Wednesday, January 12, 2005

How To Study Intuitions: Examples from the Adult and Developmental Literatures

In a previous post, I criticized the relatively new area of analytic philosophy called "experimental philosophy" on primarily methodological grounds. That critique was exclusively negative, and more of a mild rant than a well-reasoned discussion. In this post, I want to make up for that at least a little bit, by describing how I think a naturalized epistemology1 that studies people's intuitions should be done. In order to do that, I'm going to describe some examples from the study of intuitions in cognitive psycholgy. More specifically, I'm going to describe the study of folk theories in various domains. In my view, what experimental philosophers are attempting to do is to tap into peoples' folk theories of things like reference, intention, and free will, and therefore their research would benefit from the use of methodologies more similar to those used in the study of other types of folk theories.

Folk theories are collections of intuitions held by novices (i.e., non-scientists) about objects and agents in various domains. These intuitions have been studied in a wide variety of domains, including folk physics and mechanics (sometimes called naive or intuitive physics), folk biology, folk psychology (sometimes called "theory of mind"), and even folk geometry. These studies have been conducted in adult and child (including infant) populations from several cultures. Few, if any of these studies involve the sorts of self-report measures that experimental philosophers tend to use. Instead, they use more indirect measures designed to get at peoples' underlying representations and reasoning mechanisms. To illustrate these methods, I'm going to discuss some examples from the adult and developmental literature from four areas: folk physics, folk biology, ontological intuitions, and theory of mind . At the end, I'll try to say something about how similar methodologies might be used to study the intuitions that experimental philosophers find interesting.

Folk Physics and Mechanics

The empirical study of folk physics began with the Gestalt psychologists, such as Lipmann and Bogen2, who were responding to a sentiment expressed by Wolfgang Köhler toward contemporary psychology, which in many ways echoes the sentiments of experimental philosophers toward contemporary analytic philosophy. Köhler wrote3:
Psychology has not yet even begun to investigate the physics of ordinary men [Physik des naiven Menschen], which from a purely biological standpoint, is much more important than the science [of theoretical physics] itself. (p. 149)
Over the last 25 years, the study of folk physics has exploded, due largely to the influence of artificial intelligence researchers who have attempted to model peoples naive understanding of mechanics (see, e.g., this paper). Because of the influence of AI, there have been two interacting types of methodologies for studying peoples' understanding of physics and mechanics, computational modeling and experimental investigations. From the use of these two types of methodologies, we now have a fairly firm grasp of the physical and mechanical intuitions of adults. Most importantly, we know that they deviate from our scientific knowledge of physics and mechanics in certain ways. For example, one of the most widely studied errors is the "curvilinear impetus principle"4. In a typical experiment with adults or children (usually 3 years or older), participants are given a description, drawing, or computer simulation of a ball moving through a curved cylinder like the one below. They are then asked to draw or describe the motion of the ball when it leaves the cylinder. Most children and adults, across cultures, draw or describe a curved motion similar to the curve of the cylinder itself, when in fact the ball will travel in a straight line.


Fom Proffitt Perception Lab.

Similar errors are consistently observed in tasks asking people to indicate the path a pendulum will take as it falls to the ground after stopping at various points in its swing, which of two objects (usually of varying sizes and weights) dropped from the same height will hit the ground first, the trajectory an object dropped from a moving plane will take, etc. (for more examples, and drawings, see this post in PPT). In each of these experiments, participans' intuitions about physical phenomena are measured by asking them to predict the behavior of objects under certain conditions.

Folk physics and mechanics are also studied in children, including infants less than one year old. A typical question in such studies is whether young infants have intuitions about the structure and behavior of physical objects. For instance, do infants believe that solid objects will exhibit "continuity" (i.e., a single object moving along a trajectory will not break into two objects moving on two paths), "support" (unsupported objects will fall), "solidity" (no two solid objects can occupy the same point in space at the same time), etc.5? To test whether infants share these intuitions with adults, researchers utilize the knowledge that infants will tend to react (e.g., by looking, touching, sucking on a pacifier) to surprising events. Researchers will present infants with events that exhibit continuity (or one of the other principles), and events that violate the principle of continuity, and measure how surprised they are. These studies have consistently shown that infants share many of the same physical and mechanical intuitions about the behavior and composition of physical objects that adults do.

