To what extent does research support the claim that non-human animals, particular chimpanzees, have human-like understanding of mental states?
In the last three decades, cognitive approaches to the study of animal behaviour have attracted increasing research attention. Much of this interest has concentrated on social cognition and whether animals understand the mental states of others in the same way humans do. In an attempt to address this question research has predominately focused on non-human primates (hereafter primates), especially chimpanzees. This popularity is probably owing to chimpanzees’ evolutionary relatedness to humans, which is the closest of all the primate species and therefore increases the likelihood that they might share human cognitive abilities.
Human mental states are a suite of cognitive abilities which allow us to understand others’ psychological states. One example is theory of mind (ToM) which is the ability to understand that other individuals have beliefs and that such beliefs can be different from your own. This is illustrated by the ‘Smarties Test’ in which an individual is shown a tube of smarties and asked what they think is inside the tube, they correctly infer ‘smarties’ however when the tube is opened it is revealed that the tube contains only pencils. The individual is then asked what another naive person would answer if given the same test. If the individual has a ToM they should understand that the naive individual will have a false belief about the tube’s contents and therefore answer in the same way, i.e., smarties. Research shows that children under the age of 4 years have not fully developed a ToM and invariably fail the smarties test whereas children above this age pass the test.
Experiments conducted on primates have shown little evidence that they have a ToM. For example, Call and Tomasello (1999) tested chimpanzees, orangutans and children with a non-verbal false belief task. The task involved a series of finding games in which a reward was hidden in one of two identical containers by an adult (the hider), and another adult, who had witnessed the hiding process (the communicator), placed a token on the baited container which acted as a marker to indicate to the subject where the food was hidden. The subjects learnt to use the marker to locate the reward and ignore the marker when they knew it to be incorrect (during visible displacement trials). In the crucial false belief trials, the communicator watched the baiting of the container and then left the area and during the communicator’s absence the hider swapped the location of the containers. When the communicator returned she placed the marker on the container in the location she had seen the reward being hidden, which was incorrect, and therefore the communicator had a false belief about where the reward was hidden. To show an understanding of this belief the subjects would have to choose the container which the communicator did not place the marker on. The results showed that children chose the correct location of the reward demonstrating their understanding of the communicator’s false belief whereas none of the apes succeeded in choosing the correct location demonstrating their lack of false belief understanding. Other studies have also failed to find evidence of ToM in primates (for review see Povinelli 2004) and many now think that language is needed for the development of ToM which seems to have led researchers away from studying ToM in primates and other species of animals.
Another mental state that has received much research attention is intentionally. Intentionally can be defined as understanding the behaviours of other as intentional, goal directed activities. The first study to address this issue was conducted by Premack and Woodruff (1978). They presented Sarah with videotaped sequences of a human actor in several problem-solving situations that were familiar to Sarah. For example, the actor was depicted looking up toward an out-of-reach banana hanging from the ceiling, or attempting to operate a hose that was unattached to a tap. After Sarah had viewed each problem, she was shown a pair of photographs, one of which depicted the solution to the problem. For instance, in the out-of-reach banana situation, the solution consisted of the actor moving a box under the banana. Overall, Sarah performed well on these tasks from the beginning but Savage-Rumbaugh et al (1978) pointed out that Sarah might have been choosing alternatives based on the simple associations among objects formed from her experiences with caretakers and their behaviour with test items, such hoses and taps. Savage-Rumbaugh et al (1978) analysed each item Sarah was presented with and found that, overall, items for which such associative procedures were most straightforward were the ones on which Sarah performed best. In addition, Savage-Rumbaugh et al (1978) presented two language-trained chimpanzees with a matching-to-sample task in which the chimpanzees were shown, for example, a picture of a key and asked to select between pictures of a box and a padlock. Both chimpanzees performed well above chance in this task and selected the correct alternative in the majority of trials. These results therefore provided an alternative explanation to that of understanding of intention in others.
