Psychological Bulletin 1993. Vol. 113, No.

2, 211-228

Copyright 1993 by the American Psychological Association, Inc. 0033-2909/93/S3.00

Comparative Cognition: Beginning the Second Century of the Study of Animal Intelligence
E. A. Wasserman
Comparative psychology has undergone many changes since its inception in Victorian England some 100 years ago. Gone are the amusing anecdotes of pet owners and amateur naturalists, replaced by the detailed observations of behavioral scientists made under carefully controlled conditions. Yet, many of the persistent problems in the comparative analysis of intelligence remain: Are the cognitive processes of animals like those of humans? Can researchers construct a phytogeny of intelligence? What is cognition without language? This article briefly reviews the history of the study of comparative cognition. It then discusses 2 of the most active and important areas of empirical inquirymemory and conceptualizationto acquaint readers with contemporary research in the field. Given increased contact with the related areas of cognitive science, behavioral neuroscience, and behavioral ecology, comparative cognition should continue in its 2nd century to make significant contributions to the overall understanding of the principles of behavior.

How familiar a scene it is. A learned and eloquent scientist gazes upward toward the distant heavens, sighs wistfully, and plaintively asks "Is there life beyond earth? Is that life intelligent? If so, can we communicate with it?" What a noble quest: to search out alien life and to interact with it. A quest not only noble but also generously funded by governmental agencies. Radio telescopes, orbiting satellites, and both staffed and unstaffed cosmic expeditions have been and will be constructed and undertaken to find extraterrestrial lifeat great public expense and with much media fanfare. I wish not to demean these important efforts but rather to suggest that an equally challenging and probably more fruitful search for alien intelligence can take placeright here on earth. Thousands of species of animals inhabit this planet, many of them exhibiting notable craft, flexibility, and ingenuity in adapting to the numerous challenges of survival. The careful and scientific study of our planetmates may shed considerable light on the nature of intelligence, on the character of cognition without language, and on the very possibility of communication with extraterrestrial lifeif it is ever found. Perhaps most important, comparing the intelligence of many species of animals may help us know better what it means to be human. Although the study of animal intelligence has been an ongoing concern of scientists for some 100 yearsRomanes's (1883/ 1977) classic book, Animal Intelligence, was published in 1883 most people know very little about it. The present popularity of many nature programs on television plus the great publicity that several research projects on animal behavior have recently received suggest that now might be an opportune time to outI thank G. Burghardt, H. Davis, M. Rilling, D. Rumbaugh, N. Spear, and L. Van Hamme for their editorial and technical help in preparing this article. Correspondence concerning this article should be addressed to E. A. Wasserman, Department of Psychology, University of Iowa, Iowa City, Iowa 52242.
211

line for both specialists and nonspecialists the essentials of the scientific field known as comparative cognition. In addition, recent developments in human cognitive psychology have again made the study of animal intelligence central to a traditional goal of psychology: namely, "to view complex human abilities as emerging from configurations of elementary associative processes that could be studied in simple organisms" (Gluck & Bower, 1988, p. 227). This reassessment stands in stark contrast to earlier dismissals of animal behavior and cognition as irrelevant to the human species. By way of illustration, one trio of textbook authors justified their near-total disregard of nonhuman intelligence thusly:
Whenever higher mental processes are involved, we heartily disagree that human and animal behavior are necessarily governed by the same principles. We regard the human as a specialized product of evolution, as an animal whose cognition is also specialized. This means that humans and animals may share some cognitive abilities, but it is not a foregone conclusion that they do. (Lachman, Lachman, & Butterfield, 1979, p. 42)

Although it would be foolhardy to conclude on the basis of existing evidence that there are no substantial differences in human and animal cognition, it is far better to examine the matter with an open mind than from the prejudiced perspective of anthropocentrism. Thus, the present article explores the question of animal intelligence from, I hope, an objective and comparative vantage point. In pursuit of this overall objective, the present article first sketches a bit of the interesting history of comparative cognition and then introduces two of its most important and active research areas: memory and conceptualization. Additional discussion concerns numerical competence and language behavior and the place of comparative cognition within the biology of behavior. Readers should be warned that in a review such as this, it is impossible to do full justice to all areas of relevant research or to give full credit to all past and present workers in the field.

212

E. A. WASSERMAN The cat, however, with an almost incredible amount of forethought, was observed by myself, together with two other members of the household, to scatter crumbs on the grass with the obvious intention of enticing the birds, (p. 418)

Although I make no claims to comprehensiveness and thoroughness, I do believe that the present article effectively introduces the field of comparative cognition to newcomers and usefully reviews and discusses key topics and theories for individuals who already have some knowledge of the field. Readers should be further warned that clear and detailed definitions of many important termsindeed that of intelligence itselfare not given in this article; nor are they given in most others on comparative cognition. It may be far simpler to provide precise operational definitions of elements or aspects of intelligence than to give a good definition of the overarching conceptfor animals or people (for more on the definition of human intelligence, see Sternberg, 1986; for more on broader conceptions of human and animal intelligence, see Stenhouse, 1973). Historical Perspective Evolutionary Theory In the middle of the 19th century, in England, Herbert Spencer and Charles Darwin proposed that the minds as well as the bodies of animals had undergone a process of organic evolution. Living animal species can thus be comparedbehaviorally and anatomicallywith the resulting similarities and differences providing important clues to the ancestry of behavioral and anatomical traits. Since then, considerable progress has been made in elucidating the structural evolution of animals. Much less progress has been recorded in understanding the evolution of their intelligence. Beyond the obvious problem that most behavior does not fossilize, many actions that one characterizes as intelligent are not those that are common to all members of a species; instead, these intelligent actions are the often idiosyncratic responses that individual animals exhibit to specific environmental problems or circumstances. This fact means that proper documentation of animal intelligence requires not only careful recording of the relevant behavior but also precise knowledge and even control of the pertinent environmental variables. The latter point is of real historical significance, given the somewhat awkward start that Darwin and his student, George J. Romanes, gave to the field of comparative cognition.1 Methodology and Evidence Both Darwin and Romanes had a clear agenda. They believed that support for the theory of the evolution of intelligence could come by documenting a continuity of cognition among living animal species. Such documentation was initiated by collecting literally hundreds of tales of animal genius, as related by pet owners, naturalists, and zookeepers. Take, for instance, this tale of feline guile by one of Romanes's (1883/ 1977) friends:
Our servants have been accustomed during the late frost to throw the crumbs remaining from the breakfast-table to the birds, and I have several times noticed that our cat used to wait there in ambush in the expectation of obtaining a hearty meal from one or two of the assembled birds. Now, so far, this circumstance in itself is not an "example of abstract reasoning." But to continue. For the last few days this practice of feeding the birds has been left off.

As might be surmised from this vignette, the early anecdotists were not always impartial observers of behavior nor were they necessarily careful recorders of either the behavior in question or the conditions that promoted the behavior. As interesting and suggestive as these anecdotes were, they could not stand the tests of scientific scrutiny, as they were of dubious objectivity and replicability. The anecdotal method simply would not do to establish a science of comparative cognition, a point made forcefully by C. Lloyd Morgan (1894/1896). A countryman of Spencer, Darwin, and Romanes, Morgan (1894/1896) is often credited with stimulating the scientific study of animal intelligence by investigating the behaviors of newly hatched chicks reared without their mother. Morgan reasoned that with very young animals, he could more effectively limit and control the individual's prior experience; and with maternal isolation, he could remove the confounding influence of imitation on the behaviors he observed. Although Morgan's various research projects seem quaint and amateurish by today's standards (they were conducted in his own poultry yard), they set the stage for the more powerful and refined methods of Pavlov and Thorndike. As most students of behavior are aware, I. P. Pavlov, in Russia, and E. L. Thorndike, in the United States, developed highly reliable and objective methods for investigating associative conditioning in animals. It is a fair conclusion that most progress in the experimental investigation of comparative cognition has been the consequence of the creative application or modification of their two basic methodologies. Later, I review a sampling of that empirical evidence. Interpretation The Darwinian agenda not only sought to blur any sharp divisions between human and animal intelligence, but also it interpreted animal behavior in terms of human behavior and private experience. This anthropomorphic bias is implied in Darwin's (1871/1920) declaration of mental continuity between humans and animals:
The difference in mind between man and the higher animals, great as it is, certainly is one of degree and not of kind. We have seen that the senses and intuitions, the various emotions and faculties, such as love, memory, attention, curiosity, imitation, reason, etc., of which man boasts, may be found in an incipient, or even sometimes in a well-developed condition, in the lower animals. (p. 128)

The question of mental continuity was also of concern to Darwin's student, Romanes. He hypothesized that there might
' Most historians have been very harsh in their criticisms of Romanes's contributions to comparative psychology. However, more recent discussions of his ideas have cast the work of Romanes in a much more favorable light (see Rilling, 1992, for a discussion of Romanes's uncelebrated experimentation on counting in chimpanzees, and see Wasserman, 1984, for a more general consideration of Romanes's influence on comparative psychology).

COMPARATIVE COGNITION AND ANIMAL INTELLIGENCE

213

be a cognitive scale roughly paralleling the phylogenetic scale. Placement along this scale could be determined by objective behavioral tests: Can animals learn to behave in accord with specific situational demands deemed necessary to evidence some cognitive capability, such as memory, association by contiguity, reason, or communication? If so, then they must possess that mental capacity. Romanes (1884/1969) constructed an elaborate chart tracing the representation of numerous cognitive processes throughout the animal kingdom. As one progresses from more primitive to more advanced organisms, more advanced forms of cognition are assumed to be added to less advanced ones.2 Thus, although humans may sit at the summit of intelligence, humans may share with animals a number of cognitive capabilities, that number decreasing as evolutionary kinship becomes more remote. Romanes (1884/1969) also made an important effort to distinguish the objective and subjective aspects of animal behavior. As the following passage illustrates, Romanes saw that overt action and subjective experience cannot be equated. Whereas one might judge that the performance of similar responses by humans and animals implies the operation of common cognitive or emotional processes, any subjective experiences that may accompany those behaviors are inherently private; their existence can only be hypothesized by analogy with human experience.
Now, by mind we may mean two very different things, according as we contemplate it in our own individual selves, or in other organisms. For if we contemplate our own mind, we have an immediate cognizance of a certain flow of thoughts or feelings, which are the most ultimate things, and indeed the only things, of which we are cognizant. But if we contemplate mind in other organisms, we have no such immediate cognizance of thoughts or feelings. In such cases we can only infer the existence and the nature of thoughts and feelings from the activities of the organisms which appear to exhibit them. (Romanes, 1883/1977, p. 1)

jectively. No statement concerning consciousness in animals is open to refutation by observation and experiment, (pp. 335-336)

His more celebrated contemporary, J. B. Watson (1913), even more vigorously asserted that "I have virtually denied that this realm of psychics is open to experimental investigation" (p. 175). In addition, the late B. F. Skinner (1977) commented on the regressive and unproductive character of mentalistic analysis, according to which
a science of animal behavior must be replaced or supplemented by a science of animal feelings. It would be as extensive as the science of behavior because there would presumably be a feeling for each act. But feelingsare harder to identify and describe than the behavior attributed to them, and we should have abandoned an objective matter in favor of one of dubious status, accessible only through necessarily defective channels of introspection, (p. 3)

From its earliest years, comparative psychologists have found it extremely tantalizing to speculate on the nature of conscious experience in animals (for a review, see Burghardt, 1985). No less than Morgan (1894/1896) championed the use of introspection in the interpretation of animal behavior:
The wise and cautious student never forgets that the interpretation of the facts in psychical terms is based upon the inductions he has reached through introspection. The facts are objective phenomena; the interpretation is in terms of subjective experience, (p. 47)

The present state of research and theory in comparative cognition is strongly rooted in the experimental investigation of precisely controlled and recorded animal behavior. Interpretations usually center on the mechanisms and processes of cognition rather than on the nature or contents of subjective experience. In addition, considerable attention is paid to the biological substrates of cognition, with due regard for the fact that an understanding of the biology of intelligent behavior is very much a long-term goal. Considered next are two key areas of research in comparative cognitionmemory and conceptualizationso that one may gain a clearer appreciation of the progress that has been made in evaluating the thesis of mental continuity between humans and animals. It is made evident that although some of the research is truly comparative, other work seeks primarily to expand the known cognitive competencies of animals. Learning the limits of animal cognition is one of the most active avenues of contemporary investigation (Wasserman, 1981). Fully documenting the known competencies of animal cognition allows researchers to better ascertain whether similar behavioral results in humans may be due to the operation of the same biological processes or mechanisms.

Animal Memory
"The most fundamental principle of mental operation is that of memory, for this is the conditio sine qua non of all mental life" (Romanes, 1884/1969, p. 35). These lines by Romanes express in the strongest terms the centrality of memory to an understanding of cognition in behavior. Even a cursory survey of the most prestigious periodical in human cognitionthe Journal of Experimental Psychology: Learning, Memory, and Cognitionwill reveal the continued prominence of the study of memory in humans since the early work of Ebbinghaus. Yet, despite Romanes's early pronouncement and the vast body of research on human memory, the systematic study of memory has not historically been a prime focus of workers investigating the behavior of animals under controlled laboratory conditions.
Incidentally, Romanes also assumed that the same accretion of cognitive capabilities occurs as a human develops from birth to adulthood.
2

Even today, we find Griffin (1976,1978) insisting that subjective animal experience falls within the proper province of the field of comparative cognition (what he calls "cognitive ethology"see upcoming discussion). However tempting it may be to project onto animals one's own thoughts and feelings, most workers in the field of comparative psychology have deliberately resisted that temptation. One of the earliest behavioristic students of animal behavior, H. S. Jennings (1904/1976), recognized the unverifiability of consciousness as an accompaniment to action:
It is clear that objective evidence cannot give a demonstration either of the existence or of the non-existence of consciousness, for consciousness is precisely that which cannot be perceived ob-

214

E. A. WASSERMAN

Several factors probably contributed to the lack of experimental attention paid to animal memory. First, initial interest centered on associative learning as a result of the pioneering efforts of Pavlov and Thorndike. To Watson (1913) and the other early behaviorists, little else seemed to be necessary to understand the intelligence of organisms than their learning of stimulusresponse associations. Second, extreme conservatism governed the analytical efforts of the early behaviorists. Morgan (1894/ 1896) had persuasively argued against the interpretive excesses of attributing gratuitous cognitive processes to animals when simpler processes seemed to do, and most later researchers abided by his famous "canon of parsimony." Finally, whereas the basic procedures of Pavlov and Thorndike served the early investigators of animal learning exceedingly well, the more elaborate techniques necessary for studying memory were not nearly as well developed. W. S. Hunter (1913) did devise the delayed-response paradigm, in which, after a delay, an animal might identify which of several potential food sites it earlier had seen baited. Yet, successful performance in this task might not be the result of some enduring cognitive or neural process; it might as well have been due to the animal merely maintaining its bodily orientation to the baited site during the delay period (see Fletcher, 1965, for further discussion of such "behavioral mediation"). The past 2 decades have seen a veritable explosion of interest in the experimental investigation of memory in familiar laboratory animals like rats, pigeons, and monkeys. In part, this development is due to a realization that learning alone is incapable of explaining intelligent action; other cognitive processes like memory, attention, and conceptualization must also promote adaptive behavior in complex and changing environments. Also critical to the rise in the study of animal memory has been the emergence of new experimental techniques with sufficient power and reliability to disclose clearly the operation of memory and other cognitive processes. Let's now consider the contemporary study of animal memoryparticularly short-term memoryto gain a fuller understanding of the methods, results, and theories with which most workers have been concerned (see Spear, Miller, & Jagielo, 1990, for a broader review of the areas of animal memory and learning). Basic Methods and Findings Quite independently, J. Konorski (1959), in Poland, and D. S. Blough (1959), in the United States, reported two delayed discrimination methods that have come to be standard techniques in today's investigation of animal memory. In one version of Konorski's (1959) procedure (Wasserman, 1976), pigeons view two visual stimuli in succession on a nickel-sized pecking key. Each member of a pair of stimuli might be presented for 5 s, with a variable time interval between them. In the delayed matching-to-sample paradigm, food would follow the second (or test) stimulus if it was the same as the first (or sample) stimulus, but food would not follow the second stimulus if it was different from the first. By measuring the pigeon's rate of response to the test stimulus on trials with matching or nonmatching sample and test stimuli, one can assess the degree to which memory of the first stimulus controls

responding to the second. It so happens that pigeons peck the response key at high rates when food is likely but at low or zero rates when food is unlikely. With a short (1-s) interval between sample and test stimuli, the rate of response to the test stimulus on matching trials may exceed that to the test stimulus on nonmatching trials by a factor of 10 or more. However, as the sample-test delay is lengthened, the response rate differential declines in an extremely orderly way. This inverse relation between delay interval and differential test responding documents the forgetting of sample information. In Blough's (1959) procedure, two or more simultaneously presented test stimuli follow the sample stimulus.3 Only one of them is correct and leads to food if chosen; any other selection does not lead to food. In the delayed matching-to-sample paradigm, one of the test stimuli is the same as the sample and the remainder are different from the sample. With a short interval between the sample and test stimuli, pigeons showed a strong tendency to peck the correct (matching) test stimulus. However, as the sample-test delay is lengthened, choice accuracy declined, again documenting the forgetting of sample information. The essential difference between these methods is that with Konorski's (1959) procedure the animal must decide whether to respond to the test stimulus, whereas with Blough's (1959) procedure the animal must decide to which test stimulus to respond. This procedural difference notwithstanding, the two methods produce highly similar results. Thus, not only does each technique yield a decline in memory as the sample-test interval is lengthened under the delayed matching-to-sample paradigm, but also discriminative performance in each case is directly related to the duration of the sample stimulus and the time between trials (e.g., Nelson & Wasserman, 1978; Roberts & Grant, 1976). An important variant of each general method involves sample and test stimuli drawn from different item pools. Thus, animals might be shown different colors as sample stimuli and different forms as test stimuli. No physical matches are possible; only arbitrary or symbolic matches can hold. (In a particular version of Konorski's, 1959, procedure, then, food would only follow red-square and green-circle sequences but not redcircle and green-square sequences.) This so-called delayed "symbolic" matching-to-sample procedure has afforded researchers special opportunities to expand the investigation of animal memory. So, with true matching-to-sample procedures, it has been shown that pigeons can remember the color of a sample stimulus, its shape, its orientation, and its spatial location (for a review see Spear et al., 1990). However, with the symbolic matching-tosample procedure, it has also been possible to show that temporal aspects of a stimulus can be remembered. Thus, pigeons given Konorski's (1959) method remembered different durations of a red sample stimulus and reported that memory during test stimuli of differing line orientations (Wasserman, DeLong, & Larew, 1984). They remembered not only two very different stimulus durations (2 s and 16s), but also a whole range
3 Actually, his is a delayed version of Skinner's (1950) matching-tosample technique.

COMPARATIVE COGNITION AND ANIMAL INTELLIGENCE

215

of more or less different durations (1,2,3,4,5,6,7, and 8 s), and sample-test durations as long as 16 s. Beyond the attributes of single sample stimuli, pigeons given embellished versions of the symbolic matching-to-sample paradigm have also been shown to remember the temporal order (e.g., red-green) of a series of two different color sample stimuli (Weisman, Wasserman, Dodd, & Larew, 1980), the spatial order (e.g., left-right) of a series of two identically colored sample stimuli (Wasserman, Nelson, & Larew, 1980), and the relative duration (e.g., short-long) of a series of two different color sample stimuli (Dreyfus, Fetterman, Smith, & Stubbs, 1988). Furthermore, rats have successfully been trained to press one of two levers, depending on the number of prior auditory stimuli (two or four) in a simplified version of Blough's (1959) procedure (Fernandes & Church, 1982; also see Davis & Albert, 1986). (A later section of this article considers related aspects of numerical discrimination.) Yet another way in which the memory of complex information has been studied has involved sample stimuli comprising two or more manipulable attributes or elements (Riley & Roitblat, 1978). Pigeons were thus shown two-element samples composed of color (red or green) and line orientation (horizontal or vertical) elements. Using Blough's (1959) method, tests with just color comparisons or just line comparisons each yielded highly accurate performance, implying that both the color and the line orientation information of the compound samples were discriminated and remembered by the pigeons. Significantly, accuracy on these compound sample (color and line) trials was lower than on other trials involving only element samples (color or line), suggesting that the two sources of sample information on compound trials competed with one another for what in humans is called divided attention (for more on this "divided attention" notion see Brown & Morrison, 1990, and Riley & Brown, 1991).

Flexibility of Memory
Early in the modern era of memory research, it was proposed that memory might be explained by analogy with a simple electronic componentthe capacitor. In the same way that electrical charge grows in the capacitor when current is applied to it, the strength of a memory trace might grow, the longer the sample stimulus is observed; and in the same way that electrical charge dissipates from the capacitor when current is drawn from it, the strength of a memory trace might fade, the longer it has been since the sample stimulus has been observed. Although the initial evidence supported this "trace" theory of animal memory (Roberts & Grant, 1976; also see Guttenberger & Wasserman, 1985), more recent research indicates that animal memory is far more intricate and flexible than is implied by a capacitorlike trace. Cuing the delay interval. Suppose that, in the delayed matching-to-sample paradigm using either the Konorski (1959) or the Blough (1959) procedure, subjects are given additional information during the sample stimulus concerning the length of the upcoming delay interval (Wasserman, Grosch, & Nevin, 1982). With two delay intervals (e.g., 2 s and 8 s), two stimuli (e.g., a tone and white noise) can be used to cue those delays. When performance here is compared with a control condition in which the

added stimuli are completely uncorrelated with the upcoming delay intervals, the correlated condition enhances memory at the short interval but impairs memory at the long interval. Further evidence that delay interval signals affect memory performance comes from tests in which a short delay is cued but a long delay is given and in which a long delay is cued but a short delay is given. Cuing a short delay but giving a long one improved memory above that obtained at the correctly cued long delay; cuing a long delay but giving a short one impaired memory below that obtained at the correctly cued short delay. Just where in the sequence of events (sample stimulus, delay interval, test) the delay cue is exerting its effect and how it is doing so is not yet known, although one possibility is that differential attention is paid to the short- and long-cued samples (for more on this issue, see MacDonald & Grant, 1987; and Wasserman et al., 1982). What is known is that delay interval cuing similarly affects human memory performance. Humans, too, show a steeper memory functionwith better memory at short delay intervals but poorer memory at long delay intervals on a paired-associate immediate memory task when informative time tags are provided than when they are not (Hinrichs &Grunke, 1975). Cuing the test stimuli. Imagine that one were to combine true and symbolic matching-to-sample paradigms within Blough's (1959) general procedure. Here, each of the two sample colors would require choice of its same color stimulus when both colors were shown as the test stimuli; each of the two sample colors would also require choice of a specific line orientation when both line orientations were shown as the test stimuli. If color tests and line tests were equally likely and unpredictable, one might suspect that this task would pose no special challenge for the pigeon, as its memory for the color of the prior sample stimulus should afford it the means of choosing correctly on each type of test: color or line. Now, imagine that accompanying each sample stimulus is one of two forms that are perfectly correlated with the upcoming test dimension: A superimposed circle signals a color test and a superimposed triangle signals a line orientation test. Here, there could be a decided advantage to the bird's discriminating and using the form cues; advance warning of the test dimension might enable it to prepare for the upcoming test and to set itself to respond to just one particular stimulus. Without correlated test dimension cues, the animal would have to engage in more elaborate and possibly more difficult preparations and to set itself to respond to two particular stimuli depending on the nature of the test items. Note that test dimension cuing would have an advantageous effect only if there were an anticipatory or expectational aspect to animal memory. A decay or trace theory would lead one to predict that test dimension cuing would have no influence on memory. To see if cuing the test dimension has any effect on memory, Stonebraker and Rilling (1984) gave pigeons initial training under the previously mentioned conditions and then occasionally miscued the test dimension, giving a color test after the triangle and a line test after the circle. Memory on miscued trials was much poorer than was memory on correctly cued trials, thereby suggesting that subjects had come to anticipate particular test stimuli after particular prior cues. As in delay interval cuing, researchers have much more to learn about how