Folk Biology

One of the questions experimental philosophers have asked is whether people's intuitions about things like reference differ across cultures. If this is the case, they argue, then philosophical theories of reference based on the intuitions of members of one culture may not be the only theories that can be conceived. Researchers studying folk biology have asked a similar question: are people's intuitions about biological categories similar, across cultures? Do they resemble the taxonomic classifications of the biological sciences, or are they dependent on cultural and environmental factors? To answer these questions, and others, Doug Medin and his colleagues have studied biological concepts in various populations, including U.S. college undergraduates, botanists, park mainenance workers, and Itza’ and Yukatek Maya children and adults. What they have found is a complex pattern of convergences and divergences in the conceptual representations of biological kinds across cultures and levels of expertise in both children and adults.

In his research, Medin uses several different types of tasks, including sorting, induction, and comparison tasks. For example, here is a short description of the results of an indicution task6:
Suppose you are told that there is one new disease that we know affects coyotes and wolves, and another new disease that affects coyotes and cows. Now we ask which disease is more likely to affect all mammals. University of Michigan undergraduates overwhelmingly say the disease that coyotes and cows get is more likely to affect all mammals. They justify their answers by appealing to the dissimilarity of the two premises, or diversity. That is, they say that if some disease affects such different mammals as coyotes and cows, it is likely to affect all mammals. This reasoning strategy seems straightforward and the Osherson, Smith, Lopez, Wilkie, and Shafir, 1990, model for categorybased reasoning predicts that people will prefer more diverse premises in drawing inductions to a category. What is surprising is that the Itza’ Maya do not show a diversity effect. In some cases they are reliably below chance in picking the more diverse premises on these kinds of tests (p. 12).
Sorting tasks are used to develop a picture of the intuitive biological taxonomies of individuals from different cultures. In these tasks, participants are presented with cards containing pictures that represent either general kinds (e.g., pictures of mammals drawn from the local fauna), specific kinds (e.g., squirrels or bats), or a collection of biological kinds from the participant's environment (e.g., all of the local tree species)7. Such sorting tasks show a great deal of similarity between the taxonomies of American undergraduates (see a diagram of their taxonomy here) and Maya adults (see a diagram of their taxonomy here). There are differences, however. For instance, Scott Atran writes8:
An additional nonscientific dimension in Itzaj classification, which is not present in American classification, relates to ecology. For example, Itzaj form a group of arboreal animals, including monkeys as well as tree-dwelling procyonids (kinkajou, cacomistle, raccoon) and squirrels (a rodent). The ecological nature of this group was independently confirmed as follows: We asked informants to tell us which plants are most important for the forest to live. Then, we aggregated the answers into a cultural model, and for each plant in the aggregate list we asked which animals most interacted with it (without ever asking directly which animals interact with one another). The same group of arboreal animals emerged as a stable cluster in interactions with plants.
In studying the folk biological inferences of children, psychologists ask questions like, do children recognize that biological kinds have internal properties that are causally related to their external properties, and that make a biological kind what it is? Frank Keil9 addressed this question by telling children stories in which one animal (e.g., a skunk, with skunk parents) ate some strange substance that caused its appearance to transform so that it looked exactly like a squirrel. Children were then asked whether they believed the animal was now a skunk or a squirrel. By and large, children answered that it was still a skunk, despite the fact that its surface features were now much more similar to those of a squirrel. Subsequent induction studies have shown that children have similar essentialist intuitions about the internal properties of biological kinds. They are much more likely to infer common internal properties based on similarity in kind rather than surface (perceptual) similarity.

