Introduction

When an animal is suddenly attacked by a predator, it must respond with great urgency if it is to escape. The neuronal circuits that initiate such an escape response must be both straightforward and reliable in order to fulfil their biological function. A staightforward circuit is essential to ensure speed in initiating the escape, and a reliable circuit is needed not only to make sure the response occurs when required but also to avoid false alarms. These qualities of simplicity and reliability, which are of great survival value to the animal, are also of service to the neuroethologist exploring the role that nerve cells play in behaviour. Consequently, several of these startle responses have been studied in detail and they provide valuable insight into the flow of information through the nervous system from sensory inputs to muscular output.

Furthermore, these neuronal circuits often involve neurons that are exceptionally large and, because of this, are called giant neurons. The function of giant neurons is to conduct spikes rapidly along the body, but their size also makes them readily accessible to study with microelectrodes. The giant neurons therefore offer a major opportunity to investigate the role of individual nerve cells in behaviour.

Two main functions must be carried out by the neuronal circuit that initiates any behaviour pattern, including escape. First of all, a decision to initiate an activity must be made at some point in the circuit.

Introduction

Understanding the mechanisms which generate and control locomotory movements is fundamental to a complete knowledge of the neuronal control of behaviour. We can regard locomotion, such as jumping, walking or flying, as basic building blocks for much of an animal's behavioural repertoire; and we can pose three basic questions about the control of such movements. First, what mechanisms ensure that muscles contract in the appropriate sequence? In walking, for example, the basic pattern is repeated flexion and then extension of each leg, with flexion of the left leg coinciding with extension of the right. Second, how does a nervous system select, initiate and terminate a particular type of movement? For example, what initiates the pattern of walking; and how is walking rather than running or swimming selected? Third, how is the basic pattern for movement modulated appropriately? Stride pattern changes, for example, when a person walks up a flight of steps or turns a corner.

Experimental approaches to these questions have often involved work on invertebrates and lower vertebrates, animals in which the parts of the nervous system that generate programs for movement contain a limited number of neurons. This offers the opportunity to identify and characterise all the components involved in generating a particular movement. A specific question that has occupied many investigators is how to determine the source of rhythmical activity that underlies many regularly repeated movements, such as walking or flying.

INTRODUCTION

Major developments have taken place in the field of neuroscience during the last two decades. We have seen an explosion in imaging techniques used in the study of the structure and function of the living human brain, a variety of improved staining and immunohistological techniques used in post-mortem material, different tracing methods, and stereological and quantitative tools among others. Most of these innovative techniques, however, have not yet been incorporated in the comparative study of the primate brain nor in questions concerned with the evolution of the human brain.

There is a large body of knowledge about the brains of several monkey species and in particular about that of the macaque, as well as steadily growing information about the human brain. Nevertheless, very little is known about the brains of the apes and in particular the brains of gorillas and orangutans. The statement made by Tuttle (1986), more than ten years ago is still very much correct:

The title The mentalities of gorillas and orangutans: Comparative perspectives was inspired by Köhler's famous book The mentality of apes (1927), in which apes were implicitly equated with chimpanzees. This book focuses on two other great apes that were less known in Köhler's day. It is the fourth in a related series of edited volumes on cognition of great apes. The first volume, “Language” and intelligence in monkeys and apes (Parker & Gibson, 1990), emphasized the importance of using models from developmental psychology to compare the cognitive and symbolic abilities of monkeys, apes, and humans. The second volume, Self-awareness in animals and humans (Parker, Mitchell, & Boccia, 1994), used developmental frameworks to compare manifestations of self-awareness in monkeys, apes, and humans. It was aimed at broadening the scope of research in this subject to extend beyond the classical mark test for mirror self-recognition. The third volume, Reaching into thought (Russon, Bard, and Parker, 1996), focused on cognitive abilities of great apes using models from developmental psychology, but extended the scope to include studies in wild populations. This most recent volume continues the tradition of using models from developmental psychology to compare the cognitive abilities of great apes, and that of including studies from both captive and wild populations. It also expands the preview to include studies of taxonomy and phylogeny (Begun, this volume) and the brain (Semendeferi, this volume). It differs from the other volumes in focusing primarily on gorillas and orangutans. The aim of this volume is to redress the chimpocentric imbalance in attention to chimpanzees and bonobos at the expense of the other great apes.