216

E. A. WASSERMAN

precuing affects memory (see Grant & MacDonald, 1990, and Santi, Musgrave, & Bradford, 1988, for more on this issue). Yet, these and other data indicate that anticipatory or expectational processes are importantly involved in animal memory (Honig & Dodd, 1986; Honig & Thompson, 1982; Wasserman, 1986) and human memory as well (Cohen, 1989). Directed forgetting. Within the domain of research on human memory, it is customary to posit the operation of control processes: means by which memories are modulated in accordance with the needs of the individual or the demands of the task (Atkinson & Shiffrin, 1968). By their very nature, control processes comprise many of those flexibilities and intricacies of memory that fall outside the scope of trace theories. One such key control process is rehearsal. It is often said that rehearsal is a covert activity that helps to sustain the memory of some prior event. Whereas engaging rehearsal should aid in retaining earlier information, terminating rehearsal should impair retention. To investigate the role of rehearsal in human memory, workers have developed the so-called directed-forgetting paradigm (Bjork, 1972). In one version of the paradigm, shortly after presentation of the trial stimulus, subjects are given one of two cues (e.g., red or green colors) either to remember or to forget the prior stimulus. Here, it is generally found that when memory tests are given, retention is far worse on forget-cue trials than on remember-cue trials, implying that the poststimulus cues were affecting the rehearsal process and thereby modulating memory. It has further been found that postponing the forget cue in a delay interval of fixed duration leads to a loss in its effectiveness; that is, memory for earlier information improves the later into the delay interval the forget cue is given (Weiner & Reed, 1969). This result suggests that spontaneous or uncued rehearsal before the forget cue protects the memory from the decremental effect of the forget cue. Beginning with a study by Maki and Hegvik (1980), several workers in the area of animal memory have endeavored to ascertain whether directed forgetting is uniquely human. Research with both pigeons (Grant, 1984; Maki, 1981; Rilling, Kendrick, & Stonebraker, 1984) and monkeys (Roberts, Mazmanian, & Kraemer, 1984) has adapted the delayed matchingto-sample paradigm to this end by adding brief postsample cues to signal that a test for sample memory either would or would not be given. As in the case of human memory, animal memory proved to be much lower on forget-cue trials than on remember-cue trials. In addition, memory was more markedly reduced if the postsample cue was presented early than if it was presented late in the delay interval (Grant, 1981; Stonebraker & Rilling, 1981).4 Serial position function. Further evidence inconsistent with trace theory is well established in the area of human verbal memory. There, it is frequently found that memory for a series of items is a "bowed" function of input position: memory being high for initial (primacy) and terminal (recency) items but low for items in intermediate input positions (e.g., Crowder, 1976). Whatever else the bowed serial position function means for theories of memoryprimacy perhaps reflecting long-term memory and recency reflecting short-term memoryit certainly violates the monotonically increasing function predicted by trace theory. Is this bowed serial position function demonstrable in non-

verbal animals? Yes, say Wright, Santiago, Sands, Kendrick, and Cook (1985). They trained both pigeons and monkeys on a serial-probe-recognition task. Lists of color slides were projected one at a time on the upper of two rectangular screens. Each of the four list itemsall different from one another was shown for 1 s (monkeys) or 2 s (pigeons) with a 1-s interval between items. A probe item was projected on the lower screen some time after the fourth list item. If the probe item was a repeat of one of the list items ("same" trial), then a response to the right-hand manipulandum was correct and was reinforced with food or drink; if the probe item had not been presented in the prior four-item list ("different" trial), then a response to the left-hand manipulandum was correct and was reinforced. When 1 s to 2 s for pigeons or 1 s to 10 s for monkeys separated the final list item and the probe test, memory for List Items 1 and 4 exceeded memory for List Items 2 and 3, thus reproducing the classic bowed serial position memory function. At shorter list-test delays, the serial position function rose monotonically; at longer list-test delays, the serial position function fell monotonically. Not only did the primacy and recency portions of the serial position curve greatly depend on how long after the last list item the probe test was given, but also precisely the same changes in memory performance held for humans tested under similar circumstances.5 These and other results thus suggest that animal memory is far more complex and flexible than was once thought and that similar control processes may modulate both human and animal memory. Such empirical parallels are, of course, all the more significant given that the nonhuman animals in all of these investigations were nonverbal creatures.

Conceptual Behavior
Humans and other animals are constantly confronted with an extraordinarily complex array of external stimuli. Yet, sense is somehow made of these varied and varying stimuli. One way of reducing the demands on an organism's sensory and information-processing systems is for it to treat similar stimuli as members of a single class; by so doing, substantial cognitive economy can be achieved, thus freeing its adaptive machinery to deal with other competing exigencies of survival. In addition, categorical processing permits an organism to identify novel stimuli as members of a particular class and to generalize knowledge about that category to these new members. Thus, an organism need not be bound to respond to only those stimuli with which it has had prior experience, correspondingly enhancing its ability to cope with a continually changing world. Although theorists have often extolled these adaptive virtues of categorization and conceptualization, we remain far from understanding exactly how organisms process stimuli so as to
Further discussion of directed forgetting and alternative accounts of it can be found in Kendrick and Killing's (1986) book, in which they also consider a broad range of animal memory phenomena. 5 Further evidence on the serial position effect can be found in Buchanan, Gill, and Braggio (1981), in Reed, Chih-Ta, Aggleton, and Rawlins (1991), and in Kesner and Jackson-Smith (1992). The last article also considers the neural mechanisms of human and animal memory processes.
4

COMPARATIVE COGNITION AND ANIMAL INTELLIGENCE

217

partition the world into classes of related objects and events. Indeed, given early writings on the subject, we should wonder whether nonhuman animals are even capable of conceptual behavior. Nearly a century ago, Morgan (1894/1896) denied animals the ability to behave conceptually. To do so, he said, requires that
we neglect all that is variable and focus the attention on the uniform relation. [Then] we have reached a conception, and this conception is not concrete, particular, and individual, but abstract, general, and of universal application, (p. 263)

Morgan believed that only adult humans (not even children) are capable of conceptualization. Several recent lines of evidence are radically changing that initial opinion. Object Concepts One familiar instance of conceptual behavior involves the kind of open-ended categorization response one makes when one labels different natural (e.g., cat) and human-made (e.g., chair) objects with different nouns. Such verbal behaviors are occasioned by specific instances of wide variability and individuality. Indeed, accurate classification even extends to categorical exemplars never seen before. Is it at all possible for nonhuman animals lacking language to engage in this form of conceptual or classifactory behavior? To answer this question with regard to the familiar laboratory pigeon, a new technique was devised to train it concurrently to discriminate stimuli from several human language categories. The specific method (Bhatt, Wasserman, Reynolds, & Knauss, 1988, Experiment 1) was based on the technique parents often use to teach their children to label objects in a picture bookthe "name" game. When the page is turned, the child is asked to look at the object and then he or she is requested to name it. If he or she is correct, praise is the reward. If he or she is incorrect, the result is encouragement to try again. Finally, if self-correction fails, he or she is provided with the correct name. To implement this method with pigeons, a color snapshot was displayed on a 3-in. square frosted plastic screen, and the pigeon was required to peck a clear plastic key covering the screen 30 times. Completing this observing response requirement led to the illumination of four different color keys just beyond the corners of the viewing screen. A single choice response was then permitted. If it was to the correct key for reporting the stimulus on the viewing screen, all of the visual stimuli were turned off and the pigeon was fed mixed grain; if the response was to any of the three incorrect report keys, all report key lights were turned off and the trial was repeated. Only the first choice response of a trial was scored; correction trials were not considered in analyses of performance. In several studies, the 40 slides seen in each daily session depicted 10 different examples each of cats, flowers, cars, and chairs. The pictures contained one or more instances of the critical stimulus object; the objects were indoors or outdoors; near or far away; centered or off center; and in different colors, orientations, and backgrounds. In one representative experiment (Bhatt et al., 1988, Experiment IB), a group of four pigeons attained a mean level of discriminative performance of 76% correct during Days 26 to

30 of training, after beginning the investigation near the chance level of 25% correct. Also noteworthy were the results of 2 later days of test performance with the 40 original training slides and with 40 brand-new slides of cats, flowers, cars, and chairs. Mean accuracy to old slides was 81%, and to new slides it was 64%. Although test performance was highly discriminative to both sets of stimuli, accuracy was reliably higher to old than to new pictures, perhaps because the birds remembered some or all of the old slides. Finally, there was no evidence for any of the stimulus categories being harder or easier for the pigeons to discriminate (contrary to the suggestion made by Herrnstein, 1985, that pigeons cannot categorize human-made stimuli). Thus, pigeons are able concurrently to categorize stimuli from four classes of natural and artificial objects and also to extrapolate that categorization to completely novel test stimuli, albeit at a somewhat lower level of accuracy. Perhaps even more important were the results of a subsequent investigation (Bhatt et al., 1988, Experiment 3) in which a large pool of 2,000 unique snapshots (500 from each of four categories) were shown to pigeons on a one-time-only basis. Without the benefit of any stimulus repetition, the birds attained a mean accuracy level of 70% correct on Days 46 to 50. Either the pigeon has an undocumented ability to remember a rather large number of stimuli it has seen only once (cf. Vaughan & Greene, 1984, who showed that pigeons can remember up to 320 pictures seen at least 28 times each) or it can abstract some kind of generic or prototypical information from varied stimuli, as implied by Morgan's (1894/1896) earlier quotation and by several more recent models of conceptualization (Smith & Medin, 1981). It is surely no small matter to demonstrate that nonverbal animals like pigeons are so adept at categorizing snapshots of real objects (also see Herrnstein, 1985). Yet, one is bound to wonder just how similar this feat is to the conceptual behavior of humans. Here, two additional projects suggest that the similarity may be more than accidental. In one investigation (Wasserman, Kiedinger, & Bhatt, 1988, Experiment 2), two groups of four pigeons were trained to categorize the same set of 80 snapshots. The first (or true category) group had to peck one of four keys to report each of 20 stimuli from four human language categories: cats, flowers, cars, and chairs. The second (or pseudocategory) group had to classify the very same slides into random assortments, in which each of the four pseudocategories comprised equal numbers of cat, flower, car, and chair slides. Over Days 37 to 40 of training, pigeons on the true categorization task averaged 79% correct, whereas pigeons on the pseudocategorization task averaged only 44% correct (a small but reliable rise from 25%). Thus, learning proceeded far faster when the to-be-trained categories coincided with human language classes than when they did not. These and other results (Astley & Wasserman, 1992; Edwards & Honig, 1987; Herrnstein & de Villiers, 1980; Wasserman et al., 1988, Experiment 1) suggest that to pigeons, members of human language categories resemble one another more than they resemble members of other language categories. The pigeon's categorization behavior thus confirms the nonarbitrary nature of human language terms, at least for the object categories they have thus far been given. In another project (Bhatt, 1988), three groups of four pigeons

218

E. A. WASSERMAN

were trained to categorize photographic slides. The three groups of pigeons were given 48 daily training trials comprising 12 copies of 1 example from the categories cat, flower, car, and chair (Group 1); 3 copies of 4 examples from the same categories (Group 4); or 1 copy of 12 examples from the same categories (Group 12). The rate of learning to a criterion of 70% correct was an inverse function of the number of examples per category. Of additional importance were the results of a generalization test with 32 novel stimuli, 8 from each category. Here, accuracy was a direct function of the number of examples given during training. The mean percentages of correct choices on generalization test trials were 27% for Group 1,45% for Group 4, and 62% for Group 12. Thus, although increasing the difficulty of original learning, greater numbers of training examples per category enhance the accuracy of generalization performance, perhaps because of the increased likelihood that any given test stimulus will resemble one or more of the remembered training stimuli or because abstracted prototypes are more representative of the entire class the greater the number of exemplars seen (Smith & Medin, 1981). Not only are these data orderly, but they neatly correspond with a large body of research on categorization in humans (reviewed by Homa, Burrel, & Field, 1987) and with a recent report on two-category discrimination in pigeons with 5 vs. 35 examples of black-andwhite bird and mammal sketches per category (Cook, Wright, & Kendrick, 1990). Whether such interesting correspondences in conceptualization by humans and animals will continue to be found is surely to be explored in future research. Abstract Concepts One of the empirical hallmarks of conceptualization is the independence of discriminative responding on the specific details of the prevailing stimuli. Thus, it was imperative in work on object concepts to show that discriminative responding established to a set of training stimuli also extended to a set of untrained test stimuli. To have conceptualized chairs requires that new chairs occasion the same response as old ones. An even more abstract level of conceptualization may be achieved when organisms attach the terms same or different to a pair of simultaneously or successively presented stimuli. (With successively presented stimuli, the terms novel or familiar would similarly suggest an abstract level of conceptualization.) Here again, the critical test comes when new pairs of stimuli are given to see whether the organism appropriately labels untrained stimuli. Matching to sample. One setting in which acquisition of a same-different concept has been studied is the matching-tosample paradigm discussed earlier. The success of pigeons in their original mastery of this task might suggest that matching to sample would readily transfer to new stimuli. However, such a result has not been obtained (for reviews and critical analyses see DAmato, Salmon, Loukas, & Tomie, 1986; Edwards, Miller, & Zentall, 1985; D. Premack, 1978). Compared with pigeons, both new- and old-world monkeys more readily generalize their visual matching-to-sample performance to novel stimuli (e.g., DAmato & Salmon, 1984), as does the bottlenose dolphin in an auditory matching-to-sample task (Herman & Gordon, 1974). However, the clearest evidence of spontaneous transfer of matching-to-sample performance comes from chimpanzees.