Ontological Intuitions

Another area of research into children's intuitions concerns their intuitions about different kinds. At a very early age, children seem to be able to distinguish between artifacts and natural kinds, and further between humans, biological kinds, and non-biological kinds (though, in general, they tend to use teleological explanations for the properties of all of these to different degrees). For instance, infants are able to understand that living things can move without an external cause, while non-living things cannot10. These studies are performed using the surprise methods described above in the section on folk physics. Other studies have shown that infants intuitively distinguish between animals and artifacts. In this case, habituation methods in which children are habituated to objects from a category (i.e., shown over and over again until they stop responding to objects from that category), and then shown objects from the habituation category or another category, are used. When infants are habituated with pictures of animals, and then shown pictures of artifacts, they begin to respond again, indicating that their animal category does not include artifacts11.

Folk Psychology (Theory of Mind)

I talked a little bit about children's theory of mind in a recent post on attributing false beliefs to God. These studies are designed to determine whether children at different ages understand that other people have beliefs. Other studies look at children's ability to use the gaze of others to determine what they are referring to (a skill integral to language development, which appears at about 18 months) or to infer their desires (a skill that develops around age 4)12. However, since I've already described one "theory of mind" method used in developmental research in detail in the previous post, here I'll just describe an interesting study designed to test adults' theory of mind.

Here is a description of the experiment from Keysar et al.13:
A person who played the role of “director” in a communication game instructed a participant to move certain objects around in a grid. Before receiving instructions, participants hid an object in a bag, such that they but not the director would know its identity. Occasionally, the descriptions that the director used to refer to a mutually-visible object more closely matched the identity of the object hidden in the bag. Although they clearly knew that the director did not know the identity of the hidden object, they often took it as the referent of the director’s description, sometimes even attempting to comply with the instruction by actually moving the bag itself. (p. 25)
This study provides an interesting example of the differences between self-report and behavioral measures. Upon reflection, adults will quickly admit that a person who has not seen an object placed in a bag will not know what object is in it. However, when actually interacting with others, they tend to act as though the person has the same knowledge they do. In other words, even though children are able to pass the false belief test by about age 4 or 5, adults still fail a version of it during real-time interactions with others. As Keysar et al. put it:
These results show a stark dissociation between an ability to reflectively distinguish one’s own beliefs from others’, and the routine deployment of this ability in interpreting the actions of others. We propose that this dissociation indicates that important elements of the adult’s theory of mind are not fully incorporated into the human comprehension system. (p. 25)
In other words, our explicit intuitions do not always accord with the "implicit intuitions" that are reflected in our behavior. This study, and others like it (see, e.g., this post on the difference between people's concepts of God when self-reported after reflection and when used in the performance of ordinary tasks). This brings us back to the problems with many of the methods used by experimental psychologists. These involve presenting participants with scenarios, and asking them to report their intuitions about things like intention and free will. Even if we ignore for the moment the obvious potential for the influence of demand characteristics in these types of studies, we can't ignore the fact that in many cases, the "intuitions" that adults report upon reflection do not accord with their actual behavior. For this reason, we can't be sure that the methods of experimental philosophy are really getting at people's representations of concepts like intention and free will. To do this, we would need to use methods more similar to those used in the study of intuitions in the area of physics, biology, ontology, and mind.

How could these sorts of indirect and behavioral methodologies be used to study the sorts of things that philosophers are interested in? One way might be to use recognition memory tasks. Research has shown that our memories for stories and events can be systematically distorted by our prior beliefs and schemas. We would expect, then, that people's concepts of intention or free will might distort their memories for stories that activate them. In such an experiment, we would present a story that a certain philosophical theory of intention predicts should activate intention concepts, but without mentioning anything about intention. Later, during a recognition task, participants will be asked to indicate whether sentences were or were not in the original story. Some of the sentences in the recognition test that were not in the original stories will contain references to intention, or imply intention. If participants tend to mistakenly indicate that these sentences were present in the original stories, then we can infer that their intention concepts were active. Obviously, this is a pretty weak example, but what I'm trying to do is give an idea of how indirect measures might be used. Hopefully if this example isn't satisfying, then people will be able to come up with their own methods based on some of the examples I've given above, or any of the many other methods used by psychologists to study intuitions.