INTRODUCTION

Young apes have been preferred subjects in language studies for a number of reasons. As dependent creatures, young apes engage in social behaviors which provide the interactive context in which many signs are learned. Young apes appear to learn certain behaviors more rapidly than older primates (Kawai, 1965). Young apes are also easier to handle and manage than older apes, who test their strength and dominance against caretakers as they enter adolescence (personal experience). Older apes are also stronger and potentially more dangerous than younger apes. Programs requiring close interaction, such as those carried out by ape language researchers, are more readily implemented when the subjects are young apes.

For these reasons sign-learning abilities in adult apes have not been as actively examined. Nevertheless, sign-learning abilities of adult apes should be of interest to those who study great ape cognitive or linguistic abilities. Likewise, although sign learning has not been studied in free-ranging apes, their motivation and ability to learn signs is of great interest. This chapter reports Project Rinnie, a study of the early sign-learning performance of an adult orangutan free-ranging within Tanjung Puting National Park (then a nature reserve).

METHODS

Subject

Rinnie, a captive-born, free-ranging female orangutan was approximately 10 to 12 years old at the onset of the Project. She was released as a rehabilitant at the Orangutan Research and Conservation Program (Galdikas, 1979) where she established a home range in a freshwater swamp forest across the river from the Camp Leakey Research Station. Rinnie was the oldest and the most dominant among the several ex-captive orangutans.

Life-history strategy theory is the comparative study of the shape of the life cycle and the ontogeny of form and function as they relate to reproductive success in individuals and populations. According to this theory, life cycles have been shaped by selection in such a manner that life-time resources are allocated among the functions of growth, maintenance, defense, and reproduction in a manner that maximizes life-time reproductive potential (Pereira, 1993). Such design occurs within developmental and phylogenetic constraints, of course (Smith et al., 1985).

Life-history theory expands the focus of evolutionary biologists beyond morphology and behavior to include developmental and behavioral processes through the life cycle as parts of an integrated whole. Basic data for life-history studies include population data on the following features: body size, gestation length, age at weaning, birth interval, number of offspring per litter, offspring per year, neonatal body weight, adult body weight, age at onset of reproduction, lifespan, and mortality at various stages in the life cycle (Promislow & Harvey, 1990; Stearns, 1992).

Some biologists have characterized two contrasting life-history strategies. K strategists are large, long-lived animals that mature slowly and produce few offspring in each reproductive effort, but invest heavily in each offspring. They generally are adapted to ecological niches that are fairly stable over their lifespans. (Most anthropolid primates are K strategists.) In contrast, r strategists are small, short-lived animals that mature rapidly and produce large numbers of offspring in one or a few reproductive efforts, investing little energy in their offspring. They generally are adapted to unstable ecological niches that are subject to booms and busts, and are able to respond to resource booms with rapid population growth (MacArthur & Wilson, 1967).

In recent years, claims about the acquisition of sign language in chimpanzees, gorillas, and orangutans have been shrouded in controversy. On one side of this controversy have been researchers who have reported impressive language abilities in these species (Fouts, 1973; Gardner & Gardner, 1974; Patterson, 1978, 1980; Miles, 1983, 1990; Gardner, Gardner, & Van Cantfort, 1989; Patterson & Cohn, 1990). On the other side have been investigators who have dismissed these claims as largely unfounded (Seidenberg & Petitto, 1979; Sebeok & Umiker-Sebeok, 1980; Sebeok & Rosenthal, 1981; Terrace, 1985). Throughout this controversy, great effort often has been expended trying to answer the question “Do apes have language?” as if this question could be resolved unequivocally “yes” or “no.” A much more fruitful approach would be to devote greater effort determining more precisely the particular language and communicative capabilities of these species (Greenfield & Savage-Rumbaugh, 1990; Miles, 1990, 1997). An important early step in this approach to the issue of language in nonhuman primates would be to focus on how sign-language learning apes and humans both resemble and differ from each other (Rimpau, Gardner, & Gardner, 1989).

In most of the cross-species comparisons of language acquisition that have been conducted, the gestural or sign-language productions of apes were contrasted only with the spoken-language productions of young children. A major shortcoming with this approach is that it is not clear what conclusions can be drawn from such comparisons. Any differences that were observed could be attributed to differences between the species, to differences between the languages, or to both.

