Search

The stopping rule for a sequential experiment is the rule or procedure for determining when that experiment should end. Accordingly, the stopping rule principle (SRP) states that the evidential relationship between the final data from a sequential experiment and a hypothesis under consideration does not depend on the stopping rule: the same data should yield the same evidence, regardless of which stopping rule was used. I clarify and provide a novel defense of two interpretations of the main argument against the SRP, the foregone conclusions argument. According to the first, the SRP allows for highly confirmationally unreliable experiments, which concept I make precise, to confirm highly. According to the second, it entails the evidential equivalence of experiments differing significantly in their confirmational reliability. I rebut several attempts to deflate or deflect the foregone conclusion argument, drawing connections with replication in science and the likelihood principle.

The role or function of experimental and observational replication within empirical science has implications for how replication should be measured. Broadly, there seems to be consensus that replication's central goal is to confirm or vouchsafe the reliability of scientific findings. I argue that if this consensus is correct, then most of the measures of replication used in the scientific literature are actually poor indicators of this reliability or confirmation. Only meta-analytic measures of replication align functionally with the goals of replication. I conclude by addressing some objections to meta-analysis.

Machery (2020) has recently proposed a "resampling" account of experimental replication to dissolve a debate in psychology about the relative merits of direct and conceptual replication. We argue that (i) on matters of replication’s function and typology, the resampling account is not substantially different from the functional account of replication extant in the literature; (ii) on what generalizations can be drawn from replications, the resampling account is too restrictive and relies on a misunderstanding of the relation between random sampling and generalizability; and (iii) Machery’s reading of the debate on the relative importance of direct and conceptual replication elides a deeper debate about values and the distribution of research resources in science.

How can inferences from models to the phenomena they represent be justified when those models represent only imperfectly? Pierre Duhem considered just this problem, arguing that inferences from mathematical models of phenomena to real physical applications must also be demonstrated to be approximately correct when the assumptions of the model are only approximately true. Despite being little discussed among philosophers, this challenge was taken up (if only sometimes implicitly) by mathematicians and physicists both contemporaneous with and subsequent to Duhem, yielding a novel and rich mathematical theory of stability with epistemological consequences.

Based on three common interpretive commitments in general relativity, I raise a conceptual problem for the usual identification, in that theory, of timelike curves as those that represent the possible histories of (test) particles in spacetime. This problem affords at least three different solutions, depending on different representational and ontological assumptions one makes about the nature of (test) particles, fields, and their modal structure. While I advocate for a cautious pluralism regarding these options, I also suggest that re-interpreting (test) particles as field processes offers the most promising route for natural integration with the physics of material phenomena, including quantum theory.

If one is interested in reasoning counterfactually within a physical theory, one cannot adequately use the standard possible world semantics. As developed by Lewis and others, this semantics depends on entertaining possible worlds with miracles, worlds in which laws of nature, as described by physical theory, are violated. Van Fraassen suggested instead to use the models of a theory as worlds, but gave up on determining the needed comparative similarity relation for the semantics objectively. I present a third way, in which this similarity relation is determined from properties of the models contextually relevant to the truth of the counterfactual under evaluation. After illustrating this with a simple example from thermodynamics, I draw some implications for future work, including a renewed possibility for a viable deflationary account of laws of nature.

I survey from a modern perspective what spacetime structure there is according to the general theory of relativity, and what of it determines what else. I describe in some detail both the "standard" and various alternative answers to these questions. Besides bringing many underexplored topics to the attention of philosophers of physics and of science, metaphysicians of science, and foundationally minded physicists, I also aim to cast other, more familiar ones in a new light.

Merely approximate symmetry is mundane enough in physics that one rarely finds any explication of it. Among philosophers it has also received scant attention compared to exact symmetries. Herein I invite further consideration of this concept that is so essential to the practice of physics and interpretation of physical theory. After motivating why it deserves such scrutiny, I propose a minimal definition of approximate symmetry - that is, one that presupposes as little structure on a physical theory to which it is applied as seems needed. Then I apply this definition to three topics: first, accounting for or explaining the symmetries of a theory emeritus in intertheoretic reduction; second, explicating and evaluating the Curie-Post principle; and third, a new account of accidental symmetry.