The readings listed below are the foundation of this course. Where available, journal article abstracts from PubMed (an online database providing access to citations from biomedical literature) are included.

Ahn, W. K., Kalish, C., Gelman, S. A., Medin, D. L., Luhmann, C., Atran, S., Coley, J. D., Shafto, P. (2001). "Why Essences are Essential in the Psychology of Concepts." Cognition 82, 59-69.

Ahn, W. and Luhmann, C. C. (in press). "Demystifying Theory-based Categorization." Building object categories in developmental time. Edited by L. Gershkoff-Stowe, and D. Rakison.

Anderson, J. R. (1991). "The Adaptive Nature of Human Categorization." Psychological Review 98, 409-429.

Atran, S. (1998). "Folk Biology and the Anthropology of Science: Cognitive Universals and Cultural Particulars." Behavioral and Brain Sciences 21: 547-569, Cambridge University Press.

PubMed abstract: This essay in the "anthropology of science" is about how cognition constrains culture in producing science. The example is folk biology, whose cultural recurrence issues from the very same domain-specific cognitive universals that provide the historical backbone of systematic biology. Humans everywhere think about plants and animals in highly structured ways. People have similar folk-biological taxonomies composed of essence-based, species-like groups and the ranking of species into lower- and higher-order groups. Such taxonomies are not as arbitrary in structure and content, nor as variable across cultures, as the assembly of entities into cosmologies, materials, or social groups. These structures are routine products of our "habits of mind," which may in part be naturally selected to grasp relevant and recurrent "habits of the world." An experiment illustrates that the same taxonomic rank is preferred for making biological inferences in two diverse populations: Lowland Maya and Midwest Americans. These findings cannot be explained by domain-general models of similarity because such models cannot account for why both cultures prefer species-like groups, although Americans have relatively little actual knowledge or experience at this level. This supports a modular view of folk biology as a core domain of human knowledge and as a special player, or "core meme," in the selection processes by which cultures evolve. Structural aspects of folk taxonomy provide people in different cultures with the built-in constraints and flexibility that allow them to understand and respond appropriately to different cultural and ecological settings. Another set of reasoning experiments shows that Maya, American folk, and scientists use similarly structured taxonomies in somewhat different ways to extend their understanding of the world in the face of uncertainty. Although folk and scientific taxonomies diverge historically, they continue to interact. The theory of evolution may ultimately dispense with the core concepts of folk biology, including species, taxonomy, and teleology; in practice, however, these may remain indispensable to doing scientific work. Moreover, theory-driven scientific knowledge cannot simply replace folk knowledge in everyday life. Folk-biological knowledge is not driven by implicit or inchoate theories of the sort science aims to make more accurate and perfect.

Bishop, C. M. (1995). Neural Networks for Pattern Recognition. Oxford University Press. Sections 1.0, 1.1 (pages 1-4) 1.8-9 (pages 17-26), 2.0-3 (pages 33-45), 2.5 (pages 49-59).

Charniak, E. (1991). "Bayesian Networks without Tears." AI Magazine.

Gelman, S. (2003). The Essential Child. New York: Oxford University Press. Chapter 1 (pages 3-18) and Chapter 3 (pages 60-88).

Glymour, C. (2003). "Learning, Prediction and Causal Bayes Nets." Trends in Cognitive Science 7, 43-48.

PubMed abstract: Recent research in cognitive and developmental psychology on acquiring and using causal knowledge uses the causal Bayes net formalism, which simultaneously represents hypotheses about causal relations, probability relations, and effects of interventions. The formalism provides new normative standards for reinterpreting experiments on human judgment, offers a precise interpretation of mechanisms, and allows generalizations of existing theories of causal learning. Combined with hypotheses about learning algorithms, the formalism makes predictions about inferences in many experimental designs beyond the classical, Pavlovian cue-->effect design.

Goodman, N. (1955). "The New Riddle of Induction." In Goodman, Fact, Fiction, and Forecast. Cambridge, MA: Harvard University Press, Chapter 3.

