Avian Visual Cognition


Imitation in Animals:
Evidence, Function, and Mechanisms

Tom Zentall and Chana Akins
University of Kentucky

  avcrule.gif (935 bytes)  


The term imitation has been used descriptively and theoretically to characterize a broad range of animal behavior from physical antipredatory adaptations such as eye spots, which are totally under genetic control, to the human capacity for the exaggeration of individual characteristics, know as caricature, which are largely under cognitive control. In the present review many behaviors that have been called imitative are described, and are distinguished from "true imitation." It is suggested that at much of the ambiguity in the literature as to what should be called imitation can be attributed to the distinction between the function of imitation and the mechanisms responsible for it. Finally, the various mechanisms that have been proposed to account for true imitation are presented and an attempt is made to evaluate them.

  I. Introduction

When the term imitation is used by psychologists it typically implies more than the simple reproduction of behavior.  Imitation carries with it the implication of intentionality or purposiveness. When humans use the term to describe their own behavior, it implies that there is an understanding of the relation between that behavior and the behavior being modeled (or demonstrated). For example, when a child imitates the behavior of an adult, one assumes that the child understands that the two behaviors match (see photo to left). According to Freud (1933), imitation forms an integral part of the process of identification, and it is motivated by the need to identify (e.g., with the same-sex parent). But when we look for imitative learning in animals we must first pare away the mentalistic terms often used to describe and explain it, and then identify and rule out simpler mechanisms that might be involved in reports of its occurrence.

The purpose of this chapter is first to classify the various cases of social influence depending on what mechanisms are thought to underlie them. For example, among the simplest kinds of social influence are those social behaviors that are typical of the species and that happen to occur in unison (e.g., schooling, flocking, and herding). These behaviors can be thought of as genetically predisposed. Other behaviors may be made more likely to occur whenever there is an change in the motivation of the observer, and the mere presence of another animal may have a motivational effect on the observer. Still other behaviors may occur because the demonstrator or model can serve as a salient discriminative stimulus that predicts the appearance of food and allows for an account based on simple associative learning. It is also possible for the behavior of a demonstrator to draw the observer’s attention to a location or to a stimulus that is the focus of the target behavior (e.g., a bar that must be manipulated) and the observer’s mere proximity to that location or stimulus may make the target behavior more probable. Finally, it is possible for the behavior of the observer to have an effect on the environment such that the observer can learn how objects in the environment work. For example, seeing a door open outward rather than inward. In some cases more than one simpler mechanism may be involved. In many cases it may not be possible to detect the presence of true imitative learning because of the potential presence of other simpler accounts. After addressing each of these classes of social influence in turn, we will identify procedures that may separate them from what has come to be called true imitation.

A second purpose of this chapter is to distinguish between the function of imitation for the survival of the organism and mechanisms involved in its use. The implication of this analysis is that social learning in general and imitation in particular may be more wide spread than previously thought. Finally, an attempt will be made to identify the mechanisms by which organisms are able to imitate.

II. Social Influence

The broad use of the term imitation extends to any influence that an organism may have on another that results is a similarity of behavior or appearance between the two. Because biologists typically take a genetic approach to the study of organisms, for them, both physical appearance and behavior have evolutionary bases and thus, are closely related.

Species Typical Factors

Mimicry. When imitation involves copying of the physical appearance of one species by another, it is generally referred to as mimicry.

Which is the edible one?

When a relatively defenseless animals takes on the appearance of an animal that has better defenses, it is known as Batesian (or Mertensian) mimicry. The best known case of Batesian mimicry is that of the palatable viceroy butterfly (left photo) mimicking the unpalatable monarch butterfly (right photo, Turner, 1984).

A special case of mimicry involving behavior is the broken-wing display of the ground-nesting killdeer and avocet (Sordahl, 1981). When the female bird is near the nest and a predator approaches, the bird flies away from the nest while mimicking the erratic flight pattern that might be shown by a bird with a broken wing.

Contagion. When two or more animals engage in similar behavior and that behavior is species typical, the termwpe7.jpg (12370 bytes) contagion is often used (Thorpe, 1963; also called mimesis, Armstrong, 1951, or response facilitation, Byrne, 1994). Contagion can be used to describe certain courtship displays when they involve coordinated movements between the male and female that are can sometimes appear to be virtual mirror images (Tinbergen, 1960). Also, antipredatory behavior can be considered contagious when it involves the coordinated movement of a group of animals for defensive purposes. Such behavior occurs in certain mammalian (herding) and avian (flocking) species. When this coordinated behavior is aggressive (i.e., directed toward rather than away from danger), it is known as mobbing (Hoogland & Sherman, 1976). Contagion can also be shown in an appetitive context. A satiated animal in the presence of food will often resume eating upon the introduction of a hungry animal which begins eating (Tolman, 1964). In the case of contagion, the behavior of one animal appears to serve as a releaser for the unlearned behavior of others (Thorpe, 1963). For this reason, most research on animal imitation has used arbitrary, novel, to-be-acquired behavior as the focus research.

Motivational Factors

When imitative behavior is typically studied, an observing animal (e.g., a rat) is exposed to a demonstrator (rat) that is performing a novel response such as bar pressing and the rate of acquisition of the novel response is taken as a measure of imitation. But because rate of acquisition is a relative measure (it will depend on a variety of procedural factors unrelated to the conditions of observation), it must be compared with that of a control group. The choice of control group often depends on the way the experimental questions are asked. For example, early research appeared to assume that rats could learn through observation of the performance of a demonstrator, and the question was, did such learning occur more quickly than the more typically used experimenter-trained shaping procedures (i.e., by successive approximations; Corson, 1967; Jacoby & Dawson, 1969; Powell, 1968; Powell & Burns, 1970; Powell, Saunders, & Thompson, 1969). The experimental question asked appears to have been a practical one: Would observation of the demonstration of a response be a faster means of training rats to bar press. The problem with the use of a “shaped” control group is that the procedures used to shape animals to bar press are difficult to specify in ways that can be accurately reproduced. Furthermore, comparison of the rate of bar-press acquisition by an imitation group with that of a group of shaped animals may indicate little about the ability of animals to imitate, especially if the imitation group acquires the response at the same rate as, or slower than, the control group. To ask about the ability of animals to acquire a response through observation of a demonstrator, a more appropriate control group would be a “trial-and-error” control (i.e., a group of animals that acquires the response on its own).

Social facilitation. The implication of imitative learning is that information transmitted from the demonstrator to the observer has led to facilitated acquisition. There is evidence, however, that the mere presence of an animal (of the same species, i.e., a conspecific) can have effects on the motivational state of an observer (Zajonc, 1965), and that changes in the observer’s motivational state can affect the acquisition and performance of a response (Zentall & Levine, 1972; Levine & Zentall, 1974). Zajonc has referred to this effect of the “mere presence” of a conspecific on motivation as social facilitation, and he has proposed a version of Hull’s (1943) theory to account for the variety of findings of Click here to view fear reduction video clipbehavioral facilitation and inhibition that have been reported in humans and animals. According to Zajonc, the presence of a conspecific leads to an increase in arousal which can actually lead to the retardation of acquisition of a novel (to-be-learned) response. Others have suggested that the mere presence of a conspecific will facilitate the acquisition of a new response for the same reason (Gardner & Engel, 1971) or because the conspecific may have the ability to reduce fear in the observer (Davitz & Mason, 1955; Morrison & Hill, 1967). In any case, experiments concerned with true imitative learning must include a control for the possibility of facilitation or retardation of response acquisition resulting from mere presence effects.

One further potential source of demonstrator-provided motivation should be mentioned here. Although the mere presence of a conspecific may contribute to the motivational state of an observer, general (nonspecific) activity may make an additional contribution. Being in the presence of an active conspecific (i.e., one that is working for food but that is responding in a way that is irrelevant to the target response) might constitute an even better control than mere presence. We will return to this point later.

Incentive motivation. Reinforcement provided to the observer during the demonstration of a novel response (i.e., incentive motivation) may also play a role in the rate at which a novel response is acquired. Although Del Russo (1971) did not find significant evidence for imitative learning by rats that observed a demonstrator bar pressing for food (relative to a trial-and-error control) he did find significant facilitation of bar pressing by a group of observers that got fed whenever their bar-pressing demonstrators got fed. This facilitation may have involved a general increase in arousal on the part of the reinforced observer or a more specific association of the apparatus context with reinforcement. In either case, the observers would likely have been more active following observation than the nonreinforced comparison groups, and more active animals would be more likely to make accidental contact with the bar.

Observation of aversive conditioning. The imitation of a novel response being acquired or being performed by a demonstrator that is motivated by the avoidance of painful stimulation (e.g., electric shock) presents the need for special control. Emotional cues provided by a conspecific either escaping from or avoiding shock may provide emotional cues of pain or fear that could instill fear in an observer. For example, John, Chesler, Bartlett, and Victor (1968) found that cats that had observed a demonstrator being trained to jump over a hurdle to avoid foot shock learned the hurdle-jumping response faster than controls that did not observe the demonstrators. It may be, however, that being in the presence of a cat being shocked was sufficient to increase the observers’ fear (motivation) associated with the conditioning context. Under such conditions, a change in motivation may facilitate acquisition.

Under different conditions, however, although rats that observed a trained demonstrator that had acquired a discriminated shuttle avoidance response acquired that response faster than rats that observed a merely present demonstrator, rats merely exposed to the empty shuttle box acquired the shuttle response fastest (Sanavio & Savardi, 1980).

Trying to sort out mere presence and emotional motivational effects from learning effects can be quite complex. Although using well-trained demonstrators can reduce the likelihood of pain-produced cues being transmitted to the observers (Del Russo, 1975), it may not be possible to avoid the effects of demonstrator-provided, fear-produced cues

One way to avoid problems associated with differential motivational cues encountered with observation of aversively motivated conditioning is to include an observation control group exposed to performing demonstrators but with the observer's view of a critical component of the demonstrator's response blocked. Such a control was included in Click here to view Figure 3an experiment by Bunch and Zentall (1980) who used a candle flame avoidance task originally developed by Lore, Blanc, and Suedfeld (1971). Bunch and Zentall found that rats learned the candle-flame-avoidance task faster after having seen a demonstrator acquire the task, as compared with (1) a group for which a small barrier was placed in front of the candle such that the observer's view of the rat's contact with the candle was blocked and (2) a social facilitation control group (see Figure 3.). Thus, although a variety of auditory cues (a potential by-product of the demonstrator's pain), olfactory cues (e.g., potentially produced by singed whiskers, defecation, and urination), and visual cues (e.g. seeing the demonstrator approach and then rapidly withdraw from something directly behind the barrier) associated with the task should have provided comparable motivational cues to these control observers, task acquisition was not facilitated as much as for observers that could also observe the demonstrator's contact with the candle.

Another means of controlling for potential motivational cues provided by the demonstrator performing a pain- or fear-motivated task, is to expose the observers to demonstrators performing a discrimination (Kohn, 1976; Kohn & Dennis, 1972). In this research, rats that observed a demonstrator performing a relevant shock-avoidance discrimination acquired that task faster than controls for which the demonstrator's discrimination was the reverse of the observer's (i.e., what was correct for the demonstrator was incorrect for the observer and vice versa).

Although it may be difficult to control for the social transmission of motivation produced during the acquisition or performance of an avoidance response, it may be that social learning is more likely to occur under conditions of fear motivation because of its greater evolutionary value.

Perceptual Factors

When the observation of a demonstrator merely draws attention to the consequences of a response (e.g., a lever press), it may merely alter the salience of the lever (stimulus enhancement) or the place in the environment where the lever is located (local enhancement).

Local Enhancement. Local enhancement refers to the facilitation of learning that results from drawing attention to a locale or place associated with reinforcement (Roberts, 1941). For example, Lorenz (1935) noted that ducks enclosed in a pen may not react to a hole large enough for them to escape unless they happen to be near another duck as it is escaping from the pen. The sight of a duck passing through the hole in the pen may simply draw attention to the hole. Similar effects can be demonstrated with Japanese quail.

Local enhancement has also been implicated in the finding that puncturing the top of milk bottles by great tits spread in a systematic way from one neighborhood to another (Fisher & Hinde, 1949). Although the technique of pecking through the top of the bottle may be learned through observation, it is also likely that attention was drawn to the bottles by the presence of the feeding birds. Once at the bottles, the observers found the reward and consumed it. Learning to identify milk bottles as a source of food, can readily generalize to other open bottles. Finally, drinking from opened bottles can readily generalize to an attempt to drink from a sealed bottle, which in turn can lead to trial-and-error puncturing of the top.

Local enhancement may be studied in its own right. As Denny, Clos, and Bell (1988) have shown, observation by rats of merely the movement and sound of a bar being activated (by the experimenter from outside the chamber) and paired with food presentation can facilitate the acquisition of the bar press response, relative to various control procedures.

Local enhancement may also account for John et al.'s (1968, Exp. 2) finding of socially facilitated acquisition of lever pressing by cats. Cats placed in the same chamber as another cat lever pressing for food, learned to press that lever faster than other cats that observed another cat that was fed periodically without lever pressing. But observation of lever pressing may simply draw attention to the lever. Local enhancement is especially likely in this context, in which observation of the moving lever might encourage lever approach upon removal of the demonstrator (especially by a species known for its motivation to explore).

Similarly, local enhancement may play a role in the faster acquisition of lever pressing by kittens that observed their mothers as demonstrators, than by kittens that observed a strange female demonstrator (Chesler, 1969), because orientation towards the mother may be more likely than towards a strange cat.

Local enhancement may also be involved in John et al.'s (1968, Exp. 1) finding of facilitated acquisition of an aversively motivated hurdle jump response. The distinction between true imitation and local enhancement may be a subtle one in this case, but observation of the demonstrator simply may draw the observer's attention to the top of the hurdle. In other words, seeing a ball bounce over the hurdle, or even placing a flashing light at the top of the hurdle might be enough to facilitate the hurdle jumping response. In general, whenever the performance observed involves an object (i.e., a manipulandum) to which the observer must later respond, local enhancement may play a role (Corson, 1967; Denny, Clos, & Bell, 1988; Herbert & Harsh, 1944; Jacoby & Dawson, 1969; Oldfield-Box, 1970).

In other cases, it may be possible to control for local enhancement effects by including proper controls.Lefebvre and Palameta (1988), for example, found that pigeons that observed a model pierce the paper cover on a food well to obtain hidden grain, later acquired that response on their own, whereas those that observed that same response, but with no grain in the well (the model performed in extinction), failed to acquire the response.

Stimulus Enhancement. In the case of local enhancement, the attention of an observer is drawn to a particular place by the activity of the demonstrator. The term stimulus enhancement is used when the activity of the demonstrator draws the attention of the observer to a particular object (e.g., the manipulandum). Quite often in the study of imitative learning, the object in question is at a fixed location so the two mechanisms are indistinguishable. In the duplicate-chamber procedure (see Warden & Jackson, 1935; Gardner & Engel, 1971), however, a manipulandum (e.g., a lever) is present in both the demonstration chamber and in the observation chamber.Under these conditions, drawing attention to the demonstrator's lever should not facilitate acquisition of lever pressing by the observer. In fact, one could argue that it should retard acquisition of lever pressing by an observer because it should draw the observer's attention away from its own lever. In the case of stimulus enhancement, however, the similarity between the demonstrator's lever and that of the observer may make it more likely that the observer notices its own lever after having its attention drawn to the demonstrator's lever. Thus, stimulus enhancement refers to the combination of a perceptual, attention-getting process resulting from the activity of the demonstrator in the presence of the lever, and stimulus generalization between the demonstrator's and observer's levers. Because it subsumes local enhancement effects, the term stimulus enhancement may be more inclusive and thus, may be preferable (Galef, 1988b).

Stimulus enhancement may also be involved in the facilitated acquisition of an observed discrimination. If the demonstrator is required to make contact with the positive stimulus, but not the negative stimulus, the positive stimulus is likely to attract the observers attention and responding to it may be facilitated (Edwards, Hogan, & Zentall, 1980; Kohn, 1976; Kohn & Dennis, 1972; Fiorito & Scotto, 1992; Vanayan, Robertson, & Biederman, 1985).

Stimulus enhancement may also play an important role in mate-choice copying by guppies (Dugatkin, 1996). Female guppies that see a model female in the presence of a courting male will prefer that male over an alternative male (Dugatkin, 1992; Dugatkin & Godin, 1992).

The facilitation of learning through perceptual factors presents a most difficult problem for the study of imitation in animals. If the similarity between the demonstrator’s location or manipulandum and that of the observer presents an interpretational problem because of perceptual factors, making the location or the nature of the manipulandum for the observer different from that of the demonstrator is likely to interfere with the observers “understanding” of the relation between the two tasks. This problem, which will be addressed later, will require a new approach to defining adequate control procedures.

III. Social Learning

Simple Learning Factors

There are a number of cases of social learning which may be mediated by simple nonsocial learning mechanisms. Although social stimuli are present and those social stimuli may play a role in facilitating acquisition of the target behavior (perhaps because the social stimuli are more salient than nonsocial alternatives), the mechanisms by which the observer acquires the behavior may be more parsimoniously explained in terms of simpler species-typical or individual learning processes.

Imprinting.The first example of social learning that should be distinguished from true imitation is imprinting. wpe4.jpg (27793 bytes)Imprinting is a process that occurs primarily in species which do not have the luxury of a nest or den in which to protect their young. In such species (e.g., fowl and grazing mammals), the young are hatched (or born) in a precocious state that allows them to move about following a very brief period of inactivity. To compensate for their mobility (which could also put them a great risk of predation) most of these species have evolved the predisposition to follow the first moving object they see. Although this object is generally their mother, laboratory experiments show that almost any moving object can function as the object of imprinting (Hess, 1973).

Imprinting is a curious process that combines a strongly predisposed behavior (following) with considerable flexibility (learning) in the nature of the object that is followed. Although one could say, in a very general sense, that the imprinted young are imitating the mother, the act of following (or approach), is more parsimoneously interpreted as a simple conditioning process, with fear reduction serving as the reinforcer (Kovach & Hess, 1963).

Discriminated following (or matched dependent) behavior.Rats can learn to follow a trained conspecific to food in a T maze in the absence of any other discriminative stimulus (Bayroff & Lard, 1944; Church, 1957; Haruki & Tsuzuki, 1967). Although the leader rat in these experiments is clearly a social stimulus, the data are more parsimoniously interpreted in terms of simple discriminative learning. If, for example, the demonstrator were replaced with a block of wood pulled along by a string, or even an arrow at the choice point directing the rat to turn left or right, it is clear that one would identify the cue (i.e., the demonstrator, the block of wood, and the arrow) as a simple discriminative stimulus. Even if following a demonstrator led to faster learning than following a passive signal, it might merely indicate that the social cue was more salient than either a static or a moving nonliving cue (see Stimbert, 1970). For matched-dependent behavior to be analogous to imitation, the untrained animal would have to follow the demonstrator on the first trial. Even then, however, the motivation to affiliate could account for following behavior.

Observational conditioning. The observation of a performing demonstrator may not merely draw attention to the object being manipulated (e.g., the lever), but because the observer's orientation to the object is often followed immediately by presentation of food to the demonstrator, a Pavlovian association may be established. This form of conditioning has been called observational conditioning (Whiten & Ham, 1992), valence transformation (Hogan, 1988), or emulation (Tomasello, 1990) in which the observer learns the relation between some part of the environment and the reinforcer (e.g., that the top of a box can be removed to reveal what is inside). Although such conditioning would have to take the form of higher-order conditioning (because the observer would not actually experience the unconditional stimulus), there is evidence that such higher-order conditioning can occur, in the absence of a demonstrator. If, for example, the onset of a localizable light is followed shortly by the presentation of inaccessible grain, it is sufficient to produce pecking by pigeons to the light (Zentall & Hogan, 1975). The presence of a demonstrator drawing additional attention to the object to be manipulated (by pecking) and to the reinforcer (by eating) may further enhance associative processes in the absence of imitative learning.

With regard to the nature of the conditioning process, it is of interest that when reinforcement of the demonstrator's response cannot be observed (or the response-reinforcer association is difficult to make) acquisition may be impaired (Akins & Zentall, 1997; Groesbeck & Duerfeldt, 1971; Heyes, Jaldow, & Dawson, 1994). Furthermore, rats appear to acquire a bar pressing response faster following observation of a bar-pressing demonstrator if they are fed at the same time as the performing demonstrator (Del Russo, 1971). Although that result was mentioned earlier in the context of increased motivation on the part of the observer, it is also possible that feeding the observer following the demonstrator's response may result in simple Pavlovian conditioning (i.e.,the pairing of bar movement with food).

Observational conditioning may also play a role in an experiment in which observation of experienced demonstrators facilitated the opening of hickory nuts by red squirrels, relative to trial-and-error learning (Weigle & Hanson, 1980). Differential local enhancement can be ruled out, in this case, because animals in both groups quickly approached and handled the nuts, and the observers actually handled the nuts less than controls (perhaps because observers were more efficient at opening them). However, observers alone got to see the open nuts and they had the opportunity to associate open nuts with eating by the demonstrator.

Socially-transmitted food preferences (e.g., Galef, 1988a; Strupp & Levitsky, 1984) represent a special case of observational conditioning. Although food preference may appear to fall into the category of unlearned behavior subject to elicitation through contagion, consuming a food with a novel taste should be thought of as an acquired behavior. The mechanisms responsible for socially-acquired food preferences appear to have strong simple associative learning components (e.g., learned safety or the habituation of neophobia to the novel taste), for which the presence of a conspecific may serve as a catalyst. Furthermore, these specialized mechanisms may be unique to foraging and feeding systems.

One of the best examples of observational conditioning is in the acquisition of fear of snakes by laboratory-reared monkeys exposed to a wild-born conspecific in the presence of a snake (Mineka & Cook, 1988). Presumably, the fearful conspecific serves as the unconditioned stimulus, and the snake serves as the conditioned stimulus. It appears that exposure to a fearful conspecific or to a snake alone is insufficient to produce fear of snakes in the observer. For an excellent discussion of the various forms of observational conditioning see Heyes (1994).

Bird song. A special case of matching behavior by animals is the acquisition of bird song (Hinde, 1969; Marler, 1970; Nottebohm, 1970; Thorpe, 1961; see also vocal mimicry; e.g., Pepperberg, 1986; Thorpe, 1967). Although for many species of song bird the development of species typical song is regulated to a large extent by maturation and the seasonally fluctuating release of hormones, regional variations in the song appear to depend on the bird’s early experience with conspecifics (Baptista & Petrinovitch, 1984). Thus, one could say that young song birds learn their regional dialect by imitating the song of more mature conspecifics.

Acquisition of bird-song dialect is a special case of imitation for three reasons. First, although it is learned, bird song is a variation on a species typical behavior and thus, is relatively constrained. Second, according to Heyes (1994), in the acquisition of bird song, components of the matching behavior occur by chance and these components increase in frequency because they are intrinsically rewarding. Heyes refers to such behavior as copying rather than imitation. But finally, and most importantly, bird song takes place in the auditory modality. A characteristic of auditory events is that the stimulus produced by the demonstrator and that produced by the “observer” can be a close match, not only to a third party (i.e., the experimenter) but also to the observer. Thus, verbal behavior, for which comparisons between one's own behavior and that of others is relatively easy because one can hear one's own utterances with relative fidelity, may be a special “prepared” case of generalized, stimulus identity learning (in which animals that have been trained to match shapes can now use the principle of matching to match novel hue stimuli; see Zentall, Edwards, & Hogan, 1983).

Visual Matching. This analysis of the imitation of verbal behavior can also be applied to certain examples of visual imitation. Any behavior that produces a clear change in the environment, such that from the perspective of the observer there is a match between the stimulus produced by the demonstrator and that produced by the observer, may be a case of stimulus matching (e.g., observing someone turning up the volume of a radio - when the knob turns to the right, the volume increases). Such cases of visual stimulus matching can be distinguished from the perhaps more abstract and interesting case in which no visual stimulus match is possible (e.g., the imitation of a person who has his hands clasped behind his back).

Emulation of Affordances. When observation of a demonstrator allows an animal to learn how the environment works, a sophisticated form of learning is certainly involved. But one may not want to view it as imitation. For example, Bunyar and Huber (1999) allowed marmosets to observe conspecifics entering a food box through a door hinged at the top by pulling it towards themselves or pushing it away from themselves. Observers generally opened the door of the box in the way they had seen it demonstrated, but could they have learned through observation ‘how the door works’ rather than the movements required to open the door? Tomasello (1996) has referred to this kind of learning as the emulation of affordances because it may not require observation of an animal at all (see also Gibson, 1979). Would the marmosets have learned as much from having observed the door opening inward or outward by means of an invisible wire? If so, we would not want to call that learning imitation. Similarly, interpretation of the results of a number of other studies is made difficult by the fact that two different behaviors have two different effects on the environment observed. For example, Custance, Whiten, & Fredman, (1998) used “artificial fruit” to simulate the shell of fruit that must be removed by monkeys to gain access to the edible portion inside. In fact, demonstrators opened a latched clear plastic box in one of two distinctive ways. Observer monkeys showed a significant tendency to open the box by the same means as they observed it demonstrated.

Emulation of affordances can also account for findings of observation learning by Dawson and Foss (1965). They reported that budgerigars acquired a lid-removal task (by trial-and-error) in one of three different ways: Pushing the lid off with the beak, twisting it off with the beak, or grasping it with the foot and pulling it off. Observers were then exposed to these performing birds, and when they were then given the opportunity to perform themselves, each observer removed the lid in same manner as its demonstrator (see also, Galef, Manzig, & Field, 1986). But moving the lid in different ways may have provided information about different affordances. Similarly, Will, Pallaud, Soczka, and Manikowski (1974), noted a related effect in a study in which rats observed either a trained demonstrator performing a successive discrimination or an experimentally naive demonstrator. They found that the trained demonstrators typically responded with one of three distinctive patterns when the discriminative stimulus was available and that the observers learned not only to respond in the presence of one stimulus and not in the presence of the other, but they also learned the pattern of responding of their demonstrator (e.g., alternating a bar press with eating, or making a burst of bar presses followed by eating the accumulated pellets).

Heyes and Dawson (1990) have reported similar results by rats that observed demonstrators expressly trained to respond in one of two different ways. After observing demonstrators push an overhead bar either to the left or to the right, Heyes and Dawson found that observers given access to the bar tended to push the bar in the same direction as their demonstrator. Remarkably, the observers matched the demonstrators' behavior in spite of the fact that, because the observers faced the demonstrators during the period of observation, the direction of bar motion (relative to the observer's body) during observation was opposite that of the bar's motion when the observers performed. But in this experiment, too, the overhead bar moved in different directions (toward different walls of the demonstrator’s chamber) for the two observation groups. It is unlikely that the movement of the bar to a particular location was solely responsible for the reported effect, however, because Heyes, Jadlow and Dawson, (1994, Experiment 2) reported similar effects when the bar was moved between the time of observation and observer performance from the common wall between the two chambers to one of the side walls (i.e., a 90o shift in the direction of possible movement of the bar). However, it is also possible that olfactory cues, specific to the side of the bar against which the demonstrators pushed, was responsible for this imitation-like effect (Gardner & Heyes, 1998; Ray & Heyes, 1998).


True imitation has been defined as "the copying of a novel or otherwise improbable act or utterance, or some act for which there is clearly no instinctive tendency" (Thorpe, 1963, p. 135). The preceding analysis allows us to be more precise. First, one should control for motivational effects on the observer produced either by the mere presence of the demonstrator or by the mere consequences of the behavior of the demonstrator. Second, one should control for the possibility that the demonstrator's manipulation of an object merely draws the observer's attention to that object (or one like it), thus making the observer's manipulation of the object more probable. Third, one should control for the simple pairing of a novel stimulus (e.g., a lit response key or the movement of a bar) with the presentation of inaccessible food). And finally, for true imitation to be demonstrated, the target behavior should not already be part of the observing animal's repertoire (Clayton, 1978). In practice, however, it is not easy to determine the repertoire of an animal. In fact, one could argue that any behavior that an animal is capable of performing must be similar to a behavior that is already in the animals repertoire. Furthermore, ruling out the prior existence of a behavior requires the acceptance of 'never seen' as absence. Alternatively, one can define the response to be acquired in terms of the relatively low probability of the occurrence of the response in under similar conditions but in the absence of the opportunity to observe a demonstrator performing that response. Imitation is then defined as a relatively large increase in the probability of the demonstrated response, relative that of an appropriate group that controls for all of the already-noted non-imiative causes of such behavior.

The Two-Action Method and Related Variables

Click here to view treadle-pecking behavior videoIn a variation on the procedure used Dawson and Foss (1965), Akins and Zentall (1996) tried to overcome the problem ofClick here to view treadle-stepping behavior video differential environmental consequences by training quail to respond to a treadle for food either by pecking at the treadle or by stepping on the treadle. With a common manipulandum and the common movement of the manipulandum, the effect of the two response topographies on the environment should be common as well. Akins and Zentall found that observers showed a significant tendency to respond to the treadle with the same part of the body (beak or foot) as their respective demonstrator (see Click here to view Figure 6Zentall, Sutton, & Sherburne, 1996, for similar results with pigeons. Kaiser, Zentall, and Galef (1997) have found that a combination of trial-and-error learning to step together with contagious pecking was not sufficient to account for these two-action-method results.

Two important points should be made about this procedure. First, the environmental consequences of stepping and pecking were essentially the same (i.e., everything was the same except the two response topographies). And second, it is very unlikely that there was any similarity between the visual stimulus sen by the observer during observation and that seen by the observer during its own performance of the same response. Specifically, the demonstrator’s beak on the treadle must have appeared quite different to the observer from the observer’s own beak on the treadle. Similarly, though perhaps not so obviously, when the quail stepped on the treadle (located near the corner of the chamber) they pulled their head back and thrust their head forward. Thus, they could not see their foot making contact with the treadle. Once again, to the observer, the demonstrator’s response to the treadle must have appeared quite different from the observer’s own response to the treadle. Therefore, any account of the imitation found in these experiments in terms of stimulus matching is quite implausible.

Reinforcement of the demonstrated response. Earlier it was noted that the pairing of the movement of a manipulandum with a reinforcer, could increase the probability of the target behavior because of simple learning effects. Using the two-action method, the role of simple learning effects can be examined directly, rather than as an artifact to be avoided. Akins & Zentall (1998) have recently found that the correspondence between observer and demonstrator response topography disappears when the demonstrators responses are not reinforced.

A cognitive account of this finding is that through observation, the observer learns that there is no positive consequence associated with the demonstrator’s response and thus, there is no incentive for making the same response. This interpretation assumes that the lack of correspondence between observer and demonstrator response topography results from a performance decrement rather than from a learning decrement.

A simpler account of this finding suggests that with this preparation, an association between the demonstrator’s behavior and reinforcement is necessary for true imitation to occur. It does not provide an alternative account of imitative learning, however, because it cannot account for the correspondence between the observer’s and demonstrator’s response topographies. Instead, reinforcement may act as a catalyst to bring out imitative learning.

Motivation of the observer. If demonstrator reinforcement is necessary for the observer to learn through imitation, it suggests that observer incentive may play a role in imitative learning. If observers must be adequately motivated for them to imitate, it further suggests that the relevance of the demonstrator’s reinforcer to the state of the observer is also likely to be important. If this extrapolation from the data reported by Akins & Zentall (1998) is correct, food sated observers should be less inclined to learn a food-rewarded response through observation than hungry observers. Dorrance and Zentall (2001) tested this hypothesis by comparing imitative learning, using the two-action method, by quail that were either hungry or sated during observation. In support of the hypothesis, they found that hungry quail matched the demonstrator’s behavior if they had observed while hungry but not if they had observed while sated.

Proficiency of the demonstrator. One might also expect proficiency of the model to affect the rate of acquisition of the behavior by observing pigeons (Vanayan, Robertson, & Biederman, 1985). Contrary to expectation, however, Vanayan et al. found faster acquisition of a successive discrimination by observers when less proficient models were observed. It may be that observation of the consequences of incorrect (nonreinforced) responding is as important (or, in the case of aversively motivated learning perhaps, even more important) than observation of the consequences of correct (reinforced) responding. As mentioned earlier, however, observation of a discrimination being performed may result in stimulus enhancement and the demonstrator proficiency effects found may result from differential observational conditioning.

Deferred imitation. Bandura (1969) has indicated that there is an important difference between immediate imitation (which he calls imitation) and deferred imitation which he calls observational learning. For Bandura, immediate imitation may be a reflexive response that is genetically predisposed (akin to contagious behavior) whereas deferred imitation is indicative of a more cognitive process. Although it is possible that pecking and stepping behaviors could be transmitted via contagion, the fact that the demonstrated response and the observer’s performance do not occur at the same time makes that account unlikely to be correct. On the other hand, critics of this research might be more convinced if evidence could be provided for imitative learning that can be demonstrated with a significant delay between observation and observer performance. Such evidence has been reported by Dorrance and Zentall (2001). As part of a larger study, Dorrance and Zentall allowed hungry quail to observe either treadle pecking or treadle stepping. The quail were then returned to their home cage where they were fed and a half hour later were tested in the demonstration chamber. Dorrance and Zentall found clear evidence of imitation that was not distinguishable from that of observers that were tested immediately following observation. Thus, a half hour delay between observation and observer performance appears to have little effect on the expression of imitative learning by quail.

Additional Considerations

Enculturation. One of the variables that may play a role in imitative learning by primates appears to be the degree to which the animals have had extensive interactions with humans - what Tomasello (1990) refers to as enculturation. Enculturated chimpanzees and orangutans readily show signs of imitative learning (Tomasello , Gust, & Froat, 1989; Tomasello, Savage-Rumbaugh, & Kruger, 1993; Russon & Gladikas, 1993, 1995), whereas lab housed/reared chimpanzees typically do not (Whiten & Custance, 1996).

Enculturation may produce its effect in a number of ways. First, it could reduce the apes’ anxiety during test. Second, it could increases their attentiveness to social cues. Third, to could give them prior reinforced experience imitating (i.e., it could allow them to experience a form of learning to learn). A better understanding of the various components of enculturation might provide important insights into imitation by apes.

Gestural Imitation. A form of imitative learning conceptually related to the two-action method occurs when the gestures of a model are copied. Imitation of gestures has been found in chimpanzees (Custance, Whiten, & Bard, 1995, Hayes & Hayes, 1952), Dolphins (Harley, Xitco, Roitblat, & Herman, 1998; Xitco, Harley, & Brill, 1998), and a parrot (Moore, 1992). Remarkably, especially in the case of the dolphin and the parrot, the models were human rather than a conspecific. Thus, there was little similarity between corresponding body parts of the observer and the demonstrator. Because objects were not involved, local and stimulus enhancement should be irrelevant. Furthermore, each imitated gesture serves as a control for the others because it is the topography of the response that is important. In addition, the broad range of gestures that have been shown to be imitated within a few seconds of demonstration suggests that no account based on differential motivation is likely to play a role.

