Friday, August 12, 2011

The de Waal/Tomasello Debate in Primate Studies

As hard as it is to read the minds of our fellow human beings, it is even harder to read the minds of other animals.

Aristotle's biological works are full of observations about animal minds that qualify him as the first animal psychologist. His general conclusion was that there were traces of almost every human mental ability in other animals, which included emotions, parental care, social learning, communication, imagination, practical judgment, and even something close to intellect. He argued that some animals were capable of voluntary action like that of human children, although they lacked the capacity for deliberate choice that arises in human adults. Some animals are solitary and others gregarious. Of the gregarious animals, some are political. Some of the political animals have leaders. The distinguishing characteristic of the political animals is that they cooperate for some common work or function (koinon ergon). Humans, bees, ants, wasps, and cranes are all political animals in this sense (HA, 488a7-14). Human beings are the most political animals because through speech or reason (logos), they share their conceptions of the advantageous, the just, and the good (Pol, 1253a1-18). Through speech, human beings cooperate for shared ends in ways that are more complex, more flexible, and more extensive that is possible for other political animals. Through speech, human beings can deliberate about the "common advantage" (koinon sumpheron) as the criterion of justice (Rh, 1362a15-63b5). A just political community can be judged to be one that serves the common advantage of all its members, as contrasted with an unjust political community that serves only the private advantage of its ruling group (NE, 1160a13-14; Pol, 1279a17-19).

Aristotle's comparative animal psychology includes the observation that monkeys and apes belong to intermediate species close to human beings in that they "share in the nature of both a human being and the quadrupeds" (HA, 502a16). From his anatomical comparisons, which included dissections of monkeys and apes, he concluded that in their feet, legs, hands, face, teeth, and internal parts, the apes are humanlike (HA 502a17-b27; PA, 689b1-35).

And yet, in relying on direct observations of animal behavior and anatomical dissection, Aristotle, like any animal psychologist, had no direct access to the animal mind. With human beings, he had the data of speech, and thus he could study human politics through the study of political rhetoric in which human beings debate their opinions about political life. Other animals communicate in other ways, but this animal communication is often not as rich as human speech, although Aristotle observed the waggle dance of bees, which we now know to be a remarkably complex form of abstract communication.

The fundamental problem is that the social reality of animal life--including human life--is a mental construction of the animal mind that goes beyond the physical reality of the directly observable world. To understand that mental construction of social reality, we have to draw from our inward subjective experience of our minds and our intersubjective world of symbolism. From our observation of animal behavior, we can project some of our mental experience onto them, but we can never be sure how accurate this is.

We have to assume, as Aquinas said, that "the internal passions of animals can be gathered from their outward movements" (ST, I-II, q. 34, a. 4). But then, as Darwin observed in The Descent of Man, in trying to understand the evolution of human mental abilities from the mental powers of our animal ancestors, we face "the impossibility of judging what passes through the mind of an animal" (Penguin ed., 105).

Over the past hundred years, we have had more systematic study of primate behavior and cognition than was done previously. And over the last forty years, we have seen some methodologically sophisticated studies of primates both in the wild and in captivity, and the captive studies have included controlled, and often ingenious, experimentation. But with all of this primate research, we still face the same problem: primate social reality is a construction of animal minds, and judging what passes through those minds is always speculative and uncertain.

Consider, for example, the many news reports this week--in the New York Times and elsewhere--of new experiments that appear to show chimpanzee generosity. Frans de Waal and his colleagues at Emory University claim to show "spontaneous prosocial choice by chimpanzees." Experiments by other groups suggest that chimpanzees are not very helpful to one another, although observations of chimpanzees in the wild suggest the opposite. De Waal has trained his chimpanzees at the Yerkes National Primate Research Center to exchange tokens for food. In this experiment, chimpanzees were paired up and placed in adjoining cages. One chimp could choose a token from a bucket, with tokens of two different colors. One color would exchange for food for oneself but not for the other chimp. The other color would exchange for food for oneself and for the other chimp. In a majority of cases, the chimp would choose the generous option.

Notice the odd features of this experiment. The participants were seven adult female chimps. We must wonder how much we can conclude from the behavior of seven individuals. We must also wonder why only females were tested, particularly since much of the seemingly cooperative behavior among chimps in the wild is male behavior--such as group hunting and warfare. We might question how much generosity we really see here. The report is that the generous tendency for each individual chimp ranged from 52.9% to 66.7%. That suggests that for some of these individuals the choice of tokens was almost random. And even those showing the more generous tendency are not bearing any costs in their generosity.

As indicated in the New York Times article, this research report has been criticized by Michael Tomasello, Co-Director of the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany. Anyone who follows this kind of research knows that Tomasello and de Waal have been on opposite sides of a debate for many years. Most recently, Tomasello and his colleagues have published a study in Nature arguing that "collaboration encourages equal sharing in children but not in chimpanzees." They show that children around the age of three show equitable distribution of resources with those engaged in collaborative activities, but chimps do not exhibit such equal sharing with collaborative partners. They offer this as evidence for a general theory of human evolution: the uniquely human propensity for social norms of fairness and equity arose among human ancestors who shared resources after collaborative foraging, and the ancient evolutionary propensity arises early in the development of children as an evolutionarily natural trait.

De Waal and Tomasello agree in general that human nature can be explained as a product of evolutionary primate history, so that human beings are similar to their closest living relatives--chimpanzees and bonobos--but also quite different in humanly unique ways. And yet de Waal and Tomasello disagree in their emphasis--de Waal emphasizing the similarities, Tomasello emphasizing the differences--a disagreement that runs through much of the primate research.

