Can animals teach?

Active tutoring (“teaching”) can be considered the most complex form of sharing knowledge in animal communities. A working definition of teaching widely accepted among students of social learning was suggested by Caro and Hauser (1994):

“An individual actor (A) can be said to teach if it modifies its behaviour only in the presence of a naive observer, B, at some cost, or at least without obtaining an immediate benefit for itself. A’s behaviour thereby encourages or punishes B’s behaviour, or provides B with experience, or sets an example for B. As a result, B acquires knowledge or learns a skill earlier in life or more rapidly or efficiently than it might otherwise do, or that it would not learn at all”.

There are possibly two main processes of transferring information within populations in which tutoring can be involved: (1) polishing of species-specific behavioural patterns and (2) spreading of innovations. Both processes are parts of social learning.

Back-fitting of species-specific behaviour by efforts of tutoring parents is quite usual in animals, first of all, in vertebrate predators. In many species polishing of searching and hunting behaviour makes up an integral part of their ontogenetic development.

However, tutoring as an instrument for spreading innovations is a rare phenomenon in wild life. One can say that this concerns a level of complexity which is over and above the “general plan” of species-specific behaviour. The fact is that innovations can spread within populations by means that are simpler than tutoring. We have already seen this from previous sections and will consider in the last section of this chapter. Although the real role of tutoring is not great, the analysis of teaching in animals is very important for estimation of limits of their cognitive abilities. Even isolated observations on instances of teaching in the wild are valuable.

In this section we will first analyse a set of examples concerning the role of parental “teaching” in one of the most interesting and enigmatic process studied by behavioural ecologists and ethologists, namely, the process of building a character of a wild hunter. Then we proceed to general definitions and ideas concerning the possible role of teaching in spreading of innovations within populations.

Active tutoring (“teaching”) can be considered the most complex form of sharing knowledge in animal communities. A working definition of teaching widely accepted among students of social learning was suggested by Caro and Hauser (1994):

“An individual actor (A) can be said to teach if it modifies its behaviour only in the presence of a naive observer, B, at some cost, or at least without obtaining an immediate benefit for itself. A’s behaviour thereby encourages or punishes B’s behaviour, or provides B with experience, or sets an example for B. As a result, B acquires knowledge or learns a skill earlier in life or more rapidly or efficiently than it might otherwise do, or that it would not learn at all”.

There are two main processes of transferring information within populations in which tutoring can be involved: (1) polishing of species-specific behavioural patterns and (2) spreading of innovations. Both processes are parts of social learning.

Back-fitting of species-specific behaviour by efforts of tutoring parents is quite usual in animals, first of all, in vertebrate predators. In many species polishing of searching and hunting behaviour makes up an integral part of their ontogenetic development.

However, tutoring as an instrument for spreading innovations is a rare phenomenon in wild life. One can say that this concerns a level of complexity which is over and above the “general plan” of species-specific behaviour. The fact is that innovations can spread within populations by means that are simpler than tutoring. We have already seen this from previous sections and will consider in the last section of this chapter. Although the real role of tutoring is not great, the analysis of teaching in animals is very important for estimation of limits of their cognitive abilities. Even isolated observations on instances of teaching in the wild are valuable.

There is a large body of data about how predators teach their offspring to kill victims. This concerns predators of different sizes and styles of hunting, from giant whales to little jerboas. At the same time, it is known that members of many species grow up as self-made hunters. For instance, polecats (Putorius putorius) learn very quickly how a mouse must be grasped by the neck so that it cannot bite back. During normal ontogenesis complex hunting behaviour can mature with the assistance of individual experience. It can be considered a complex process because innate releasing mechanisms mature as well and they become increasingly more selective through individual learning. This scenario is widely distributed in animal species which can be illustrated by Eible-Eibesfeldt’s (1970) examples of the prey catching behaviour of toads, frogs and others. Even in such advanced hunters as Mustellidae parents’ instructions have little part. If so, is that necessary for predators to teach the young? Is it possible that parents’ instruction run idle in animals? Special investigation is needed in each case if we want to know whether parents take part in shaping of a character of a wild hunter. Let us consider several examples.

It has been known for a long time that in felids and other carnivores’ mothers modify their predatory behaviour in a series of stages. It was described by Leyhausen and Tinkin (1979) in domestic cats Felis catus. Adult females pursue, capture, kill and eat prey in a smooth sequence with little hesitation between acts. However, when cats become mothers and their kittens start walking out of the nest, mothers alter their behaviour and carry prey to their kittens to eat it in front of them. Next, they carry live prey directly to their offspring and allow them to play with it but recapture it if it escapes. Finally females take little part in prey catching at all, merely moving toward prey initially while kittens chase, capture and dispatch it efficiently. Mothers give characteristic mewing calls to their kittens in all of these situations.

