Primates who think or ape?

In this book, each of the various contributors' attempts to arrive at some means of characterising, classifying, describing, or theorising about the form of thought in apes and monkeys. Part one of the book deals with the scope of primate intelligence and part two with its organisation, development and evolution. Topics that are deemed relevant to the understanding of nonhuman thought include tool use, imitation, teaching of others, imputing of mental states to others, reconciliation, deception, pretending, empathy, reciprocal altruism, self-awareness, culture, social attention, social referencing, joint attention, numerical skills, intentional communication, social learning, cooperation, role playing, drawing, rehearsal, transfer of learned behaviours to new contexts, pointing, eye contact, food-processing strategies and object manipulation with set construction and alteration, and of course, evolution.

Notably absent from this extensive list are studies of consciousness, language skills or vocal communication and social structure. Since language, consciousness and sociopolitical groupings are the major factors of human societies which seemingly set us apart from other animals, their absence in a book on nonhuman thought is remarkable. The exclusion of such topics leaves the unknowledgeable reader wondering why nonhuman primates seem so intelligent, while seemingly lacking the critical capacities that would enable them to construct rich and communicative mental and cognitive worlds. The overall impression one draws from this book is that yes, apes surely must think, but about what? There is little hope that investigations limited to the topics above will provide answers.

The few suggestions on the mental life of apes come from a short chapter by Daniel Hart and Mary Pat Karmel, who suggest that episodic memory (memory of specific events in one's past) is absent from monkeys, but that a gorilla is able to remember its capture as an infant. Most humans cannot recall events before one year of age, yet we have little difficulty remembering a variety of specific events throughout our lives. How can it be that monkeys remember nothing and that gorillas recall things that even we cannot remember?

For the reader who is prepared to set aside any interest in the mental life of apes and monkeys, the book provides a wealth of information on the topics noted above. Although Josep Call and Michael Tomasello express some scepticism, the general conclusion, strongly supported from a variety of different perspectives by the remaining authors, is that apes are capable, in some form or fashion, of all the different capacities listed above.

One of the most interesting chapters in the book is provided by Anne Russon, the first editor. Russon observed semi-captive orangs at the rehabilitation center Camp Leakey in Borneo. The behaviour that orangs imitated is impressive and attests both to the intelligence of the orangs themselves and to the functional quality of actions that exist in such a field laboratory. Russon's observations are nicely bolstered by those of Lynn Miles, with Chantek, a captive-reared orangutan who was taught sign language. After reading these two chapters, it is difficult to see how anyone could conclude anything other than that imitative learning is well within the capacity of the orang. Russon's description of an orang stealing a canoe, dumping out the extra water so that it would float, and using it to sneak past a guard so that she could get to a group of people washing their clothes and then steal the soap and the clothes and begin washing them herself is compelling.

Call and Tomasello suggest that it is human rearing that leads to a process they term enculturation and makes it possible for apes to differentiate between means and ends in the behaviour of others and to view these others as intentional agents. Yet this view is strongly called into question by the work of Frans De Waal, who provides the only chapter that deals with the social intelligence of apes and monkeys. De Waal's excellent and long-term studies of a variety of social phenomena in macaques and chimpanzees provides some of the clearest evidence of all that human rearing is not needed for the chimpanzee to attribute intentions to others. In his intriguing analysis of postconflict resolutions, De Waal clearly illustrates the need for reconciliation with an aggressor and for comforting the victims of aggression. It is difficult to explain such actions if one thinks that the actors have no communicative intent and are unable to view others as intentional agents. De Waal observes that both macaques and chimpanzees follow what seems a general rule among primates: reconciliation aims to restore valuable relationships. To have such a rule it is necessary to understand what a valuable relationship is, how it differs from a nonvaluable relationship and the kinds of things that are needed to maintain such a relationship.

Although little information is provided about social rearing and how group norms are imposed or transferred from adults to youngsters in this volume, Tetsuro Matsuzawas's chapter leaves no doubt that chimpanzees have distinct cultural traditions and that they transmit these in a cross-generational manner. While simple by human standards, the material cultures of chimpanzees nonetheless show great variation.

A recurrent question that must emerge in any book looking at thought in nonhuman primates is why it is that chimpanzees seem to be such proficient tool users in the wild, while monkeys and other apes seem not to do so, or to do so only sporadically. A cogent analytic comparison of tool use in capuchin monkeys and apes is presented by Visalberghi. Unlike others who have looked closely at this mysterious South American monkey, which shows such a penchant for object manipulation, Elisabetta Visalberghi does not just attempt to detail what it is that capuchins can do. Instead she asks what it is that they comprehend about their tool manipulations. In a series of clever studies, Visalberghi reveals that capuchins lack an understanding of the cause-effect relationships between what they are doing and how it works. Just as you and I may not understand why an "a" appears when we punch an "a" on our computer keyboard, capuchin monkeys seem not to understand that the reason a stick is used to push an object out of tube is the principle of physical displacement. They know that using a stick in a tube gets them food, but do not understand why this occurs.