Generalized imitation. Imitation of a particular response can be thought of as one example of a broad class of imitative behavior. One can then ask if an animal can learn to match any behavior of another "on cue". (i.e., can an animal learn the general concept of imitation and then apply it when asked to do so in a "do-as-I-do" test). Hayes and Hayes (1952) found that a chimpanzee (Viki) learned to respond correctly to the command "Do this!" over a broad class of behavior. More recently, Custance, Whiten, and Bard (1995) have replicated this result under more highly controlled conditions. Furthermore, Custance and Bard (1994) using the “do as I do” procedure, have found that actions on parts of the body that cannot be seen by the performer were just as readily copied as those that could be seen. The importance of behavior that cannot be seen by the performer (e.g., touching the back of one’s head) is that it rules out the possibility that some form of visual stimulus matching might account for the behavioral match. The establishment of a “do as I do” concept not only verifies that chimpanzees can imitate, but it also demonstrates that they are capable of forming a generalized behavioral-matching concept (i.e., the chimpanzees have acquired an imitation concept).

Intentionality. Interest in imitation research can be traced, at least in part, to the assumption that true imitation involves some degree of intentionality. This is certainly the case in many of the higher order forms of imitation, such as the human dancer who repeats the movements of the teacher. Unfortunately, intentionality, because of its indirect nature, can only be inferred, and evidence for it appears most often in the form of anecdote rather than experiment. Ball (1938), for example, noted the case of a young rhesus monkey that, while kept with a kitten, was observed to lap its water in the same way as a cat. Ball noted further that lapping is extremely rare in rhesus monkeys. Similarly, Mitchell (1987), in an analysis of various levels of imitation, provides a number of examples of imitation at these higher levels. For example, he discusses the young female rhesus monkey who seeing her mother carrying a sibling, walks around carrying a coconut shell at a same location on her own body (Breuggeman, 1973). Such anecdotes, by their very nature, are selected and are difficult to verify. If there were some way to bring these examples of intentional imitation under experimental control, it would greatly increase their credibility.

Symbolic imitation. In the highest level of imitative behavior, what Mitchell (1987) refers to as fifth-level imitation, not only does the behavior of the observer not match that of the demonstrator, but the differences are explicit and they are produced for the purpose of drawing attention to certain characteristics of the model. Examples of such symbolic imitation can be found in the human use of parody and caricature. Such forms of imitation are mentioned primarily for completeness and to note the degree of subtlety that can be involved in imitation.

IV. Biological Function

Although imitation has been viewed by psychologists as purposive, intentional, and reflective, when imitation is viewed from the perspective of biological function, it can take on very different characteristics. These functional characteristics may account for the broad range of behavior that has been classified as imitative. For example, rather than viewing imitation as higher form of learning that requires conceptual ability, Boyd and Richerson (1988) have examined it in terms of its potential specific costs and benefits to an organism, as compared with two other evolutionary strategies: species typical behavior (primarily under genetic control) and individual (trial-and-error) learning (primarily under control of the direct consequences of the behavior to the organism).

The benefit of species typical behavior is its certainty. Birds do not have to rely on trial-and-error learning to build a nest. They are genetically predisposed or programmed to build them. On the other hand, there is a cost to such inflexible behavior. Should the environment change in a way that is inconsistent with an animal’s predispositions, there may not be enough flexibility in the system to allow for survival. The giant panda lives almost entirely on bamboo shoots. There has been little competition for this resource and thus, there has been little need for the panda to develop a varied diet. When bamboo is plentiful the panda can thrive. But with the encroachment of human populations the bamboo forests have shrunk in size and the panda has become an endangered species.

Trial-and-error learning allows an animal to adjust to a changing environment. For example, in an unpredictable environment where ideal foods may not be available or where others animals may compete for that food, an animal may be predisposed to follow a more general rule. In humans, the preference for sweet food is an example of such a rule. When such categorical rules are insufficient, more arbitrary rules based on individual experience (learning) may be necessary. For example, many animals (including humans) are predisposed to follow the rule, “If one encounters a novel taste, one should eat only a small amount; if one then gets sick, one should stay away from that taste; if one doesn’t get sick one should eat more” (see Garcia & Koelling, 1966). Thus, animals can learn which foods are good to eat and which are not.

But trial-and-error learning has its cost as well. For an animal to learn what not to eat, it must sometimes risk the negative consequences. Rats can learn what foods to eat by trial-and-error but such behavior can result in them getting poisoned. Learning can be beneficial, but it can also be costly. According to Boyd and Richerson (1988), social learning may serve an intermediate role between species typical behavior and individual trial-and-error learning. If an animal learns from watching another animal, it can benefit from the trial-and-error learning of the other without having to suffer the consequences of errors. For example, if a rat encounters two novel flavors of food, it will prefer the flavor that it sees another rat eat (Galef, 1988a). Although the mechanism by which food preference are acquired socially is much simpler than true imitation, imitative learning may have evolved for similar reasons - it allows the experience of one individual to be passed on to others more efficiently than would be the case with trial-and-error learning, yet retaining much of the flexibility of individual learning.

In human societies, culture and tradition are the means of passing on the experiences of group members to the benefit of other members of the group. Individual learning can last a lifetime, social learning can be passed on for many generations, but it allows for considerably more flexibility that does genetically predisposed behavior.

A more functional, inductive approach to the study of imitative learning may also provide insights into the conditions under which it would be most likely to occur (see Davis, 1973; Howard & Keenan, 1993). The relatively asocial laboratory rat or pigeon used in psychological research may not be most appropriate subject for the study of imitation. Instead, animals with better visual acuity than the rat and which are more precocious and social than the pigeon may be more likely to imitate (e.g., the Japanese quail, Akins & Zentall, 1996).

Although we have argued that animals that engage in social learning are likely to benefit from it by way of increased fitness over those than engage in only species typical behavior and individual learning, not all species may benefit from imitative learning. Even highly social species that are capable of considerable behavioral plasticity (in the form of individual learning), such as monkeys, may not benefit from imitation because of the nature of resources in their environment. For example, they may be more likely to find and eat novel food in the presence of others who are eating (Fragazy & Visalberghi, 1996) but they may not benefit from observing how others find and eat novel food. Thus, although imitative learning appears to be maximized in highly social species, it may not be an inevitable consequence of living in social groups. Imitation appears to be widely scattered among species, with humans and great apes being the most prolific imitators, but dolphins and a number of avian species including parrots, pigeons, and Japanese quail show evidence of imitative learning as well. Imitation by a number of bird species, together with the relative absence of imitation in monkeys (Fragazy & Visalberghi, 1989, 1990; Whiten & Ham, 1992; but see also Custance et al., 1998) suggests that a high degree of behavioral plasticity and sociality may be neither necessary nor sufficient for the development of imitative learning.

V. Psychological Mechanisms

There has been much speculation about the meaning of true imitative learning. Some have described it as a reflective (conscious?) process ( Morgan, 1900) or as “the purposive, goal-directed copying of the behavior of one animal by another” (Galef, 1988b, p. 21), whereas others see it as an extension of simple learning principles (Gewirtz, 1969).

Associative Learning Accounts

The simplest account of imitation has been provided by social learning theorists (Bandura, 1969; Gewirtz, 1969; Miller & Dollard, 1941). For most social learning theorists, imitation can be explained as a special case of simple instrumental learning. Gewirtz, for example, proposed that initially, when a demonstrator (or model) engages in a particular behavior, the young observer responds in a variety of ways that are unrelated to the behavior being modeled. Occasionally, and only by chance, a correspondence might occur between the behavior of the model and that of the observer. According to Gewirtz, those instances of behavioral correspondence are typically accompanied by reinforcement. For example, a parent may say “daddy” many times to an young child and on those occasions on which it happens to be followed by “dada” spoken by the child, positive reinforcement (perhaps socially, through the parent’s excitement and attention) will often be provided. The word "daddy" then comes to serve as a conditioned stimulus that signals the opportunity to obtain reinforcement for emitting the response "dada." Thus, conditioning theory can explain individual cases of response copying, especially when verbal behavior is involved. But every imitated word does not go through such a processes of reinforcement by successive approximation (i.e., trial-and-error shaping).

To account for the extensive use of imitative learning by children, Gewirtz (1969) proposed that copied responses that occur initially through selective reinforcement, come to generalize to other behavior, without the need for additional consistent reinforcement. If generalization, as it is used here, is meant to be explanatory, rather than merely descriptive, however, it requires a more complex mechanism than simple associative learning can provide.Stimulus generalization theory (Spence, 1937) is based on the principle of physical stimulus similarity. Specifically, a reinforced response in the presence of a particular stimulus will tend to occur in the presence of other stimuli, to the extent that those other stimuli are physically similar to the training stimulus. But when applied to imitation, how does an infant generalize from repeating the word "daddy" to repeating the quite different sounding word "ball"? Furthermore, the concept of generalization refers to the probability of occurrence of the trained response, rather than a matching response. Thus, according to such a conditioning model, the response “daddy” should occur to other stimuli to the extent that they sound like “daddy.”