The sequence of mothers’ acts is so logical from a human point of view that it undoubtedly looks like successive shaping of the youngs’ hunting behaviour. Under controlled laboratory conditions Caro (1981) tested the alternative hypothesis about the role of mother’s teaching in shaping hunting behaviour in kittens. In Caro’s study kittens between the ages of 4 and 12 weeks were exposed to domestic mice as a live prey. In one series of trials mother cats were present, while in other trials the mothers were absent. Control kittens received identical exposures but without their mothers being present. Behaviour of mothers was also recorded. The obtained results suggest that maternal behaviour reduces the age at which kittens acquire predatory skills. For instance, when six month–old kittens were tested on their predatory abilities, experimental subjects delivered significantly more bites to the nape but not to other regions of the mouse’s body than did control kittens, that is, they more easily applied the method by which adult cats dispatch rodent prey. However, Caro (1981) found it hard to say whether maternal behaviour is sensitive to developmental changes in kitten behaviour. These two processes seem to go in two parallel courses. Indeed, the timing of each step in the mothers’ predatory sequence might not be contingent upon improvement in their cubs’ predatory skills, but rather change according to their individual time course. Certain mothers started to leave prey with their kittens at very early stages when the young were not able to react to it. Many aspects of mother’s predatory behaviour were significantly negatively correlated with increasing skills of their offspring. Caro (1987, 1994, 2005) then has continued with studying of interactions between mothers, cubs and prey on wild cheetah in the Serengeti National Park, Tanzania. The researcher found a very slow progress in young cheetah’s education seemingly as a graduate result of maternal great efforts. It turned out that cub’s hunting skills remain poor up to and beyond independence from their mother, showing surprisingly little improvement in the ten months after first being introduced to prey. In sum, Caro’s data enable us to be careful with the conclusion that this is just maternal tutoring which make cats skilled hunters.

A good example that illustrates how difficult it is to judge about the roles of maturation and investments of parents in shaping of hunting behaviour, is the behaviour of ospreys Pandion haliactus. Meinertzhagen (1954) provided rich description of adult ospreys encouraging their fledglings to catch fishes. At first, the adult perched away from the nest with fish in their talons but would not feed the young, despite their screaming for food, repeatedly flying away with an apparent attempt to encourage the young to follow. On the first day the fledglings did not leave the nest, but on the next two days when the young flew off the nest to a rock, they were fed. On the following day, the young followed the parents to hunt over a lake. Each fledgling caught a fish, carried it toward the young and then dropped it, but caught it again and secured it before it hit the water. After having repeated this many times, one of the young finally caught the fish in a stoop and carried it to the rock to eat it. The less successful sibling now flew to the rock to share the catch, but the parent arrived and literally pushed this offspring off the rock forcing it to take wing again. The process of dropping a fish was repeated until the second fledging finally caught it and went back to the rock to eat. On the fifth day the same procedure was observed with each fledging following a parent around and unsuccessfully attempting to catch fish that were dropped for it in midair. When the fish reached the surface, the parents would retrieve it until eventually youngsters descended to the water and picked up the fish. On the seventh day, the adults drove the offspring away from the lake and they were not seen again. So, the period of education took six days.

Seemingly, this study provides a strong evidence of a definite role of parental teaching in the building of the hunting behaviour in osprey. Nevertheless, observations of hand-raised young ospreys showed that they successfully caught fishes within three days to three weeks of being released into the wild, in the absence of parental instruction (Schaadt and Rymon, 1982).

These data enables us to suggest that parental “instructions” run in parallel with maturation of hunting behaviour of young, and that at least one aspect of the use of parental teaching in animals is to awake dormant behavioural patterns; repetition of instructions in mammals and numbers of encounters with successful hunters in ants possibly have a cumulative effect. This does not mean that teaching should be excluded from consideration of ontogenesis of hunting behaviour.

Transmission of innovations by teaching. Active teaching as a mean for sharing new experience would seem to be very rare in animal kingdom, even in apes. The great role of imitation in social life of anthropoids does not necessarily means that they can teach each others. Young chimps learn how to break twigs from trees, strip away the leaves, and insert them into termite holes by observing adults. The steps required to extract termites in this manner are lengthy and complex. Without the demonstrations of adults, many chimps would probably never become very successful termite fishers. However, part of the acquisition of this tool use appears to relate to innate characteristics of chimpanzee behaviour. All young chimpanzees amuse themselves by playing with sticks and poking them into holes. It seems as though the chimps are able to observe the more skilled adults and translate their juveniles play into a successful means for securing food. But this does not mean that young chimpanzees follow instructions of adults (Goodall, 1986).

During ten years of investigations Boesch (1991, 1995) observed interactions between mothers and their young among Tai chimpanzees in the context of nut cracking. He divided his observations of mother-offspring inter-relations into " stimulation", “facilitation” and " active teaching." Observations of stimulations and facilitation included such things as mothers' leaving intact nuts for their infants to crack (which they never did for other individuals) or placing hammers and nuts in the right position near the anvil for their infants to use. Stimulations were observed on 387 occasions of interactions between mothers and their children. Stimulation differs from the common behavioural pattern for adult chimpanzees, when they carry their hummers during nut collection and consume the nuts that they have placed on an anvil. Mothers incurred a foraging cost by having to find more nuts and another tool for opening them.