THE STUDY

This chapter reports on the development of social play during the first three years of life of an infant lowland gorilla living in a social group at the San Francisco Zoo (Parker, 1977a). It presents a hierarchical model for describing the temporal structure of play and its development in a single subject: at the highest level are the longest units, episodes of play. These are composed of shorter units, bouts of play. Bouts, in turn, are composed of games. Games are composed of the shortest, lowest-level units, schemes. Play bouts and episodes are punctuated by play regulators including enticements, terminations, and interventions. It also focuses on the relationship between social play and cognitive development.

Subjects

The subject of this study is an infant lowland gorilla (Gorilla gorilla gorilla) born and reared in a social group at the San Francisco Zoo: Mkumbwa was born May 1, 1975 to a wild-born mother, Jacqueline, and a wild-born father, Bwana. He is the younger brother of Koko, a gorilla who was taught to sign by Francine Patterson. At the time of the study, the social group was also composed of Mrs, a wild-born female, her male infant, Sunshine, also fathered by Bwana, and a nulliparous captive-born female, Pogo (referred to as Mkumbwa's “aunt”). Mkumbwa is currently the silverback male of the group which still includes Bwana, now a grandfather, and Pogo, as well as two younger females, Zura and Bawang. The group also includes Bawang's infant sons, Shango and Barney, fathered by Mkumbwa.

INTRODUCTION

Chimpanzees have long been famous for their tool use in the wild (Goodall, 1964), as well as in captivity (Köhler, 1927). Orangutans are now known to use tools in the wild (Van Schaik, Fox & Sitompul, 1996) as well as in captivity (Lethmate, 1982). Even bonobos, who are not known to use tools in the wild (Kano, 1986), are widely known for their tool use in captivity (Savage-Rumbaugh & Lewin, 1994). Of the great apes, only gorillas have been described as rarely using tools in captivity (McGrew, 1992), as well as never using tools in the wild (Schaller, 1963). In his review of tool use in great apes, McGrew (1992) cites only one report of tool use – use of a stick to rake in food – in captive gorillas. The idea that gorillas lack problem-solving capacities traces back to Yerkes' study of the gorilla Congo (Yerkes, 1927). Recently, this idea was reinforced by a study of the development of tool use in an infant gorilla (Natale, 1989)

On the other hand, at least two factors suggest that gorillas should be capable of tool use. First, gorillas display 5th- and 6th-stage sensorimotor cognitive abilities similar to those of other great apes (Chevalier-Skolnikoff, 1977; 1983), and tool use – at least, purposeful use of a variety of tools to solve problems – is an index of 5th-stage sensorimotor intelligence similar to that of 2–year-old human children (Piaget, 1952). Second, orangutans, gorillas, chimpanzees, bonobos, and humans are closely related sister species that share a common ancestor roughly dating back 14 to 18 million years ago (see Begun, this volume).

This chapter is based on the premise that it is impossible to understand human evolution unless we study gorillas and orangutans as well as our closest living relatives, chimpanzees and bonobos. It is designed to redress the chimpocentric bias of primatologists (Beck, 1982) by reviewing recent data on the cognitive abilities of gorillas and orangutans, especially data presented in this volume. It is also designed to highlight aspects of gorilla and orangutan behavior and abilities heretofore neglected, misrepresented, or unknown. Finally, it is designed to review the similarities and differences among the four species of great apes using the cladistic method to show the evolutionary implications of these comparative data.

THE CLADISTIC APPROACH TO PHYLOGENETIC RECONSTRUCTION

The cladistic method for reconstructing the evolution of behavior involves comparing the taxonomic distribution of behaviors among all the species closely related to our target species. (In order to avoid circularity, it is important to use a cladogram previously constructed from molecular data.) Through this systematic process of comparison, it is possible to determine which characteristics are (1) shared with members of other taxonomic groups (shared characters), (2) shared only by a group of sister species (shared–derived characters), and (3) unique to a given species (derived characters). From this it is possible to identify the putative common ancestors in whom particular shared derived characteristics must have arisen.

These patterns can only be determined through comparisons among the ingroup species, and between those species and more distantly related outgroup species Brooks & Mclennan, 1991; Harvey & Pagel, 1991).