Oden, Thompson, and Premack (1988) taught four infant chimpanzees to match to sample using a set of only two objects, a lock and a cup. In the simultaneous matching-to-sample procedure they used the chimpanzee was handed a sample object; it was then required to choose from the set of two test objects the one that was the same as the sample object on that trial. A correct choice resulted in social and gustatory reinforcement, whereas an incorrect choice did not. After reaching a criterion of 83% correct, the animals were given a series of tests with novel objects and fabrics. Test accuracy on these trials averaged 85%. Thus, the chimpanzees transferred their matching-tosample performance without decrement to brand-new stimuli, suggesting that they had strongly conceptualized the same-different relation. It is surely noteworthy that dolphins and primates more readily generalize their matching-to-sample behavior than do pigeons. However, does this mean that pigeons are completely unable to appreciate the abstract relation of sameness-difference? Other evidence suggests not. Paired comparison. Yet another procedure has been used to assess control over behavior by same and different stimuli. In the paired-comparison procedure, two stimuli are either simultaneously or successively exposed. Then, two response alternatives are afforded to subjects: one for reporting that the stimuli were the same and the other for reporting that they were different. Correct choices occasion reinforcement, whereas incorrect choices do not. In a project by Santiago and Wright (1984), pigeons were trained on a simultaneous visual paired-comparison procedure. During original training, 105 color slides of fruit, flowers, animals, people, and other natural and human-made objects were shown in pairs on a split screen. After making an observing response to a clear panel covering the split screen, two choice keys were lighted: the left for reporting that the two slides were different and the right for reporting that they were the same. Half of the trials involved same stimuli and half involved different stimuli. The large number of training stimuli that were used guaranteed that no stimulus occurred on more than one trial in a session, in the hope that the birds' behavior would more likely come under control of the same-different relation than under control of the specific features of the stimuli. After training to over 80% correct on the original set of slides, the pigeons were shown 105 brand-new slides. First-session transfer performance averaged 70% correct. This score was a bit lower than transfer performance to the training slides; but, it was much higher than the 50% score expected by chance, and it compared quite favorably with the 72% first-session transfer score of rhesus monkeys trained and tested under virtually identical circumstances (Wright, Santiago, & Sands, 1984). Familiarity-novelty discrimination. Further evidence that even pigeons can appreciate abstract stimulus relations comes from a recent report by Macphail and Reilly (1989). In this project, pigeons were shown a series of color slides depicting indoor scenes, outdoor scenes, objects, faces, and so on. Each slide was shown twice in each daily 48-trial session, and slides were never reused from one session to another. Pecks to the first presentation of a given slide were reinforced with food, whereas pecks to the second presentation were not. After only four sessions of discrimination training, pigeons pecked much

COMPARATIVE COGNITION AND ANIMAL INTELLIGENCE

219

more often on the first presentation of a slide than on its second presentation. Because of the continually changing composition of the slide arrays, these results suggest that the pigeons were readily able to discriminate familiar from novel stimuli, quite apart from the specific attributes of each stimulus display. Macphail and Reilly proposed that earlier difficulties in training highly general same-different or familiar-novel reports in pigeons may have been due to procedural factors rather than to any cognitive limitations of the species. Oddity discrimination. Whereas the familiarity-novelty discrimination of Macphail and Reilly (1989) necessitates successively presented stimuli, a related problem, also requiring abstract conceptualization, does not. In an oddity concept, choice is made of the odd stimulus in an array of three or more simultaneously presented novel stimuli. Lombardi, Fachinelli, and Delius (1984) first established an oddity discrimination in pigeons given problems arranged from a pool of either 5 or 20 white-onblack visual forms. After reaching nearly equivalent levels of performance (the group given fewer stimuli learning faster than the group given more stimuli; cf. Bhatt, 1988), all animals were tested with brand-new forms. Highly reliable transfer was observed, that transfer being better for pigeons given more training stimuli than those given fewer training stimuli (cf. Bhatt, 1988). Interestingly, an analogous project with rats given visual and olfactory oddity tasks failed to find evidence of an oddity concept (Thomas & Noble, 1988). Equivalence Class Concepts The quest for clear evidence of conceptual behavior has also brought about the development of new analytical ideas and experimental procedures. For instance, the object concepts discussed earlier could be and probably were based solely on physical similarity. Despite our present inability to isolate and manipulate the relevant physical features of such complex stimuli as snapshots of cats, flowers, cars, and chairs, theorists can easily appeal to the well-documented principle of primary stimulus generalization to account for transfer to novel instances from the training categories. Yet, other theorists (e.g., Lea, 1984) have insisted that true conceptual behavior must be based on something more than mere physical resemblance. Learning a response to some members of a heterogeneous set of stimuli should ideally "propagate to all members of the set, without regard to similarity" (Herrnstein, 1990, p. 150). This requirement reduces to the demonstration of transfer of control through secondary stimulus generalization (Hull, 1943), and sets of such physically dissimilar yet functionally substitutable stimuli are called equivalence classes (see Sidman, 1986, for one specific rendering of this general idea). Initial research by Sidman et al. (1982) using matching-tosample procedures suggested that although children could form equivalence classes, neither rhesus monkeys nor baboons could do so. More recent research using different procedures is more encouraging to the idea that animals other than humans can form equivalence classes (Vaughan, 1988; Wasserman, DeVolder, & Coppage, 1992; Zentall, Steirn, Sherburne, & Urcuioli, 1991). In Vaughan's (1988) experiment, for example, pigeons were reinforced with food for pecking at a set of 20 out

of 40 slides, all 40 of which depicted trees. Pigeons were shown the 40 slides in different random orders each session. The slides were divided into two arbitrary assortments: one positive (reinforcement in their presence, 1+) and the second negative (no reinforcement in their presence, 2). Different birds were given different 1+, 2- slide assortments to reduce the possibility that some unknown physical feature of the slides was associated with category membership. After the pigeons learned to discriminate the slides (responding to 1+ but not to 2- stimuli), the contingencies were reversed (1, 2+) and then reversed again, repeatedly. After dozens of reversals, the pigeons were able to discriminate the positive from the negative collections after presentation of only the first few stimuli in each. Evidently, the birds had formed a common stimulus class for each of the two slide collections; by sampling a few slides, the pigeons could tell which equivalence class had a positive or a negative valence in any particular session. In the familiar terminology of Rosch and Mervis (1975), it might be said that Vaughan's (1988) birds had learned a subordinate concept, in which particular instances of a basic-level concept, like trees, were segregated in subgroups. In higher level groupings, is there any evidence that animals can learn to collect stimuli from different basic-level concepts into superordinate concepts? Yes. Wasserman et al. (1992) used a three-step procedure with pigeons, in which collections of perceptually dissimilar stimuli, like chairs and cars, were first associated with a common response. Then, a new response was learned to just one of those classes of stimuli, like chairs. Finally, subjects were tested for their tendency to make the new response to the class of stimuli (cars) not given during the second step. Pigeons showed a strong propensity to make the new response to stimuli with which that response was never before associated. Thus, merely by being associated with a common response in the first step, classes of perceptually dissimilar stimuli appear to amalgamate into a new superordinate category of functionally equivalent stimuli. Although not the end of the story, it does appear that diverse concepts are learnable by common laboratory animals, like pigeons, with different experimental methods being more or less conducive to disclosing those conceptual abilities.

Other Active Research Areas

In the previous discussions of animal memory and conceptual behavior, I tried to provide readers with a broad overview of two of the most active and systematic areas of research in comparative cognition. Of course, there is no way to discuss the full range of research in the field in one review article. Interested readers should consult several edited volumes (e.g., Hulse, Fowler, & Honig, 1978; Kendrick, Rilling, & Denny, 1986; Roitblat, Bever, & Terrace, 1984; Weiskrantz, 1985) and textbooks (Pearce, 1987; Roitblat, 1987) for more comprehensive coverage of comparative cognition. Before leaving specific research domains within comparative cognition, however, two additional topics are considered, with a special eye toward matters of experimental strategy and evidential interpretation. These two areasnumerical competence and language behaviorhave fostered rather more de-

220

E. A. WASSERMAN

bate, but less systematic data, than the areas of animal memory and conceptual behavior.

Numerical Competence
Does number control the behavior of animals? If so, then does such discriminative control resemble counting and other mathematical behaviors in humans? Recent research suggests a clear yes to the first question and a guarded perhaps to the second (for a review of the literature and additional critical commentary, see Davis & Perusse, 1988). As is often the case in comparative research, matters of definition can be critical to deciding if the behaviors of different organisms are alike. So, the seemingly simple question "Do animals count?" requires operational definition of counting. Unfortunately, considerable debate exists concerning that very definition (see, for example, Davis & Perusse, 1988, and Gallistel, 1989, for different viewpoints). Rather than trying to solve this definitional dispute directly, some investigators have undertaken behavioral analyses to identify more basic cognitive skills that appear to be necessary for any organism to show counting or higher mathematical abilities. Washburn and Rumbaugh (1991) thus sought evidence on the more definite question "Can animals learn that different symbols are associated with different quantities of food?" An affirmative answer to this more limited query would indicate that the animal under study has a prerequisite skill to engage in more complicated mathematical performances, like counting and adding.6 In one set of experiments (Experiments 1, 2, and 3), Washburn and Rumbaugh (1991) gave two rhesus monkeys the choice of responding to two arabic numerals; whichever numeral was chosen determined the number of food pellets given. For example, in a 2-9 pairing, choosing the 2 resulted in receiving 2 pellets, whereas choosing the 9 resulted in receiving 9 pellets. The monkeys showed a clear tendency to choose the larger numeral, that tendency increasing as the difference between the numerals increased. Also, one of the monkeys showed clear generalization to pairs of numerals that had never before been presented during training. In a final experiment (Experiment 4), Washburn and Rumbaugh further examined the monkeys' symbolic capabilities by presenting five different numerals on each trial. Each selection delivered the appropriate number of food pellets and eliminated that numeral from the next choice opportunity. Here, both animals tended to choose the largest numeral available, with their performance varying somewhat as a function of both the number of numerals available and the differences among the numerals. As a sample of performance, one monkey first chose the 8, then the 6, then the 5, and finally the 4 when given successive choices of 8-6-5-4-1,6-5-4-1,5-4-1, and 4-1, respectively. These results led Washburn and Rumbaugh to conclude that "rhesus monkeys learned . . . that symbols (i.e., arabic numerals 0-9) were associated with different quantities of pellets and ordinally sequenced the symbols that indexed those quantities" (p. 192). A further claim by the authors implies that this ability may be limited to primates: "These rhesus monkeys displayed a proficiency at discriminating, representing, and ordering quantities beyond that yet demonstrated with any nonhuman species" (Washburn & Rumbaugh, 1991 p. 192). However, Bhatt and

Wasserman (1987) had earlier found that pigeons could learn to associate four different color keys with four different quantities of food reinforcement and that theirchoice of a reference schedule of food reinforcement was a clear function of the alternative quantity of reinforcement they were given, choice of the reference quantity declining as the alternative quantity was increased. In addition, Hulse and O'Leary (1982) had even earlier found that rats given access to a four-arm radial maze with different quantities of food at the ends of each arm learned to order their within-trial selections in accord with those quantities. After 5 weeks of training, at least 75% of the rats' choices were ordered 18-6-1-0 (0 = no choice to the 0-pellet arm). A wide range of animal species can thus associate different arbitrary stimuli (e.g., numerals, colors, or spatial locations) with different quantities of food. Furthermore, their choice among those arbitrary stimuli clearly preserves the ordinal relations those stimuli have with different food quantities. Now, researchers can logically pursue questions concerned with numerical operations, such as counting and adding, which might be performed on those symbols (see Boysen & Capaldi, 1992, for much more work on this question).

