CAPÍTULO II MARCO TEÓRICO
B. Daños extra patrimoniales
Great Apes are of special interest in regards to imitation. Humans are, of course, a kind of ape and imitation is very pronounced in humans. Meltzoff (1988) goes so far as to suggest imitation is one of our defining characteristics and he re-dubs
Homo sapiens as "'Homo imitans'*. One might, therefore, expect those species most closely related to humans to be likely to manifest a similar ability in imitation. Byrne (in press, a) states that, "among the apes it is now certain that humans are much more closely related to some species than others." Whiten (1993) further explains:
"The two species of chimpanzee represent our closest living relatives in the animal kingdom. Current estimates put the date of evolutionary divergence between our ancestors and those of these apes at only 4-7 million years ago. This may be contrasted with the split between our line (i.e. the ape lineage) and that of the Old World
monkeys, which occurred as long as 20-30 million years (Sibley and Ahlquist 1984; Hasagawa et al. 1989). The ancestors of the other Great Apes (the orangutan and the gorilla) and the Lesser Apes (gibbons), split off in between" (p. 373).
Therefore, if one is interested in identifying appropriate subjects for comparative studies in imitation, the Great Apes and especially chimpanzees would appear to be very
suitable candidates.
In the present chapter, I shall discuss the observational and experimental evidence for imitation in each of the Great Ape species (examples are included from
Gorilla gorilla, Pongo pygmaeus, and Pan paniscus, but the majority of reports come from Pan troglodytes). I shall not discuss Lesser Apes simply because I have not been able to find any reports of imitation in these primates: although, one should bare in mind that "absence of evidence is not evidence of absence" (Byrne & Whiten 1990, p. 5). In Chapter 5, I shall consider the evidence for imitation in monkeys.
UNCONTROLLED OBSERVATION OF IMITATION IN GREAT APES
There are a great many anecdotes or uncontrolled observations of apparent imitation in Great Apes (see chapter 3 for definitions of anecdotes or uncontrolled observations, controlled observations and experiments). Reports of these observations can be found scattered throughout scientific and popular (factual) literature. I have made a search of this literature and compiled a data-base of anecdotal reports of imitative-like behaviour in non-human primates. A similar data-base was compiled by Byrne and Whiten (1986/1990) on tactical deception in primates. They used anecdotes that came from "highly experienced primatologists, each with years of familiarity with the species concerned, and training in behavioural recording" (Byrne 1993). Whiten and Byrne (1988) argue that single anecdotes can never be more than "a jumping off point for more systematic work" (p. 243). They suggest that, "no single observation can be regarded as definitive evidence in support of a hypothesis ... There are two major reasons for scepticism. One is that in many cases the relevant evidence involves very fine distinctions in behavior ... Second, a single instance may in many cases simply represent a coincidence" (p. 243). Instead, the authors recommend that only multiple records be considered. Byrne (1993) states that, "The lesson is this: careful and unbiased recording of unanticipated or rare events, followed by collation and an attempt at systematic analysis, cannot be harmful. At worst, the exercise will be superseded and made redundant by methods that give greater control; at best, the collated data may add to theory".
Even Whiten and Byrne's careful treatment of anecdotes have been highly criticised with some commentators arguing that anecdotes are intrinsically "unscientific" (e.g., Berstein 1988, Thomas 1988). There is a very genuine fear that such work will lead other researchers to revert to the sloppy use of anecdotes that abounded at the turn of the century. Burghardt (1988), however, indicates that our modern day aversion to anecdotes may be over-zealous. He points out, "Lloyd Morgan (1894), who with his
"canon" supposedly brought down the anecdotal method, actually called for more careful, critical use of anecdotes and his own books abound in them" (p. 248).
One way to avoid Romanes' uncritical approach is to evaluate each anecdote using strict criteria. In the present imitation catalogue each observation was judged using six criteria which are partially based on the work of Moore (1992) and Russon and Galdikas (1993):
"Perhaps reports should simply be considered on their individual merits, with greatest weight given to
(1) first-hand accounts by (2) competent observers who
(3) demonstrate adequate knowledge of the species in question" (Moore 1992, p. 236). Moore (1992) and Russon and Galdikas (1993) also argue that the observers should: (4) describe the context of the episode and the precise actions performed by the subject in detail and,
(5) be acquainted with the relevant details related to the ontogeny of the apparently imitative behaviour.
(6) Finally, of course, the details of the behaviour and its ontogeny should indicate that imitation was indeed the learning process involved and that no other less complex process can adequately explain the subject's behaviour.
Washburn (1908) discussed two further potential pitfalls in anecdotal data, both of which are related to observer bias. The observer may:
(I) have "a personal affection for the animal concerned, and a desire to show its superior intelligence" (p. 5)., and
(II) have "the desire, common to all humanity, to tell a good story" (p. 5).