Alternatively, generalized imitation may be related conceptually to identity learning with visual stimuli, but in the case of imitation, it is the matching of behavior rather than stimuli. When Baer, Peterson, and Sherman (1967) trained retarded children to match several behaviors of a model when instructed to, “Do this,” they found that the children continued to match in the absence of reinforcement. They proposed that the children had formed a functional stimulus class defined by the correspondence between the stimulus output of the child’s behavior and the stimulus output of the model’s behavior. Thus, to account for such stimulus/response matching, a child must, at a minimum, have a concept of identity (i.e., the child must understand what it means that two things are the same, see Zentall, Edwards, & Hogan, 1983) or what Gewirtz (1969) calls a matching-response class. Such an analysis implies processes that go beyond simple learned associations and their generalization. But children do show evidence of identity learning at a very early age (Tyrrell, Zingaro, & Minard, 1993) and pigeons too show the capacity for identity learning (Zentall, Edwards, Moore, & Hogan, 1981).

The copying of verbal behavior may be explained in this way because one can hear one's own utterances with relative fidelity, and one can compare them directly with those of a model (i.e., stimulus matching). However, such matching of response-produced stimuli to target stimuli cannot account for imitation by a young child when an adult model says, "Do this," as the model places his hands over his eyes. In this type of imitation, from the perspective of the observer, there is no match between the stimulus provided by the behavior of the model and that provided by the observer’s own behavior.

Cognitive Accounts

According to Piaget (1955), true imitation involves sensory-motor assimilation. It is the coordination of first, the sensory-motor system of the individual (or the self), which occurs at an early age and then, an appreciation of the similarities of between the individual and others. In Piaget’s view, this process occurs first through the similarity between the seen body parts of others and the corresponding seen body parts of child. Later, an acquired cross-modal matching process allows the child to understand the correspondence between its own unseen body parts (e.g., the eyes, nose, and head) and those of others. This cross-modal matching process is based on (1) touch (the felt parts of self and others), (2) the correspondence of touch and sight (the felt and seen body parts of others), and (3) the inference that because one’s own unseen body parts feel like those of others, they must look like those of others. Finally, this assimilation results in a schema of the individual (i.e., an image, in our mind’s eye, of ourselves). In explaining the development of imitation in children, Guillaume (1971) views imitation as linked to the child’s notion of self and he proposes that “imitation enables the child to see himself in the person of another” p. 207. Thus, according to this view, the mechanism that makes imitation possible is the ability to take the perspective of another.

However, if imitative learning occurs in species as varied as rats, pigeons, and Japanese quail, as it appears to, the responsible mechanism is not likely to be theory of mind or perspective taking. But in cases in which stimulus matching is inadequate to account for imitation, some precursor of perspective taking is likely to be involved.

But how does a pigeon infer the similarity between its own beak (seen only as a gross distortion) and the beak of a demonstrator? Such an inferential process would seem to be beyond the capacity of rats, pigeons, and quail.

VI. Biological Mechanisms

Alternatively, Metzoff (1996) has proposed that in the case of human infants, there is an “inbuilt drive to ‘act like’ their conspecifics (p. 363). Metzoff bases his conclusion on data suggesting that infants (and even newborns, Metzoff & Moore, 1989; Reissland, 1988) imitate a wide range of adult demonstrated gestures, including lip, cheek, brow, head, and finger movements, as well as emotional expression. Although Jones (1996) suggested that early research on infant imitation involving tongue protrusion may be accounted for more parsimoneously in terms of very early attempts at the oral exploration of objects, the range of imitated gestures as well as the number of independent reports of such imitation (Metzoff, 1996) suggest that these effects cannot easily be explained away.

The implications of infant imitation are important because if true imitation can occur in newborns, it suggests that the mechanisms responsible for imitation are probably not cognitively based. Clearly, even the most rudimentary cognitive structures involved in perspective taking would not have had time to develop in newborns.

The data suggest that infants are born with the ability to engage in “a matching-to-target process in which they actively compare the visual information about the seen body movements [of the adult] with the proprioceptive feedback from their own movements in space” Metzoff (1996, p. 351). Such innate cross-modal matching must be quite different from Piaget’s experienced-based process. Metzoff’s data suggest that infants do not have to learn the correspondence between the behavior of others and their own, they just appear to do so reflexively. According to this view, infants are “prewired” to imitate the behavior of conspecifics. But the mechanism cannot be so general. Generalized imitation is a category of matching behavior that is defined by the third party (e.g., the experimenter). To the imitator, there is no match, especially if learning is not involved. Thus if imitation (in humans at least) is an innate response, then each demonstrated behavior that is imitated (e.g., tongue protrusion and brow movement) must be a releaser for the same behavior in the infant. Given the wide range of imitated behavior, the list of releasers must be quite long.

It is difficult to imagine the evolution of such an elaborate set of releasers to account for imitation by humans, however, in species for which a more cognitive perspective-taking account seems even less probable (e.g., quail and pigeons) the existence of such a set of predisposed releasers definitely may be involved.

VII. Conclusions

Procedures have now been developed which are capable of separating true imitative learning from other social influences on behavior. Early results indicate that imitative learning can be found in a variety of species. Such findings should not be surprising because social learning, by imitation and otherwise, provides clear benefits to many organisms over genetically based behavior and trial-and-error learning. The mechanism by which animals are able to match their behavior to that of a demonstrator may involve some form of coordination of visual and tactile sensory modalities, and in some species, such coordination may be predisposed. However, a more complete account of these processes will have to await research to determine the necessary and sufficient conditions for obtaining the various forms of imitative learning in animals.

avcrule.gif (935 bytes)

VIII. References

       Akins, C. K., & Zentall, T. R. (1996). Imitative learning in male Japanese quail (Coturnix japonica) using the two-action method. Journal of Comparative Psychology, 110, 316-320.

     Akins, C. K., & Zentall, T. R. (1998). Imitation in Japanese quail: The role of reinforcement of the demonstrator’s response. Psychonomic Bulletin & Review, 5, 694-697.

     Armstrong, E. A. (1951). The nature and function of animal mimesis. Bulletin of Animal Behaviour, 9, 46-48.

     Baer, D. M., Peterson, R. F., & Sherman, J. A. (1967). The development of imitation by reinforcing behavioral similarity to a model. Journal of the Experimental Analysis of Behavior, 10, 405-416.

     Ball, J. (1938). A case of apparent imitation in a monkey. Journal of Genetic Psychology, 52, 439-442.

     Bandura, A. (1969). Social learning theory of identificatory processes. In D. A. Goslin (Ed.). Handbook of socialization theory and research (pp. 213-262). Chicago, IL: Rand-McNally.

     Baptista, L. F., & Petrinovich, L. (1984). Social interaction, sensitive phases, and the song template hypothesis in the white-crowned sparrow. Animal Behaviour, 32, 172-181.

     Bayroff, A. G., & Lard, K. E. (1941). Experimental social behavior of animals. III. Imitational learning of white rats. Journal of Comparative and Physiological Psychology, 37, 165-171.

     Boyd, R. & Richerson, P. J. (1988). An evolutionary model of social learning: The effect of spatial and temporal variation. In T. R. Zentall & B. G. Galef, Jr. (Eds.), Social learning: Psychological and biological perspectives (pp. 29-48). Hillsdale, NJ: Erlbaum.

     Breuggeman, J. A. (1973). Parental care in a group of free-ranging rhesus monkeys. Folia Primatologica, 20, 178-210.

     Bunch, G. B., & Zentall, T. R. (1980). Imitation of a passive avoidance response in the rat. Bulletin of the Psychonomic Society, 15, 73-75.

     Byrne (1994). The evolution of intelligence. In P. J. B. Slater & T. R. Halliday (Eds.), Behavior and evolution (pp. 223-265). Cambridge, MA: Cambridge University Press.

     Chesler, P. (1969). Maternal influence in learning by observation in kittens. Science, 166, 901-903.

     Church, R. M. (1957). Two procedures for the establishment of imitative behavior. Journal of Comparative and Physiological Psychology, 50, 315-318.

     Clayton, D. A. (1978). Socially facilitated behavior. Quarterly Review of Biology, 53, 373-391.

     Corson, J. A. (1967). Observational learning of a lever pressing response. Psychonomic Sciences, 7, 197-198.

     Custance, D. M., & Bard, K. A. (1994). The comparative and developmental study of self-recognition and imitation: the importance of social factors. In S. T. Parkar, R. W. Mitchell, & M. L. Boccia (Eds.), Self-awareness in humans and animals: developmental perspective (pp. 207-226). Cambridge, MA: Cambridge University Press.

     Custance, D. M., Whiten, A., & Bard, K. A. (1995). Can young chimpanzees imitate arbitrary actions? Hayes and Hayes revisited. Behaviour, 132, 839-858.

     Custance, D. M., Whiten, A., & Fredman, T. (1998, June). Social learning of "artificial fruit" processing in capuchin monkeys (Cebus apella). Paper presented at the Napoli Social Learning Conference, Naples, Italy.