Language Behavior
The popular and scientific attention paid to recent research in language behavior in animals is unprecedented. Many of the apes, dolphins, and avians that have been studied in this work are as well or better known by name than their human investigators. In addition, the rivalry among those investigators has occasionally sparked rancorous debates, which some would claim have done rather little to advance the understanding of language behavior in nonhuman animals. As most readers are already aware, a focal question of this general line of work is "Is language a uniquely human phenomenon?" Some of the most exciting and controversial research addressing this question involves teaching animals various human-made communication systems, including vocal (Pepperberg, 1981), gestural (Gardner & Gardner, 1984; Herman, Morrel-Samuels, & Pack, 1990), token-based (A. J. Premack & Premack, 1972), and computer-based (Rumbaugh, 1977) schemes. No effort is made here to rule on whether the behavioral feats heretofore performed by the many animals in these diverse projects proves them to be capable of humanlike language. Rather, as in the case of numerical competence, the question is whether animals possess any of the component skills necessary for language behavior. This more limited approach yields some clear and unequivocal conclusions. Central to language is the idea of reference: "To the extent that the communication of information depends on the arbitrary pairing of terms with conceptual categories, then that biological function of a natural language depends on the rote learning of paired associates" (Gardner & Gardner, 1984, p. 401). Thus, in Project Washoe and in follow-up studies, the Gardners successfully trained chimpanzees to make one of
Such a "componential" analysis of counting can be seen to accord with the general approach of Gluck and Bower (1988) to complex cognitive performance mentioned in the introduction.
6

COMPARATIVE COGNITION AND ANIMAL INTELLIGENCE

221

many manual gestures in American sign language to refer to members of such human language categories as cats, flowers, balls, and shoes.
The chimpanzees transferred the signs they had learned for a few balls, shoes, flowers, or cats to the full range of the categories whenever they found them and however represented, as if they divided the world into conceptual categories just as humans do. (Gardner & Gardner, 1984, p. 400)

Thus, the chimpanzees had passed the definitional criterion of reference wherein the animal "has a lexicon of arbitrary signals that symbolically stand for objects, events, concepts, and features" (Roitblat, 1987, p. 278). Interestingly, the Gardners (Gardner & Gardner, 1984) entitled their article describing this work "A Vocabulary Test for Chimpanzees." Note that Bhatt et al. (1988) found essentially the same result with pigeons, albeit with only four motor behaviors being used for their reporting four categories of objects. Although use of the word vocabulary for either pigeons or chimpanzees may be controversial, the findings from each species clearly demonstrate the operation of reference in nonhuman behavior, at least for some kinds of conceptual categories.7 Also critical to language is the notion of grammar: "a finite list of rules that can be used to produce an infinitely large number of expressions" (Roitblat, 1987, p. 278). Allied to the notion of grammar is that of symbol sequence or syntax: "wherein the same symbols in different orders. . . can express different meanings" (Roitblat, 1987, p. 278). Many of the language-learning projects in animals have been criticized for failing to show grammatical competence. For instance, Terrace, Petitto, Sanders, and Bever (1979) concluded that although apes can learn many isolated symbols, these projects "yielded no evidence of an ape's ability to use a grammar" (p. 891). Lawful regularities in symbol order were observed in those projects; however, such ordered responding "can, in each case, be explained by reference to simpler nonlinguistic processes" (p. 900). It is no small matter to show that even pigeons can be taught to produce an ordered string of up to five different responses in the absence of external discriminative feedback (Terrace, 1991). Such sensitivity to response order is obviously a necessary skill for grammatical performance. However, it does not easily allow for the production of novel grammatical sequences entailing the same or new responses. The issue of symbol order has been also investigated in animal behavior with regard to the question of different orders conveying different meanings. Here, it has been found that different sequences of colors can effectively signal different contingencies of reinforcement to pigeons (e.g., Weisman et al., 1980, Experiment 3). Such early results suggest that there is good reason to believe that at least the comprehensive aspect of syntax is operative in animal behavior. No clear separation of the comprehensive and productive aspects of syntax has yet been achieved in monkeys (Devine, Burke, & Rohack, 1979) and pigeons (Parker, 1984), each having succeeded in reproducing two-item sequences previously shown to them.8 Obviously, there are many component cognitive skills that are concatenated in human language. Which individual and combined skills appear in the behavior of animals may provide important clues in evaluating the uniqueness of human lan-

guage. Real insights may still result from direct efforts to teach human communication systems to animals, or they may emerge from more oblique inquiries into the limits of animal cognition. In either event, comparative cognition should continue to contribute to researchers' understanding the biological origins of this most notable form of human behavior. The recent remarks of Gisiner and Schusterman (1992) provide a fitting finale for this discussion of language behavior: "It seems highly implausible that the linguistic abilities of humans have arisen in complete ontogenetic and phylogenetic isolation from nonlinguistic learning abilities" (p. 90). Overview After a long, fallow period, researchers are again exploring the cognitive processes of animals with renewed vigor. Much has been learned from the experimental investigation of animal behavior since Spencer and Darwin first ignited interest in animal intelligence and its relation to human cognition. Use of richer and more refined methods of inquiry are now beginning to pay real dividends in disclosing that such cognitive processes as memory, attention, and conceptualization importantly participate in animal behavior. In addition, as the previous review indicates, there are good reasons to believe that at least some cognitive processes may be common to animals and human beings (also see Wasserman, 1990). This rather optimistic evaluation notwithstanding, a few remarks are in order considering the fact that, just 20 years ago, many had proclaimed comparative psychology to be either terminally ill or dead (see Hodos & Campbell, 1969; Lockard, 1971; Tobach, Adler, & Adler, 1973, for presentation and discussion of this thesis). Thus, the concluding sections of this article raise and, I hope, answer many questions that have arisen concerning the place of comparative cognition in the biology of behavior. fs Comparative Cognition Comparative? Even a cursory examination of the research reviewed in this article will reveal that only a small fraction of it entails explicit comparisons of different species within the same study. How then can one claim that comparative cognition is truly comparative? The answer to this query is that most comparisons among species are conducted across different studies. With a few notable exceptions (see Bitterman, 1975), most researchers of cognition in animals have concentrated on one species. Such concen7 Others would also insist that reference is contingent on the effective communication between "speakers" about things that are not necessarily present; this additional requirement appears to have been passed by chimpanzees (Savage-Rumbaugh, Murphy, Sevcik, & Rumbaugh, in press). 8 Recent research by Greenfield and Savage-Rumbaugh (1990,1991) has reconsidered the issues of grammar and syntax by examining the self-generated sequences of lexigram use by a chimpanzee. They concluded that this ape had the potential to invent a rudimentary grammar. Gisiner and Schusterman (1992), studying a language-trained sea lion, have also made interesting observations on syntactical control. They concluded that this animal can learn a number of syntactic relations from a limited set of standard combinatorial sequences.

222

E. A. WASSERMAN The analysis is plausible because it places cognitive process and cognitive behavior within the framework of a functional and experimental analysis of behavior. . . . There is nothing magical or mysterious about the relevant experimental or criterion behaviors, and thus processes remain within the realm of behavioral identification and analysis. We do not need a new kind of psychology to deal with cognitive events, (p. 11)

tration enhances the chances of that research incisively and comprehensively elucidating the cognitive processes of that species. Of course, the risk of overspecialization is that rather much may be learned about rather few species. Beyond rats, pigeons, monkeys, and apes, researchers know rather little about cognition in nonhuman animals. The field welcomes the systematic study of underrepresented species. All of this discussion, of course, ignores the most salient of all comparisons, namely, the comparison of animal and human cognition. Here, the rich empirical literature on cognitive processes in humans affords countless comparative opportunities to investigators of animal behavior. Also, as the present review clearly discloses, researchers have frequently evaluated the performance of their animal subjects in the context of human cognition. This most natural of all comparisons can be expected to command a good measure of future research in comparative cognition. Is Comparative Cognition Cognitive? Today, the term cognitive psychology has come to denote a particular domain of human behavior and the term cognitivism a specific approach to investigating cognitive processes in human behavior. Most who advocate this cognitive approach have curiously found it easier to accept functional and structural parallels between human beings and digital computers than between human and nonhuman animals (Haugeland, 1978; Lachmanetal., 1979). Indeed, the information-processing analogy of humans to computers has spawned "the computational view of thought, which sees thinking as the manipulation of an internal representation ('mental model') of an external domain" (Hunt, 1989, p. 604). Most workers in comparative cognition do not ascribe to this computational view for fear of substituting what they see as a form of mental ism for behavioral analysis. Thus, as to the role of internal representations in behavior, Schnaitter (1987) follows Skinner's (1985) lead in observing that
there is no independent and direct way to determine how this internal environment works. . . . The behaviorally relevant aspects of the inner environment are neither independently assessable nor are they directly controllable. Consequently it is impossible to state any constraining generalizations about internal context, and without knowledge of constraining generalizations it is impossible to establish their relationship to performance generalizations. The whole project for an analysis of behavior fails, (p. 10)

To study cognitive processes in behavior, then, in no way forces one to follow the theoretical lead of cognitivists, although some researchers of animal behavior have explored the heuristic value of one of its most controversial notionsrepresentation (e.g., Roitblat, 1982). Is Comparative Cognition Relevant to Evolutionary Biology? Comparative psychologists have often been criticized for their choice of species to study. For instance, Ratner (1970) accused many of engaging in "capricious comparison," an idea more fully developed by Hodos and Campbell (1969):
Much of the current research in comparative psychology seems to be based on comparisons between animals that have been selected for study according to rather arbitrary considerations and appears to be without any goal other than the comparison of animals for the sake of comparison. This rather tenuous approach to research has apparently been brought about by the absence of any broad theoretical foundation for the field, (p. 337)

Further contributing to confusion over the nature of comparison in comparative psychology is the claim by King and Nichols (1960) that "the concept of evolution, which is basic to the zoological system of classification, is probably not the concept most useful for a classification of behavior" (p. 22). How can researchers inaugurate the field of comparative cognition with Darwin's theory of evolution and not adopt phylogenesis as the framework for classifying behavior and for selecting animals to study? One answer is given in a recent article by Gottlieb (1984; also see Campbell & Hodos, 1991; and Yarczower & Hazlett, 1977). After reviewing past and present practices in evolutionary biology, Gottlieb concluded that
the discernment of evolutionary trends has always gone on both within and without strict phylogenetic lineages since the inception of the theory of evolution in the early 19th century and . . . this practice continues with fruitful intellectual results to the present day. (p. 448)

Segal (1978) added the following:

There is a great danger of a return to the worst fallacies of mentalism and dualism in the current rise of cognitive concepts. Awarenessor a homunculus within the brain that controls information processingis not a necessary part of cognition, (p. 214)

For Gottlieb, it is not a matter of comparative psychology lacking a unifying theory, as some have claimed (Hodos & Campbell, 1969); rather, it is the nature of that theory.
There is a theory of comparative psychology and that theory is, and always has been, founded on a psychological concept of anagenesis; the progressive evolution of adaptive behavior, learning ability, or intelligence (pp. 448-449).

Such concerns with cognitivism have not, however, prevented investigators of animal behavior from researching many of the most complex and challenging aspects of cognition. Within a behavioristic framework (Wasserman, 1981, 1982, 1983), researchers have studied such benchmarks of cognition as memory, attention, conceptualization, and language. One of the initiators of modern research on comparative cognition, W K. Honig (1978), has assessed this approach thusly:

Many readers may be familiar with the evolutionary concept of cladogenesis, the splitting of organisms into distinct and reproductively isolated populations over time. Whereas some evolutionary psychologists have insisted that only the latter is the appropriate subject matter of comparative psychology, Gottlieb (1984) argued that the former is also a legitimate concern for comparative psychology.

