Content Pages
Social Cognition
Social Inference
Symbolic Inference
Alan B. Bond
Alan C. Kamil
Russell P. Balda
Jeffrey R. Stevens
Docs & Links



Social Cognition

Group living is prevalent among birds. It is often restricted to particular times of year – solitary species often forage socially, at least during migration – or it can be characteristic of particular developmental stages. After breeding season, many corvids and parrots form persistent flocks of juveniles and non-breeding adults. The main advantages of group living are well known. Most obviously, there is safety in numbers. A group has many watchful eyes, alert to the approach of danger, so social foragers can pay more individual attention to finding food, while distributing the demands for security across a wider set of flock members (Hamilton 1971). The second benefit is foraging efficiency. Food tends to be patchily distributed, so it pays to keep watch on the successes of other individuals. Where those birds have had good results, the watchers probably will, too (Vickery et al. 1991; reviewed in Krause & Ruxton 2002). When these ecological influences are not present, group living is much less common: predatory birds seldom forage together, presumably because they are not worried about being attacked, and when they do obtain food, there is not usually enough of it to allow others to partake.

But a diverse set of avian species go beyond opportunistic aggregations. They hang out together throughout much of the year, foraging in flocks, breeding communally, and interacting repeatedly with the same group of birds for long periods of time. Under these circumstances, the group ceases to be a society of strangers. The birds come to know one another as individuals, to recognize the virtues, strengths, and weaknesses of others, and to make use of this knowledge in developing and refining their own life decisions. There is a popular belief that living in such “individualized” societies imposes greater cognitive demands. As Satchel Paige remarked, “the social ramble ain’t restful (1953).” And because of this burden of additional memory and information processing, individualized social systems are thought to provide an important selective force for the evolution of intelligence. This has been termed the “social complexity” or “social brain” hypothesis. The accumulated literature is enormous (reviewed in part by de Waal & Tyack 2009; Emery, Clayton & Frith 2007), but there has been surprisingly little substantive research on the underlying cognitive mechanisms. Just how challenging is it really to live in a persisting, individualized social group? What are the primary capabilities that must be acquired or enhanced?

There is no single module for social cognition, no brain lobe specifically dedicated to handling representations of other individuals and their relationships. Social life draws on a disparate collection of capabilities, all of which have overlapping applications in other cognitive domains. Highly social birds need to be able to recognize a large number of other individuals, both by appearance and by voice. And they need excellent episodic memory to encode and recall salient interactions – aggression, mating, allofeeding, allogrooming, play – enabling them to construct a cognitive representation of the network of relationships among flock members. They need detailed representations of at least the most important individuals in their group, modules that will link to indications of how trustworthy they are, how reliably they reveal locations of food, what novel foraging techniques they have developed. And because social groups are dynamic, constantly changing in membership and internal structure, highly social birds need to be cognitively flexible, able to change their representations on the fly to accord with recent events.

To study this assemblage of faculties under controlled conditions, Balda, Kamil, and Bednekoff (1996) suggested that highly social pinyon jays (Gymnorhinus cyanocephalus) might be an appropriate experimental species. A great deal is known about their natural history (Marzluff & Balda, 1992). They are one of the most highly social corvids in North America, foraging and breeding in permanent groups of up to several hundred individuals. And their social system contrasts strongly with that of other, related jays, enabling comparative studies of social cognition across species. The most experimentally accessible aspect of their social system is their dominance hierarchy. Females are dominant to all juveniles, males are dominant to all females, but within male pinyon jay society, there is a well-established, linear dominance structure that determines priority in feeding opportunities. In a group of 200 jays, that still makes for a large number of other birds to recognize and to remember their relationships. In such large social groups, individual birds cannot rely solely on their own interactions with other group members to learn their position in the hierarchy. For groups larger than about 20, the probability of having had an extended direct encounter with any other, randomly chosen male becomes vanishingly small (Seyfarth & Cheney 1990, especially pp. 80-86).

So we would expect that highly social species should be very good at making use of heuristics to fill in the blanks in their social network (Watts 1999). Given that the dominance hierarchy is linear, the simplest heuristic for extrapolating relative dominance is transitive inference. The mechanism works like this: Consider three pinyon jays in the same flock. Let’s call them Arthur, Basil, and Charlie. Suppose that sometime in the past, Charlie was defeated in an aggressive interaction with Basil. Charlie now behaves submissively whenever he and Basil encounter one another. Later on, Charlie sees Basil behaving submissively to Arthur, a bird with whom Charlie has never interacted. When Charlie subsequently meets up with Arthur in person, Charlie uses his known relationship to Basil and Basil’s observed relationship to Arthur to predict the likely outcome of the interaction. So on the basis of prior information alone, Charlie initially behaves submissively to Arthur, even without a confirmatory fight. If a jay knows that C < B, and observes that B < A, it can infer that C < A. Birds don’t employ mathematical logic, but knowledge of a social network can be used for the same purpose.

Social Inference

We conducted three lines of investigation related to social inference of dominance status. In the first, we constructed virtual dominance hierarchies in contests between pairs of captive male pinyon jays, testing for their ability to infer their status relative to that of strangers (Paz-y-Miño, Bond, Kamil & Balda 2004; see the Social Inference page). Prior to this work, transitive social inference had never been demonstrated under controlled conditions in animals. In this experiment, we found that pinyon jays did, in fact, draw sophisticated inferences about their own dominance status based on interactions that they observed, the first direct demonstration that animals use transitive inference in social contexts.