Visalberghi's experiments in this field are supported by the data of Jonas Langer on first- and second-order cognitive operations in cebus and chimpanzees. Langer's work analyses the components of cognitive actions by looking at the types of mental sets that infants and/or nonhuman primates are able to form, as well as the ensuing operations that can occur on such sets once they are constructed. Langer's work reveals that cebus do not go beyond the construction of first-order operations, while chimpanzees are clearly capable of second-order operations. Human rearing can increase the capacity to construct second order operations with objects, but it is possible even for apes who are not human-reared. Second-order operations require the simultaneous existence of more than one set at a time. The fact that cebus monkeys neither recognise themselves in mirrors nor show simultaneous cognisance of what they are doing with a tool and what the tool itself is doing, illustrate the deficits that can result when two sets and the operations on and between them cannot be simultaneously present in the mind space at one time.

Langer's analysis also paves the way for a more sophisticated understanding of what have, for so long, proved to be anomalous results with regard to ape-monkey comparisons. Langer's work suggests that monkeys can learn causal effects, for example, the effect of a stick moving on a displaceable object, or the effect of their hand displacing a stick. However, it is difficult for both schemas to be simultaneously operable on the mental plane at one time. The monkey is thus forced to alternate between these two perspectives, never grasping what can be clearly seen if both were operable in the mind simultaneously. Apes, by contrast, can cognitively represent both schemas at once, thus permitting them to become efficient tool users as well as capable of mirror recognition.

Andrew Whiten's contribution raises the topic of mind reading, most notably with the introduction of the topic of deception. Deception would seem to require, at minimum, three perspectives: one's own, that of the recipient of the intended communication, and the deceptive communication that is presented. There is no doubt that apes can do this. De Waal cites examples of females pretending to intend to reconcile, only to attack another ape who is deceived into coming close enough to permit the surprise aggression. Miles cites a similar example for Chantek, who is able to deceive a human caretaker. Chantek's success in this endeavour suggests that human minds, may, in some cases, not be a full match for apes'.

Sarah Boysen attempted to get an experimental grasp on this phenomena by offering apes a chance to deceive another ape, but was not able to do so since the paradigm that she adopted required one ape to select which of two portions of food to give to another ape. The ape always chose the larger portion, seemingly unable to ignore the fact that the experimenter always gave the portion selected first to the other ape. Boysen found that apes could overcome the tendency to select the larger portion first if she replaced the food items with numerical symbols. Such symbols, it seemed, helped the apes to keep in mind the fact that the portion selected first was to go to another ape, while the portion selected next would go to them. Without the numbers, the apes focused on the largest portion, and continued to select it, even though it was given to the other ape. One would surmise that with repeated practice, apes would acquire the capacity to select the larger food portion for themselves, even though the numbers were absent. This view is supported by observations of Sherman and Austin, two language-trained apes from Georgia State University, which performed the original food-sharing tasks. Sherman always selected the smaller portion of food to give to Austin, with or without the use of symbols. Sherman was the larger and more dominant of the two individuals. Austin often, but not always, selected the larger portion to give to Sherman, Sherman monitored Austin's actions closely and on occasions when Austin tried to keep the larger portion for himself, Sherman intervened and switched portions.

In the final chapter, the editors suggest that all can be understood if the reader adopts a Piagetian point of view. The goal, it seems, is to take everything that nonhuman primates do and to classify it somewhere within Piaget's schema of infant development so that we will know where other nonhuman primates stand relative to ourselves. I find this discomforting. Piaget's schema was developed with, by, for and on humans. It cannot and does not take into account even simple physical differences in hands and feet, much less perceptual differences in sensory systems. It has often been argued, for example, that apes do not point unless they are raised by humans. But apes walk on their hands, and it is difficult to point while walking. Young apes, carried by humans, point often as their hands are not occupied. Older apes point far less. This is not a cognitive difference. Older apes point with their eyes, something we humans do not do as often, since our hands are free. Piaget's schema is species biased and is bound to lead to an underestimation of the cognitive skills of other animals. Nonetheless Reaching into Thought is a tour de force in the areas it purports to cover and will serve as a thorough graduate introductory text to these phenomena for some years to come. Hopefully, it will help us move beyond such questions as whether or not apes imitate, whether or not they have culture and whether or not they understand that the minds of others are not registering all events exactly as their own are doing. This book should leave little doubt that the answer to each of these questions is a resounding yes.

Sue Savage-Rumbaugh is professor of biology and psychology, Georgia State University, Atlanta, USA.