11-2: Observational Learning in Animals

Psychology of Learning

Module 11: Observational Learning 1

Part 2: Observational Learning in Animals

Looking Back

Part 1 introduced observational learning as a qualitative leap in adaptive capacity—learning from others’ experiences rather than only through direct personal experience. We distinguished simpler mechanisms like social facilitation & stimulus enhancement from true imitation requiring cognitive representation. Now we explore observational learning across diverse animal species, examining Bennett Galef’s groundbreaking research on social transmission & the factors that influence social learning.

Social Transmission of Food Preferences in Rats

Bennett Galef & colleagues at McMaster University conducted a quarter century of research examining social learning in Norway rats (Rattus norvegicus). Their work represents some of the most rigorous experimental analyses of social learning in animals, with careful attention to eliminating alternative explanations (Galef, 1996; Galef & Laland, 2005).

Here’s a critical fact about rat biology that makes social learning essential for survival: rats cannot vomit. Vomiting serves a valuable protective function in most mammals—it allows expulsion of potentially toxic substances. Since rats cannot vomit, if a rat ingests a toxic substance, it is too late; the rat will die. Therefore, rats are incredibly averse to new flavors. When a rat encounters a novel food, it will take the tiniest nibble possible, then wait. If it feels ill at any point in the next 24 hours or so, the rat will never eat that food item again. This is why rat poison must be extremely potent—if the initial nibble isn’t lethal, rats become “bait shy” & will never touch that flavor again.

Given this severe neophobia (fear of new things), how do wild rats safely expand their diets? The answer is social learning. Galef demonstrated that rats acquire information about safe foods through interaction with conspecifics.

The Demonstrator-Observer Paradigm

In Galef’s basic experimental procedure, a demonstrator & an observer rat are first housed together & allowed to eat standard rat chow, establishing familiarity. The rats are then isolated & food deprived. The demonstrator rat, which is now quite hungry, is given access to a novel flavored food (such as cinnamon-flavored mashed potatoes). The demonstrator & observer rats are then caged together again for a brief interaction period, during which the observer rat can smell the breath of the demonstrator rat. Then the observer rat (who is also hungry) is given a choice between two different novel foods—one of which the demonstrator has eaten (cinnamon) & one which the demonstrator has not eaten (cocoa). The experimenter measures how much of each food the observer rat eats.

The results are striking: observer rats consistently eat significantly more of the “demonstrated” flavor—the food their cage-mate had recently eaten. This preference develops after a single brief interaction & can last for weeks (Galef & Whiskin, 2003). The rats are using olfactory information from their companion’s breath to identify safe foods. Galef identified carbon disulfide, a component of rat breath, as the critical signal that makes food-related odors transmitted from demonstrator to observer rats behaviorally significant (Galef, Mason, Preti, & Bean, 1988).

Robustness & Generality of the Effect

Galef’s research demonstrated that this social transmission of food preferences is remarkably robust. The effect occurs in both laboratory & wild-strain rats, in food-deprived & non-deprived observers, in familiar & unfamiliar demonstrator-observer pairs, in both juvenile & adult observers, & with both solid foods & liquids. The preference effect persists for at least a month following a single 30-minute interaction with a demonstrator (Galef & Whiskin, 2003). This long-lasting influence of a brief social interaction provides an efficient model system for studying the neural & behavioral substrates of long-term social memory.

Perhaps most remarkably, social information can override even conditioned taste aversions. If a rat has been made ill after eating cinnamon-flavored food, it will normally never touch cinnamon again—this conditioned taste aversion is extremely powerful & persistent. However, Galef & Whiskin (2008) demonstrated that interaction with a demonstrator who had eaten cinnamon could partially restore cinnamon consumption in the aversive rat. The social information competed with the rat’s direct experience of illness. This “peer pressure” effect demonstrates the powerful influence that social information exerts on feeding behavior.

Galef’s Perspective on Animal Social Learning

Galef has been a consistent advocate for rigorous experimental methodology & cautious interpretation of animal social learning. He argues that true social learning involving complex cognitive factors should not be casually attributed to animals—doing so would be to anthropomorphize (Galef, 1992). Many behaviors that appear to involve sophisticated observational learning can be explained by simpler mechanisms like local enhancement—when a demonstrator’s presence near an object increases its salience to an observer, leading to increased interaction with that object without requiring understanding of the demonstrator’s behavior or goals.