1 For more on what what naturalized epistemology is, and how it connected to the experimental philosophy movment, check out Lindsay Beyerstein's wonderful essay on the topic, which can be found here.
2 Lipmann, O., & Bogen, H. (1923) Naive Physik. Arbeiten aus dem Institut für angewandte Psychologie in Berlin. Theoretische und experimentelle Untersuchungen über die Fähigkeit zu intelligentem Handeln. Leipzig: Johann Ambrosius Barth.
3 Köhler, Wolfgang 1921 "Intelligenzprüfungen an Anthropoiden", Abhandlungen der Preussischen Akademie der Wissenschaften. Jahrgang 1917, Physikal.-Mathem. Klasse, 1; Eng. trans. of 2nd German ed. as The Mentality of Apes, London: Kegan Paul, Trench, Trübner, 1927. As quoted in Smith, B., & Casati, R. (1994). Naive physics: An essay on ontology. Philosophical Psychology, 7(2), 225-244.
4 Caramazza, A., McCloskey, M., & Green, B. (1981). Naive beliefs in "sophisticated" subjects: Misconceptions about trajectories of objects. Cognition, 9(2), 117-123. McCloskey, M. (1983) Intuitive physics. Scientific American. 248(4), 122-130. McCloskey M.,% Kohl D.(1983). Naive physics: the curvilinear impetus principle and its role in interactions with moving objects. Journal of Experimental Psychology: Learning, Memory, & Cognition, 9(1), 146-156.
5 Boyer, P. (2000). Natural epistemology or evolved metaphysics? Developmental evidence for early-developed, intuitive, category-speciŽ c, incomplete, and stubborn metaphysical presumptions. Philosophical Psychology, 13(3), 277-297.
6 Medin, D.L & Atran. S. (Submitted). The Native Mind: Biological Categorization, Reasoning and Decision Making in Development Across Cultures. Psychological Review.
7 Atran, S. (1998). Folk biology and the anthropology of science: Cognitive universals and cultural particulars. Behavioral & Brain Sciences, (4):547-69.
8 Ibid.
9 Keil, F. C. (1989). Concepts, kinds, and cognitive development. Cambridge, MA: MIT Press.
10 Premack, D. (1990). The infant’s theory of self-propelled objects. Cognition, 36, 1–16.
11 Eimas, P.D. (1994). Categorization in early infancy and the continuity of development. Cognition, 50, 83–93.
12 Lee, K., Eskritt, M., Symons, L.A., & Muir, D. (1998). Children's use of triadic eye gaze information for "mind reading." Developmental Psychology, 1998, 34(3), 525-539.
13 Keysar, B., Lina, S., & Barr, D. J. (2003). Limits on theory of mind use in adults. Cognition, 89, 25-41.

3 comments:

Anonymous said...

Chris, I would like to see whether you might be interested in posting some of your stuff on the experimental philosophy blog that I run. I have found a number of your comments and suggestions helpful. If you are interested in becoming a contributor/author, email me.
Best wishes,
Thomas Nadelhoffer 

Posted by tnadelhoffer

Anonymous said...

Just though I might add that the url for the experimental philosophy blog that Thomas mentioned is
http://experimentalphilosophy.typepad.com/ 

Posted by Jonathan Weinberg

Anonymous said...

http://blog.oricon.co.jp/misspoloshirt
http://blog.qlep.com/blog.php/gillette
http://blog.goo.ne.jp/pumafr
http://www.pumafr.com/blog
http://www.clothessports.com
http://chinaclothes.net