Qualifying quests, 1: Causation & Explanation

Approaches to explanation, levels of analysis and incomplete knowledge

There has been a tension in the philosophy of science between theories of causation and explanation that rely on law-covering and stastical information between types of events, and theories that rely on mechanistic and interventional information about individual events. To what extent have these views been reconciled in light of computational level analyses of causal inference (esp. Bayesian rational accounts) and to what extent has the conflict persisted in our understanding of human causal and explanatory reasoning? To what extent does the incompleteness of any one individual's causal knowledge undermine these accounts of causal and explanatory reasoning?

Explanatory virtues, Mechanism and Hidden Causes

Recent empirical work emphasizes a variety of explanatory virtues that stem from philosophy of explanation (Lombrozo, Gopnik, Read & Marcus-Newall,Preseton & Epley). what are some of those virtues and how do they relate to our understanding of human explanation? If space/time permits, please explore how these virtues bear on mechanistic accounts of explanation and causation, which often postulate unobservable entities to fill these mechanistic roles.

Dimensions of Theories of Causal Inference, emph. Time

What are some of the dimensions along which theories of human causal inference differ? In particular how have different cognitive scientists approached the temporal dimension of causal inference, and how can these distinctions be analyzed at the computational level, and how temporal considerations bear on other issues that differentiate accounts of human causal inference, explanation and understanding?

Qualifying quests, 2: Rational Probabilistic Models of Higher Level Cognition

Learning causal theories, continuity and time and their role in the level at which we analyze human theories

Describe some of the recent work on rational models of theory learning about real-world phenomena (e.g., causal theories). How might this work be strengthened by incorporating temporal and continuous-dimensional concerns to these models of theories? How does adding this complexity affect our interpretation of the meaning of theories? More specifically, can time be understood at a computational level(as a part of the overall goal of theory-making), or is it relegated to the algorithmic level(where it merely affects the dynamics of theories, not their "meaning") and how does our answer to that question change our picture of theory-craft?

Generalization: Explanations, Categories, Causes, and Theories

Generalization is agreed to be a key feature of explanation, categorization, causal inference and (at the most general level) theory learning. Generalization also has been posed as a potential universal law of cognition (Shepard, Griffiths & Tenenbaum, Chater & Vitányi). to what extent can we see explanation, categorization, causal inference, and theorizing as forms generalization? In what ways do would this require a departure from generalization as described originally, and what are some of the benefits and drawbacks from taking this more general perspective on generalization?

The merits and demerits of probabilistic, rational models of higher level cognition, and how they change with time

In what ways do Bayesian rational, computational analyses of cognition carry empirical meaning and how do they add to our understanding of the human mind? What are some of the major critiques of Bayesian theorizing? How might both the positive and negative arguments be affected by considering the role of time in either the content or the development of these theories?

Qualifying quests, 3: Science as a Computational, Cognitive Process

For and against: cognitive development and the philsophy of science, and their companionship on the road from rationality

Piaget and the Gestalt-psychologists inspired Kuhn; Kuhn inspired Carey and Gopnik and others. The feedback loop between the study of the ontogeny of knowledge (cognitive development: Piaget, Carey, Gopnik, Keil, Koslowski, Schulz, Spelke, etc.) and the study of the phylogeny of knowledge (philosophy, history, and science of science: Kuhn, Lakatos, Hacking, Hull, Solomon, Salmon) has always been very tight. Cognitive development and the philosophy of science at times have travelled on the same road and influences have flown readily between them, but at times they have diverged. What are some of the analogies and disanalogies between the study of cognitive development and the philosophy of science In particular if time/space allow: Of particular interest to the philosophy of science in the wake of Kuhn's Structure of Scientific Revolutions, has been to chronicle the role of 'non-rational' considerations (rationality usually as construed in the perspective of logical positivism, deductivism, and falsificationism in the vein of Carnap, Hempel and Popper) in determining the path of scientific progress. What are some of these considerations, and to what extent can work by cognitive scientists bear on these aspects of science? Or to put it another way, to what extent is the child as scientist and scientist as child mutual analogy capable of enhancing our understanding of both.

With and without: Theories and thinkers in a larger scheme of things

One of the central aspects that recent philosophy of science has focused on is the role of social aspects of science, whether that is because of the organizational/motivational role of research groups (Kuhn, Hull, Hacking, Galison) or a cognitive benefit to distributing the labor of research (Kitcher, Keil, Weisberg, Strevens). Additionally, different theories fit together in a larger collection of theories, experiments and data (Lakatos, Hacking, Leksovec). But models of cognitive development and rational theory change often are proposed in terms of an child learning an individual theory (Carey, Vosniadou & Brewer, Gopnik, Schulz) or an individual scientist choosing a theory among a set of theories (Kuhn, Henderson Goodman Tenenbaum, Salmon, Solomon, Strevens) without explicitly representing the role of social support and the interactions of different theories. However, this has been a matter of emphasis, not a strict contradiction --- to what extent can these approaches be brought together(e.g, Keil, Gopnik Schulz)? What can be gained by a closer look at social aspects in our understanding of both cognitive development and science? What can be gained from considering the larger theoretical context in which theory learning occurs, and how might that be formalized?

Beyond and about: Drawing connections between computational models and the developments of science and cognition

Considering that many models of Bayesian theory learning motivate themselves from humanity's skill in "getting the world 'right'"(Schulz, 2012), to what extent can we understand scientific progress using these models? Does a computational level account suffice, or, in order to explain interesting phenomena identified by philosophers and historians of science, do we need to appeal to algorithmic (or other) levels of analysis? Perhaps most importantly, what might the empirical content of these theories as models of human cognition and especially human cognitive development relate to theories of scientific progress.