This disagreement was evident a few years ago when de Waal and his colleague Sarah Brosnan gained wide publicity for a report in 2003 entitled "Monkeys Reject Unequal Pay," which seemed to show that capuchin monkeys had a "sense of fairness." Having been trained to exchange tokens for food, the monkeys seemed to be engaged in economic exchange, which is why this kind of research has gained attention from economists. In an experiment to test for "inequity aversion," monkeys who offered a token for food would receive either a slice of cucumber or a grape, and the monkeys clearly favored the grape as more desirable. If one monkey got a slice of cucumber, while seeing that another monkey nearby got a grape, the monkey apparently protested against this, either by refusing to exhange the tokens or by throwing away the cucumber. It seemed as though the monkey was protesting "unequal pay."

But again we might wonder about some of the features of this research. The report was based on the behavior of only five females. Apparently, males had not shown "inequity aversion." We have to wonder then how far we can go in drawing general conclusions from this experiment.

Moreover, as some critics noticed, it was possible that the monkeys weren't showing "inequity aversion" but only frustrated expectations. If a monkey receives a cucumber slice, while noticing that grapes are available, she might feel frustration at not getting the more desirable food, but this would have nothing to do with "unequal pay." Other critics suggested that a monkey perceiving inequity in the distribution of food would want more cucumber slices, not less, to compensate for the inequity.

We might also ask whether these monkeys were really showing a sense of fairness. Even if we are persuaded that the monkey receiving the less favored food was showing indignation in protesting the inequity from the other monkey getting the more favored food, there is no evidence that the monkey getting the unfair advantage felt any guilt. If the monkeys receiving grapes were to throw away their grapes to show sympathy for those receiving only cucumbers, that would be a far more impressive display of a sense of justice.

A few years later, de Waal and Brosnan tried to answer some of these criticisms with new experiments using chimpanzees. They reported that chimps also showed inequity aversion, although chimps in close social relationships were more tolerant of inequity. And yet, once again, Tomasello and other critics pointed out weaknesses in this report. For example, of the 20 individuals studied, 14 refused inequitable exchanges in less than 2% of the trials, which is not very impressive.

Tomasello and his colleagues did a study of their own with 7 orangutans, 6 gorillas, 4 bonobos, and 13 chimpanzees in which the behavior seemed to show that there was no inequity aversion, and thus contradicting de Waal and Brosnan's research.

Some of the details in this debate are surveyed in an article by Kenneth Krause for eSkeptic.

As compared with what de Waal does, Tomasello's research is more interesting in that he performs similar experiments with chimpanzees and young children to see how and at what age the children surpass the chimpanzees. This allows him to argue that the early development of human children replicates human evolutionary history: we see the children starting out with chimpanzee-like abilities but then quickly moving to the uniquely human capabilities that arose early in human evolution. In this way, ontogeny might recapitulate phylogeny.

Some good videos of this research with both chimpanzees and children can be found here. There is also a good PBS Nova video on this and related research by Tomasello and others.

Tomasello's work is embedded within a general social theory of cooperation, which is well summarized in his book Why We Cooperate (MIT Press, 2009). One can see the influence on Tomasello of John Searle's The Construction of Social Reality (Free Press, 1995). Searle argues that we need to understand how social reality differs from physical reality, because social reality is a construction of the mind through the "collective intentionality" of "we consciousness" as opposed to the "I intentionality" of "I consciousness." In social life, we collaborate with one another as we create institutional practices in acting intentionally for shared goals. We thus create "institutional facts" that are just as real as "brute facts" or physical facts, although these institutional realities depend on the work of our minds, and they are not directly observable the way physical reality is directly observable. Tomasello picks up this idea in arguing that human beings are unique in their capacity for "shared intentionality," which allows human beings through their symbolic capacities to engage in cultural niche construction in ways that far surpass chimpanzees and other primates.

What Tomasello is doing is that he's exploring in experimental ways the evolutionary basis for what Aristotle and Aquinas saw--that we are similar to other social and political animals, and yet we are unique in our capacity to use our conceptual and linguistic abilities to create social worlds of shared intentionality that go beyond anything seen in the rest of the animal world.

Tomasello's comparative studies of chimpanzees and children follows in the tradition of Darwin, who methodically studied one of his infant children and compared the child with monkeys and apes to see when the child showed the moral sense that Darwin thought was uniquely human.

If you really want to probe into de Waal's mind, you'll have to go to the video of his interview on "The Colbert Report."

Some posts on related topics can be found here, here, here, here, and here.

REFERENCES
Brauer, Juliane, Josep Call, and Michael Tomasello, "Are Apes Really Inequity Averse?" Proceedings of the Royal Society B 273 (2006): 3123-3128.

Brosnan, Sarah F., and Frans de Waal, "Monkeys Reject Unequal Pay," Nature 425 (18 September 2003): 297-99.

Brosnan, Sarah F., Hillary C. Schiff, and Frans de Waal, "Tolerance for Inequity May Increase with Social Closeness in Chimpanzees," Proceedings of the Royal Society B, 272 (2005): 253-58.

Hamann, Katharina, Felix Warneken, Julia Greenberg, and Michael Tomasello, "Collaboration Encourages Equal Sharing in Children But Not in Chimpanzees," Nature, early online publication, July, 2011.

Horner, Victoria, J. Devyn Carter, Malini Suchak, and Frans de Waal, "Spontaneous Prosocial Choice by Chimpanzees," Proceedings of the National Academy of Sciences, early edition, August, 2011.

No comments: