Observational learning

1.The *acquisition of novel behaviour by observing its *performance by a model.

2. The generation or modification of lasting *internal representations of actions, or actions and their consequences, by observing the behaviour of a model.

'Whatever you see me do, do like-wise', said Gid'on to his selected three hundred men, blew the shofar, waved the torch, and surprised the Midyanites in their camp (Judges 7: 17-22). He was a master demonstrator in an observational learning class. A lot of what we learn in our lifetime, from others we learn, frequently by observing a model and adapting its actions. In "reallife, the model is a conspecific; in the lab it could be an individual of another species or a behaving inanimate.1 This capability, to gain from the experience of others, clearly expands the behavioural repertoire of individuals much beyond the limits of their own innate ("a priori) responses and solo experience combined.

Behaviours emitted or acquired via social interaction are termed 'socially dependent'. They are widespread and observational learning is but one type. Socially dependent behaviours include: (a) socially released behaviour; (b) socially facilitated behaviour; (c) social learning, which includes imitation and observational learning, instructed learning, and collaborative learn-ing.2 In socially released behaviour an innately predisposed response pattern, such as courtship or attack, is triggered by the "perception of a specific "stimulus (Lorenz 1981). In socially facilitated behaviour, there is an increase in the frequency or intensity of response that is already in the individual's active repertoire, e.g. eating or locomotion, when in the presence of others that are engaged in the same behaviour (Clayton 1978). In contrast, in social learning, the interaction results in new behaviour. This interaction could be between a model and a learner (imitation and observational learning), an intentional teacher and a learner (instructed learning), or two or more learners (collaborative learning; Tomasello et al. 1993). The differentiation among observational, collaborative, and instructed learning is appropriate in the discussion of social learning in humans, but not necessarily in other species, in which the parties in the learning process might be unaware of their 'formal' role in it. Collaborative and instructive learning will not be further discussed here.

Many authors use the terms 'imitation' (or 'imitative learning') and 'observational learning' interchangeably. This should not, however, blur the wide spectrum of complexity of the behaviours involved. Others prefer to reserve 'imitation' to describe a subtype of observational learning, in which the learner faithfully duplicates the behavioural performance of the model. In the more flexible forms of observational learning, the learner emulates the behavioural strategy involved, not only the overt motor acts, and is clearly capable of adapting and improving the learned behaviour to attain the goal. Another point to note is that 'observational learning' connotes visual learning, whereas 'imitation' covers all the sensory modalities. In the rest of this discussion, for the sake of convenience, the term used will be 'imitation and observational learning' (abbreviated IOL).

Our knowledge of IOL draws from research in multiple disciplines: behavioural psychology and ethology; education, developmental and social psychology; and cognitive psychology and the philosophy of mind. The systematic discussion of IOL in behavioural psychology and ethology was initiated already in the nineteenth century; examples from this period are provided in Darwin (1871, 1872), Romanes (1882), and Morgan (1896). Most contemporary reports on species other than humans were mostly anecdotal ("anthropomorphism). The interpretation of the factual or the alleged data was at first rather shaky. In this context, even Darwin erred: he suggested that in evolution, the dog has started to bark in an attempt to imitate its talkative human master (Darwin 1872). Since then, IOL has been documented in a great variety of species, ranging from guppies and octopi, via birds and cats, to "monkeys and apes (John et al. 1968; Griffin 1984; Anderson 1990; Cheney and Seyfarth 1990; Fiorito and Scotto 1992; Barresi and Moore 1996; Whiten et al. 1996; Marler 1997; Laland and Williams 1997; Templeton 1998). And, reassuringly enough, not only humans find it more rewarding to learn from other's mistakes than from their successes (Templeton 1998).

Research in education and social psychology has contributed tremendously to our understanding of the role of IOL in human "culture (Deahl 1900; Miller and Dollard 1941; Bandura 1962; Bandura and Walters 1963; Bandura 1986). A major question is how much of children's behaviour is moulded by observing their parents, their siblings, their classmates, or TV? The answer is simple: a lot. IOL starts already in the neonate (Meltzoff and Moore 1977), and plays a central part in shaping behaviour throughout the critical periods of emotional and cognitive "development (Piaget 1962). For example, in a "classic, influential set of studies, it was found that children who observe a model rewarded for aggressive behaviour tend to exhibit more aggressive responses than children who see the model punished (Bandura et al. 1963) (Figure 51). This is a case of 'vicarious learning', so called because the "subject sympathizes with the reward or punishment of the model without experiencing it itself (Bandura and Walters 1963). It is difficult to overemphasize the importance of the findings on the role of IOL in children, especially in a TV-dominated society. Some kids may be misled to think that jumping in front of a car is safe and kicking another guy is good, because on TV stuntmen/women are never hurt and villains live happily forever (Potts et al. 1996). IOL keeps working throughout life, even in situations in which we are utterly unaware of it; for example, in a restaurant, what we order is influenced by incidental observation of the reaction of others to the food (Bayenes et al. 1996).

