Four-year-old Rodney and his 2-year-old sister Margie had always been told to keep away from the stove. Being a mischievous imp, however, Rodney decided one day to place his hand over a burner until the singeing of his flesh led him to recoil, shrieking in pain. Rodney was more scared than hurt, really—and he learned something important that day. But little Margie, who stood by watching these events unfold, also learned the same lesson. Rodney’s story is a behaviorist’s textbook example: The administration of punishment led to a learned change in the boy’s behavior. But how can we explain Margie’s learning? She received neither punishment nor reinforcement, and yet it’s arguable that she’s just as likely to keep her hands away from stoves in the future as Rodney is.

Margie’s is a case of observational learning, in which learning takes place by watching the actions of others. In all societies, appropriate social behavior is passed on from generation to generation, not only through deliberate training of the young, but also through young people observing the patterns of behaviors of their elders and each other (Flynn & Whiten, 2008). Tasks such as using chopsticks or learning to operate a TV’s remote control are more easily acquired if we watch these activities being carried out before we try ourselves. Even complex motor tasks, such as performing surgery, are learned in part through extensive observation and imitation of models. And anyone who is about to undergo surgery is grateful for observational learning. Just the thought of a generation of surgeons acquiring their surgical techniques using the trial-and-error techniques studied by Thorndike or the shaping of successive approximations that captivated Skinner would make any of us very nervous.

In a series of studies that have become landmarks in psychology, Albert Bandura and his colleagues investigated the parameters of observational learning (Bandura, Ross, & Ross, 1961). The researchers escorted individual preschoolers into a play area filled with toys that 4-year-olds typically like. An adult model, someone whose behavior might serve as a guide for others, then entered the room and started playing with a Bobo doll, which is a large inflatable plastic toy with a weighted bottom that allows it to bounce back upright when knocked down. The adult played quietly for a bit but then started aggressing toward the Bobo doll, knocking it down, jumping on it, and kicking it around the room. When the children who observed these actions were later allowed to play with a variety of toys, including a child-size Bobo doll, they were more than twice as likely to interact with it in an aggressive manner as a group of children who hadn’t observed the aggressive model (see FIGURE 7.13).

231

The children in these studies also showed that they were sensitive to the consequences of the actions they observed. When they saw the adult models being punished for behaving aggressively, the children showed considerably less aggression. When the children observed a model being rewarded and praised for aggressive behavior, they displayed an increase in aggression (Bandura, Ross, & Ross, 1963). The observational learning seen in Bandura’s studies has implications for social learning and cultural transmission of behaviors, norms, and values (Bandura, 1977, 1994).

Observational learning is important in many domains of everyday life. Sports provide a good example. Coaches rely on observational learning when they demonstrate critical techniques and skills to players, and athletes also have numerous opportunities to observe other athletes perform. Studies of athletes in both team and individual sports indicate that all rely heavily on observational learning to improve their performance (Wesch, Law, & Hall, 2007). But can merely observing a skill result in an improvement in performing that skill without actually practicing it? A number of studies have shown that observing someone else perform a motor task, ranging from reaching for a target to pressing a sequence of keys, can produce robust learning in the observer. In fact, observational learning sometimes results in just as much learning as practicing the task itself (Heyes & Foster, 2002; Mattar & Gribble, 2005; Vinter & Perruchet, 2002).

232

Humans aren’t the only creatures capable of learning by observing. In one study, for example, pigeons watched other pigeons get reinforced for either pecking at the feeder or stepping on a bar. When placed in the box later, the pigeons tended to use whatever technique they had observed other pigeons using earlier (Zentall, Sutton, & Sherburne, 1996).