This parsimonious approach has been valuable in distinguishing genuine social learning from simpler processes. However, it also highlights that even without complex cognition, social influences on behavior can be powerful & adaptive. The rat food preference work demonstrates that animals need not understand why a food is safe—the mere association between a conspecific’s odor & a food flavor is sufficient to modify behavior in adaptive ways.

Mate Choice Copying in Japanese Quail

Galef & colleagues extended their research on social learning to reproductive behavior, studying mate choice copying in Japanese quail (Coturnix japonica). In these experiments, a focal female observed two males: one she initially preferred (spent more time near) & one she initially did not prefer. The focal female then watched her non-preferred male interact with another female (the model female). Following this observation, the focal female was again given a choice between the two males (Galef & White, 1998).

Results revealed that focal females increased their preference for the non-preferred male after watching him with another female. This shift occurred even when the model female & non-preferred male were separated by an opaque barrier preventing direct interaction—suggesting that the mere proximity of a female to a male enhanced his attractiveness. This form of social learning could provide females with information about male quality: if other females find a male attractive, he may possess desirable traits that are not immediately apparent.

Interestingly, male quail showed the opposite pattern. Males avoided females they had observed mating with other males, perhaps reflecting a strategy to avoid sperm competition. These sex differences make adaptive sense given the different reproductive strategies of males & females.

Further research demonstrated that females lay more fertilized eggs after mating with a male they have seen mate with another female than after mating with a male they did not watch mate. This remarkable finding suggests that mate choice copying has genuine reproductive consequences—females who copy the choices of other females may obtain higher-quality mates. Female quail’s preferences among males are also affected by observation of males’ aggressive interactions, with virgin females preferring dominant males & sexually experienced females preferring subordinates (Galef, 2008).

Social Learning of Fear & Predator Recognition

Social transmission of fear responses has been documented across diverse species. Griffin & Evans (2003) demonstrated that a Tammar wallaby’s fear of foxes (a wallaby predator) acquired through direct experience could be socially transmitted to predator-naive companions. Wallabies that had never encountered foxes developed fear responses after observing the fearful reactions of experienced companions. Similarly, Ferrari & Chivers (2008) demonstrated social transmission of fear of salamanders in frogs, with the amount of fear transmission increasing as the tutor-to-observer ratio increased.

Research with fish reveals similar patterns. Many fish species have specialized epidermal cells that release “alarm substance” when mechanically damaged. This chemical diffuses through the water & triggers predator escape responses in surrounding individuals (Gariépy et al., 2014). Naive fish can learn to recognize novel predators by observing the fearful responses of experienced fish—after witnessing conspecifics responding fearfully to a particular predator, observers develop lasting avoidance of that predator. This socially acquired predator recognition can persist for weeks & transfer across contexts.

Environmental Stimulation, Social Rearing, & Conspecific Modeling

Research on environmental enrichment has shown that rats raised in stimulating environments with opportunities for social interaction & exploration show increased brain weight, enhanced blood flow, greater exploratory behavior, & improved learning ability compared to rats raised in impoverished environments. But how do different components of enrichment contribute to these outcomes? A study by Eyck & Vergara (1988) examined how environmental stimulation, social rearing, & conspecific modeling independently & interactively affect rat behavior.

Sixty-four female Long-Evans hooded rats were subjects in a 3-way factorial design examining combinations of enrichment experiences not previously studied together. The three independent variables were: (1) environmental stimulation (raised in enriched vs. standard cages), (2) social rearing (raised with littermates vs. isolated), & (3) conspecific modeling (opportunity to observe a trained demonstrator rat vs. no demonstration). The dependent variables measured exploratory behavior in an open field—time to leave the start square, number of line crosses, & number of object contacts.

Results revealed that all three enrichment factors contributed to exploratory behavior, but they combined in interesting ways. For line crosses (a measure of general locomotor exploration), there was a significant correlation of r = .52 between the number of enriching conditions present (0, 1, 2, or 3) & the amount of exploration—a cumulative effect where each additional enrichment factor added to exploratory behavior. Rats with all three enrichment conditions showed the most exploration, while rats with none showed the least, with intermediate conditions falling between these extremes.

Importantly, conspecific modeling had effects even after controlling for environmental complexity & social experience. Rats who observed a trained demonstrator navigating the open field showed faster latencies to leave the start square & more object contacts than rats who received no demonstration. This demonstrates that observational learning contributes to behavioral development above & beyond the effects of physical & social enrichment—watching a competent model provides information that improves performance.

