Dogs Don't, But Maybe Chimps Remember

Dogs Don't, But Maybe Chimps Remember

Posted May 07, 2010

The chimpanzee was always my favorite character in those old Tarzan movies. I'm convinced he was the smartest one involved. Chimps have demonstrated cognitive abilities similar to human abilities. Perhaps chimps, unlike dogs, can remember what happened yesterday.

As I noted in my previous post (Dogs Don't Remember), Endel Tulving has forcefully argued that human cognition is unique because only humans have episodic memory. Two cognitive abilities underlie episodic remembering: a clear sense of self and mental time travel. Episodic remembering involves recollecting the self in a different time and place.

Since chimpanzees are our closest evolutionary relatives, they may have the best chance of demonstrating episodic remembering. Chimpanzees demonstrate a variety of high level cognitive abilities. They can recognize themselves in mirrors. They work together in social groups. They engage in observational learning. They can learn to use symbols to communicate with caregivers. Chimps are impressive cognitive beings.

Charles Menzel, in his chapter in "The Missing Link in Cognition" (edited by Terrace and Metcalf), presented compelling evidence of location memory by chimps. In one study, a chimp watched while several items, such as preferred foods, were hidden in a very large area. When later released into the area, the chimp would take an efficient route to retrieve most of the items. In addition, Menzel described research with Panzee, a chimp skilled in the use of a symbol board to communicate (see picture of Panzee). One experimenter would hide an item in a visible, but unreachable, location. Later, often more than one day later, Panzee would direct a different caregiver to retrieve the item by using her symbol board (pointing to the item symbol) and pointing to the location. She gave excellent pointing directions and almost always used the correct symbol.


Bennett Schwartz, in the same edited volume, reported on a series of studies with King, a gorilla (see picture). King had previously solved mirror identification problems and so had demonstrated the self awareness needed for episodic remembering. King was able to report past events, via cards representing people and objects. King selected cards denoting which food had been to him by which experimenter. He performed at above chance levels at both 5 minute and 24 hour delays. Pretty impressive: King reported the past information without using it to solve a current problem.


Other animals also display excellent location memory. For example, Clayton and Dickinson have shown that Scrub Jays can remember where they cached items. In addition, they demonstrate an understanding of what is hidden because they have preferred foods (such as grubs, no accounting for taste). In addition, the Jays must use time in some fashion because they avoid preferred foods if it has been long enough for the food to go bad (grubs don't stay fresh forever). Even the lowly dog can retrieve a buried bone.

So is location memory evidence for episodic remembering? Animals remember information from a single experience. They remember where and what information. They independently retrieve the information. In addition, Panzee and King communicated the information to someone else.

We can't, however, be sure if they are relying on episodic memory. When I'm at home, I know where my car keys are even if I don't remember the experience of putting my keys away. I have a pretty good idea of the foods stored in my fridge and know one or two that should be tossed even without the sniff test (so I'm at least as smart as a Scrub Jay). Sometimes I recall the food information without bothering to recollect the experience of putting the leftovers in the containers. I know where an object is relying only on semantic memory. Similarly, the animals may solve location memory problems without using episodic memory. For these reasons, Tulving has continued to doubt that location memory is evidence for episodic memory in non-humans.

I'm worried about anthropomorphism no matter how one interprets the evidence. Some people think they personally solve location memory problems by episodic remembering. They find objects by recollecting the experience of putting the objects in locations. When they see an animal perform a similar task, they believe the animal has a similar mental experience. Other people realize they can solve location memory problems without episodic remembering. They know where something is, without remembering putting the object there. They argue animals do the same. Both answers rely on attributing a personally experienced mental state to a non-human animal.

Non-human animals share many cognitive abilities with us. Many of the differences are in quantity, not quality. Some animals may engage in simple forms of episodic remembering. I think part of the problem is approaching episodic remembering as an all or none cognitive ability. Toss away that underlying assumption and the answer to animal cognition may be less controversial.