The words “mimicry” and “imitation” often have a negative connotation in English of being an unoriginal fake. A mimic is often an annoying copy cat, while an imitation can be a second-rate copy of a more valuable original. However, mimicry and imitation are based on special cognitive abilities that are rare among animals and that form the basis of culture. Humans learn most cultural traits—from the words in our language to the way we prepare food or hold a tool— through observational learning and imitation. We must learn thousands of these cultural traits through imitation before we can make an original contribution to our culture.
Imitation is a form of social learning—it requires an animal to observe a “demonstrator” performing a behavior and then to be able to perform that behavior itself (Galef, 1988; Whiten and Ham, 1992). Many psychologists distinguish between vocal learning, in which an animal modifies the sounds it produces based on the sounds it hears, and motor imitation, which involves an animal watching a posture or movement of another animal and then copying that movement—”monkey see, monkey do.” Some also distinguish between motor imitation of something like clapping, where the animal can watch its hands in the same wav that it watches the demonstrator, versus facial gestures or whole body movement, where the actor cannot receive sensory input about its own performance that directly parallels the observation of the demonstrator.
There is strong evidence for vocal learning and imitation among a variety of animals, including marine mammals and birds (Janik and Slater, 1997). Perhaps the simplest evidence involves vocal mimicry where an animal demonstrates the ability to produce sounds after exposure to model sounds that were not part of its pre-exposure repertoire. Animals such as parrots (Todt. 1975; Pepperberg, 2000). starlings (West et al, 1983), and a harbor seal (Phoca vitulina) named Hoover (Ralls et al, 1985) have demonstrated abilities to imitate the sounds of human speech. Other animals such as dolphins have been trained to imitate acoustic features of artificial sounds (Richards et al, 1984).
Humpback whales, Megaptera novaeangliae, have not been kept in captivity where one could most easily study imitation of manmade sounds, but their songs have a structure that cannot be explained by any mechanism other than vocal imitation. At any one time, most singing humpback whales within a population sing songs that are similar (Tyack, 1999). These songs change week by week, month bv month, and year by year, and individual whales have been shown to track these changes. Humpback songs have been recorded for decades, with no suggestion that the songs repeat. This suggests that humpback whales learn to produce the current song and to track the progressive changes that make up such a distinctive feature of this signal.
Vocal imitation has also been reported for the natural sounds of bottlenose dolphins (Tursiops truncatus). Most dolphins studied in either captive or wild settings develop an individually distinctive signature whistle. Dolphins can imitate the whistles of social partners (Tyack, 1986). Three male Indian Ocean bottlenose dolphins (T. adunctis) that formed a strong social bond were reported to modify their whistles over 3 years as the bond formed, such that all three converged on a shared whistle (Smolker and Pepper, 1999). The functions of this imitation are currentlv not known, but similar imitation has been suggested to function as a name for reference (Tyack, 1999). as a threat (Janik and Slater. 1997). or as an affiliative signal (Smolker and Pepper, 1999).
It has been controversial whether animals can perform motor imitation (Galef, 1988). This stems in large part from the difficult)’ of proving that a display of which the demonstrator was capable was not part of the pre-exposure repertoire of the animal. Ethologists expect that many animal displays represent fixed action patterns that are inherited. If an action pattern is simply triggered by sensing a conspecific performing the same action, that does not demonstrate observational learning. There are many anecdotes about motor imitation in marine mammals that are difficult to explain via any mechanism other than observational learning. Tayler and Saayman (1973) provided some of the most interesting examples. They reported captive bottlenose dolphins swimming with postures and motor patterns similar to those of seals, turtles, fishes, and penguins that were housed in the same pool. These postures and swimming patterns are so awkward and different from normal dolphin locomotion that it is scarcely credible that they represent anything other than learned behaviors. The most striking example of imitation involved a calf Indian Ocean bottlenose dolphin that observed through an underwater window a human blowing out a cloud of cigarette smoke. The calf swam over to its mother, suckled, swam back to the window, and expelled a mouthful of milk into a cloud that looked similar to exhaled tobacco smoke! This land of anecdote clearly suggests that it would be worth conducting careful experimental tests of motor imitation in cetaceans.
A good teacher can shape our behavior in ways that may look like imitation, but only demand associative learning on the part of the student. Animal trainers can shape the behavior of animals in the same way. If wild animals were to train one another in this way. this could create a faulty appearance of imitation, but this is thought not to be a problem for there is little evidence that one animal will train another (Caro and Hauser, 1992). However, there are some indications of what might be called teaching in cetaceans. Tyack and Sayigh (1997) provided suggestive evidence for possible teaching of signature whistles in wild bottlenose dolphins, and Guinet and Bouvier (1995) suggested that killer whales (Orcinus orca) teach the young how to strand in order to catch pinnipeds on the beach. Rather than being an alternative to passive observational learning, teaching appears at least in our own species to function in tandem with observational learning. This potential synergy between teaching and imitation would be most likely to benefit highly social animal groups, such as carnivores, primates, and cetaceans, in which cultural traditions for foraging provide a strong selective advantage. These kinds of observations in wild cetaceans, coupled with careful experimental tests with captive cetaceans, suggest that cetaceans are promising subjects for the study of mimicry in the development of cultural traits in animals.