Iolab Iol Model 103

Fig. 51 A *classic study of observational learning: children imitate the aggressive behaviour of an adult model they had observed torturing an inflated doll on film. In this study, on the average, gender had a clear effect: boys displayed more acute aggression toward the doll, whereas girls were more inclined to sit on the doll rather than punch it. (From Bandura et al. 1963.)

Fig. 51 A *classic study of observational learning: children imitate the aggressive behaviour of an adult model they had observed torturing an inflated doll on film. In this study, on the average, gender had a clear effect: boys displayed more acute aggression toward the doll, whereas girls were more inclined to sit on the doll rather than punch it. (From Bandura et al. 1963.)

There used to be in the literature a temptation to use complex forms of IOL as evidence that subhuman species have, similarly to "Homo sapiens, a 'theory of mind', i.e. that the individual of the species can impute mental states to itself and to others.3 This is because in some situations it seems as if the observer really 'reads the mind' of the model. This is why cognitive psychologists and philosophers of the mind discuss IOL (e.g. Gallese and Goldman 1998). A caveat is, however, appropriate: even a remarkable capability to learn from the other is not sufficient to prove a 'theory of mind', as the behaviour might still be explained by "associative conditioning, rendering the theory of mind explanation superfluous ("criterion, "Ockham's razor). The acid test for a 'theory of mind' is the ability to compute what the other subject will do on the basis of false belief, not physical reality, because "cues in physical reality could govern the behaviour of the observer without the need to access the mental state of the other individual. Possibly, in addition to humans, only great apes have a (rudimentary) 'theory of mind' (Barresi and Moore 1996; Frith and Frith 1999).

What brain mechanisms subserve IOL? Given the rich repertoire of IOL performances in various species, no single solution should be expected, at neither the computational, nor the neuronal hardware "level. The problem is easier to approach in the laboratory in tasks that involve elementary motor acts, such as imitation of reaching or grasping in humans and monkeys. The "methods used to analyse the imitating brain are cellular physiology in the behaving monkey, and "functional neuroimaging in humans. A useful conceptual starting point is that IOL of simple motor movements exploits elementary motor subroutines of the learner. The learning therefore does not involve de novo generation of the entire movement on the basis of the information received by observation of the model. In other words, here is another example of the rule 'to learn something, you must already know a lot' ("a priori). An appealing hypothesis is that the learner matches the "percept with its own motor representations of motor action, in order to understand (Jeannerod 1994) and learn (Rizzolatti et al. 1996; all this is done implicitly; no "declarativity required).

Evidence from both humans and the monkey concur with this 'matching' or 'resonance' hypothesis. In the monkey, neurons were detected in the pre-motor area in the frontal cortex, which fire both when the monkey performs an action and when it observes a similar action made by another monkey or by the experimenter (di Pellegrino et al. 1992; Rizzolatti et al. 1996). These neurons are termed 'mirror neurons'. Similarly, in humans, areas were identified in the frontal and parietal lobe that are activated both when a movement is performed by instruction or by imitation (Iacoboni et al. 1999). In another study, brain regions involved in "planning and generation of action, including frontal cortex, were activated in humans that observed an action with the intention to imitate it, as opposed to observation only (Decety et al. 1997). These data indicate that the neuronal circuits that subserve the perception and learning of the action-to-be-imitated overlap with circuits that subserve the preparation and the execution of the same action. If we "generalize a bit, and look at the other facet of the coin, this conclusion echoes the so-called 'motor theories' of speech perception ("birdsong, "performance): in order to understand an action, we must be able to perform it.4

Selected associations: Acquisition, Birdsong, Cerebral Cortex, Internal representation, Planning

1Using inanimate models could be tricky, as animals often distinguish between animates and their inanimate replicas, and might fail to imitate in the absence of the live model; e.g. see discussion in Rizzolatti et al. (1996).

2Some socially dependent behaviours do not fit easily into this 'taxonomy. For example, the transfer of food information from one rat to another via odours (Galef et al. 1984) is social learning, but could hardly be considered observational, or bona fide collaborative, or instructive. Also, relevant to socially dependent behaviours is biological communication, which includes many types of vocal, visual, olfactory, and somatosensory species-specific communication systems, and, of course, human language.

3The 'theory of mind' is termed 'theory' because the mental states are not directly observable but inferred, and this inference is used to make predictions. The term in this meaning was introduced by Premack and Woodruff (1978), in the context of their studies of the chimpanzee. Similar notions are 'mentalizing', and an 'intentional stance'; having an 'intentional stance' means treating the other subject as an agent with intentionality (Dennett 1987; on intentionality as a 'criterion for mental systems, see 'system).

4This also echoes a maxim of scholastic philosophy, resurging in Vico (1710): only the one who makes something can fully uderstand it. Alas, the generalization of this wisdom to most human behaviours is clearly questionable.

Happiness Today

Happiness Today

Unhappy With Your Current Situation? Follow The Principles Within This New audio series and Guide and Create True Happiness Everyday Of Your Life. Download To Discover How To Live A More Fulfilling And Happy Life Starting Today.

Get My Free Ebook


Post a comment