Currently, COMIC lab scientists are working on six major programs of research, all of which focus on comparative cognition with an emphasis on nonhuman primates, particularly chimpanzees, capuchin monkeys, and rhesus monkeys, and human children and adults. We sometimes work with other species, such as elephants, bears, and orangutans, but these are our primary research subjects.
The first research program focuses on numerical cognition. Specifically, we are interested in counting and arithmetic skills in nonhuman primates, human adults, and human children. With regard to the question of whether animals are capable of counting behavior, we have approached the question through use of computerized tests of what is called constructive enumeration.
In these tests, chimpanzees have learned to select items on a computer screen, one-at-a-time, until they have accumulated (or constructed) a set equal to a presented target numeral. Chimpanzees are successful on such tasks for numerals up to 8, but performance is indicative of a process more similar to estimation than to formal counting.
The chimpanzees show decreasing performance levels as a function of increasing set size, and they also show greater variability in the size of the constructed set as numeral values increase. This pattern indicates that the enumerative process used by the animals is approximate in its representation of set size.
We have assessed estimation skills in chimpanzees, monkeys, human adults, and human children. In these tests, participants observe as items are placed into and removed from opaque containers so that additive and subtractive operations can be presented. This sequential presentation method has produced compelling similarities in the performance of the nonhuman primates and human children. In addition, recent studies have indicated that articulatory suppression methods can block subvocal counting routines in adult humans, producing data sets that are very similar to those of nonhuman primates and children who have not yet mastered the counting routine. Thus, these comparative data strongly suggest a shared mechanism for the approximate representation of set size for sequentially presented arrays. Modifications of the testing paradigm have demonstrated that chimpanzees also are capable of responding to addition and subtraction manipulations on these arrays, and also that chimpanzees can retain numerical information resulting from enumeration processes for extended time periods (up to 20 minutes). We have examined the link between these types of numerousness judgments and the formal counting skills acquired by children. We have postulated that counting skill is not required for children to be sensitive to arithmetic manipulations as such a sensitivity appears to be widespread phylogenetically. This research is supported by the National Institute of Child Health and Human Development (HD-38051).
Our second research program focuses on metacognition in nonhuman animals. Here, the question is whether animals may experience some sense of knowing what they do or do not know when faced with a decision. This is a difficult thing to assess in animals given the role of verbal reports in our understanding of human metacognition. However, certain experimental procedures have provided some insight into the metacognitive skills of nonhuman animals. This research is in collaboration with David Smith who developed one of the first experimental tasks to assess metacognition in nonhuman animals. In these tasks, monkeys are presented with various psychophysical and memory tasks for which stimuli can be categorized objectively as more difficult or less difficult for the animals based on task performance. Animals also are given an additional response option, called the uncertainty response, that acts in various ways to remove the contingencies of the primary response to the stimulus (e.g., to remove a trial from the screen rather than force the animal to classify a quantity of dots as large or small). In many cases, the animals use that response on exactly those trials for which the primary response (e.g., "large" or "small") is made least efficiently. This suggests that the monkeys may monitor their own knowledge states when faced with decisions about how to respond to stimuli. However, alternate explanations also remain, and we are currently working to separate associative and cognitive explanations for these patterns of results (for example, to determine whether the uncertainty response is used as a result of uncertainty that is felt by the animal, or if it is used because the animal is tracking the reinforcement history of its primary responses in the presence of different kinds of stimuli). We also are examining the extent to which the use of the uncertainty response generalizes to new tasks. This also will provide an indication of the extent to which these responses reflect metacognition.
Other projects, with chimpanzees, have shown that these animals seek-information based on what they know that they know, and they also provide measures of confidence in their own memory abilities (click here for a summary of one recent study). They do this by anticipating food reward for correct responses even before any feedback is given to them. This research is supported by the National Science Foundation (BCS - 0956993 and BCS - 0634662), the National Institute of Child Health and Human Development (RO1-HD061455) and the European Science Foundation (as part of a Eurocore Programme entitled Consciousness in a Natural and Cultural Context). To see a video of a rhesus monkey performing a computerized metacognition task like the one shown above, click here.
Our third research project is conducted in collaboration with Sarah Brosnan (Georgia State) and Bart Wilson (George Mason University) and is concerned with understanding how apes, monkeys, and humans engage in strategic economic interactions. All species play cooperative games that allow us to verify the hypothesis that nonhuman primates and humans share basic economic decision making strategies which take both partner identity and agency, and the potential reward, into account. We have designed more intuitive “hands-on” versions of classic economic experimental tasks which involve exchanges with human experimenters or cooperative barpull paradigms, as well as computerized versions that involve joystick computer tasks. A secondary goal of the project is to use these two methodologies to clarify how nonhuman primates comprehend the “partner” with whom they interact on a computer screen. This research will also help clarify the similarities and differences between nonhuman primates and humans with respect to cooperation in a strategic game. It is vital to understand these interactions in both nonhuman primate and humans in order to more properly place nonhuman behavior in context with human behavior and to understand the roots from which human economic decision making emerged. This research is supported by the National Science Foundation (SES - 0729244, SES - 1123897, and SES - 1425216).
