Can dogs pretend? This is the question I asked yesterday, prompted by Sheril's story:

...this afternoon Happy did something unusual. She carried a toy frog over to her water bowl, and gently put it down as pictured. Given its orientation, I’m skeptical that her placement was an accident.

The frog continues to sit like this (20 minutes later) as Happy arranges other toys nearby. Now I admit I may be anthropomorphizing, but her behavior sure reminds me of a child playing “make believe” with stuffed animals.

Any pet owner has, from time to time, been charmed and entertained by their furry critters. And that's a good thing! But the empirical question of whether or not animals in general (and dogs in particular) can engage in pretend play was particularly interesting to me. More precisely, does Happy understand that the toy frog isn't really a frog, but instead is an object that represents a category of agents (frogs) - and then, is she pretending that the toy really is a frog, even though she knows it isn't?

Pretend Play in Human Children

Young children do this with ease. They know that a banana isn't a telephone, for example, but since bananas share some common characteristics with telephones, they can pretend that one is the other. In other words, in addition to the primary mental representation of the true identity of the object (banana), they also represent the object's pretend identity (telephone). At what age are humans able to engage in this sort of play?

Young infants try to grasp the objects in pictures, but by nineteen months, their grasping is replaced by pointing - this suggests that by that age, human infants understand that a picture is a representation of something, rather than the thing itself. However, another study found that three year-olds, but not four year-olds, believed that a balloon represented on a TV screen would fly to the ceiling of a room if the top of the TV was removed. So, children between the ages of 18 months and three years might be able to pretend, but they seem to be limited in their ability to fully understand the concept of pretending. Many developmental psychologists would argue that the child need not have the meta-awareness of pretending in order to pretend in the first place; they don't need to know that they are pretending.

Can Animals Pretend?

Empirically investigating pretend play is exceptionally hard, because pretend play is by its nature spontaneous. It's only through direct observation and case studies that we have any evidence for pretend play in animals at all - and so far, only really in apes. Even still, the scant evidence that exists for pretend play in non-human apes is fairly consistent.

In 1993, primatologists Jensvold and Fouts analyzed fifteen hours of video footage of five chimpanzees, specifically looking for instances of "imaginary play." These particular chimpanzees (including the famous Washoe - pictured left) had been trained in the use of American Sign Language, which must have helped the researchers in understanding their actions. In the fifteen hours of footage, six instances of imaginary play were documented. Several of them were actions in which the chimps gave living characteristics to an inanimate objects. For example, the chimpanzee Dar signed "tickle" to a stuffed bear. Similarly, Washoe used a natural action called a "hit away," which is usually used with other chimpanzees. The researchers note that "when Washoe made the gesture she had a play face which also indicated the playfulness and nonliteral nature of her actions." The rest of the instances of pretend play in these chimps featured substitution, in which one object was treated as if it was something different. For example, Washoe treated as brush as if it was a book. This was confirmed by the fact that "Washoe often had been observed placing a book under her arm and holding it there which was the same action she used with the brush. Thus her actions and her signing of "book" afterwards were indicative of her substitution." And Mosha treated a purse as if it was a shoe, including signing the word "shoe."

Another researcher described a chimpanzee using a shell first as a cigarette, then as an earring. Jensvold and Fouts recount Hayes's 1952 description of the actions of a home-reared chimpanzee named Viki:

Very slowly and deliberately she was marching around the toilet, trailing the fingertips of one hand on the floor. Now and then she paused, glanced back at her hand, and then resumed her progress... During the next couple of days Viki often played this new game, but now she paused frequently to make sure that I was not watching her...Viki was at the pulltoy stage when a child is forever trailing some toy on a string, when everything with a string attached becomes a pulltoy. Dragging wagons, shoes, dolls, or purses, her body assumed just this angle. She trudged along just this busily on two feet and one hand, while the other arm extended backward this way to pull the toy. Viki had an imaginary pulltoy!

Hayes described times when Viki acted as though the toy had gotten stuck on something and she tugged on the invisible string until it came undone as evinced by a little jerk and Viki continuing on her way. On one occasion Hayes reported that Viki acted as if the toy had become stuck. Hayes wrote "she sat down abruptly with her hands extended as if holding a taut cord. She looked up at my face in the mirror and then she called loudly, 'Mama! Mama!'." Hayes pretended to unwrap the cord. Viki then tore off again in this same imaginary pulltoy position.