Other research on chimpanzees’ understanding of intentions has produced mixed results. Povinelli et al (1998), for example, tested six chimpanzees who could choose to point to one of two caretakers to provide them with a cup of juice. However, prior to the choice the apes had experience of the caretakers’ behaviour in which one caretaker would accidentally drop the drink on the floor and the other would purposely empty the cup of juice on the floor. The authors reasoned that an understanding of the intentionally of the caretakers would result in the apes choosing the caretaker to provide them with juice who had previously acted with good intention but accidentally spilled the drink. The results showed that the apes made no preference for the ‘clumsy’ caretaker indicating that the apes were unable to understand the experimenters’ intentions. However, Call and Tomasello (2004) argue that there was no motivation to choose between the two caretakers because whomever they chose they would receive no juice. This possibility can be tested by repeating the experiment and allowing the caretakers to sometimes drop or purposely spill the drink on some trials but not on others. This would allow the apes to receive juice in some of the trials e.g., 50%, and therefore raise the apes’ motivation whilst maintaining the essence of the experiment.
In contrast to the Povinelli et al’s (1998) study, Call et al (2004) found evidence that chimpanzees do show understanding of intentional behaviour in humans. The authors conducted an ‘unwilling’ and ‘unable’ test in which an experimenter would feed an ape a grape through a hole in the Plexiglas window. On some trials the ape would not receive the grape from the experimenter who in one condition would accidentally drop the grape and in a second condition would tease the ape by repeatedly pulling the grape back from the hole. When the apes’ behaviour was analysed it was found that they behaved differently in the two conditions. When the grapes were accidentally dropped the ape remained in the room longer than when the experimenter teased the ape. Moreover, during the teasing condition the apes often showed frustrated behaviour, such as spitting and punching the plexi-glass panel. The authors suggest this demonstrates the apes’ understanding of the experimenter’s intentions of being unable to give the grape in one condition and unwilling to give the grape in the other condition. However, it would be interesting to know exactly how the teasing behaviour was executed by the experimenter. For example, if the experimenter used loud vocalisations in the teasing condition but not in the clumsy condition it is possible the apes were just reacting to this rather than the experimenter’s intention.
Knowing what others can see is another mental state humans possess and one that chimpanzees have shown evidence in possessing. For example, Hare et al (2001) used a competitive paradigm to test whether a subordinate chimpanzee would take into account how a dominant rival would behave depending if the rival had seen food being baited behind a barrier. The two chimpanzees faced each other across an ape enclosure which had one barrier in the middle of the enclosure. An experimenter hid one piece behind the barrier (on the subordinate’s side) and one piece in the open. In some conditions the dominant saw the reward being placed behind the barrier and in other conditions did not see the food being placed there. After the baiting procedure the subordinate and dominant were then allowed access to rewards. When the subordinate had observed that the dominant had not seen the food being hidden behind the barrier the subordinate would approach the food behind the barrier leaving the dominant to retrieve the reward that was out in the open. However, when the dominant had seen the food placed behind the barrier the subordinate showed no preference for the food behind the barrier. In a follow-up study Hare et al (2001) extended these findings by repeating the experiment but in some conditions the dominant was switched (after seeing the food being hidden) with another dominant who had not seen the food being hidden. The results showed that the subordinate approached the reward placed behind the barrier more often when the dominant had been switched with the naive dominant and the authors suggest that this demonstrates chimpanzees can predict who had seen what. In addition, a control condition ruled out the possibility that the subordinate was acting on simple rules such as preferring food lying close to the barriers. For example, when the opaque barrier was replaced by a transparent barrier, so that the dominant would see the reward even when he had not seen the food being placed there, the subordinate’s preference for the food behind the barrier disappeared.