Gopnik, A. and Glymour, C. (2002). "Causal Maps and Bayes Nets: A Cognitive and Computational Account of Theory-formation." The Cognitive Basis of Science. Edited by Carruthers et al. Cambridge: Cambridge University Press.

Gopnik, A., Glymour, C., Sobel, D., Schulz, L. E., Kushnir, T., and Danks, D. (in press). "A Theory of Causal Learning in Children: Causal Maps and Bayes Nets." Psychological Review.

Heckerman, D. (1995). A Tutorial on Learning Bayesian Networks. Technical Report MSR-TR-95-06, Microsoft Research.

Jefferys, W. and Berger, J. (1992). "Ockham's Razor and Bayesian Analysis." American Scientist 80 , 64-72.

Keil, F. C. (2003). "Folkscience: Coarse Interpretations of a Complex Reality." Trends in Cognitive Sciences 7, 368-373.

PubMed abstract: The rise of appeals to intuitive theories in many areas of cognitive science must cope with a powerful fact. People understand the workings of the world around them in far less detail than they think. This illusion of knowledge depth has been uncovered in a series of recent studies and is caused by several distinctive properties of explanatory understanding not found in other forms of knowledge. Other experimental work has shown that people do have skeletal frameworks of expectations that constrain richer ad hoc theory construction on the fly. These frameworks are supplemented by an ability to evaluate and rely on the division of cognitive labour in one's culture, an ability shown to be present even in young children.

Kemp, C. and Tenenbaum, J. B. (2003). "Theory-based Induction." In Proceedings of the Twenty-Fifth Annual Conference of the Cognitive Science Society.

Laurence, S. and Margolis, E. (1999). "Concepts and Cognitive Science." In Concepts: Core Readings. Edited by E. Margolis and S. Laurence. Cambridge, MA.: Bradford Books/MIT Press, pages 3-81.

Margolis, E. and Laurence, S. (2002). "Radical Concept Nativism." Cognition 86, 22-55.

PubMed abstract: Radical concept nativism is the thesis that virtually all lexical concepts are innate. Notoriously endorsed by Jerry Fodor, radical concept nativism has had few supporters. However, it has proven difficult to say exactly what's wrong with Fodor's argument. We show that previous responses are inadequate on a number of grounds. Chief among these is that they typically do not achieve sufficient distance from Fodor's dialectic, and, as a result, they do not illuminate the central question of how new primitive concepts are acquired. To achieve a fully satisfactory response to Fodor's argument, one has to juxtapose questions about conceptual content with questions about cognitive development. To this end, we formulate a general schema for thinking about how concepts are acquired and then present a detailed illustration.

McClelland, J. L. and Rogers, T. T. (2003). "The Parallel Distributed Processing Approach to Semantic Cognition." Nature Reviews Neuroscience 4, 310-322.

Medin, D.L., Coley, J.D., Storms, G., and Hayes, B. (in press). "A Relevance Theory of induction." Psychonomic Bulletin and Review.

Osherson, D.N., Smith, E.E., & Shafir, E. 1986."Some Origins of Belief." Cognition 24, 3, 197-224.

Osherson, D. N., Smith, E. E., Wilkie, O., Lopez, A., and Shafir, E. (1990). "Category-based Induction." Psychological Review 97, 185-200.

Pearl, J. (2000). Causality: Models, Reasoning, and Inference. New York: Cambridge University Press, pages 1-40.

Quine, W. V. O. (1969). "Natural Kinds." Chapter 5 in Ontological Relativity and Other Essays. New York: Columbia University Press.

Rehder, B. (in press). "A Causal-model Theory of Conceptual Representation and Categorization." Journal of Experimental Psychology: Learning, Memory, and Cognition.

Russell, S., and Norvig, P. (2002). Artificial Intelligence: A Modern Approach. Excerpt on relational probability models (pages 519-522).

Shepard, R. N. (1980). "Multidimensional Scaling, Clustering, and Tree-fitting." Science 210, 390-398.

Shepard, R. N. (1987). "Towards a Universal Theory of Generalization for Psychological Science." Science 237, 1317-1323.

Sivia, D. S. (1996). Data Analysis: A Bayesian Tutorial. Oxford University Press. Pages 1-22.

Sloman, S., and Ahn, W. K. (1998). "Feature Centrality and Conceptual Coherence." Cognitive Science 22, 189-228.

Smith, E. E., Shafir, E., and Osherson, D. (1993). "Similarity, Plausibility, and Judgments of Probability." Cognition 49, 2, 67-96.

PubMed abstract: Judging the strength of an argument may underlie many reasoning and decision-making tasks. In this article, we focus on "category-based" arguments, in which the premises and conclusion are of the form All members of C have property P, where C is a natural category. An example is "Dobermans have sesamoid bones. Therefore, German shepherds have sesamoid bones." The strength of such an argument is reflected in the judged probability that the conclusion is true given that the premises are true. The processes that mediate such probability judgments depend on whether the predicate is "blank"--an unfamiliar property that does not enter the reasoning process (e.g., "have sesamoid bones")--or "non-blank"--a relatively familiar property that is easier to reason from (e.g., "can bite through wire"). With blank predicates, probability judgments are based on similarity relations between the premise and conclusion categories. With non-blank predicates, probability judgements are based on both similarity relations and the plausibility of premises and conclusion.

Strevens, M. (2000). "The Essentialist Aspect of Naive Theories." Cognition 74, 149-175.

PubMed abstract: Recent work on children's inferences concerning biological and chemical categories has suggested that children (and perhaps adults) are essentialists - a view known as psychological essentialism. I distinguish three varieties of psychological essentialism and investigate the ways in which essentialism explains the inferences for which it is supposed to account. Essentialism succeeds in explaining the inferences, I argue, because it attributes to the child belief in causal laws connecting category membership and the possession of certain characteristic appearances and behavior. This suggests that the data will be equally well explained by a non-essentialist hypothesis that attributes belief in the appropriate causal laws to the child, but makes no claim as to whether or not the child represents essences. I provide several reasons to think that this non-essentialist hypothesis is in fact superior to any version of the essentialist hypothesis.

Tenenbaum, J. B. (2003). Introduction to Probability and Bayesian Inference.

Tenenbaum, J. B., and Griffiths, T. L. (2001). "Generalization, Similarity, and Bayesian Inference." Behavioral and Brain Sciences 24, 629-641.

PubMed abstract: Shepard has argued that a universal law should govern generalization across different domains of perception and cognition, as well as across organisms from different species or even different planets. Starting with some basic assumptions about natural kinds, he derived an exponential decay function as the form of the universal generalization gradient, which accords strikingly well with a wide range of empirical data. However, his original formulation applied only to the ideal case of generalization from a single encountered stimulus to a single novel stimulus, and for stimuli that can be represented as points in a continuous metric psychological space. Here we recast Shepard's theory in a more general Bayesian framework and show how this naturally extends his approach to the more realistic situation of generalizing from multiple consequential stimuli with arbitrary representational structure. Our framework also subsumes a version of Tversky's set-theoretic model of similarity, which is conventionally thought of as the primary alternative to Shepard's continuous metric space model of similarity and generalization. This unification allows us not only to draw deep parallels between the set-theoretic and spatial approaches, but also to significantly advance the explanatory power of set-theoretic models.

Tversky, A. (1977). "Features of Similarity." Psychological Review 84, 327-352.

Tenenbaum, J. B. (2000). "Rules and Similarity in Concept Learning." In Advances in Neural Information Processing Systems 12. Edited by S. A. Solla, T. K. Leen, and K.-R. Muller. Cambridge, MA: MIT Press, pages 59-65.

Tenenbaum, J. B., and Griffiths, T. L. (2001). "Structure Learning in Human Causal Induction." Advances in Neural Information Processing Systems 13.

Tenenbaum, J. B., and Griffiths, T. L. (2003). "Theory-based Causal Inference." Advances in Neural Information Processing Systems 15.

Tenenbaum, J. B., and Niyogi, S. (2003). "Learning Causal Laws." Proceedings of the Twenty-fifth Annual Conference of the Cognitive Science Society.

Wellman, H. M., and Gelman, S. A. (1992). "Cognitive Development: Foundational Theories of Core Domains." Annual Review of Psychology 43, 337-75.