Comment on Michael Tomasello s Origins of Shared Intentionality 1 The demarcation problem What owners of pets always knew, became only recently a well established scientific view, namely that at least higher non-human animals are cognitively very similar to humans. Empirical findings strongly suggest that these animals possess all the cognitive abilities that especially philosophers considered to be unique features of humans. Aristotle believed that only humans are rational beings. Descartes claimed that animals don t have any kind of consciousness. Linguistic philosophers like Sellars or Davidson claimed that animals don t possess a language, don t entertain thoughts or don t have knowledge. Many philosophers in the continental tradition strongly hesitate to ascribe something like self-consciousness to non-human animals. But great apes show a complexity in their behaviour that can only be explained by assuming that they are conscious creatures who possess belief-like states, knowledge and at least rudimentary capacities to communicate through symbolic behaviour. Moreover, the Rouge-test (testing whether creatures are able to recognize that their mirror-images are reflections of themselves by looking at the behaviour of these creatures) has established that non-human animals like dolphins, elephants and primates display all the typical forms of behaviour of self-conscious creatures. Other experiments suggest that higher animals are quite smart and rational beings. So it seems as if there were no in principle, but only a gradual difference between nonhuman animals and humans. Of course, humans possess a more sophisticated language than non-human animals, especially because of its productivity and systematicity. And furthermore, humans are still much cleverer than non-human animals. But this merely gradual difference does not explain the unique cultural accomplishments of human beings. After all, only humans have social institutions, traditions and norms. There seems to be a huge explanatory gap between these unique human phenomena and the view that human cognition is only gradually different from non-human cognition. 2 Tomasello s solution to the demarcation problem I label the problem of explaining the uniqueness of human cultural accomplishments the demarcation problem. This problem is the starting point of Michael Tomasello s considerations. He formerly held the view that the real difference between humans and non-human animals comes down to the fact that humans but not animals are able to 1
understand the intentions of their conspecifics. Humans but not animals were supposed to possess a theory of mind. In the meantime, Tomasello has changed his mind. For, new empirical findings show that at least great apes are able to distinguish mere movements from intentional actions. And they recognize quite well whether someone intentionally withholds their food or whether he is unsuccessfully trying to feed them, in which case they are patiently waiting for being fed. These findings suggest that great apes even possess a rudimentary form of a theory of mind. For this reason, Tomasello has recently offered a new explanation of the in principle difference between humans and non-human animals. We are now told that the cognitive difference that makes all the difference is the uniquely human motivation for sharing psychological states with others. The human animal would then be the only animal doing real mental teamwork. Even in very young children, Tomasello finds a tendency to imitate others, to build up joint attention towards external objects by pointing gestures and looking behaviour which only humans can display, to provide others with helpful information, to share emotions and experiences with others, to reengage the adult in a common activity on which he has given up, to have shared goals (which are known to be shared by all interactors) and to coordinate action plans and roles within a group of interactors. In many cases, children are interested in the common activity as such and do not see it as a mere means to their own benefit. Tomasello points out that even very young children display genuine cooperative and collaborative behaviour, whereas great apes, according to Tomasello, completely lack motivation for even the most basic forms of sharing psychological states. Non-human animals are not at all interested in taking part in true collaboration and cooperation. They are not team-players. I quote from one of Tomasello s recent articles: In general, it is almost unimaginable that two chimpanzees might spontaneously do something as simple as carrying something together or help each other to make a tool, that is, do something with the commitment to do it together and help each other with their role if needed. (Tomasello 2005, p. 685) Thus, Tomasello defends the three following claims: T1 Ontogenetically, in the first year of their live, children build up the ability to understand the intentions of others (mind reading) and develop an intrinsic motivation to share psychological states with others. These are the two sources of their collaborative behaviour. 2
T2 T3 Their collaborative behaviour provides humans with a selective advantage over the more competitive and individualistic behaviour of other primates. The ability for mind reading and the intrinsic motivation to share psychological states with others are the two biological origins of the uniqueness of human culture. Before I present two objections to Tomasello s view, let me set the stage by a few general remarks. One might feel tempted to oppose to Tomasello s claim that there are biological sources of our having a culture. Why, one might ask, can t human culture just be the product of pure contingency and then has continued to exist not in virtue of our genetic disposition to live as cultural beings but in virtue of raising our offspring within a socio-cultural environment (acculturation)? However, this alternative story about the origin of our culture seems to be falsified by the fact that there are biological limits of acculturation. Apes don t become humans just by living within a human environment. There is, of course, the fascinating story of the bonobo Kanzi. Having been raised by human researchers he shows quite a lot of characteristic human behaviour. He is able to make and handle fire by himself; he communicates with humans by using abstract symbols; he is interested in playing music, looking at the yellow press etc. Nevertheless, there still remain important differences to humans. So it seems reasonable to look for a at least partly biological explanation of the unique human capacity to create a complex culture. However, I am not so sure whether there has to be a unique cognitive skill to explain this feature. Why shouldn t gradual cognitive differences between humans and apes add up to some completely new phenomenon? 