Although the above two studies seem to show that chimpanzees understand what others see, several researchers suggest that chimpanzees are using simpler mechanisms to solve the task. Karin-D’Arcy and Povinelli (2002), for example, argue that subordinate chimpanzees might simple prefer food that is next to barriers as this provides some safety from attack by the dominant. The authors repeated Hare et al’s 2002 study and found that subordinate chimpanzees preferred food next to the barrier even when both chimpanzees could see both pieces of food (the barrier was positioned vertically in relation to both chimpanzees). In addition, Povinelli and Vonk (2003) argue that even if the chimpanzees did not have a preference per se for the barriers in Hare et al’s 2002 studies the chimpanzees can understand which food the dominant will be likely to go for without evoking any mental states. For example, behavioural abstraction taken from previous experiences of similar interactions can be used to predict the dominant’s behaviour. This can occur if the subordinate understands that when the dominant is facing the food the dominant will go towards that food, so use the rule ‘don not go for the food if the dominant is orientated toward the food’. Povinelli and Vonk (2003) further argue that experiments testing mental states in animals will only be of value if the possibility of behavioural abstraction is removed from the experimental equation. The authors therefore advocate that mental attribution can only be tested in chimpanzees and other animals by showing evidence of an individual extrapolating its own experiences to the mental states of others, a theory first put forward by Heyes (1998). Povinelli and Vonk (2003) suggest the following as a suitable experiment to address Heyes’ (1998) idea. First, allow an ape to interact with two buckets, one red and one black. The red bucket is placed over the ape’s head and the ape can clearly see its surroundings through the bucket, however when the black bucket is placed over the ape’s head he can see nothing. Now if the ape is confronted with two experimenters, one wearing the black bucket over her head and the other the red bucket over her head, and the ape begs for food only from the experimenter wearing the red bucket then it is likely that the ape is using mental attribution as the only experience the ape has had of the buckets is through its own experience. Povinelli and Vonk (2003) argue that experiments of this nature can provide evidence of mental states and rule out alternative interpretations such as behavioural abstractions.
The mental states of humans include a suite of cognitive functions which allow us to understand others’ knowledge and beliefs such as theory of mind and intentionally. Many argue that only humans have these types of mental states and other animals use lower cognitive functions based on learning and experience, such as behavioural abstractions. However, there is increasing evidence that other animals, especially chimpanzees, might also have human-like mental states, such as understanding intentions and understanding the relationship between seeing and knowing. One major caveat, however, is, to date, the most convincing research on chimpanzees’ mental states can not distinguish if the chimpanzees’ actions are based on human-like mental states or based on mechanisms not involving mental attribution such as behavioural abstractions. What is therefore needed is further research which can distinguish between these two possibilities and therefore increase our understanding of the psychological abilities of chimpanzees and other animals.
Call, J., & Tomasello, M.(1999) A nonverbal theory of mind test. The performance of children and apes. Child Development, 70, 381-395
Call, J., Hare, B., Carpenter, M.,& Tomasello (2004) Unwilling or unable: Chimapnzees’ understanding of human intentional action. Developmental Science, 7, 488-498
Hare,B., Call, J., Agnetta, B.,& Tomasello, M.(2000) Chimpanzees know what conspecifics do and do not see. Animal Behaviour, 59, 771-785.
Hare,B., Call, J., & Tomasello, M.(2001). Do chimpanzees know what conspecifics know? Animal Behaviour, 61, 139-151.
Heyes,CM (1998). Theory of mind in nonhuman primates. Behavioural and Brain
Sciences, 21, 101-148.
Povinelli, D.J., Perilloux, H.K., Reaux, J.E. & Bierschwale, D.T. (1998) Young and Juvenile chimpanzees reactions to intentional versus accidental and inadvertent actions. Behavioural Processes, 42, 205-218
Povinelli, D.J. (2004) Behind the ape’s appearance: escaping anthropocentrism in the study of other minds. Dadalus 29-41
Povinelli, D. J.,& Giambrone, S. (2001). Reasoning about beliefs: A human specialization? Child Development, 72, 691-695.
Povinelli, D.J. & Vonk, J. (2003) Chimpanzees minds: Suspiciously human? Trends
in Cognitive Sciences, 7, 157-160.
Povinelli, D.J. & Vonk, J. (2004). We don’t need a microscope to explore the chimpanzee mind. Mind and Language, 19, 1-28.
Tomasello, M., Call, J., and Hare, B. (2003b). Chimpanzees understand
psychological states- the question is which ones and to what extent. Trends in Cognitive Sciences, 7, 153-156.
Savage-Rumbaugh, E.S., Rumbaugh, D.M., & Boysen, S.T. (1978). Sarah’s problems in comprehension. Behavioral and Brain Sciences, 1, 555-557.