When zoo-living gorillas perform communicative physical motions that seem to resemble those of signing gorillas, what does it mean? What are the processes involved in developing these gestures? What communicative behavior is universal for the species and what is individually learned? Though some gestures are shared by all gorillas and others are unique to individuals, such a simple dichotomy does not really tell the whole story. We have found that, in the gorilla group at the San Francisco Zoo, “species typical” expressions such as slapping, clapping, pounding, and chest beating develop quite differently in each individual. Individuals also create gestures, some of which have not been described for other gorillas. Other gestures are shared by a few, but not all, individuals in a group. Gestural repertoires of different individuals at the same ages vary both in type and quantity. There is variation in individual gorillas' usage of gestures over time and in accord with changing social conditions. The purpose of the present research is to describe this variation, explore the diverse physical and functional properties of these gestures, and learn why and how they have developed in this particular captive group of gorillas.

Before beginning the observations at the San Francisco Zoo that are described here, the first author (JT) had worked for 8 years with the signing gorillas Koko and Michael of the Gorilla Foundation, developing a particular interest in untaught signs and “natural” gestures of sign language tutored gorillas (Patterson, Tanner, & Mayer, 1988; Patterson & Tanner, 1988).

NATURAL HISTORY

In his monograph on the mountain gorilla of the Virunga Volcanoes, Schaller describes how gorillas eat over thirty different foods (Schaller, 1963, pp. 156–165). The accounts are brief, and sometimes based on only a very few observations. Nevertheless, it is immediately clear to the reader that interestingly different techniques are involved with each plant.

Some, such as the vine Droquetia iners, are “merely pushed into the mouth.” Almost as little care is taken with small ferns Polypodium sp.: to eat them a gorilla “reaches below the branch and without looking grabs a handful of the hanging ferns which it pulls in. After severing and discarding the roots with one bite, it stuffs the greens into its mouth.” Biting off encasing material is a method used in several different ways to detach inedible, contaminated, or, occasionally edible parts. Bark from the tree fern Cyathea deckenii is bitten off and discarded in small piles, and only the tender inside is eaten; taproots of the herbs Cynoglossum amplifolium and C.geometricum are hauled out of the ground and their tough bark is bitten off before consumption; in contrast, the dry bark of Hagenia and Hypericum trees is bitten off, but consumed.

Other plants evidently present more challenge to preparation. Often, this is a consequence of the relatively greater hardness and toughness of the encasing material compared with the fragile but edible interior. Schaller noted that feeding remains of bamboo Arundinaria alpine suggest that the tough and hairy outer layers are “peeled back to expose the pith, much as a human prepares a banana.”

In his books Animal Intelligence and Mental Evolution in Animals, Romanes (1882/1906; 1884/1900) argued that apes, being phylogenetically close to humans, should be capable of intentional deception. Until recently, scientists have had little more than Romanes' faith and a few observations upon which to base an expectation of skill in intentional deception in apes (Mitchell, 1999). But in the past two decades, descriptions of deceptive behavior in apes and other nonhuman primates, as well as discussions of its psychological requisites, have become commonplace, lending greater support to the expectations of intentional deception by apes and of commonalities in the underlying mental states in humans and apes (see, e.g., Menzel, 1974; De Waal, 1982, 1986; Miles, 1986; Mitchell, 1986, 1993, 1997; Mitchell & Thompson, 1986a,c; Smith, 1987; Whiten & Byrne, 1988; Byrne & Whiten, 1990, 1992; Byrne, 1997). As might be expected on phylogenetic grounds, many deceptions used by humans in sports, play, and teasing (e.g., Mawby & Mitchell, 1986; Mitchell & Thompson, 1986b, 1993; LaFrenière, 1988; Leekam, 1991; Reddy, 1991; Mitchell, 1996) are strikingly similar to those used by nonhuman primates. Just as the signing skills of great apes are comparable to those of 2- to 3-year-old children (Brown, 1970; Savage-Rumbaugh et al., 1993; Miles, 1997; this volume; Parker & McKinney, in press), their deceptions are also sometimes similarly comparable (e.g., Reddy, 1991), but often great apes' deceptions seem much more skillful than those of even 4-year-old children (e.g., LaFrenière, 1988).