3 First objection: Tomasello s explanation is not sufficiently justified Tomasello s explanation of the uniqueness of human culture can only be correct if the following three constraints are met: (i) Shared goals and joint action plans with coordinated roles within group behaviour must be a unique feature of humans. (ii) The intrinsic motivation for sharing intentionality must not be epiphenomenal. It has to provide a basic explanation. (iii) The intrinsic motivation for sharing intentionality must (together with the capacity for mind reading) be sufficient to explain the human cultural 3
accomplishments. If it were not, then Tomasello could no longer claim that motivation for sharing intentionality is the one mark of human animals. Moreover, the sharing of intentionality also has to be the best explanation for the uniqueness of human culture. Pr. (i) Observations of wild chimpanzees show that during group hunts these animals perform a variety of different complementary and coordinated roles, some of them having only very low individual success rates. This suggests that every group member is participating in a common activity with a shared goal. Other activities have no other aim as sharing experiences or conforming to a group norm (spread of hand-clasp grooming). There seems also to be imitation behaviour without any direct reward (unsuccessful early nut-cracking attempts). But if we find at least minimal traits of collaboration with shared goals among apes, then the in principle difference between humans and apes vanishes. We are back to yet another merely gradual difference. I just don t see why the mere fact that great apes (even completely) lack sensitivity to pointing gestures and looking behaviour, which is established by some of Tomasello s spectacular experiments, proves that great apes completely lack the capacity for sharing intentions. Just as well, one might argue that they simply use different means to communicate and collaborate than humans do. Pr. (ii) Tomasello seems to assume that the capacity for mind reading and the early human interest in sharing intentionality are two independent sources of human collaboration. However, simulation theorists hold that human understanding relies on empathy realized in the neural activity of the so-called mirrorneurons. Simulation theory would suggest that understanding and engagement is one and the same thing. We understand someone else by activating the same mental state that she herself is having. So sharing intentionality and understanding someone else comes down to the very same thing. Isn t that the true source of shared intentionality among humans? Pr. (iii) I have some doubts about Tomasello s claim that our early motivation for sharing intentionality, together with our ability to understand others, is sufficient to explain our cultural accomplishments. Within human social 4
practice and institutions, the fact of social norms and normativity is of greatest significance. But how can sharing intentions explain normativity? Whenever someone violates a norm he simply seems to opt out of the shared activity. Shared intentionality is, of course, a precondition of social norms to which all members of the society are committed. But it does not sufficiently explain normativity. On the other hand, one might come up with a different explanation of the basic features of our culture. Let us assume, with Tomasello, that collaboration and true cooperation is one of these features. Since collaboration is a very good means to the achievement of individual benefit, we could just as well explain cooperative behaviour as the rational choice of a group of radical egoists. We need not refer to any intrinsic motivation for sharing intentions. So even if young children show this motivation, it need not be the final explanation of adult collaboration. 4 Second objection: the over-intellectualization of collaborative engagement According to Tomasello, truly collaborative activities (even at the age of 12 to 14 months) require a cognitive representation not only of the shared goals but also of the often complementary action plans and roles of all the participants. This cognitive representation ( dialogical cognitive representation ) must be internal to any participant of the activity. Now here is a problem for this view: It assumes that children at the age of about one year are able to ascribe properties to themselves and to others. But do children at that age possess an explicit knowledge of themselves and others? Moreover, if they represent action plans and roles of others, they must be able to ascribe instrumental beliefs to others. But as the false-belief test shows, children cannot attribute beliefs to others before the age of 3 or 4. So how can they be able to represent the plans of others at the age of 1 year, as Tomasello seems to assume? So, perhaps Tomasello s understanding of early collaborative activities or early sharing of intentionality among humans is the product of a massive overintellectualization of small children. Maybe there is participation involved in these kinds of activities without the need of any prior cognitive representation of the different roles of all the participants. Maybe knowing-how precedes knowing-that at least in these cases, as the philosopher Gilbert Ryle would put it. 5