Penn has considerable strength in philosophy of science and related areas of science studies. We are especially strong in the philosophy of the life and social sciences, the relations between the history of philosophy and the history of science, and the history of the philosophy of science.

Philosophy of the Natural Sciences and Mathematics

Our faculty work on diverse topics in the philosophy of natural sciences, but areas of special interest interest include philosophy of biology (Spencer, Weisberg), psychology and vision (Hatfield), learning theory (Bicchieri, Weinstein), the history of biology and psychology (Detlefsen, Hatfield), philosophy of chemistry (Weisberg), and Public Understanding of Science (Weisberg). A number of faculty also work directly in areas of the natural and formal sciences including cognitive science (Bicchieri, Hatfield, Weinstein, Weisberg), evolutionary and ecological modeling (Bicchieri, Weisberg), and computer science (Weinstein). Spencer, and Weisberg also work on many central topics of philosophy of science including explanation, the structure of theories, confirmation, and the social structure of science.

In addition, William Ewald (Law) teaches history and philosophy of mathematics, Steve Kimbrough (Wharton) teaches modeling, machine learning, and induction, Alan Kors and Ann Moyer (Department of History) offer courses in early modern intellectual history and history of science. History and Sociology of Science regularly offers courses in the history of biology (Lindee), the Scientific and Romantic revolutions (Kucuk, Tresch), and the history of technology (Voskuhl).

Philosophy of Social Science

Our faculty are also interested in some of the central questions in contemporary social science, such as: Are social beings with intentions producing collective outcomes nobody planned or predicted? Can groups of rational agents act in collectively beneficial ways? Can we explain features of the social world, like conventions and social norms, as the result of individuals’ beliefs and desires? How institutions evolve, and how can we model their dynamics?  

Cristina Bicchieri is interested in how norms may emerge and become stable, why an established norm may suddenly be abandoned, how is it possible that inefficient or unpopular norms survive, and what motivates people to obey norms. In order to answer some of these questions, she has combined evolutionary and game-theoretic tools with models of decision making drawn from cognitive and social psychology. For example, she has developed a theory of context-dependent preferences that explains the observed variability in norm compliance and is testing it in experimental games that involve pro-social norms of fairness and reciprocity.

The emergence of norms can be modeled in several ways, depending upon the type of norm that is investigated. Bicchieri and her students have studied how unpopular descriptive norms such as "bad" fashions and fads may occur as the result of negative informational cascades when agents are in the grip of 'pluralistic ignorance'. Often what we call a social norm is a stable behavioral disposition that is supported by a variety of strategies. Impersonal trust, for example, can evolve as a stable disposition in a population of conditionally "nice" agents. A surprising result of this evolutionary model is that what we take to be unconditional moral norms can only survive in populations of conditional choosers.

Michael Weisberg has developed agent-based models that explains how scientific communities coordinate in producing scientific results, and how the division of cognitive labor affects the success of the scientific enterprise. In particular, Weisberg has developed an agent-based computational model in which different research approaches (strategies) are distributed within an epistemic landscape, each approach having its own epistemic payoff. A natural question to ask is how much attention should scientists pay to what other scientists are doing, and what are the costs/benefits of doing so. Weisberg shows that the most effective communities (i.e., those with the highest payoff) are a combination of trendsetters (who open new research paths) and followers (who imitate the most successful members of the scientific community).