Unsure Minds

People may not be the only ones who know when they don't know

A cat crouches on a kitchen floor, gazing up at a glass of milk high on a counter. The animal’s muscles tense. Its tail bobs from side to side like a metronome. The distance from floor to counter is a long way to cover in a single feline leap, perhaps too long. With a slightly cocked head, the cat emits a tentative meow. Finally, the animal springs for the counter. Half a world away, African monkeys moving through a stand of trees spot a few chimpanzees in the distance. The monkeys freeze, staring intently at the powerful apes, which sometimes kill and eat monkeys. Staccato chattering erupts in the monkey troop as the chimps move closer. Suddenly, the monkeys scatter.

SCREEN TIME. Videotape image shows a monkey perusing a computer screen challenge and reaching for a joystick to indicate its response. Washburn

WILD THOUGHTS. The late Donald Griffin spent decades defending the controversial idea that conscious thoughts occur in animals other than people. Kris Snibbe/Harvard News Office

As these two cases illustrate, situations arise in which animals act as if they need to make decisions but are uncertain what to do. From a scientific perspective, though, it’s hard to know whether cats, monkeys, or any other creatures truly experience a sense of uncertainty. The capacity to think about one’s own thoughts has long been regarded as a trait unique to humanity. In this view, only we creatures with a gift for gab can truly appreciate life’s uncertainties and wallow in self-doubt.

Some researchers studying other big-brained animals, however, now have qualms about that view. In experiments, at least two rhesus monkeys and one bottlenose dolphin have shown signs that they experience uncertainty when trying to make difficult perceptual distinctions or to recall hazy memories, says psychologist John David Smith of the State University of New York at Buffalo.

Smith and his colleagues, Wendy E. Shields of the University of Montana in Missoula and David A. Washburn of Georgia State University in Atlanta, survey their own and others’ research on uncertainty among animals in the December 2003 Behavioral and Brain Sciences. In 21 commentaries on that article, published in the same journal, scientists present a wide range of opinions about the survey’s findings.

One camp views the data, as Smith and his colleagues do, as a sign that some nonhuman animals consciously think about uncertain situations and perhaps experience some form of self-doubt.

Another camp asserts that animals never reflect on their actions. If they run into a situation where the best response isn’t obvious, they may simply walk away from it, just the way they might respond to unpalatable food. Human observers may mistakenly consider that behavior as an expression of doubt.

A third group doubts that scientists will ever devise studies that unravel the minds of nonverbal creatures. Monkeys, dolphins, and other animals can’t say, “I don’t know,” much less, “Hmm, your guess is as good as mine.” So, scientists are left to rely on inherently arguable interpretations of nonverbal behaviors.

In their article, Smith and his coauthors argue that recent experimental approaches tap into mental states that animals frequently experience in the wild. “Responses to some situations, such as finding a regular water hole empty or a seeing a predator positioned near one’s [offspring], can’t be left to automatic pilot,” Smith says. “We don’t know that a monkey or a dolphin feels like a self who is uncertain, but these animals feel a sense of uncertainty.”

Think again

For more than a century, psychologists have theorized that moments of mental conflict and uncertainty evoke reflection in people, and possibly in many animals as well. While people can tell you whether they’re unsure about something, animals can’t. That has been the experimentalist’s challenge.

In one study by Smith and his coworkers, two rhesus monkeys learned to use a joystick to position a cursor on a square riddled with illuminated dots that was displayed on a computer monitor.

Using squares containing considerably fewer than 3,000 dots, the researchers trained the animals to move the cursor away from the square itself to a letter S on the same computer screen.

Each correct response yielded a food pellet. A mistake resulted in a wait lasting about 20 seconds before the next square appeared with a different dot configuration.

Once the number of dots in the square reached 2,600 or so, the monkeys started to look at the screen longer and showed signs of uneasiness. For squares with more than 2,600 but fewer than 3,000 dots, they usually placed the cursor on a star at the bottom of the screen. This response elicited no food reward but did immediately bring on the next trial. The researchers regard pointing to the star as an indication of uncertainty.

The researchers then used the same test on people, but substituted money for pellets. The eight volunteers showed the same pattern of responses that the monkeys did. Again, squares that contained slightly fewer than 3,000 dots generated the bulk of uncertainty responses.

In another animal experiment, a bottlenose dolphin was trained to press one lever when it heard a low-frequency tone and another lever when it heard a high-frequency tone. A third lever immediately brought the next trial.

As the low tone approached the frequency of the high tone, the dolphin started to hesitate and move back and forth between levers. At that point, the animal usually pressed the third lever.

A group of people asked to make the same audio judgments performed much as the dolphin did. They reported having experienced much uncertainty before pressing the third lever.

Another experiment examined uncertainty about memory rather than sensory cues. The same monkeys that had previously performed the dots-in-the-box test were trained on a new task. The animals viewed four objects presented one at a time on a computer screen, after which the screen briefly went blank and a single item appeared. Each monkey used the cursor either to indicate that the item had been part of the original group or to choose the star.

The monkeys often remembered items that had appeared first or last in a previous series. For those items shown second or third in that series, the monkeys often chose the star.

People performing a similar task showed the same pattern of responses and reported feeling uncertain about the middle items.