     Davis, J. M. (1973). Imitation: A review and critique. In P. P. G. Bateson, & P. H. Klopfer, (Eds.), Perspectives in ethology: Vol. 1 (pp. 43-72). New York, NY: Plenum.

     Davitz, J. R. & Mason, D. J. (1955). Socially facilitated reduction of a fear response in rats. Journal of Comparative and Physiological Psychology, 48, 149-151.

     Dawson, B. V. & Foss, B. M. (1965). Observational learning in budgerigars. Animal Behaviour, 13, 470-474.

     Del Russo, J. E. (1971). Observational learning in hooded rats. Psychonomic Science, 24, 37-45.

     Del Russo, J. E. (1975). Observational learning of discriminative avoidance in hooded rats. Animal Learning and Behavior, 3, 76-80.

     Denny, M. R., Clos, C. F., & Bell, R. C. (1988). Learning in the rat of a choice response by observation of S-S contingencies. In T. R. Zentall & B. G. Galef, Jr. (Eds.), Social learning: Psychological and biological perspectives (pp. 207-223). Hillsdale, NJ: Erlbaum.

     Dugatkin, L. A. (1992). Sexual selection and imitation: females copy the mate choice of others. American Nature, 139, 1384-1389.

     Dugatkin, L. A. (1996). Copying and mate choice. In C. M. Heyes & B. G. Galef (Eds.), Social learning in animals: The roots of culture (pp. 85-105). San Diego, CA: Academic Press.

     Dugatkin, L. A., Godin, J.-G. J. (1992). Reversal of female mate choice by copying. Proceedings of the Royal Society of London: Series B, 249, 179-184.

     Edwards, C. A., Hogan, D. E., Zentall, T. R. (1980). Imitation of an appetitive discriminatory task by pigeons. Bird Behaviour, 2, 87-91.

     Fiorito, G., & Scotto, P. (1992). Observational learning in Octopus vulgaris. Science, 256, 545-546.

     Fisher, J. & Hinde, R. A. (1949). Further observations on the opening of milk bottles by birds. British Birds, 42, 347-357.

     Fragasy, D. M., & Visalberghi, E. (1989). Social influences on the acquisition of tool using behavior in capuchin monkeys (Cebus apella). Journal of Comparative Psychology, 103, 159-170.

     Fragasy, D. M., & Visalberghi, E. (1990). Social processes affecting the appearance of innovative behaviors in capuchin monkeys. Folia Primatologica, 54, 155-165.

     Fragasy, D. M., & Visalberghi, E. (1996). Social learning in monkeys: Primate "primacy" reconsidered. In C. M. Heyes & B. G. Galef (Eds.), Social learning in animals: The roots of culture (pp. 65-84). San Diego, CA: Academic Press.

     Freud, S. (1933). New introductory lectures on psycho-analysis. New York, NY: Norton.

     Galef, B. J., Jr. (1988a). Communication of information concerning distant diets in a social, central-place foraging species: Rattus norvegicus. In T. R. Zentall & B. G. Galef, Jr. (Eds.), Social learning: Psychological and biological perspectives (pp. 119- 139). Hillsdale, NJ: Erlbaum.

     Galef, B. J., Jr. (1988b). Imitation in animals: History, definition, and interpretation of data from the psychological laboratory. In T. R. Zentall & B. G. Galef, Jr. (Eds.), Social learning: Psychological and biological perspectives (pp. 3-28). Hillsdale, NJ: Erlbaum.

      Galef, B. J., Jr., Manzig, L. A., & Field, R. M. (1986). Imitation learning in budgerigars: Dawson and Foss (1965) revisited. Behavioral Processes, 13, 191-202.

     Garcia, J., & Koelling, R. A. (1966). Relation of cue to consequence in avoidance learning. Psychonomic Science, 4, 123-124.

     Gardner, E. L., & Engel, D. R. (1971). Imitational and social facilitatory aspects of observational learning in the laboratory rat. Psychonomic Science, 25, 5-6.

     Gardner, M., & Heyes, C. M. (1998, June). Limitations of the bidirectional control procedure. Paper presented at the Napoli Social Learning Conference, Naples, Italy.

     Gewirtz, J. L. (1969). Mechanisms of social learning: Some roles of stimulation and behavior in early human development. In D. A. Goslin (Ed.). Handbook of socialization theory and research (pp. 57-211). Chicago, IL: Rand-McNally.

     Gibson, J. (1979). The ecological approach to visual perception. Boston, MA: Houghton Mifflin.

     Groesbeck,R. W., & Duerfeldt, P. H. (1971). Some relevant variables in observational learning of the rat. Psychonomic Science, 22, 41-43.

     Guillaume, P. (1971). Imitation in children (E. P. Halperin, Trans.). Chicago: University of Chicago Press.

     Harley, H. E., Xitco, M. J., Jr., Roitblat, H. L., & Herman, L. M. (1998, June). Imitation of human models by bottlenose dolphins. Paper presented at the Napoli Social Learning Conference, Naples, Italy.

     Haruki, Y., & Tsuzuki, T. (1967). Learning of imitation and learning through imitation in the white rat. Annual of Animal Psychology, 17, 57-63.

     Hayes, K. J., & Hayes, C. (1952). Imitation in a home-reared chimpanzee. Journal of Comparative and Physiological Psychology. 45, 450-459.

     Herbert, M. J., & Harsh, C. M. (1944). Observational learning by cats. Journal of Comparative Psychology, 37, 81-95.

     Hess, E. H. (1973). Imprinting. New York, NY: Van Nostrand Reinhold.

     Heyes, C. M. (1993). Imitation, culture, and cognition. Animal Behavior, 46, 999-1010.

    Heyes, C. M. (1994). Reflections on self-recognition in primates. Animal Behavior, 47, 909-919.

     Heyes, C. M., & Dawson, G. R. (1990). A demonstration of observational learning in rats using a bidirectional control. Quarterly Journal of Experimental Psychology, 42B, 59-71.

     Heyes, C. M., Jaldow, E., & Dawson, G. R. (1994). Imitation in rats: Conditions of occurrence in a bidirectional control procedure. Learning and Motivation, 25, 276-287.

     Hinde, R. A. (Ed.). (1969). Bird vocalizations. Cambridge, England: Cambridge University Press.

     Hogan, D. E. (1988). Learned imitation by pigeons. In T. R. Zentall & B. G. Galef, Jr. (Eds.), Social learning: Psychological and biological perspectives (pp. 225-238). Hillsdale, NJ: Erlbaum.

     Hoogland, J. L., & Sherman, P. W. (1976). Advantages and disadvantages of bank swallow (Riparia riparia) coloniality. Ecological Monographs, 46, 33-58.

     Howard, M. L., & Keenan, M. (1993). Outline for a functional analysis of imitation in animals. Psychological Record, 43, 185-204.

     Hull, C. L., (1943). Principles of behavior. New York, NY: Appleton.

     Jacoby, K. E., & Dawson, M. E. (1969). Observation and shaping learning: A comparison using Long-Evans rats. Psychonomic Science, 16, 257-258.

     John, E. R., Chesler, P., Bartlett, F., & Victor, I. (1968). Observational learning in cats. Science, 159, 1489-1491.

     Jones, S. S. (1996). Imitation or exploration: Young infants’ matching of adults’ oral gestures. Child Development, 67, 1952-1969.

     Kaiser, D. H., Zentall, T. R., & Galef, B. G., Jr. (in press). Can imitation in pigeons be explained by local enhancement together with trial-and-error learning? Psychological Science.

     Kohn, B. (1976). Observation and discrimination learning in the rat: Effects of stimulus substitution. Learning and Motivation, 7, 303-312.

     Kohn, B., & Dennis, M. (1972). Observation and discrimination learning in the rat: Specific and nonspecific effects. Journal of Comparative and Physiological Psychology, 78, 292-296.

     Kovach, J. K. & Hess, E. H. (1963). Imprinting: Effects of painful stimulation upon the following response. Journal of Comparative and Physiological Psychology, 56, 461-464.

     Lefebvre, L. & Palameta, B. (1988). Mechanisms, ecology, and population diffusion of socially learned food-finding behavior in feral pigeons. In T. R. Zentall & B. G. Galef, Jr. (Eds.), Social learning: Psychological and biological perspectives (pp. 141-164). Hillsdale, NJ: Erlbaum.

     Levine, J. M. & Zentall, T. R. (1974). Effect of conspecific's presence on deprived rats performance: Social facilitation vs. distraction/imitation. Animal Learning and Behavior, 2, 119-122.

     Lore, R., Blanc, A., & Suedfeld, P. (1971). Empathic learning of a passive avoidance response in domesticated Rattus norvegicus. Animal Behaviour, 19, 112-114.

     Lorenz, K. (1935). Der kumpanin der umvelt des vogels: die artgenosse als ausloesendesmoment socialer verhaltensweisen. Journal fur Ornithologie, 83, 137-213, 289-413.