Symbolic Inference

Our second research thread involved comparing the transitive inference abilities of pinyon jays to that of related corvids that do not live in large, persistent flocks. Several of the comparison species could not be tested reliably in pair-wise encounters: they would not establish tolerant relationships when confined together in a small cage. So we transposed the task to an operant design, using transitive inference of the reward relationships among a set of colored symbolic targets (von Fersen et al. 1991; reviewed in Lazareva 2012). We first compared pinyon jays to relatively nonsocial western scrub jays (Aphelocoma californica). Pinyon jays learned to track multiple dyadic relationships more rapidly than scrub jays and displayed a more robust and accurate mechanism of transitive inference (Bond, Kamil, & Balda 2003; see the Symbolic Inference page). This work was later extended to additional comparison species with more nuanced results: both social complexity and reliance on cached food independently promoted high accuracy in symbolic transitive inference (Bond, Wei, & Kamil 2010). The results provided a clear demonstration of the association between social complexity and cognition in birds and confirmed that similarly challenging ecological processes can convergently promote some of the same cognitive abilities.

Cognitive Flexibility

The third line of research explored whether complex sociality in jays was associated with higher flexibility. Because social networks are plastic structures, changing in both composition and organization over time, we predicted that species that experience higher levels of social complexity should also be more flexible in dealing with modifications and reversals in hierarchical order. We first tested pinyon jays, scrub jays, and Clark’s nutcrackers (Nucifraga columbiana) in serial reversal learning, a classical operant design in which subjects learn to respond differentially to two stimuli. When the task is fully acquired, reward contingencies are reversed, requiring the subject to relearn the altered associations. This alternation of acquisition and reversal can be repeated many times, and the ability of a species to adapt to this regimen has been considered as a primary indication of behavioral flexibility. Pinyon jays displayed significantly lower error rates than did nutcrackers or scrub jays after reversal of reward contingencies for both spatial and color stimuli (Bond, Kamil & Balda 2007).

In a later study, we used a list-linking procedure (Treichler & Van Tilburg, 1996) to explore the processes by which pinyon jays flexibly assemble cognitive structures from fragmentary and often contradictory data. The jays were trained to a high level of accuracy on two implicit transitive lists -- A > B > C > D > E and 1 > 2 > 3 > 4 > 5. They were then trained on just E > 1, informing them of the relationship of the single pair that would link the two lists into a composite, 10-item hierarchy. Following linkage training, the birds were tested on nonadjacent probe pairs drawn both from within (B‑D and 2‑4) and between (D‑1, E‑2, B‑2, C‑3) each of the original lists. Linkage training resulted in a significant transitory disruption in performance, and the adjustment to the resulting implicit hierarchy was hardly instantaneous. But the birds all recovered their original accuracy, forming a reliable 10-item sequence, and they did so much more rapidly than would be expected from a simple associative learning process. Detailed analysis of the course of the disruption and its subsequent recovery provided important insights into the cognitive mechanism underlying symbolic inference (Wei, Kamil & Bond 2014).

References from Other Sources:

Cheney, D.L., & Seyfarth, R.M. (1990). How Monkeys See the World. Chicago, IL: U. of Chicago Press.

Emery, N.J., Clayton, N.S., & Frith, C.D. (Eds.) (2007). Social Intelligence: From Brain to Culture. Oxford: Oxford U. Press.

von Fersen, L., Wynne, C.D.L., Delius, J.D., & Staddon, J.E.R. (1991). Transitive inference formation in pigeons. Journal of Experimental Psychology: Animal Behavior Processes 17: 334-341.

Hamilton, W.D. (1971). Geometry for the selfish herd. Journal of Theoretical Biology 31: 295-311.

Krause, J. & Ruxton, G.D. (2002). Living in Groups. Oxford: Oxford U. Press.

Lazareva, O.F. (2012). Transitive inference in nonhuman animals. In: T.R. Zentall & E.A. Wasserman (Eds.), The Oxford Handbook of Comparative Cognition (pp. 718-735). Oxford: Oxford U. Press.

Marzluff, J.M., & Balda, R.P. (1992). The Pinyon Jay: Behavioral Ecology of a Colonial and Cooperative Corvid. London: T&AD Poyser.

Paige, S. (1953). How to stay young. Collier's Magazine, 13 June 1953.

Treichler, F.R., & Van Tilburg, D. (1996). Concurrent conditional discrimination tests of transitive inference by macaque monkeys: List linking. Journal of Experimental Psychology: Animal Behavior Processes 22: 105–117.

Vickery, W.L., Giraldeau, L.A., Templeton, J.J., Kramer, D.L., & Chapman, C.A. (1991). Producers, scroungers, and group foraging. American Naturalist 137: 847-863.

De Waal, F.B., & Tyack, P.L. (Eds.). (2009). Animal Social Complexity: Intelligence, Culture, and Individualized Societies. Cambridge, MA: Harvard U. Press.

Watts, D.J. (1999). Small Worlds. Princeton, NJ: Princeton U. Press.