Research
(For a complete list of previously published research, see my CV here. Drafts of work in progress available upon request.)
Overview
My research primarily focuses on questions in the philosophy of science concerning the nature of model-based representation, explanation, and understanding, with a particular focus on how those questions relate to recent debates in cognitive science. This work in strongly interdisciplinary, requiring me to engage with the ideas of both philosophers and scientists. Additionally, I'm interested in contributing to the scholarship of teaching and learning, especially how recent philosophical and empirical approaches to the nature of "understanding" can inform philosophical pedagogy.
Publications & Reviews
Kelly, Paul J. (2023) Helping Novice Students Understand How Logic Relates to Philosophy. American Association of Philosophy Teachers Studies in Pedagogy, Volume 8.
Kelly, Paul J. (2023) Review of Peter Godfrey-Smith’s Theory and Reality: An Introduction to the Philosophy of Science (2nd Edition). Teaching Philosophy, Volume 46, Issue 2.
Under Review (Titles & Abstracts Removed for Blind Review)
Article on how functional models in cognitive science can explain
I advocate for a variety of model-based representation that indicates a way in which functional models in cognitive science can at least partially explain some cognitive phenomena.
I advocate for a variety of model-based representation that indicates a way in which functional models in cognitive science can at least partially explain some cognitive phenomena.
Works in Progress
Project 1: Model-based Representation & Explanation
What Justifies the Model-to-Mechanism-Mapping Requirement?
Mechanists claim that dynamical models in cognitive science are nonexplanatory because they fail to satisfy the model-to-mechanism mapping requirement (3M). I argue that while it is clear that the adoption of 3M entails that dynamical models are nonexplanatory, it is much less clear why we should accept 3M. Some passages by mechanists seem to indicate that the adoption of 3M is justified based on a particular account of what makes models explanatory: namely, that a model is explanatory in virtue of accurately representing a phenomenon and its ontic explanation. I argue that this account cannot justify the adoption of 3M, since 3M and the account issue conflicting judgments about which models are explanatory. If I am correct, then (in the absence of additional rationales) the adoption of 3M is unjustified, and mechanists can no longer use it to exclude dynamical models from the class of explanatory models.
Mechanists claim that dynamical models in cognitive science are nonexplanatory because they fail to satisfy the model-to-mechanism mapping requirement (3M). I argue that while it is clear that the adoption of 3M entails that dynamical models are nonexplanatory, it is much less clear why we should accept 3M. Some passages by mechanists seem to indicate that the adoption of 3M is justified based on a particular account of what makes models explanatory: namely, that a model is explanatory in virtue of accurately representing a phenomenon and its ontic explanation. I argue that this account cannot justify the adoption of 3M, since 3M and the account issue conflicting judgments about which models are explanatory. If I am correct, then (in the absence of additional rationales) the adoption of 3M is unjustified, and mechanists can no longer use it to exclude dynamical models from the class of explanatory models.
Project 2: Model-based Understanding
Dynamical Models, Scientific Understanding, and Explanatory Unification
It is often claimed that dynamical models are capable of, at least partially, explaining some cognitive phenomena. Mechanists disagree, arguing that such models fail to satisfy their proposed model-to-mechanism mapping requirement (hereafter, 3M), and are thereby nonexplanatory. I propose three independent reasons why the mechanist’s judgment that dynamical models are nonexplanatory is mistaken: first, because it is inconsistent with recent work in computational neuroscience that is prima facie explanatory, second, because parity of reasoning would require that we also view equilibrium models as nonexplanatory, and third, because dynamical models can facilitate a kind of understanding that is possible only if such models are, at least partially, explanatory. I then conclude with some thoughts about what theory of explanation best accommodates the possibility of explanatory dynamical models.
It is often claimed that dynamical models are capable of, at least partially, explaining some cognitive phenomena. Mechanists disagree, arguing that such models fail to satisfy their proposed model-to-mechanism mapping requirement (hereafter, 3M), and are thereby nonexplanatory. I propose three independent reasons why the mechanist’s judgment that dynamical models are nonexplanatory is mistaken: first, because it is inconsistent with recent work in computational neuroscience that is prima facie explanatory, second, because parity of reasoning would require that we also view equilibrium models as nonexplanatory, and third, because dynamical models can facilitate a kind of understanding that is possible only if such models are, at least partially, explanatory. I then conclude with some thoughts about what theory of explanation best accommodates the possibility of explanatory dynamical models.
Project 3: Philosophically & Empirically Informed Philosophical Pedagogy
Philosophy of Understanding for Philosophy Teachers (with Michael Bruckner)
The concept of “understanding” is commonly appealed to in course objectives in philosophy classes. Instructors want students to understand philosophical texts, understand competing lines of argumentation, understand how to appropriately respond to objections advanced by their peers, and so on. Recently, a substantial philosophical literature has emerged focusing on the concept of “understanding.” Different varieties of understanding have been distinguished, and different proposals about the specific inferential and reasoning abilities that constitute those varieties have been advanced. There have also been recent attempts to integrate these philosophical accounts with empirical work in psychology and neuroscience. Somewhat surprisingly, there has been comparatively little discussion about how these developments can, and should, inform philosophical pedagogy. We seek to remedy this, by proposing several significant ways in which the recent philosophical and empirical literature on the nature of understanding can fruitfully inform future discussions of philosophical pedagogy.
The concept of “understanding” is commonly appealed to in course objectives in philosophy classes. Instructors want students to understand philosophical texts, understand competing lines of argumentation, understand how to appropriately respond to objections advanced by their peers, and so on. Recently, a substantial philosophical literature has emerged focusing on the concept of “understanding.” Different varieties of understanding have been distinguished, and different proposals about the specific inferential and reasoning abilities that constitute those varieties have been advanced. There have also been recent attempts to integrate these philosophical accounts with empirical work in psychology and neuroscience. Somewhat surprisingly, there has been comparatively little discussion about how these developments can, and should, inform philosophical pedagogy. We seek to remedy this, by proposing several significant ways in which the recent philosophical and empirical literature on the nature of understanding can fruitfully inform future discussions of philosophical pedagogy.