COMPARATIVE COGNITION AND ANIMAL INTELLIGENCE What distinguishes the two classificationsand causes unease in some zoologists and psychologistsis that the grade or levels notion [of anagenesis ] deals with the ranking of behavioral organization (capacity), not kinds of animals. . . ; that is, it is most often not a phyletic [or cladistic] ordering, (p. 453)

223

sis is compatible with what we referred to as the analysis of adaptation, which does not confine comparisons to within-lineage comparisons, . . . such studies [being] both valid and valuable, (pp. 216,220)

Thus, behavioral analysis may suggest a particular hierarchical organization of cognitive processes, some being relatively basic and others being rather more advanced (Razran, 1971; Romanes, 1884/1969). Those different grades or levels of cognitive processes may represent a meaningful evolutionary progression, one that might be recapitulated by species in different and even distant evolutionary lineages. Trying to reconstruct such behavioral hierarchies, if they exist, is unlikely to succeed by studying a few closely related species, although this phylogenetic strategy is useful for other evolutionary purposes:
When one reads the biological literature in search of examples of biological anagenesis, one finds that the instances recited are usually the readily discerned ones that come from comparing larger (supraspecific) taxonomic units, as, for example, in the evolution of homoiotherms (birds and mammals) from poikilotherms (reptiles), the evolution of the three-cone retina from the single-cone retina, and the like (Gottlieb, 1984, p. 451) Why?

This discussion thus places comparative cognition squarely within the realm of the evolutionary biology of behavior. What Is the Relation Between Comparative Cognition and Cognitive Ethology? From the foregoing, one might define comparative cognition as the comparative analysis of cognition in human and animal behavior. Researchers in this field generally adopt experimental methods of investigation, thereby allowing careful control and manipulation of relevant variables, precise and unbiased measurement of behavior, and replicability of experimental results. They endeavor to abide by Morgan's (1894/1896) canon in explaining the processes of cognition with a small number of operationally defined theoretical notions. Set apart from comparative cognition is cognitive ethology, a self-proclaimed antibehavioristic approach to problems of cognition in behavior. Initiated by D. R. Griffin (1976), the principal reason for creating this field of inquiry is "to learn as much as possible about the likelihood that nonhuman animals have mental experiences, and insofar as these do occur, what they entail and how they affect the animals' behavior, welfare, and biological fitness" (Griffin, 1978, p. 528). Recalling previous discussion of interpretation in the early history of comparative psychology, Griffin's (1978) call for a cognitive ethology appears to be a throwback to a prescientific analysis of behavior in terms of conscious experience. Writing in 1928, Warden (Warden, 1928) described the influence that the emergence of behaviorism had on comparative psychology: "Since that time, comparative psychology has been attempting to readjust itself to a strictly natural science position as the logical outcome of the Darwinian conception of psychology as a biological science" (p. 508). It is evident that the adjustment process is far from complete if researchers are still debating the usefulness of such notions as mind, mental experience, awareness, consciousness, and other ideas Griffin believes can be proper subjects for scientific study (see Ristau, 1991, for further discussion of cognitive ethology).9 That one can imagine animal behavior to be accompanied by subjective experiences cannot be the issue, for it is equally possible to imagine those behaviors to be performed without conscious accompaniments, a point recognized nearly a century ago by Jennings (1904/1976; see his earlier quotation). There is simply no clear or necessary role for subjective experience to play in behavior, as has been observed by Segal (1978):

Here, one might suggest that most phylogenetic or cladistic reconstructions have involved behavioral traits of remarkable narrowness and inflexibility, such as the reproductive behaviors of avians (Tinbergen, 1959) and rodents (Dewsbury, 1975). This suggestion is supported by "a wealth of material showing that interspecies differences in motor patterns of the kind most commonly used in comparative [evolutionary] studies are almost invariably innate (Hinde & Tinbergen, 1958, p. 255). Also, because "most of the studies made hitherto have dealt with relatively small behavior elements within groups of closely related species. . .[the] conclusions drawn. . . refer at most to microevolution" (Hinde & Tinbergen, 1958, p. 253). Different issues and strategies are bound to arise when one is trying to compare not the behaviors themselves but the adaptive processes manifested by those behaviors, such as intelligence and cognitionthe very processes that concerned Darwin and Spencer and that launched the field of comparative psychology. This contrast between behavior and process was not made by self-serving comparative psychologists but by two leading experts in ethology, R. A. Hinde and N. Tinbergen (1958):
Ultimately it will be desirable to make comparative studies not only of overt behavior but also of the causal mechanisms underlying it. However, since the motor patterns are directly observable, it is these which have been studied most often, (p. 253)

Whereas Hinde and Tinbergen (1958) surmised that its greater ease might make cladistic reconstruction a forerunner of anagenic reconstruction, there is no reason not to consider these as two parallel and complementary lines of evolutionary inquiry. In addition, although they are strong critics of many interpretations of anagenesis, even Campbell and Hodos (1991) conceded that
if anagenesis is considered to be a temporal sequence of grades (stages in the improvement of an organic design) exhibited in a number of different lineages. . . , then this functional anagene-

Concern with mental experience, although it might not be a fitting topic for scientific inquiry, may nevertheless be an important source of investigable hypotheses. Thus, within the experimental analysis of animal psychophysics, numerous studies have been conducted to confirm the existence of various perceptual illusions in humans and animals; for an excellent recent example, see Fujita, Blough, and Blough (1991), who reported that pigeons see the Ponzo illusion.

224

E. A. WASSERMAN nition (e.g., Kamil & Roitblat, 1985), thus connecting the study of naturalistic contingencies with complex and modifiable patterns of action.10 These increased interactions and interdisciplinary efforts should also reduce the intellectual isolation that has heretofore characterized the field. Thus, the second century of work in comparative cognition is off to a fast start. Its strong commitment to carefully controlled experimental methods and its clear focus on objectively verifiable processes of cognition should help the field avoid the methodological and interpretive traps that hampered progress during its first century of inquiry. The debt to the early evolutionists can be no better paid than by advancing the field of comparative cognition in accordance with the best methods of behavioral science. Within this area of inquiry, there has been a great deal of interest in adaptive specializations of intelligence: Are the processes of learning, memory, and cognition strongly shaped by the specific ecological niche an animal occupies? Although opinions differ dramatically on this question, the research it has spawned is of considerable interest and importance. See Balda and Kamil (1989) and Shettleworth (1992) for evidence on the remarkable spatial memories of food-storing avians; see Macphail (1985) for a critical introduction to the general question of adaptive specialization; and reexamine the initial quotation by Lachman et al. (1979) for a presentation of this thesis with regard to human intelligence (one that seems diametrically opposed to Morgan's, 1894/1896, canon). References Astley, S. L., & Wasserman, E. A. (1992). Categorical discrimination and generalization in pigeons: All negative stimuli are not created equal. Journal of Experimental Psychology: Animal Behavior Processes, 18, 193-207. Atkinson, R. C, & Shiffrin, R. M. (1968). Human memory: A proposed system and its control processes. In K. W Spence & J. T. Spence (Eds.), The psychology of learning and motivation (pp. 89195). San Diego, CA: Academic Press. Balda, R. P., & Kamil, A. C. (1989). A comparative study of cache recovery by three corvid species. Animal Behaviour, 38, 486-495. Bhatt, R. S. (1988). Categorization in pigeons: Effects of category size, congruity with human categories, selective attention, and secondary generalization. Unpublished doctoral dissertation, University of Iowa. Bhatt, R. S., & Wasserman, E. A. (1987). Choice behavior of pigeons on progressive and multiple schedules: A test of optimal foraging theory. Journal of Experimental Psychology: Animal Behavior Processes, 13, 40-51. Bhatt, R. S., Wasserman, E. A., Reynolds, W E, Jr., & Knauss, K. S. (1988). Conceptual behavior in pigeons: Categorization of both familiar and novel examples from four classes of natural and artificial stimuli. Journal of Experimental Psychology: Animal Behavior Processes, 14, 219-234. Bitterman, M. E. (1975). The comparative analysis of learning: Are the laws of learning the same in all animals? Science, 188, 699-709. Bjork, R. A. (1972). Theoretical implications of directed forgetting. In A. W Melton & E. Martin (Eds.), Coding processes in human memory (pp. 217-235). Washington, DC: Winston. Blough, D. S. (1959). Delayed matching in the pigeon. Journal of the Experimental Analysis of Behavior, 2, 151-160. Boysen, S. J., & Capaldi, E. J. (Eds.). (1992). The development of numerical competence: Animal and human models. Hillsdale, NJ: Erlbaum. Brown, M. E, & Morrison, S. K. (1990). Element and compound
10

Animal cognition does not imply awareness. To say that an animal learns about environmental relationships or relationships between its behavior and its consequences and then acts on that knowledge is not to say that the animal knows it has knowledge, or knows what it is doing. It may simply be that organisms with complex brains react to stimulus input in complex ways. At no time is it necessary that the organism be an active, conscious participant in its information-processing functions; biological matter may be sufficient to do the jobas physical matter is sufficient to do comparable problem-solving tasks in the digital computer, (pp. 213-214) One possible reason for the resurrection of mental istic terms at this particular time is the keen public and scientific interest that surrounds many of the animal language projects mentioned earlier. Because those projects have succeeded in establishing two-way communication between a human and an animal, Griffin (1978) saw a possibility for using such a communication system as a "window" on the minds of animals, through which they might themselves make known their thoughts, feelings, and emotions. This idea is indeed enchanting; but can it achieve its objective? How can researchers be sure that this and other so-called windows on the minds of animals are not in reality mirrors, reflecting back the thoughts, feelings, and emotions of the humans? In addition, if researchers must resort to anthropomorphism to interpret the behaviors of animals (see Burghardt, 1985), must not researchers also now solve the persistent problems of introspectionism, whose intractability spurred the rise of behaviorism? Perhaps it is best to conclude this discussion of comparative cognition and cognitive ethology with these sobering reflections of Mason (1976): What has been achieved [by behavioristic investigations of animal cognition] may seem pale in comparison with the vision of sitting down like Dr. Doolittle for an informal and revealing chat with an animal friend. But even if this were possible, how much would necessarily remain unsaid? If we have learned one thing from years of effort devoted to the problem, it is that there is no "window" that will allow us to gaze directly on another mind, even that of another human being, and to see its workings clearly and to see them whole. Mind, after all, lacks "thing quality"; it is but a construct, hardly more than a label, really, for complex processes and functions that we are still far short of understanding in any creature, including ourselves. We have learned what is perhaps the hardest lesson of all: There is no royal road to mind; we are forced to approach along the only paths that are open to us, through the tortuous byways of analysis, inference, hypothesis, and reconstruction, (p. 931)

Prospectus
From the preceding review and discussion, it is clear that research on the comparative psychology of cognition is entering a period of real growth and accomplishment. As research in the area of comparative cognition continues, there is likely to be much greater contact with the areas of cognitive science, on the one hand, and behavioral neuroscience, on the other. Fuller elucidation of the similarities and differences between human and animal cognition plus greater appreciation of the biological mechanisms of cognition will surely come from these contacts. Also underway is an important effort to investigate the role of ecological factors in the evolution of adaptive behavior and cog-