Our fourth research program focuses on the planning skills of primates and the use of prospective memory by primates including humans. People spend a lot of time thinking about the past and the future (what is sometimes called mental time travel). Being able to remember the past, including what, how, and when things happened, can be very helpful in new situations when one is not sure how to behave. Planning for the future, and remembering to carry out those plans, helps people prepare for things that are not immediately important but could be important hours, days, or even years from now. This ability to flexibly plan for the future has long been reserved for humans. In fact, it has been argued that animals are “stuck in time,” and they cannot think about the past or future because their behavior is affected only by their current needs and surroundings. If true, this would indicate a unique aspect of human memory and behavior. However, animals may show capacities for mental time travel, and such evidence would provide a better understanding of the evolutionary foundations of human memory and behavior. This project includes new tests of future-oriented thought and behavior in humans and three primate species (chimpanzees, rhesus monkeys, and capuchin monkeys). This project will test each species’ ability to anticipate future situations and plan future actions so as to determine continuities and discontinuities in the prospective memory and planning abilities of humans and other primates. In some cases, primates may show that they can plan for future situations that are different from present ones, and this performance will be directly compared to human performance.
Failures of prospective memory and failures to plan for the future can have profound consequences for humans. Understanding the causes of such failures is important and can benefit from a broad scientific approach that includes a comparative perspective. This project will provide a better understanding of the evolutionary emergence, as well as the limits, of planning and future-oriented thought and memory in humans and primates. These studies offer new ideas about the nature of primate memory, the beginnings of planned behavior, and the nature of prospective memory. This research will determine similarities and differences between primates and humans in their planning behavior and prospective memory and will help determine whether any of these abilities are unique to humans. The research is supported by the National Science Foundation (BCS - 0924811).
Our fifth research program focuses on the self-control behavior of nonhuman primates. Here, we use delay of gratification tasks in which animals can obtain a more preferred or larger reward by waiting to make a response whereas they obtain a smaller or lesser preferred reward if they make that response. In the most recent studies, we have used a technique in which food items accumulate as long as an animal inhibits consumption of those items. Thus, the longer the animal waits to eat the food items that are accessible, the more food items it can acquire. This rather simple technique has provided compelling evidence that chimpanzees show excellent delay of gratification (sometimes for periods in excess of 10 minutes with very highly preferred food accumulating in front of them). Even rhesus monkeys, traditionally viewed as a highly impulsive species, show some success with this task, and recent projects with a new apparatus, the rotating tray task, has shown that capuchin monkeys with very poor delay of gratification skills can improve when given that task. We also have examined the relation between attentional allocation either to the food items or to other available stimuli and delay maintenance (continued inhibition of the impulsive response). For children, attention to the reward is highly detrimental to delay maintenance, but this appears not to be true for chimpanzees. In some cases, attention to the food items may even facilitate greater delay maintenance, and so we continue to probe this relation as well as other aspects of self-control in these species. For example, we have shown that chimpanzees will use self-distraction to help aid delay of gratification, as in the photo at left where Sherman is looking through a magazine during the delay interval. This research is supported by the National Institute of Child Health and Human Development (HD-060563). To see a story about this ongoing research, click here. For a video discussing self-control using the accumulation test and self-distraction (as Sherman demonstrates above), click here.
Our sixth program of research is concerned with how primates (including children and adults) see the world, and how and when they fall prey to perceptual and cognitive illusions. We are interested in many illusions, including the Solitaire illusion (top left image, where many people would respond when asked that there are more white dots in the array at the left, even though both arrays have 16 dots of each color), the Delboeuf illusion (bottom left image), where central dots (or servings of food as at bottom right) appear larger when surrounded by a close ring or small plate compared to a larger ring or larger plate, and the less-is-more effect where containers that are more overflowed appear to hold more (top right image, where the number of marshmallows in each container is actually the same, but many people think the one at the left has more - something chimpanzees also think!). For a recent news release on some of this research, click here. We also are interested in how context affects perception and choice behavior. Specifically, we are studying the role of anchoring effects, decoy effects, and similar effects that may generate heuristics in primates and generate choice biases.
Other ongoing projects include assessing the long term retention of symbol vocabularies in three language-trained chimpanzees. This project (now entering its second decade) is one of the longest running systematic evaluations of symbolic vocabulary retention in nonhuman animals. It seeks to provide an objective, consistent evaluation of exactly how many symbols these animals know and remember throughout their lives. It offers empirical evidence of this long-term retention of symbolic capacity rather than generic statements about vocabulary size and breadth that may overestimate (or underestimate) the true symbolic capacities of animals. For a paper summarizing this long term project, please click here.