Psychologist Robert R. Hampton of the National Institute of Mental Health in Bethesda, Md., has conducted another set of monkey studies that he says indicates that animals recognize their own uncertainty. After viewing objects on a computer screen, the animals could participate in memory tests to attempt to win a peanut. If they gave a wrong answer, they got no reward. However, if they pressed a key to decline participation, they received a food pellet, which monkeys consider less desirable than a peanut. Ensuing tests confirmed that the monkeys rarely remembered items that they had declined to be tested on.

Further experiments on monkeys are under way. So far, rats and pigeons haven’t shown evidence of experiencing uncertainty, Smith adds.

Uncertain or not

Among scientists who agree that nonhuman animals may experience uncertainty, there’s disagreement about whether that uncertainty would indicate a rich mental life. Smith and his colleagues offer an intriguing new way to look for mental similarities between people and other animals, remarks psychologist Lori Marino of Emory University in Atlanta. Since monkeys and dolphins appear to report on their own uncertainty when given a means to do so, it’s reasonable to regard conscious thought as a “highly adaptive general property of many species,” she holds.

Marino has presented evidence that dolphins may possess a sense of self, an example of high-level consciousness. Marino and a coworker reported in 2001 that two dolphins used a mirror placed next to their pool to inspect temporary ink marks on parts of their bodies that they couldn’t see directly. There’s much debate about whether apes and monkeys are capable of this behavior, which is considered a cardinal sign of having a self-concept (SN: 1/20/96, p. 42).

Although nonhuman animals and human infants may experience conscious feelings of uncertainty, only human adults think about those feelings, proposes psychologist John H. Flavell of Stanford University. Researchers currently know “next to nothing” about what infants do think about, he says.

The same paltry state of knowledge applies to monkeys and dolphins, contends psychologist Janet Metcalfe of Columbia University.

Research on how people think typically considers what psychologists call feeling-of-knowing judgments, in which volunteers who can’t answer a question—say, “Who is the prime minister of Canada?”—rate the likelihood they will in fact recognize the answer if they see it later among a number of choices. Participants make these predictions with considerable accuracy.

Most of the studies cited by Smith and his colleagues in the Behavioral and Brain Sciences article tested animals’ recognition of previously learned material but not their internal feeling of whether they knew the material or not, Metcalfe says.

However, in one study by Shields, the two monkeys were expected to indicate whether they knew that they could remember unseen shapes previously associated with those shown on a screen.

Each monkey in this test viewed pairs of various shapes and learned to associate the members of each pair. The animal next saw one shape and either put the cursor on it to indicate memory of the associated shape or put the cursor on a star, theoretically showing uncertainty.

The monkey then saw the associated shape along with a shape that had not been previously paired with it or its mate. If the monkey pointed to the previously associated shape, it received food pellets. Errors resulted in a brief wait before getting another shot at a food pellet, unless the animal had first signaled uncertainty.

There was little difference in the accuracy of the subsequent response when the monkey had or hadn’t pointed to the star. Therefore, that action couldn’t be taken as evidence that the animal had thought about what it did and didn’t know, Shields concluded.

These findings support the view that nonhuman animals rarely apply a conscious analysis to uncertain situations, argues Peter Carruthers, a philosophy professor at the University of Maryland at College Park.

A feeling of anxiety or an awareness of raw reactions to uncertainty, such as tilting the head at an angle or wrinkling a brow, could trigger rote responses, he says. For instance, a dog offered a treat by a stranger might stop in its tracks and then flee as an automatic response.

The possibility that animals don’t think through their responses to uncertainty leaves Smith unfazed. He hopes to expand his research on monkeys and perhaps conduct experiments with chimpanzees.

There’s little doubt in his mind that the cross-species study of uncertainty is heating up.

The Thinking-Animal’s Man

Science loses a champion of creature consciousness

In one of his last published scientific writings, zoologist Donald R. Griffin of Rockefeller University in New York endorsed the proposal that animals other than people reflect on uncertain situations. “If animals are aware of anything, the many uncertainties that are critical for survival must often require conscious attention,” Griffin reasoned.

This controversial opinion was hardly surprising coming from Griffin, who died on Nov. 7, 2003, at age 88. For 25 years, he had championed the view that many animal species think consciously about their actions and feelings, even if such thoughts differ in many ways from those of people. Scientists often rejected Griffin’s argument, especially in the early years. Many theorized that nonhuman animals react to their surroundings without actually thinking. Even if animals think, others said, it’s possible to study only behavior, not minds.

Griffin would have none of it. In the wild, he contended, animals’ social behaviors betray a wide range of conscious thought inspired by uncertainty. Even honeybees exhibit a basic form of awareness, Griffin held. For example, those that have scouted reasonably suitable cavities for a swarm to move into will, on returning home, waggle about using gestures to indicate the distance, direction, and desirability of an inspected site. A process of communal decision making then follows, as scouts peruse the gestures of fellow scouts and drop their own to copy those that describe better candidate cavities. In Griffin’s view, this behavior reflects each scout’s awareness, at some fundamental level, of not being sure of which bee has found the best location.

Griffin first made his scientific mark 60 years ago with another winged creature. He and a coworker discovered that bats listen to the echoes of their own voices to orient themselves and to detect objects. Griffin dubbed the phenomenon echolocation.

Griffin’s influence continues to echo among scientists who study animals’ mental lives. “I knew him and greatly admired him,” remarks chimp researcher Frans B.M. de Waal of Emory University in Atlanta. “He was a great scientist [who] always preferred what animals do naturally over the many tricks that we humans can teach them.”

Bruce Bower has written about the behavioral sciences for Science News since 1984. He writes about psychology, anthropology, archaeology and mental health issues.