COMPARATIVE COGNITION AND ANIMAL INTELLIGENCE matching-to-sample performance in pigeons: The roles of information load and training history. Journal of Experimental Psychology: Animal Behavior Processes, 16, 185-192. Buchanan, J. P., Gill, T. V, & Braggio, J. T. (1981). Serial position and clustering effects in a chimpanzee's "free recall." Memory & Cognition, 9,651-660. Burghardt, G. M. (1985). Animal awareness: Current perceptions and historical perspective. American Psychologist, 40, 905-919. Campbell, C. B. G., & Hodos, W (1991). The Scala Naturae revisited: Evolutionary scales and anagenesis in comparative psychology. Journal of Comparative Psychology, 105, 211-221. Cohen, G. (1989). Memory in the real world. Hillsdale, NJ: Erlbaum. Cook, R. S., Wright, A. A., & Kendrick, D. F. (1990). Visual categorization by pigeons. In M. L. Commons & R. J. Herrnstein (Eds.), Quantitative analyses of behavior: Pattern recognition, Vol. VIII (pp. 187214). Hillsdale, NJ: Erlbaum. Crowder, R. G. (1976). Principles of learning and memory. Hillsdale, NJ: Erlbaum. DAmato, M. R., & Salmon, D. P. (1984). Processing of complex auditory stimuli (tunes) by rats and monkeys (Cebus apella). Animal Learning and Behavior, 12, 184-194. DAmato, M. R., Salmon, D. P., Loukas, E.,&Tomie, A. (1986). Processing of identity and conditional relations in monkeys (Cebus apella) and pigeons (Columba livia). Animal Learning and Behavior, 14, 365-373. Darwin, C. (1920). The descent of man, and selection in relation to sex (2nd ed.). New York: Appleton. (Original work published 1871) Davis, H., & Albert, M. (1986). Numerical discrimination by rats using sequential auditory stimuli. Animal Learning and Behavior, 14, 5759. Davis, H., & Perusse, R. (1988). Numerical competence in animals: Definitional issues, current evidence, and a new research agenda. Behavioral and Brain Sciences, 11, 561-615. Devine, J. V, Burke, M. W, & Rohack, J. J. (1979). Stimulus similarity and order as factors in visual short-term memory in nonhuman primates. Journal of Experimental Psychology: Animal Behavior Processes, 5, 335-354. Dewsbury, D. A. (1975). Diversity and adaptation in rodent copulatory behavior. Science, 190, 947-954. Dreyfus, L. R., Fetterman, J. G, Smith, L. D., & Stubbs, D. A. (1988). Discrimination of temporal relations by pigeons. Journal of Experimental Psychology: Animal Behavior Processes, 14, 349-367. Edwards, C. A., & Honig, W K. (1987). Memorization and "feature selection" in the acquisition of natural concepts in pigeons. Learning and Motivation, 18, 235-260. Edwards, C. A., Miller, J. S., & Zentall, T. R. (1985). Control of pigeons' matching and mismatching performance by instructional cues. Animal Learning and Behavior, 13, 383-391. Fernandes, D. M., & Church, R. M. (1982). Discrimination of the number of sequential events by rats. Animal Learning and Behavior, 10, 171-176. Fletcher, H. J. (1965). The delayed response problem. In A. M. Schrier, H. F Harlow, & F. Stollnitz (Eds.), Behavior of nonhuman primates (pp. 129-165). San Diego, CA: Academic Press. Fujita, K., Blough, D. S., & Blough, P. M. (1991). Pigeons see the Ponzo illusion. Animal Learning and Behavior, 19, 283-293. Gallistel, C. R. (1989). Animal cognition: The representation of space, time, and number. Annual Review of Psychology, 40, 155-189. Gardner, R. A., & Gardner, B. T. (1984). A vocabulary test for chimpanzees (Pan troglodytes). Journal of Comparative Psychology, 98, 381-404. Gisiner, R., & Schusterman, R. J. (1992). Sequence, syntax, and semantics: Responses of a language-trained sea lion (Zalophus california-

225

nus) to novel sign combinations. Journal of Comparative Psychology, 706,78-91. Gluck, M. A., & Bower, G. H. (1988). From conditioning to category learning: An adaptive network model. Journal of Experimental Psychology: General, 117, 227-247. Gottlieb, G. (1984). Evolutionary trends and evolutionary origins: Relevance to theory in comparative psychology. Psychological Review, 91 _ 448-456. Grant, D. S. (1981). Stimulus control of information processing in pigeon short-term memory. Learning and Motivation, 12,19-39. Grant, D. S. (1984). Rehearsal in pigeon short-term memory. In H. L. Roitblat, T. G. Bever, & H. S. Terrace (Eds.), Animal cognition (pp. 99-115). Hillsdale, NJ: Erlbaum. Grant, D. S., & MacDonald, S. E. (1990). An evaluation of the role of dual coding in mediating the effect of incorrectly cuing the comparison dimension in delayed matching in pigeons. Animal Learning and Behavior, 18,151 -156. Greenfield, P. M., & Savage-Rumbaugh, E. S. (1990). Grammatical combination in Pan paniscus: Processes of learning and intervention in the evolution and development of language. In S. T. Parker & K. R. Gibson (Eds.), Language and intelligence in monkeys and apes: Comparative developmental perspectives (pp. 540-578). Cambridge, England: Cambridge University Press. Greenfield, P. M., & Savage-Rumbaugh, E. S. (1991). Imitation, grammatical development, and the invention of protogrammar by an ape. In N. A. Krasnegor, D. M. Rumbaugh, R. L. Schiefelbusch, & M. Studdert-Kennedy (Eds.), Biological and behavioral determinants of language development (pp. 235-258). Hillsdale, NJ: Erlbaum. Griffin, D. R. (1976). The question of animal awareness: Evolutionary continuity of mental experience. New York: Rockefeller University Press. Griffin, D. R. (1978). Prospects for a cognitive ethology. Behavioral and Brain Sciences, 1, 527-538. Guttenberger, V T, & Wasserman, E. A. (1985). Effects of sample duration, retention interval, and passage of time in the test on pigeons' matching-to-sample performance. Animal Learning and Behavior, 13,121-128. Haugeland, J. (1978). The nature and plausibility of cognitivism. Behavioral and Brain Sciences, 2, 215-226. Herman, L. M., & Gordon, J. A. (1974). Auditory delayed matching in the bottlenose dolphin. Journal of the Experimental Analysis of Behavior, 21, 19-26. Herman, L. M., Morrel-Samuels, P., & Pack, A. A. (1990). Bottlenosed dolphin and human recognition of veridical and degraded video displays of an artificial gestural language. Journal of Experimental Psychology: General, 119, 215-230. Herrnstein, R. J. (1985). Riddles of natural categorization. Philosophical Transactions of the Royal Society of London, B308, 129-144. Herrnstein, R. J. (1990). Levels of stimulus control: A functional approach. Cognition, 37, 133-166. Herrnstein, R. J., & de Villiers, P. A. (1980). Fish as a natural category for people and pigeons. In G. Bower (Ed.), The psychology of learning and motivation (pp. 59-95). San Diego, CA: Academic Press. Hinde, R. A., & Tinbergen, N. (1958). The comparative study of species-specific behavior. In A. Roe & G. G. Simpson (Eds.), Behavior and evolution (pp. 251-268). New Haven, CT: Yale University Press. Hinrichs, J. V, & Grunke, M. E. (1975). Control processes in short-term memory: Use of retention interval information. Journal of Experimental Psychology: Human Learning and Memory, 104, 229-237. Hodos, W, & Campbell, C. B. G. (1969). Scala naturae: Why there is no theory in comparative psychology. Psychological Review, 76, 337350. Homa, D, Burrel, L., & Field, D. (1987). The changing composition of abstracted categories under manipulations of decisional change,

226

E. A. WASSERMAN Macphail, E. M., & Reilly, S. (1989). Rapid acquisition of a novelty versus familiarity concept by pigeons (Columba livid). Journal of Experimental Psychology: Animal Behavior Processes, 15, 242-252. Maki, W S. (1981). Directed forgetting in animals. In N. E. Spear & R. R. Miller (Eds.), Information processing in animals: Memory mechanisms (pp. 199-225). Hillsdale, NJ: Erlbaum. Maki, W S., & Hegvik, D. K. (1980). Directed forgetting in pigeons. Animal Learning and Behavior, 8, 567-574. Mason, W A. (1976). Windows on other minds. Science, 194, 930-931. Morgan, C. L. (1896). An introduction to comparative psychology. London: Scott. (Original work published 1894) Nelson, K. R., & Wasserman, E. A. (1978). Temporal factors influencing the pigeon's successive matching-to-sample performance: Sample duration, intertrial interval, and retention interval. Journal of the Experimental Analysis of Behavior, 30, 153-162. Oden, D. L., Thompson. R. K. R., & Premack, D. (1988). Spontaneous transfer of matching by infant chimpanzees (Pan troglodytes). Journal of Experimental Psychology: Animal Behavior Processes, 14,140145. Parker, B. K. (1984). Reproduction memory of two-event sequences in pigeons. Journal of the Experimental Analysis of Behavior, 41, 135141. Pearce, J. M. (1987). Introduction to animal cognition. Hillsdale, NJ: Erlbaum. Pepperberg, I. M. (1981). Functional vocalizations by an African grey parrot (Psittacus erithacus). Zeitschrift fur Tierpsvchologie, 55,139160. Premack, A. J., & Premack, D. (1972). Teaching language to an ape. Scientific American, 227, 92-99. Premack, D. (1978). On the abstractness of human concepts: Why it would be difficult to talk to a pigeon. In S. H. Hulse, H. Fowler, & W K. Honig (Eds.). Cognitive processes in animal behavior (pp. 423451). Hillsdale, NJ: Erlbaum. Ratner, S. C. (1970). Comparative psychology. In A. Gilgen (Ed.), Contemporary scientific psychology (pp. 115-146). San Diego, CA: Academic Press. Razran, G. (1971). Mind in evolution. New York: Houghton-Mifflin. Reed, P., Chih-Ta, T, Aggleton, J. P., & Rawlins, J. N. P. (1991). Primacy, recency, and the von Restorff effect in rats' nonspatial recognition memory. Journal of Experimental Psychology: Animal Behavior Processes, 17, 36-44. Riley, D. A., & Brown, M. F. (1991). Representation of multidimensional stimuli in pigeons. In G. Lockhead & J. R. Pomerantz (Eds.), The perception of structure (pp. 227-245). Washington, DC: American Psychological Association. Riley, D. A.. & Roitblat, H. L. (1978). Selective attention and related cognitive processes in pigeons. In S. H. Hulse, H. Fowler, & W K. Honig (Eds.), Cognitive processes in animal behavior (pp. 249-276). Hillsdale, NJ: Erlbaum. Rilling, M. (1992). Invisible counting animals: A history of contributions from comparative psychology, ethology, and learning theory. In S. T. Boysen & E. J. Capaldi (Eds.), The development of numerical competence: Animal and human models. Hillsdale, NJ: Erlbaum. Rilling, M., Kendrick, D. F, & Stonebraker, T. B. (1984). Directed forgetting in context. In G. Bower (Ed.), The psychology of learning and motivation (pp. 175-198). San Diego, CA: Academic Press. Ristau, C. A. (Ed.). (1991). Cognitive ethology: The minds of other animals. Hillsdale, NJ: Erlbaum. Roberts, W A., & Grant, D. S. (1976). Studies of short-term memory in the pigeon using the delayed matching-to-sample procedure. In D. L. Medin, W A. Roberts, & R. T. Davis(Eds.), Processes of animal memory (pp. 79-112). Hillsdale, NJ: Erlbaum.

choice difficulty, and category size. Journal of Experimental Psychology: Learning, Memory, and Cognition, 13, 401-412. Honig, W K. (1978). On the conceptual nature of cognitive terms: An initial essay. In S. H. Hulse, H. Fowler, & W K. Honig (Eds.), Cognitive processes in animal behavior (pp. 1-14). Hillsdale, NJ: Erlbaum. Honig, W K., & Dodd, P. W D. (1986). Anticipation and intention in working memory. In D. F. Kendrick, M. E. Rilling, & M. R. Denny (Eds.), Theories of animal memory (pp. 77-100). Hillsdale, NJ: Erlbaum. Honig, W K., & Thompson, R. K. R. (1982). Retrospective and prospective processing in animal working memory. In G. Bower (Ed.), The psychology of/earning and motivation (pp. 239-283). San Diego, CA: Academic Press. Hull, C. L. (1943). Principles oj behavior. New York: Appleton-CenturyCrofts. Hulse, S. H., Fowler, H., & Honig, W K. (1978). Cognitive processes in animal behavior. Hillsdale, NJ: Erlbaum. Hulse, S. H., & O'Leary, D. K. (1982). Serial pattern learning: Teaching an alphabet to rats. Journal of Experimental Psychology: Animal Behavior Processes, 8, 260-273. Hunt, E. (1989). Cognitive science: Definition, status, and questions. Annual Review of Psychology, 40, 603-629. Hunter, W S. (1913). The delayed reaction in animals and children. Behavior Monographs, 2, 1-86. Jennings, H. S. (1976). Behavior of the lower organisms. Bloomington: Indiana University Press. (Original work published 1904) Kamil, A. C, & Roitblat, H. L. (1985). Foraging theory: Implications for animal learning and cognition. Annual Review of Psychology, 36, 141-169. Kendrick, D. F., & Rilling, M. E. (1986). AIM: A theory of active and inactive memory. In D. F. Kendrick, M. E. Rilling, & M. R. Denny (Eds.), Theories of animal memory (pp. 129-152). Hillsdale, NJ: Erlbaum. Kendrick, D. F, Rilling, M. E., & Denny, M. R. (Eds.). (1986). Theories of animal memory. Hillsdale, NJ: Erlbaum. Kesner, R. P., & Jackson-Smith, P. (1992). Neurobiology of an attribute model of memory: Role of prefrontal cortex. In I. Gormezano & E. A. Wasserman (Eds.), Learning and memory: The behavioral and biological substrates (pp. 251-273). Hillsdale, NJ: Erlbaum. King, J. H., & Nichols, J. W (1960). Problems of classification. In R. H. Waters, D. A. Rethlingshafer, & W E. Caldwell (Eds.), Principles of comparative psychology (pp. 18-42). New York: McGraw-Hill. Konorski, J. (1959). A new method of physiological investigation of recent memory in animals. Bulletin de VAcademic Polonaise des Sciences Serie des Sciences Biologiques, 7, 115-117. Lachman, R., Lachman, J. L., & Butterfield, E. R. (1979). Cognitive psychology and information processing. Hillsdale, NJ: Erlbaum. Lea, S. E. G. (1984). In what sense do pigeons learn concepts? In H. L. Roitblat, T. G. Bever, & H. S. Terrace (Eds.), Animal cognition (pp. 263-276). Hillsdale, NJ: Erlbaum. Lockard, R. B. (1971). Reflections on the fall of comparative psychology: Is there a message for us all? American Psychologist, 26, 168179. Lombardi, C. M., Fachinelli, C. C, & Delius, J. D. (1984). Oddity of visual patterns conceptualized by pigeons. Animal Learning and Behavior, 12, 2-6. MacDonaid, S. E., & Grant, D. S. (1987). Effects of signaling retention interval length on delayed matching-to-sample in pigeons. Journal of Experimental Psychology: Animal Behavior Processes, 13, 116125. Macphail, E. M. (1985). Vertebrate intelligence: The null hypothesis. Philosophical Transactions of the Royal Society of London, B308, 3751.

