MOVOUS COUTURE 2017秋冬高定营造个性幻梦！

Krupenye et al. (1) report that great apes pass a false belief task (2) by anticipating an agent will return to a location at which they last saw an object. Krupenye et al. (1) acknowledge that their data are "open to an abstract behavior rule–based explanation" (e.g., 3, 4), yet suggest apes have a theory of mind (ToM) because minimalist accounts "cannot easily accommodate" an agents' insight into goals (5) or their ability to "infer whether others can see through objects that look opaque, based on their own experience" (6, 7).

We note that minimalist accounts can easily explain such findings by positing that apes think only about agents and outcomes (4). Regarding goals, apes retrieve an object that an agent reaches for and are said to predict goals even when an agent fails to complete an action. Yet, they need only think about the likely outcome that will ensue from an agent's action. If the outcome might ensue they remain patient. If the outcome won't ensue they complete it themselves or thwart a competitor's outcome, thinking about outcomes but not explicitly about goals. Regarding what others can see, apes know when to gesture to others because they have learned not to gesture when one's eyes/face are blocked, they know to take a hidden piece of food because they project connections between eyes/faces and actions, and need only gain experience of opaque versus transparent barriers to consider whether an agent looking through such is connected to an object at a location and predict the agent's actions.

In Krupenye et al.'s (1, 2) false belief task and all other tasks (5-7), success can ensue without computing mental states of perception, desire, intention, ignorance or belief. Great apes are skilled at predicting behaviour but not ToM.

References and Notes
1. C. Krupenye, F. Kano, S. Hirata, J. Call, M. Tomasello, Great apes anticipate that other individuals will act according to false beliefs. Science 354, 110 (2016).
2. V. Southgate, A. Senju, G. Csibra, Action anticipation through attribution of false belief by 2-year-olds. Psychol. Sci. 18, 587 (2007).
3. J. Perner, T. Ruffman, Infants' insight into the mind: How deep? Science 308, 214 (2005).
4. T. Ruffman, To belief or not belief: Children's theory of mind. Dev. Rev. 34, 265 (2014).
5. J. Call, M. Tomasello, Does the chimpanzee have a theory of mind? 30 years later. Trends Cogn. Sci. 12, 187 (2008).
6. K. Karg, M. Schmelz, J. Call, M. Tomasello, The goggles experiment: can chimpanzees use self-experience to infer what a competitor can see? Anim. Behav. 105, 211 (2015).
7. A. Senju, V. Southgate, C. Snape, M. Leonard, G. Csibra, Do 18-month-olds really attribute mental states to others? A critical test. Psychol. Sci. 22, 878 (2011).

This study (1) was interpreted by its authors, by Frans de Waal (2), and widely by the media as showing or at least strongly supporting the claim that apes can ascribe false beliefs to agents. I think the study provides next to no support to this claim.

Krupenye and his colleagues conducted two experiments. In one of them they made apes watch a scene in which a person in ape custom – King Kong – stole a stone from a man, put it under one box, frightened the man away, and then, while the man was away, first moved the stone to under another box and then took it and left. There were slight variations on this scenario, as can be seen on the article's accompanying online videos. When the man came back and made as if to retrieve the stone, 17 out of 30 apes first looked at the box under which the man last saw the stone being hid (5 first looked at the other stone and 8 did not look at either). Similar results were obtained in the second experiment, which is similar in the relevant respect to the first: 20 apes first looked at the location in which the returning man last saw the relevant thing, 10 first looked at the other location, and 10 didn't look at either.

Krupenye and his colleagues concluded from the results of these experiments that 'great apes anticipate that other individuals will act according to false beliefs'. However, already the term 'anticipation' involves an unwarranted interpretation: what did the 17 apes anticipate, according to the authors? That the agent will try to retrieve the stone from under the box at which they looked first? The results do not support this at all. Unlike related experiments that were conducted with children (4, 5), Krupenye and his colleagues did not show that any violations of expectations were involved in the apes' case when an agent first looked at where an object is and not where he falsely believed it to be. The experiments show that the apes ASSOCIATE a location with an agent, but they contain no measure to indicate that they anticipate any action consequential on this association.

In addition, although the location at which 17 out 30 apes first looked was that in which the agent was supposed to falsely believe the stone to be, that location can also be characterised as the one in which the agent last saw the stone. This is the last location in which the drama enfolded with the agent present and involved, so this is the location most strongly ASSOCIATED with the agent. Again, association can explain the apes' looking pattern. So the results can be described as: when an agent returns to the scene, apes reliably look at a location where he last saw an object. We need not refer either to anticipation or to the agent's false belief, and mere association – which reflects very little understanding – can explain the results.

Toward the end of their article, the authors acknowledge 'that apes could solve the task by relying on a rule that agents search for things where they last saw them', referring to (5), where such an interpretation of related results is found. But even this acknowledgment contains too much of an unacceptable interpretation. Unlike standard false belief studies with children, the apes had no task to solve. They were not asked anything and they didn't have to do anything. And as I claimed earlier, the looking pattern is better explained not by anticipation of agents' search but by some sort of reflexive pattern due to an association of a location with an agent.

Earlier studies have recurrently failed to show that apes can ascribe false beliefs to agents (6–9). To make their results compatible with the earlier failures, the authors suggest that although 'apes have not yet succeeded on tasks that measure false-belief understanding based on explicit behavioural choices', the apes operate on an IMPLICIT level with an understanding of false beliefs. But if the looking pattern reflected understanding, then this IS an explicit behavioural choice: the apes that looked at the location in which the agent last saw the stone could have looked elsewhere, as almost half of the apes participating in the study actually did. I see no coherent notion of implicit belief supported by this study that can resolve the incompatibility between the authors' interpretation of their results and the results obtained in earlier studies. By contrast, the 'mere association' interpretation suggested here does not involve any such apparent incompatibility.

I therefore think that the new study does not provide us with a good reason to hold that apes can ascribe false beliefs to agents. The authors' title, 'great apes anticipate that other individuals will act according to false beliefs' is therefore unjustified by their results. And stronger claims, like de Waal's 'apes know what others believe; understanding false beliefs is not unique to humans', are surely unwarranted.

References and Notes
1. C. Krupenye, F. Kano, S. Hirata, J. Call, M. Tomasello, Science 354, 110–114 (2016).
2. F. B. M. de Waal, Science 354, 39–40 (2016).
3. K.H. Onishi, R. Baillargeon, Science 308, 255–258 (2005).
4. R. Baillargeon, R. M. Scott, Z. He, Trends Cogn. Sci. 14, 110–118 (2010).
5. J. Perner, T. Ruffman, Science 308, 214–216 (2005).
6. B. Hare, J. Call, M. Tomasello, Anim. Behav. 61, 139–151 (2001).
7. J. Kaminski, J. Call, M. Tomasello, Cognition 109, 224–234 (2008).
8. C. Krachun, M. Carpenter, J. Call, M. Tomasello, Dev. Sci. 12, 521–535 (2009).
9. J. Call, M. Tomasello, Trends Cogn. Sci. 12, 187–192 (2008).