Tool Use Learning in Primates

Observational learning plays a crucial role in the acquisition of tool use among primates. In wild chimpanzee populations, complex tool use behaviors—such as using sticks to extract termites from mounds, stones to crack nuts, or leaves as sponges to collect water—are passed from generation to generation through social learning. Young chimpanzees spend years observing their mothers & other skilled group members before mastering these techniques themselves.

Recent research by Malherbe & colleagues (2024) followed three communities of wild western chimpanzees over seven & a half years, documenting 1,460 instances of stick tool use. They found that tool-use skill development extends well into adulthood—adult chimpanzees were more adept at choosing the most effective grip for different tasks (power grip for pounding, precision finger grip for insertion). This protracted developmental timeline for tool mastery parallels extended learning periods in humans & suggests that observational learning enables the accumulation of complex skills over many years.

A 2024 study by van Leeuwen & colleagues provided direct experimental evidence that chimpanzees use social learning to acquire skills they fail to independently innovate. Chimpanzees in a sanctuary were given a novel foraging task requiring a specific sequence of actions. Animals who had opportunity to observe trained demonstrators were significantly more likely to acquire the solution than those who attempted to solve the problem through individual trial & error. Twenty-one percent of chimpanzees who observed knowledgeable models successfully learned the task, while virtually none discovered it independently—demonstrating that social learning is necessary for acquiring complex skills that exceed individual innovation capacity.

Social tolerance also influences tool use learning. A 2025 study examining over 2,300 peering events (close-range observation of conspecifics) in wild chimpanzees found that immatures observe many different role models, favoring older & more tolerant individuals. Learning peaked around weaning age & increased with food processing complexity. This suggests that chimpanzees learn from multiple tolerant group members across protracted development, which likely enables acquisition of diverse complex skills including tool use (Malherbe et al., 2025).

Social Learning in Birds: Song Learning & Foraging

Birds demonstrate diverse forms of social learning. Song learning in songbirds represents one of the most extensively studied examples—young birds must hear conspecific songs during a critical period to develop species-typical vocalizations. Interestingly, birds preferentially learn from tutors with similar songs to their own population, & developmental stressors can affect song quality, which in turn affects mate attraction.

The spread of milk bottle opening in British tits (described in Part 1) represents a classic example of foraging innovation spreading through social learning. Slagsvold & colleagues conducted cross-fostering experiments with blue tits & great tits, rearing eggs of one species with parents of the other. Early learning from foster parents caused lifelong shifts in foraging sites—birds preferentially foraged in locations typical of their foster species rather than their biological species. These effects persisted across seasons & were reflected in actual prey selection, demonstrating that social learning fundamentally shapes foraging niches.

Recent research by Zurek & colleagues (2024) demonstrated that social demonstration of color preferences improves learning of associated actions in birds. When birds observed a demonstrator preferring food of a particular color, they more rapidly learned to perform actions associated with that color. This suggests that social information about preferences can facilitate subsequent learning—social attention guides what observers attend to & learn about.

Social Learning in Non-Grouping Animals

While social learning has been extensively studied in group-living species, Webster (2023) reviewed evidence showing that social learning also occurs in non-grouping animals including solitary arthropods, fish, reptiles, & mammals. This finding challenges the assumption that social learning requires complex social structures. Even relatively solitary animals encounter conspecifics during mating, territorial interactions, or aggregation around resources—and these brief social encounters can support observational learning. The benefits of social learning—avoiding costly trial-and-error learning, acquiring information about food, predators, & mates—apply to non-social species as well.

Distinguishing Social Learning Processes

Research distinguishes several processes underlying social learning in animals. Stimulus enhancement increases attention to objects that demonstrators interact with. Local enhancement draws observers to locations where demonstrators have been active. Observational conditioning involves learning stimulus-stimulus associations by observing others’ responses—as when young monkeys develop fear of snakes after observing their parents’ fearful reactions. Emulation involves learning about environmental affordances (what can be done with objects) without necessarily copying the demonstrator’s specific movements. Imitation involves copying the specific motor patterns used by demonstrators.

Determining which process underlies observed social learning requires carefully designed experiments. The “two-action test” presents observers with demonstrators performing one of two different actions to achieve the same outcome. If observers match the specific action they observed (not just the outcome), this suggests imitation rather than mere emulation. Ghost control conditions, in which objects move without a visible agent, help distinguish learning from observing actions versus learning from environmental changes alone.

Looking Forward

Part 3 examines psychological explanations of observational learning, focusing on Bandura’s social-cognitive theory & the four processes required for observational learning to translate into performance: attention, retention, reproduction, & motivation. Understanding these processes helps explain when observational learning succeeds or fails & how it can be enhanced.