     Marler, P. (1970). A comparative approach to vocal learning: Song development in white-crowned sparrows. Journal of Comparative and Physiological Psychology, 71, 1-25.

     Meltzoff, A. N. (1996). The human infant as imitative generalist: A 20-year progress report on infant imitation with implications for comparative psychology. In C. M. Heyes & B. G. Galef (Eds.), Social learning in animals: The roots of culture (pp. 347-370). San Diego, CA: Academic Press.

     Meltzoff, A. N., & Moore, M. K. (1989). Imitation in newborn infants: Exploring the range of gestures imitated and the underlying mechanisms. Developmental Psychology, 25, 954-962.

     Miller, N. E., & Dollard, J. (1941). Social learning and imitation. New Haven, CT: Yale University Press.

     Mineka, S. & Cook, M. (1988). Social learning and the acquisition of snake fear in monkeys. In T. R. Zentall & B. G. Galef, Jr. (Eds.), Social learning: Psychological and biological perspectives (pp. 51-75). Hillsdale, NJ: Erlbaum.

     Mitchell, R. W. (1987). A comparative-developmental approach to understanding imitation. In P. P. G. Bateson & P. H. Klopfer (Eds.). Perspectives in ethology: Vol. 7  (pp. 183-215). New York, NY: Plenum.

     Moore, B. R. (1992). Avian movement imitation and a new form of mimicry: Tracing the evolution of a complex form of learning. Behaviour, 122, 231-263.

     Morgan,C. L. (1900). Animal behavior. London, U. K.: Arnold.

     Morrison, B. J., & Hill, W. F. (1967). Socially facilitated reduction of the fear response in rats raised in groups or in isolation. Journal of Comparative and Physiological Psychology, 63, 71-76.

     Nottebohm, F. (1970). Ontogeny of bird song. Science, 167, 950-956.

     Oldfield-Box, H. (1970). Comments on two preliminary studies of "observation" learning in the rat. Journal of Genetic Psychology, 116, 45-51.

     Pepperberg, I. M. (1986). Acquisition of anomalous communicatory systems: Implications for studies on interspecies communication. In R. Schusterman, J. Thomas, & F. Wood, (Eds.). Dolphin behavior and cognition: Comparative and ethological aspects (pp. 289-302). Hillsdale, NJ: Erlbaum.

     Piaget, J. (1955). The child’s construction of reality. London: Routledge.

     Powell, R. W. (1968). Observational learning vs. shaping: A replication. Psychonomic Science, 10, 263-364.

     Powell, R. W., & Burns, R. (1970). Visual factors in observational learning with rats. Psychonomic Science, 21, 47-48.

     Powell, R. W., Saunders, D., & Thompson, W. (1969). Shaping, autoshaping, and observational learning with rats. Psychonomic Science, 13, 167-168.

     Ray, E. D., & Heyes, C. M. (1998, June). Implications of a contemporary learning theory for social learning research. Paper presented at the Napoli Social Learning Conference, Naples, Italy.

     Reissland, N. (1988). Neonatal imitation in the first hour of life: Observations in rural Nepal. Development Psychology, 24, 464-469.

     Roberts, D. (1941). Imitation and suggestion in animals. Bulletin of Animal Behaviour, 1, 11-19.

     Russon, A. E., & Galdikas, B. M. F. (1993). Imitation in free-ranging rehabilitant orangutans. Journal of Comparative Psychology, 107, 147-161.

     Russon, A. E., & Galdikas, B. M. F. (1995). Constraints on great apes’ imitation: Model and action selectivity in rehabilitant orangutan (Pongo pygmaeus) imitation. Journal of Comparative Psychology, 109, 5-17.

     Sanavio, E., & Savardi, U. (1980). Observational learning of a discriminative shuttlebox avoidance by rats. Psychological Reports, 44, 1151-1154.

     Sordahl, T. A. (1981). Slight of wing. Natural History, 90, 42-49.

     Spence, K. W. (1937). The differential response in animals to stimuli varying within a single dimension. Psychological Review, 44, 430-444.

     Stimbert, V. E. (1970). A comparison of learning based on social and nonsocial discriminative stimuli. Psychonomic Science, 20, 185-186.

     Strupp, B. J., & Levitsky, D. A. (1984). Social transmission of food preferences in adult hooded rats (Rattus norvegicus). Journal of Comparative Psychology, 98, 257-266.

     Thorpe, W. H. (1961). Bird song: The biology of vocal communication and expression in birds. Cambridge, MA: Harvard University Press.

     Thorpe, W. H. (1963). Learning and instinct in animals (2nd ed.). Cambridge, MA: Harvard University Press.

     Thorpe, W. H. (1967). Vocal imitation and antiphonal song and its implications. In D. W. Snow (Ed.). Proceedings of the XVI International Ornithological Congress (pp. 245-263). Oxford, U. K.: Blackwell.

     Tinbergen, N. (1960). The herring gull’s world. Garden City, NY: Doubleday.

     Tolman, C. W. (1964). Social facilitation of feeding behaviour in the domestic chick. Animal Behaviour, 12, 245-251.

     Tomasello, M. (1990). Cultural transmission in the tool use and communicatory signaling of chimpanzees? In S. Parker & K. Gibson (Eds.), "Language" and intelligence in monkeys and apes: Comparative developmental perspectives  (pp. 271-311). Cambridge, UK: Cambridge University Press.

     Tomasello, M., Gust, D., & Frost, T. (1989). A longitudinal investigation of gestural communication in young chimpanzees. Primates, 30, 35-50.

     Tomasello, M., Savage-Rumbaugh, S., & Kruger, A. C. (1993). Imitative learning of actions on objects by children, chimpanzees, and enculturated chimpanzees. Child Development, 64, 1688-1705.

     Turner, J. R. G. (1984). Mimicry: The palatability spectrum and its consequences. In R. I. Vane-Wright & P. R. Ackery (Eds.), The biology of butterflies (pp. 141-161). New York, NY: Academic Press.

     Tyrrell, D. J., Zingaro, M. C., & Minard, K. L. (1993). Learning and transfer of identity-difference relationships by infants. Infant Behavior and Development, 16, 43-52.

     Vanayan, M., Robertson, H., & Biederman, G. B. (1985). Observational learning in pigeons: The effects of model proficiency on observer performance. Journal of General Psychology, 112, 349-357.

     Warden, C. J., & Jackson, T. A. (1935). Imitative behavior in the rhesus monkey. Journal of Genetic Psychology, 46, 103-125.

     Weigle, P. D., & Hanson, E. V. (1980). Observation learning and the feeding behavior of the red squirrel (Tamiasciurus hudsonicus): The ontogeny of optimization. Ecology, 61, 213-218.

     Whiten, A., & Custance, D. (1996). Studies of imitation in chimpanzees and children. In C. M. Heyes & B. G. Galef (Eds.), Social learning in animals: The roots of culture (pp. 291-318). San Diego, CA: Academic Press.

     Whiten, A., & Ham, R. (1992). On the nature and evolution of imitation in the animal kingdom: Reappraisal of a century of research. In P. J. B. Slater, J. S. Rosenblatt, C. Beer, & M. Milinski (Eds.), Advances in the Study of Behavior: Vol. 21 (pp. 239-283). New York, NY: Academic Press.

     Will, B., Pallaud, B., Soczka, M., & Manikowski, S. (1974). Imitation of lever pressing "strategies" during the operant conditioning of albino rats. Animal Behaviour, 22, 664-671.

     Xitco, M. J., Jr., Harley, H. E., & Brill, R. L. (1998, June). Action level imitation by bottlenose dolphins. Paper presented at the Napoli Social Learning Conference, Naples, Italy.

     Zajonc, R. B. (1965). Social facilitation. Science, 149, 269-274.

     Zentall, T. R., Edwards, C. A., & Hogan, D. E. (1983) Pigeons' use of identity. In M. L. Commons, R. J. Herrnstein, & A. Wagner (Eds.), The quantitative analyses of behavior: Vol. 4. Discrimination processes (pp. 273-293). Cambridge, MA: Ballinger.

     Zentall, T. R., Edwards, C. A., Moore, B. S., & Hogan, D. E. (1981). Identity: The basis for both matching and oddity learning in pigeons. Journal of Experimental Psychology: Animal Behavior Processes, 7, 70-86.

     Zentall, T. R. & Hogan, D. E. (1975). Key pecking in pigeons produced by pairing key light with inaccessible grain. Journal of the Experimental Analysis of Behavior, 23, 199-206.

     Zentall, T. R. & Levine, J. M. (1972). Observational learning and social facilitation in the rat. Science, 178, 1220-1221.

     Zentall, T. R., Sutton, J. and Sherburne, L. M. (1996). True imitative learning in pigeons. Psychological Science, 7, 343-346.


Preparation of this article was supported by Grant MH 55118 from the National Institute of Mental Health and by Grant IBN 9414589 from the National Science Foundation.