Burghardt (1988) argues that these two potential flaws in anecdotal data (I and II above) are particularly intractable. Yet, if the observer is vigilant and shows him or herself
willing to point out weaknesses or flaws in the subject's imitation, this might go some way towards counteracting the two kinds of observer bias described by Washburn.
One difficult problem related to assessing an anecdote is how to objectively judge who qualifies as a "competent observer" (see criterion 2 above). Washburn
(1908) suggests that the observer should be, "scientifically trained to distinguish what he sees from what he infers" (p. 5). Yet, it is possible that "the careful observations of experienced pet-owners, unfamiliar with behaviourist theory but knowing their animals well" (Byrne, in press a and b) can also be of merit. The best way to illustrate this point is to consider an actual observation made by an experienced pet-owner. Let us closely examine the observation made by Kearton (1927) of his pet chimpanzee, Mary (see report 18 in the data-base).
In order to assess Kearton's competence as an observer, I propose to consider how well the anecdote satisfies the other four criteria listed above, and to what extent Kearton avoids Washburn's two types of observer bias. Kearton's report is clearly a first hand account (criterion 1). Also, one can be confident that Kearton possessed a great knowledge of chimpanzees, since Mary was the second chimpanzee he had raised in his home (criterion 3). Since Kearton had raised Mary from an early age, we can assume that he was very knowledgeable of her past experiences (criterion 5). Kearton was particularly careful in describing the context in which the observation took place. He wished to demonstrate to a sceptical onlooker, a certain Mr. Jones, that Mary was not forced to learn "tricks", but performed a number of human-like activities through a combination of natural intelligence, curiosity and imitation. Kearton set up the situation of presenting Mary for the first time with a bucket and spade on the beach, in order to demonstrate her natural ability to Mr. Jones. Hence, he described the demonstration and Mary's actions in great detail (criterion 4).
It is true that Kearton was trying to demonstrate to Mr. Jones that Mary was possessed of natural intelligence (observer bias I), and since he was writing a children's book, telling a good story was one of his main aims (observer bias II). Yet, Kearton did not only point out Mary's strengths, he also pointed out her weaknesses. He
comments that despite several demonstrations Mary did not learn to turn the sand out of the bucket to form individual structures, she simply filled the container then tipped the sand out and filled it again. The fact that Kearton does not fail to mention Mary's shortcomings, helps to convince one of the unbiased nature and accuracy of his report. All these factors combined should convince the reader that Kearton was indeed a competent observer, even if he was not a trained scientist. Each anecdote in the data base has been evaluated in this way, in order to judge the relative competency of the observers.
Let us now consider the content and structure of the data-base itself. It is not always easy to decide what is an observation versus an experiment. For example, Hayes and Hayes (1952) demonstrated hammering a stake into the ground and then rubbing it with sandpaper, with the expressed intention of trying to elicit an imitative response from their chimpanzee, Viki. Although these acts were deliberately
demonstrated they have been treated as observations rather than experiments because no formal controls or measures were used. The behaviour was simply demonstrated and then Viki was allowed to respond. Later tests with Viki do seem to constitute
experiments and hence these do not appear in the database. In one experiment Viki was systematically taught to imitate novel arbitrary actions on command. In a second study her imitative problem-solving ability on specially designed tasks was compared with young children and a young laboratory- (rather than home-) raised chimpanzee. Experiments are distinguished from observations by assessing the degree to which controls and formal measures are used on specific tasks.
Observations were included in the data-base if: (a) the original authors interpreted the behaviour as imitative, (b) the description of the subject's behaviour indicates that imitation (or at least a complex form of social learning) was involved, but the original authors offer no explicit interpretation and (c) the original authors clearly state that no evidence of imitation was found despite attempts to promote its occurrence.
(1) the reference,
(2) whether the subject was observed in the wild or captivity,
(3) the interpretation provided by the original author (I = imitation; D.L = delayed imitation; O.L. = observational learning (with no single social learning process stipulated); 0 = no interpretation given),
(4) the present author's interpretation of the observation (I = program-level and/or exact action imitation; D.I. = delayed imitation; E = emulation; D.E. = delayed emulation; S.E. = stimulus enhancement; R.E. = reward enhancement; C =
contagion; Exp - exposure; I.E. = individual learning; O.C. = operant conditioning, such as when an experimenter inadvertently shapes or molds the response; ? = an independent judgement was not possible.)
(5) which of the six criteria for evaluating the validity of anecdotes were not satisfied, (6) the species of the subject,
(7) the name of the subject.
The details of the observation itself follow these preliminary notes. Each report was, as far as possible, quoted in full. Only if the report was particularly long was it paraphrased with the most relevant sections quoted in the original words. After each report the present author provides a brief explanation of her interpretation of the subject(s)' behaviour. The observations from each species of primate are clumped together with the chimpanzees appearing first, followed by gorillas and then finally orangutans. Table 4.1, at the end of the database, summarizes each report.
OBSERAVATIONAL REPORTS OF IMITATIVE-LIKE BEHAVIOUR IN GREAT