Research

Research in the Concepts & Cognition lab focuses on a variety of questions concerning conceptual representation and reasoning. Some of the major topics addressed by on-going research include categorization, judgments about causation, and the nature of explanations. You’ll find brief descriptions of some on-going projects below.

Explanatory preferences. When confronted with something we don’t understand, most of us can’t help but wonder “why?” Moreover, we have strong intuitions about what counts as a satisfying explanation. We seem to prefer explanations that are simple and provide a reason for what we’re trying to explain. This line of work investigates what constitutes an explanation, and in particular why explanations in terms of reasons, functions, or goals seem to be preferred. What’s the relationship between these so-called functional explanations and our causal beliefs about the world? How does reasoning about an object’s function change the way we categorize and reason about it? Might a preference for functional explanations help explain the appeal of religious ideas like creationism?

Explanation and probability. Many real-world decisions involve assessing probability: Is it more likely a Republican or a Democrat will win the next presidential election? Is my congestion due to allergies or a cold? This line of work investigates how we assess the probability of claims, and in particular the hypothesis that probability judgments are informed by evaluating explanations. First, is a claim judged more probable if it is easy to explain? And second, is a claim judged more probable if it provides a good explanation for something else? These issues relate to Inference to the Best Explanation in philosophy.

Explanation and learning. Explanation and learning are intimately related. We not only learn by receiving explanations, but also by generating explanations, whether for oneself or for someone else. Research in educational psychology has found that explaining can increase understanding, a finding known as the “self-explanation effect.” This line of work investigates how explaining affects learning. Why and when is explaining an effective mechanism for learning? Are there conditions under which explaining doesn’t help learning, and if so, why? So far, our research has focused on explaining in the context of learning about novel categories, but we plan to extend this line of work to other contexts, applying the tools of cognitive psychology to help account for the relationship between explanation and learning.

Explanation and inference. The ability to generalize from the known to the unknown is fundamental to learning and inference. For example, given that a particular species of mushroom contains a fatal toxin, accurately inferring which other species are poisonous has obvious advantages. But the basis for making such generalizations is far from obvious. In this project, we investigate the hypothesis that explanations guide the generalization of properties from known to unknown cases. If a mushroom’s toxin is explained as the product of a particular metabolic process, for example, other mushrooms sharing that metabolic process are likely to be judged poisonous. But if the toxin is instead explained as a biological adaptation to deter fungivores, other mushrooms facing similar predators are likely to be judged poisonous. How do explanations guide this process of inference?

Explanation and causation. Within philosophy and psychology, two very different ways of thinking about causation have been proposed. According to one approach, A is a cause of B if B counterfactually depends upon A in the appropriate way. According to a second approach, A is a cause of B if there was an appropriate physical connection or transfer of force between A and B. Might both of these ways of thinking about causation have some psychological reality? In this line of work, we’ve investigated whether different kinds of explanations promote causal judgments that are more or less consistent with these two approaches to causation. In particular, we’ve examined whether teleological explanations are more consistent with a dependence approach, and mechanistic explanations with a physical connection approach.

Explanation and categorization. Several researchers have proposed that explanation and categorization are closely related — deciding which category an item belongs to may be a matter of identifying which category membership would best explain its features, different items may be grouped into a common category by virtue of sharing common explanations, and the explanatory relationships between category features may influence the relative importance of those features. This line of work considers how and why explanations influence categorization and conceptual representation.

Theory of mind and moral reasoning. Traditionally, theory of mind — our capacity to reason about people’s beliefs, desires, and other mental states — has been studied relatively independently of moral reasoning — our capacity to evaluate obligations, rights, and moral responsibility. More recently, several researchers have proposed that theory of mind and moral reasoning are intimately related, and perhaps even that moral evaluations influence how theory of mind judgments are made. In this line of work, we’re exploring the relationships between theory of mind and moral judgments.

Intuitions about reference. How do words and thoughts refer to aspects of the world? In particular, what determines the individual or category a word or thought refers to? These questions have traditionally been explored within philosophy, where so-called descriptivist and causal theories of reference have predominated. More recently, experimental philosophers have examined folk intuitions about reference and considered their implications for philosophical claims. In this line of work, we’re investigating the conditions under which intuitions are more descriptivist or more causal, and how these two approaches might be reconciled into a more unified theory.

Why study explanation?

As you’ll note, a majority of the projects in the lab involve explanations. Studying explanations often, well…requires explanation. Explanation is a relatively new area of study in cognitive psychology. In many ways, cognitive psychology is a latecomer to explanation. Philosophers of science have been thinking about explanation for decades, and social psychology has a large literature on how we explain our own and other people’s behavior.

One argument for studying explanation comes from its ubiquity. We spend a great deal of time seeking, generating, and evaluating explanations. If you eavesdrop on others or attend to your own conversations, it won’t be long before you hear an explanation, or at least a request for one. Why is your roommate angry? Why did the cake turn out too dry? Why is there traffic at 3pm? And yes, even “why are we here?” The tendency to seek explanations is so pervasive that some psychologists have posited an “explanation urge” (Kosslyn, 1995), or a “drive to explain” (Gopnik, 2000). The evolutionary biologist Stephen Jay Gould allegedly characterized humans as “the primates who tell stories,” and many of these stories take the form of explanations.

Thinking about explanation raises a number of important questions. Why are we so driven to explain? What counts as an explanation, and what makes some explanations better than others? Are there different kinds of explanations? If so, what are they? These are the kinds of questions addressed in the Concepts & Cognition lab. For example, in one line of research we consider the role of simplicity in evaluating explanations. Why do people prefer simpler explanations, and what are the consequences of this preference? One finding is that people tend to treat simpler explanations as if they’re more likely to be true, even when there’s evidence to the contrary. In a second line of research, we’ve explored the distinction between mechanistic explanations (those that cite causes and causal mechanisms) and teleological explanation (those that cite functions and goals). On-going work suggests that these two kinds of explanations correspond to different ways of reasoning about objects and events: depending on which kind of explanation you spontaneously entertain, you might generate different responses to the very same questions.

While explanation is fascinating in its own right, another motivation for studying explanation comes from the potential to learn about other areas of cognition. Explanation is at the core of basic cognitive processes like learning, inference, and categorization. To illustrate, consider the relationship between explanation and learning. Anyone who’s ever tutored, taught a course, or simply explained something to a friend has had the experience that explaining can lead to greater understanding. This is pretty mysterious. When you request an explanation from someone else, you gain new information. But when you explain to someone else, you’re not gaining new information in the same way–you’re just reorganizing what you already know. Why should this lead to greater understanding? More generally, what’s the role of explanation in learning, and what can we learn about learning by explaining explanation? Based on everyday observations and evidence from cognitive psychology, there’s reason to think explanation is intimately related to inference and categorization as well as learning.

If you would like to learn more about explanation, you can read a short review paper in Trends in Cognitive Sciences here. Many of the other publications from the lab are also about explanation. And finally, it’s worth saying that not every project in the Concepts & Cognition lab is about explanation! Some projects are related to explanation, but focus on causation or categorization. Other projects aren’t particularly related to explanation at all. You can read about the members of the lab and their interests here.