Human infants have one important job during the first years of life, and that is to learn about the world and their culture from their parents and other caregivers. But what is learning? I've previously written that Hungarian developmental psychologists Gergely and Csibra have defined learning as the acquisition of new, generalizable knowledge that can later be used within a new context. Further, they have posited that evolution has prepared humans to learn generalizable knowledge from their caregivers. They proposed an elegant hypothesis: that a specialized innate pedagogy mechanism - called the pedagogical learning stance - is in place that allows an infant to retain generic information. This means that they are able to learn information in a given instructional setting, that they can later apply to wide range of potential new situations.
This sort of cognitive system requires at least three things. First, the learner must understand the communicative intent of the teacher via ostensive cues. One such cue for humans is the use (by a parent or teacher) of infant-directed speech, or baby-talk. Second, the teacher and learner must be able to jointly use referential signals, such as eyegaze and pointing, in order to facilitate joint attention. Third, the learner must be able to understand the information content of the pedagogical interaction; that is, they must realize that they are getting relevant information for the given task.
There is good evidence that humans do, indeed, have innate pedagogy. But Gergely and Csibra take their claim a bit further. They claim: (1) that natural pedagogy is human-specific, (2) that natural pedagogy is universal among human cultures, and (3) that this sort of human social communication was explicitly selected for in evolution, rather than having emerged as a by-product of some other selection.
Pedagogy is, in its simplest form, a form of social learning that occurs via communication between two (or more) individuals, which has as its outcome the transfer of knowledge or skills. There are abundant cases of both social learning and of communication in non-human animals, of course. Are there ever any cases where social learning and communication combine in any other species that allows for knowledge transfer similar to human teaching? The short answer is: no. But let's examine this in some more depth.
One question we might ask is whether there are any forms of animal communication that go beyond the providing of episodic information (related to the here-and-now, immediate context) and approach the transfer of generalizable information? Certainly, there are cases of non-human animal signals that are referential, inasmuch as they are meant to facilitate joint attention on an external stimulus or event, even if that stimulus or event is not within view. For example, various animals use alarm calls that refer to predators or to food, or ones that are used for other social purposes, such as to recruit group members for an antagonistic interaction with another individual. Bees use elaborate "dances" to indicate the location and quantity of food available in a given location. And some animals, such as howler monkeys, even have predator-specific calls, such as "aerial predator," like the one they use for the fearsome harpy eagle.
However, these signals don't identify something like "aerial predator," they actually transmit a message more like "aerial predator approaching!" There is no alarm call for "hey, did you see how I totally beat up that aerial predator?" or "the aerial predator is approaching from the northeast but his vision is based on movement so remain still," or even "aerial predator gone, the coast is clear." More important for the question of pedagogy, there is no alarm call that goes something like, "aerial predators tend to hunt during the daytime." It is in this sense that communication among non-human animals only transmits episodic information, and not generic information.
In developing this argument, Gergely and Csibra are quick to point out that the absence of generic information transmission among non-human animals does not mean that non-human animals can not learn, in a more general way, from episodic information in the first place. For example, using a simple statistical learning mechanism, an infant bird could learn how to find food by learning a correlation between a certain chirp made by its mother, certain features of her location, and the resulting feast. In this case, however, the generic information (where to find food) comes not from the communicative interaction, per se, but from a learning process wholly contained within the infant bird (even if the learning process is initiated by a response to a food call).
A promising example of natural pedagogy in non-human animals might have been the meerkat. Meerkats hunt and eat scorpions, which are extremely dangerous. It is really important for a juvenile meerkat to learn how to safely hunt and consume a scorpion, because just one mistake could prove fatal. What adult meerkats do is they provide the juveniles over time with scorpions, beginning with dead ones with stingers removed, then dead ones with the stingers intact, and finally living and lethal scorpions. The progression from safe to more and more dangerous scorpions is determined according to the age and skills of the juvenile meerkat. For example, in the stage where the adult provides a living scorpion with stinger removed, the juvenile can successfully learn how to kill a scorpion without the danger of dying. Two important criteria for teaching are met here: first, the teacher adjusts the curriculum (and therefore, his or her behavior) as the learner progresses and gains knowledge. This is strong indication of pedagogy. However, there is no explicit communication involved in this process. In fact, this process more resembles a teaching strategy termed "scaffolding" by developmental psychologists, than the sort of natural pedagogy that Gergely and Csibra have hypothesized. One other important distinction between pedagogy and this case is that teaching in meerkats is domain specific. This means that scorpion eating is the only form of instructional learning in meerkats. Pedagogy in humans, on the other hand, is domain general: humans can teach and learn just about anything.
Another promising case of teaching in non-human animals seemed to be tandem-running ants (Temnothorax albipennis). This occurs when one individual knows the location of food, and he communicates that information to a second individual, who then learns that information. When Nigel Franks and Tom Richardson published this finding in Nature, they announced that, "tandem running is an example of teaching, to our knowledge the first in a non-human animal." Even if this is not natural pedagogy, as Gergely and Csibra would argue, it is a remarkable example of a simpler form of teaching. Specifically, the leader in the pair not only directs the follower to the food, but modifies its behavior based on the behavior of the learner. For example, it slows down or stops in order to allow the learner to memorize the path, and if for some reason the teaching is interrupted, the leader will wait for the learner to return before commencing with the instruction again. The leader takes the front position, and requires consistent tapping on his rear end from the antennae of the learner in order to continue the lesson. While this form of teaching does, therefore, include bidirectional signals and information transfer, the information is decidedly episodic in nature, and not generalizable. The information gained by the learner is how to get from specific location A to specific location B, and not how to find food more generally. And, as with the meerkats, this teaching is domain-specific. There is no evidence yet that this sort of bidirectional communication is used by ants for anything other than indicating the location of food.
In general, the evidence is mounting that natural pedagogy is unique to humans. That does not mean that it is impossible that we might one day discover natural pedagogy in non-human animals. Indeed, there are some examples that come tantalizingly close. Critically, however, there are no known teaching processes in non-human primates, beyond simple learning mechanisms such as imitation and emulation.
Therefore, if pedagogy is absent in non-human primates, and ends up being present in some other animal, this would more likely be a case of convergent, rather than homologous, evolution. This would be consistent with Gergely and Csibra's hypothesis. With respect to the possibility for non-human pedagogy, they write:
If this is proven, it will show that natural pedagogy is not human-specific. Nevertheless, this would not disconfirm our hypothesis that the cognitive systems that enable pedagogical knowledge transmission in humans represent a hominin adaptation, because analogous adaptations can emerge independently in distinct lineages. Crucially, no convincing example of teaching has been found in non-human apes or other primates. Thus, if natural pedagogy is an evolutionary adaptation, it must have emerged in the hominin lineage.
At least provisionally, the conclusion can be reasonably drawn that natural pedagogy is a uniquely human cognitive mechanism. Count this is one point in favor of Gergely and Csibra's pedagogy-as-adaptation argument.
The Pedagogy Series
For more on social learning:
For more on animal communication:
Csibra G, & Gergely G (2011). Natural pedagogy as evolutionary adaptation. Philosophical transactions of the Royal Society of London. Series B, Biological sciences, 366 (1567), 1149-57 PMID: 21357237
Csibra, G., & Gergely, G. (2009). Natural pedagogy Trends in Cognitive Sciences, 13 (4), 148-153 DOI: 10.1016/j.tics.2009.01.005
Boston Common ducks: Flickr/sfgamchick
Harpy eagle: Flickr/ZakVTA