The bonobo Kanzi (pictured right, with Savage-Rumbaugh) engaged in a different sort of pretend play. Savage-Rumbaugh and McDonald wrote:

Kanzi frequently pretends to hide invisible objects in piles of blankets or vegetation. Later he will take them out and pretend to eat them... Kanzi also engages the participation of others in the "invisible objects" games by giving them the pretend object and then watching to see what they do with it.

Is language a prerequisite for imagination?

In 1898, psychologist Karl Groos wrote, "If we could be certain that apes treat lifeless objects as dolls, this act would be in the foremost rank of illusion plays....But we cannot be sure of these things, for speech is wanting in these creatures." But is this really true? Is there not a way to determine if an animal is pretending outside of direct linguistic evidence?

Some possible instances of play have been observed in wild chimpanzees. Jensvold and Fouts recount a story told by Jane Goodall:

...a 4-year-old chimpanzee, Wanda, who had been watching her mother, who was perched on a branch above a termite hill, dip a stick into the insects' hole and pull it out loaded with termites. Wanda then picked up a small twig, perched herself on a sapling branch, and poked her stick in a downward direction. A similar instance of imaginary play is very common in human children using cups, saucers, pots, and toy stoves to pretend to prepare and serve a meal. In these instances a child uses adult tools to go through the motions of a common adult activity, be it pots for cooking or twigs for dipping, these are analogous behaviors.

They also recount an instance of pretend play in a wild chimpanzee from Richard Wrangham:

An 8-year-old chimpanzee (Kakama), traveling with his pregnant mother, picked up a log and carried it for hours, treating it suspiciously like a baby (e.g., he made a nest and placed the log in it). Four months later, two field assistants, who were not aware of the previous incident, observed similar behavior by the same chimpanzee, collected the new log, and labeled it "Kakama's toy baby."

Language is probably not necessary for imaginary play, since wild chimpanzees appear to engage in at least some forms of pretending as well.

But what about dogs?

Comparative psychologist Marc Bekoff (blog) notes that

Dogs and cats often chase their tails and bite them as if the tail was prey or a competitor, and they frequently attempt to kill their food bowl or mate with it. Some dominant animals also engage in self-handicapping or role-reversing during play, in which they seem to pretend that they are not really dominant, perhaps to get others who would otherwise not play with them to play with them.

In addition, canids such as dogs, wolves, and coyotes, use behaviors taken from mating, hunting, and fighting in order to play. However, it is not clear if they are simply calling upon a set of familiar behaviors, or if they are actually thinking about attacking or mating with their play partner. However, as I noted above, the criterion by which an individual must be aware of the nature of pretend play in the first place is not applied to human children - therefore, it ought not be applied to animals.

Taken together, it seems as if some animals, under some circumstances, do indeed engage in pretend play. They treat some objects as if they were different objects (as with Washoe and her brush), they act is if objects are present where they are not (as with Kanzi and his food), or they use behaviors relevant to one activity (fighting or mating, in dogs) and repurpose them for play. However, animals might not possess the meta-awareness that they are pretending. They might not understand the very concepts of "pretend" or "imagination," the way we as humans do. In one study, researchers asked children if a troll that hopped like a rabbit was pretending to be a rabbit, even if he had never seen or heard of rabbits in the first place. Only older children, beginning around age 4, understood that pretending is a function of the mind, rather than of the body. This is consistent with research on the development of theory of mind. So the extent to which animals are able to make believe is probably similar to the way that a child younger than age four can pretend.

Still, some animals seem able to imagine, even if in a limited way. And that's pretty cool.

Jensvold, M., & Fouts, R. (1993). Imaginary Play in chimpanzees (Pan troglodytes) Human Evolution, 8 (3), 217-227 DOI: 10.1007/BF02436716

Suddendorf T, & Whiten A (2001). Mental evolution and development: evidence for secondary representation in children, great ages, and other animals. Psychological bulletin, 127 (5), 629-50 PMID: 11548971

Lillard AS (1993). Pretend play skills and the child's theory of mind. Child development, 64 (2), 348-71 PMID: 8477622

Flavell, J., Flavell, E., Green, F. L., & Korfmacher, J. E. (1990). Do young children think of television images as pictures or real objects?. Journal of Broadcasting & Electronic Media, 34(4), 399-419.

Bekoff, Marc (1997) Playing with Play: What Can We Learn About Cognition, Negotiation, and Evolution? D. Cummins and C. Allen (eds.) The Evolution of Mind. New York. Oxford University Press

DeLoache, J., Pierroutsakos, S., Uttal, D., Rosengren, K., & Gottlieb, A. (1998). Grasping the Nature of Pictures Psychological Science, 9 (3), 205-210 DOI: 10.1111/1467-9280.00039