One of the most important questions about observational learning in animals concerns whether monkeys and chimpanzees can learn to use tools by observing tool use in others, the way young children can. In one of the first controlled studies to examine this issue, chimpanzees observed a model (the experimenter) use a metal bar shaped like a T to pull items of food toward him or her (Tomasello et al., 1987). Compared with a group that did not observe any tool use, these chimpanzees showed more learning when later performing the task themselves. In a later experiment, the researchers introduced a novel twist (Nagell, Olguin, & Tomasello, 1993). In one condition, a model used a rake in its normal position (with the teeth pointed to the ground) to capture a food reward, which was rather inefficient because the teeth were widely spaced and the food sometimes slipped between them. In a second condition, the model flipped over the rake so that the teeth were pointed up and the flat edge of the rake touched the ground—a more effective procedure for capturing the food. Both groups that observed tool use performed better when trying to obtain the food themselves than did a control group that did not observe a model using the tool. However, the chimpanzees that observed the more efficient procedure did not use it any more often than did those that observed the less efficient procedure. By contrast, 2-year-old children exposed to the same conditions used the rake in the exact same way that they had seen the model use it. The chimpanzees seemed only to be learning that the tool could be used to obtain food, whereas the children learned something specific about how to use the tool.

The chimpanzees in these studies had been raised by their mothers in the wild. Could chimpanzees that had been raised in environments that also included human contact learn to imitate the exact actions performed by a model? The answer is a resounding yes: Chimpanzees raised in a more humanlike environment showed more specific observational learning than did those that had been reared by their mothers, and thus they performed much like human children (Tomasello, Savage-Rumbaugh, & Kruger, 1993). This finding led Tomasello et al. (1993) to suggest that being raised in a human culture has a profound effect on the cognitive abilities of chimpanzees, especially their ability to understand the intentions of others when performing tasks such as using tools, which in turn increases their observational learning capacities.

More recent research has found something similar in capuchin monkeys that are known for their tool use in the wild, such as employing branches or stone hammers to crack open nuts (Boinski, Quatrone, & Swartz, 2000; Fragaszy et al., 2004) or using stones to dig up buried roots (Moura & Lee, 2004). Fredman and Whiten (2008) studied monkeys that had been reared in the wild by their mothers or by human families in Israel as part of a project to train the monkeys to aid quadriplegics. A model demonstrated two ways of using a screwdriver to gain access to a food reward hidden in a box. Some monkeys observed the model poke through a hole in the center of the box, whereas others watched him pry open the lid at the rim of the box (see FIGURE 7.14). Both mother-reared and human-reared monkeys showed evidence of observational learning, but the human-reared monkeys carried out the exact action they had observed more often than the mother-reared monkeys.

Although this evidence implies that there is a cultural influence on the cognitive processes that support observational learning, the researchers noted that the effects on observational learning could be attributed to any number of influences on the human-reared monkeys, including more experience with tools, more attention to a model’s behavior, or increased sensitivity to the intentions of others. Thus, more work is needed to understand the exact nature of those processes (Bering, 2004; Tomasello & Call, 2004).

233

Observational learning involves a neural component as well. As you read in the Neuroscience and Behavior chapter, mirror neurons are a type of cell found in the frontal and parietal lobes of primates (including humans). Mirror neurons fire when an animal performs an action, such as when a monkey reaches for a food item (FIGURE 7.15). Mirror neurons also fire when an animal watches someone else perform the same task (Rizzolatti & Craighero, 2004). For example, monkeys’ mirror neurons fired when they observed humans grasping for a piece of food, either to eat it or to place it in a container (Fogassi et al., 2005). Although the exact functions of mirror neurons continue to be debated (Hickok, 2009), it seems likely that mirror neurons contribute to observational learning.

Mirror neurons exist in humans too. Studies of observational learning in healthy adults have shown that watching someone else perform an action engages some of the same brain regions that are activated when people actually perform the action themselves. In one study, participants practiced dance sequences to unfamiliar songs and watched videos containing other dance sequences accompanied by unfamiliar songs (Cross et al., 2009). The participants were then given fMRI scans while viewing videos of sequences that they had previously danced or watched, as well as videos of untrained sequences.

Results showed differences between watching the untrained sequences and watching the previously danced or watched sequences. Viewing the previously danced or watched sequences caused activity in brain regions considered to be part of the mirror neuron system. The results of a surprise dancing test given to participants after the conclusion of scanning showed that performance was better on sequences previously watched than on the untrained sequences, demonstrating significant observational learning, but it was best of all on the previously danced sequences (Cross et al., 2009). So, while watching Dancing with the Stars might indeed improve your dancing skills, practicing on the dance floor should help even more.

234