisms related to predicting (e.g., working memory) operating in addition to intentional inferences.

-- An imitiation study was also suggested (perhaps along the lines of the Meltzoff). The problem in using this kind of design experiment in primates (or chimpanzees) is that they do not imitate. In a study by Call and Tomasello, they looked at whether chimpanzees and orangutans will imitate an experimenter's behavior to solve a problem. In the study, food was place at a distant location out of reach. Subjects (chimpanzee and children) were given a pole to help them reach for the food. The pole has two ends: flat or rack. Results showed that children would use the same end of the pole (flat or rack) that the experimenter used in his/her succession in reaching for the food. Chimpanzees and orangutans, on the contrary, used either ends of the pole randomly. This is also true in the wild. When even after one primate observes how another primate used a stick to get nuts out of a tree, the novice will use the stick but randomly, without imitating behavior.

2. Out of this frustration grew a somewhat nihlistic issue. What the heck are we hoping to find in the study of theory of mind in non-human primates? A couple of viewpoints were offered. One was that we are hoping to find whether there is a capacity for this ability in our close evolutionary ancestors, as this could give us a notion of where it came from, and perhaps the kinds of cognitive mechanisms which support it in us. An analogue was drawn with language. We know that chimps don't use language the way we do, but it is nonetheless an interesting endeavour to determine if they can. A second viewpoint was offered suggesting that we may be shooting ourselves in the foot a little here by insisting that we use our own ToM as a guideline. It could be that ToM develops out of a sensitivity to certain information that is helpful to predict future actions. Chimps could have that sensitivity, but the information that is relevant to non-human primates might be different and thus lead to the development of a quite different ToM. Issues about cross-cultural differences in ToM development within a species were also brought up in this light.

3. Why did ToM evolve in humans and not in primates?

One reason for evolving ToM is that we need it to speed up the processing required to engage in complex social interactions. Are our interactions more complex than those of non-human primates? Well, we live in bigger groups which may mean more complexity. This discussion gave rise to the inevitable chicken/egg question (and we don't really know how to phrase this): Did complex social interaction demands force the evolution of ToM or is it the evolution of ToM that allowed for complex social interactions. Issues regarding the nature of how things are selected for maintenance and how changes occur were brought up (which gave rise to either vigorous head shaking or confused looks).