COMPARATIVE COGNITION AND ANIMAL INTELLIGENCE Roberts, W A. Mazmanian, D. W, & Kraemer, P. J. (1984). Directed forgetting in monkeys. Animal Learning and Behavior, 12, 29-40. Roitblat, H. L. (1982). The meaning of representation in animal memory. The Behavioral and Brain Sciences, 5, 353-406. Roitblat, H.L. (1987). Introduction to comparative cognition. New\brk: Freeman. Roitblat, H. L, Bever, T. G., & Terrace, H. S. (Eds.). (1984). Animal cognition. Hillsdale, NJ: Erlbaum. Romanes, G. J. (1969). Mental evolution in animals. New York: AMS Press. (Original work published 1884) Romanes, G. J. (1977). Animal intelligence. New York: Appleton. (Reprinted in D. N. Robinson [Ed.], Significant contributions to the history of psychology 1750-1920 [Series A: Orientation, Vol. H. G. J. Romanes]. Washington, DC: University Publications of America.) (Original work published 1883) Rosch, E., & Mervis, C. B. (1975). Family resemblances: Studies in the internal structure of categories. Cognitive Psychology, 7, 573-605. Rumbaugh, D. M. (1977). Language learning by a chimpanzee: The Lana project. San Diego, CA: Academic Press. Santi, A., Musgrave, S., & Bradford, S. A. (1988). Utilization of cues signaling different test stimulus dimensions in delayed matching to sample by pigeons. Learning and Motivation, 19, 87-98. Santiago, H. C, & Wright, A. A. (1984). Pigeon memory: Same/different concept learning, serial probe recognition acquisition, and probe delay effects on the serial-position function. Journal of Experimental Psychology: Animal Behavior Processes, 10, 498-512. Savage-Rumbaugh, E. S., Murphy, J., Sevcik, R. A., & Rumbaugh, D. M. (in press). Language comprehension in ape and child. Monographs of the Society for Research in Child Development. Schnaitter, R. (1987). Behaviorism is not cognitive and cognitivism is not behavioral. Behaviorism, 15, 1-11. Segal, E. F. (1978). Where behavioral psychology and cognitive psychology meet. Revista Mexicana de Analisis de la Conducta, 4, 203-216. Shettleworth, S. J. (1992). Spatial memory in hoarding and nonhoarding tits (Paraded). In I. Gormezano & E. A. Wasserman (Eds.), Learning and memory: The behavioral and biological substrates (pp. 2544). Hillsdale, NJ: Erlbaum. Sidman, M. (1986). Functional analysis of emergent verbal classes. In T. Thompson & M. D. Zeiler (Eds.), Analysis and integration of behavioral units (pp. 213-245). Hillsdale, NJ: Erlbaum. Sidman, M., Rauzin, R., Lazar, R., Cunningham, S., Tailby, W, & Carrigan, P. (1982). A search for symmetry in the conditional discriminations of rhesus monkeys, baboons, and children. Journal of the Experimental Analysis of Behavior, 37, 23-44. Skinner, B. F. (1950). Are theories of learning necessary? Psychological Review, 57,193-216. Skinner, B. F. (1977). Why I am not a cognitive psychologist. Behaviorism, 5, 1-10. Skinner, B. F. (1985). Cognitive science and behaviorism. British Journal of Psychology, 76, 291-301. Smith, E. E., & Medin, D. L. (1981). Categories and concepts. Cambridge, MA: Harvard University Press. Spear, N. E., Miller, J. S., & Jagielo, J. A. (1990). Animal memory and learning. Annual Review of Psychology, 41,169-211. Stenhouse, D. (1973). The evolution of intelligence: A general theory and some of its implications. Winchester, MA: Allen & Unwin. Sternberg, R. J. (1986). What is intelligence? Contemporary viewpoints on its nature and definition. Norwood, NJ: Ablex. Stonebraker, T. B., & Rilling, M. (1981). Control of delayed matchingto-sample performance using directed forgetting techniques. Animal Learning and Behavior, 9, 196-201. Stonebraker, T. B., & Rilling, M. (1984). Retrospective versus prospec-

227

tive processes in delayed matching to sample. Bulletin of the Psychonomic Society, 22, 372-375. Terrace, H. S. (1991). Chunking during serial learning by a pigeon: I. Basic evidence. Journal of Experimental Psychology: Animal Behavior Processes, 17, 81-93. Terrace, H. S., Petitto, L. A., Sanders, R. J., & Bever, T. G. (1979). Can an ape create a sentence? Science, 206, 891-902. Thomas, R. K., & Noble, L. M. (1988). Visual and olfactory oddity learning in rats: What evidence is necessary to show conceptual behavior"? Animal Learning and Behavior, 16,157-163. Tinbergen, N. (1959). Comparative studies of the behavior of gulls (Laridae): A progress report. Behaviour, 15,1-70. Tobach, E., Adler, H. E., & Adler, L. L. (Eds.). (1973). Comparative psychology at issue. Annals of the NewYork Academy of Sciences, 223, 1-197. Vaughan, W, Jr. (1988). Formation of equivalence sets in pigeons. Journal of Experimental Psychology: Animal Behavior Processes, 14, 3642. Vaughan, W, Jr., & Greene, S. L. (1984). Pigeon visual memory capacity. Journal of Experimental Psychology: Animal Behavior Processes, 70,256-271. Warden, C. J. (1928). The development of modern comparative psychology. The Quarterly Review of Biology, 3, 486-522. Washburn, D. A., & Rumbaugh, D. M. (1991). Ordinal judgments of numerical symbols by macaques (Macaco mulatto). Psychological Science, 2, 190-193. Wasserman, E. A. (1976). Successive matching-to-sample in the pigeon: Variations on a theme by Konorski. Behavior Research Methods and Instrumentation, 8, 278-282. Wasserman, E. A. (1981). Comparative psychology returns: A review of Hulse, Fowler, and Honig's Cognitive processes in animal behavior. Journal of the Experimental Analysis of Behavior, 35, 243-257. Wasserman, E. A. (1982). Further remarks on the role of cognition in the comparative analysis of behavior. Journal of the Experimental Analysis of Behavior, 38, 211-216. Wasserman, E. A. (1983). Is cognitive psychology behavioral? The Psychological Record, 33, 6-11. Wasserman, E. A. (1984). Animal intelligence: Understanding the minds of animals through their behavioral "ambassadors." In H. L. Roitblat, T. G. Bever, & H. S. Terrace (Eds.), Animal cognition (pp. 45-60). Hillsdale, NJ: Erlbaum. Wasserman, E. A. (1986). Prospection and retrospection as processes of animal short-term memory. In D. F Kendrick, M. Rilling, & M. R. Denny (Eds.), Animal memory (pp. 53-75). Hillsdale, NJ: Erlbaum. Wasserman, E. A. (1990). Attribution of causality to common and distinctive elements of compound stimuli. Psychological Science, 1, 298-302. Wasserman, E. A., DeLong, R. E., & Larew, M. G. (1984). Temporal order and duration: Their discrimination and retention by pigeons. Annals of the New York Academy of Sciences, 423,103-115. Wasserman, E. A., DeVolder, C. L., & Coppage, D. J. (1992). Nonsimilarity-based conceptualization in pigeons via secondary or mediated generalization. Psychological Science, 3, 374-379. Wasserman, E. A., Grosch, J., & Nevin, J. A. (1982). Effects of signalled retention intervals on pigeon short-term memory. Animal Learning and Behavior, 10, 330-338. Wasserman, E. A., Kiedinger, R. E., & Bhatt, R. S. (1988). Conceptual behavior in pigeons: Categories, subcategories, and pseudocategories. Journal of Experimental Psychology: Animal Behavior Processes, 14, 235-246. Wasserman, E. A., Nelson, K. R., & Larew, M. B. (1980). Memory for

228

E. A. WASSERMAN delay effects. Journal of Experimental Psychology: Animal Behavior Processes, 70,513-529. Wright, A. A., Santiago, H. C, Sands, S. E, Kendrick, D. E, & Cook, R. G. (1985). Memory processing of serial lists by pigeons, monkeys, and people. Science, 229, 287-289. Yarczower, M., & Hazlett, L. (1977). Evolutionary scales and anagenesis. Psychological Bulletin, 84, 1088-1097. Zentall, T. R., Steirn, J. N., Sherburne, L. M., & Urcuioli, P. J. (1991). Common coding in pigeons assessed through partial versus total reversals of many-to-one conditional and simple discriminations. Journal of Experimental Psychology: Animal Behavior Processes, 17, 194-201. Received November 13,1991 Revision received March 17,1992 Accepted March 17,1992

sequences of stimuli and responses. Journal of the Experimental Analysis of Behavior, 34, 49-59. Watson, J. B. (1913). Psychology as the behaviorist views it. Psychological Review, 20, 158-177. Weiner, B., & Reed, H. (1969). Effects of the instructional sets to remember and to forget on short-term retention: Studies of rehearsal control and retrieval inhibition (repression). Journal of Experimental Psychology, 79, 226-232. Weiskrantz, L. (1985). Animal intelligence. Oxford, England: Oxford University Press. Weisman, R. G., Wasserman, E. A., Dodd, P. W D., & Larew, M. B. (1980). Representation and retention of two-event sequences in pigeons. Journal of Experimental Psychology: Animal Behavior Processes, 6, 312-325. Wright, A. A., Santiago, H. C, & Sands, S. F. (1984). Monkey memory: Same/different concept learning, serial probe acquisition, and probe

AMERICAN PSYCHOLOGICAL ASSOCIATION SUBSCRIPTION CLAIMS INFORMATION

Today's Date:_

We provide this form to assist members, institutions, and nonmember individuals with any subscription problems. With the appropriate information we can begin aresolution. If you use the services of an agent, please do NOT duplicate claims through them and directly to us. PLEASE PRINT CLEARLY AND IN INK IF POSSIBLE.

PRINT FULL NAME OR KEY NAME OF INSTITUTION

MEMBER OR CUSTOMER NUMBER (MAY BE FOUND ON ANY PAST ISSUE LABEL) DATE YOUR ORDER WAS MAILED (OR PHONED): P.O. NUMBER: PREPAID

STATE/COUNTRY

ZIP

CHECK CHARGE CHECK/CARD CLEARED DATE:

YOUR NAME AND PHONE NUMBER

(If possible, send a copy, front and back, of your cancelled check to help us in our research of your claim.) ISSUES: MISSING DAMAGED

TITLE

VOLUME OR YEAR

NUMBER OR MONTH

Thank you. Once a claim is received and resolved, delivery of replacement issues routinely takes 4-6 weeks. DATE RECEIVED: ACTION TAKEN: _ STAFF NAME: (TO BE FILLED OUT BY APA STAFF) DATE OF ACTION: _ INV. NO. & DATE: LAB EL NO. & DATE:

SEND THIS FORM TO: APA Subscription Claims, 750 First Street, N.E., Washington, DC 20002 PLEASE DO NOT REMOVE. A PHOTOCOPY MAY BE USED.