Active teaching was observed only twice and involved direct intervention on the part of the mother in her offspring’s attempt to crack open a nut. In one example, a six-year-old male had taken a majority of his mother’s nuts, as well as her stone-hummer. After the young male placed a nut on the anvil, but prior to opening it, his mother approached, picked up the nut, cleaned the anvil, and put the nut back in a different position, more suitable for opening. The young male cracked the nut and ate the kernel. In the second example, another mother reoriented the hummer for her five-year-old daughter who then succeeds in opening several nuts by maintaining the same grip on the hummer that her mother had used.

One more example of isolated cases of teaching in animals came from the experimental investigations on free-living scrub jays, Aphelocoma coerulescens, by Midford et al. (2000). Experimenters trained models (demonstrators) and then followed them as they modelled the task in the presence of naï ve (observer) animals. Jays had to learn that a class of objects (bright plastic rings) indicated the presence of buried food (peanut peaces) in a specific location, the centre of the ring. Birds were trained in their family groups to perform the task during summer season, and were allowed to perform the task in the presence of juveniles in later years. Jays living in 18 control families received partial exposure to the training situation, but received no exposure to the ring before being presented with the task in the presence of their young. Juveniles in 16 families with trained jays were able to witness demonstrations and to scrounge peanut peaces from the models as they completed the task. These 41 juveniles learned much more of the task than the 33 juveniles in control families. What is important for our narrative is that the authors observed three cases of active teaching in two separate families. In each case, the highly ranked bird (the breeder of the family) dug in a centre of a ring until it uncovered the food. Then, rather than taking the pieces of nuts, the adult either departed or stood over the depression it dug, pointed its bill downwards, towards the pieces, until the juvenile took them. This differs markedly from the usual behaviour of jays after finding bits and, as the authors give this, fall within the definition of teaching.

27. CULTURE IN ANIMAL SOCIETIES

In this chapter I briefly analyse a complex and fascinating problem of what is culture in animals. The relationship between cultural and genetic evolution was identified by Wilson (1998) as one of the “great remaining problems of the natural sciences”. Several studies, adopting Dawkins’ (1976) concept of the “meme” as the unit of cultural evolution, have examined factors influencing the transmission and success of memes in animal cultures. The presence of cultural processes within animal societies is an area of some controversy (see: McGrew, 1992, 2004; Wrangham, McGrew and De Waal (eds.), 1994; Boesch, 1996, 2003; van Schaik, 2004; DeWaal, 2001, 2003; Tomasello, 1999, 2000; Whiten, 1989; Fragaszy and Perry (eds.), 2003). How to treat cultural behaviour in animals much depends on its definition. Many definitions in literature attribute cultural traits only to humans. At the other end of the scale is considering culture as a “meme pool” in populations which can include all cases of the regular use of public information in populations basing on very simple forms of social learning (Laland and Brown, 2002).

Arguing about animals’ variants of culture, some authors concentrate on clearly delineated customs within social groups, considering them elementary units of culture, whereas others insist on including into consideration “ethnographic patterns” in animal societies and complexes of unique inter-relations within groups (“pieces of unique atmosphere”) transmitted by means of social learning. Complexes of characteristic behaviours in whales and dolphins (Rendell and Whitehead, 2001), as well as very specific inter-relations in a troop of olive baboons (Sapolsky and Share, 2004), both transferred through social learning but not genetically, serve as good examples.

Many cognitive ethologists agree now that human beings are biologically adapted for culture in ways that other primates are not, as evidenced most clearly by the fact that human cultural traditions accumulate modifications over historical time (the ratchet effect: Tomasello, 1999). As we have already seen above and will meet further in this book, our species is likely to possess some uniquely powerful forms of cultural learning, enabling the acquisition of language, discourse skills, tool-use practices, and other conventional activities.

Not arguing about restricted notions of “culture” in animals, I would rather adhere to broad definitions basing on social learning as the main mechanism of “cultural transmission” of behavioural patterns in animal societies. I see here a fascinating perspective of estimating limits of power of social learning in non humans that allow some species to improve adaptiveness of behaviour by non genetic means. In this context, I consider the broad definitions as acceptable such as “culture is information or behaviour acquired from conspecifics through some form of social learning” (Boyd and Richerson, 2005), and “animal tradition that rests either on tuition of one animal by another or on imitation by one animal of acts performed by another” (Galef, 1992). Imanishi (1952) defined culture as " socially transmitted adjustable behaviour". In this broad sense, cultural learning is widely accepted by students of animal behaviour. Defining culture as a package of behaviours, the working description given by Nishida (1987) is useful: “Cultural behaviour is defined as behaviour that is (a) transmitted socially rather than genetically, (b) shared by many members within a group, (c) persistent over generations and (d) not simply the result of adaptation to different local conditions “.