Are objects in convex passenger-side mirrors “closer than they appear”? If one adapts to inverting goggles, does the world go back to looking the way it was before, or does the world look approximately the same throughout the course of adaptation, only losing its normative sense of wrongness? The answers to these empirical, introspective questions might help cast light on the classic philosophical debate about the degree of resemblance between our visual experience of reality and things as they are in themselves.

Considerable methodological difficulties abound in neuroimaging, and several philosophers of science have recently called into question the potential of neuroimaging studies to contribute to our knowledge of human cognition. These skeptical accounts suggest that functional hypotheses are underdetermined by neuroimaging data. I apply Mayo’s error-statistical account to clarify the evidential import of neuroimaging data and the kinds of inferences it can reliably support. Thus, we can answer the question “What can we reliably learn from neuroimaging?” and make sense of how this knowledge can contribute to novel construals of cognition.

Neuroimaging studies of the resting state continue to gather philosophical and scientific attention. Most discussions assume an identification between resting-state activity and activity in the so-called default mode network. I argue we should resist this identification, structuring my discussion around a dilemma first posed by Morcom and Fletcher. I offer an alternative view of rest as a state dominated by long-term processes and show how interaction effects might thereby let rest shed light on short-term changes in activation.

I argue that ductal carcinoma in situ (DCIS), the tumor most commonly diagnosed by breast mammography, cannot be confidently classified as cancer, that is, as pathological. This is because there may not be dysfunction present in DCIS—as I argue based on its high prevalence and the small amount of risk it conveys—and thus DCIS may not count as a disease by dysfunction-requiring approaches, such as Boorse’s biostatistical theory and Wakefield’s harmful dysfunction account. Patients should decide about treatment for DCIS based on the risks it poses and the risks and benefits of treatment, not on its disease status.

Although cases are central to the epistemic practices utilized within clinical medicine, they appear to be limited in their ability to provide evidence about causal relations because they provide detailed accounts of particular patients without explicit filtering of those attributes most likely to be relevant for explaining the phenomena observed. This paper uses a series of recent case reports to explore the role of cases in causal attribution in medical diagnosis. It is argued that cases are brought together by practitioners to generate causal attributions and testable predictions using a manipulability view of causation.

This paper addresses the problem of how scientific knowledge, which is always locally generated, becomes accepted in other sites. The analysis suggests that there are a small number of strategies that enable scientists to resituate knowledge and that these strategies are generic: they are not restricted to specific disciplines or modes of doing science but rather are found in a variety of different forms across the sciences.

It has become customary among philosophers and biologists to claim that folk racial classification has no biological basis. This paper attempts to debunk that view. In this paper I show that ‘race’, as used in current US race talk, picks out a biologically real entity. I do this by first showing that ‘race’, in this use, is not a kind term, but a proper name for a set of human population groups. Next, using recent human genetic clustering results, I show that this set of human population groups is a partition of human populations that I call ‘the Blumenbach partition’.

This paper distinguishes three concepts of “race”: bio-genomic cluster/race, biological race, and social race. We map out realism, antirealism, and conventionalism about each of these, in three important historical episodes: Frank Livingstone and Theodosius Dobzhansky in 1962, A. W. F. Edwards’s 2003 response to Lewontin’s 1972 paper, and contemporary discourse. Semantics is especially crucial to the first episode, while normativity is central to the second. Upon inspection, each episode also reveals a variety of commitments to the metaphysics of race. We conclude by interrogating the relevance of these scientific discussions for political positions and a post-racial future.

In this paper, I argue against Andreasen and Sesardic, who have both claimed that recent cluster results in population genetics serve as evidence that the human species contains, or at least once contained, subspecies. I show that the cluster results are in fact evidentially inert relative to each author’s preferred subspecies concept. I then sketch the kinds of biological facts that could be used to push the debate further.

Here, we clarify the relationship among three space-time conditions of interest: geodesic completeness, hole-freeness, and inextendibility. In addition, we introduce a related fourth condition: effective completeness.

I discuss singular space-times in the context of the geometrized formulation of Newtonian gravitation. I argue first that geodesic incompleteness is a natural criterion for when a model of geometrized Newtonian gravitation is singular, and then I show that singularities in this sense arise naturally in classical physics by stating and proving a classical version of the Raychaudhuri-Komar singularity theorem.

The view that it is symmetries, not particles, that are fundamental to nature is frequently expressed by physicists. But comparatively little has been written either on what this claim means or whether it should be regarded as true. After placing the claim into a general fundamentality framework, I consider whether the priority of symmetries over particles can be defended. The conclusions drawn are largely negative.

It is shown that three common conditions for scientific explanations are violated by a widely used class of domain-independent explanations. These explanations can accommodate both complex and noncomplex systems and do not require the use of detailed models of system-specific processes for their effectiveness, although they are compatible with such model-based explanations. The approach also shows how a clean separation can be maintained between mathematical representations and empirical content.

Large changes of financial market prices without exogenous causes deviate significantly from the Gaussian behavior of random variables. This indicates that financial markets should be treated as complex systems, for which nonlinear interactions of its subunits/agents are crucial. I focus on how the complex systems perspective impacts the notion of explanations in economics. The mechanistic model seems to fit the bill, but problems surface on closer scrutiny. One characteristic of complex systems is that their behavior is surprisingly independent from microscopic details. Thus, mechanistic explanations in the microreductionist manner seem unavailable. Despite these conflicts, I defend a modified structural mechanistic approach.

The robust behavior of the patent citation network is a complex target of recent bottom-up models in science. This paper investigates the purpose and testing of three especially simple bottom-up models of the citation count distribution observed in the patent citation network. The complex causal webs in the models generate weakly emergent patterns of behavior, and this explains both the need for empirical observation of computer simulations of the models and the epistemic harmlessness of the resulting epistemic opacity.

Despite the close connection between the central limit theorem and renormalization group (RG) methods, the latter should be considered fundamentally distinct from the kind of probabilistic framework associated with statistical mechanics, especially the notion of averaging. The mathematics of RG is grounded in dynamical systems theory rather than probability, which raises important issues with respect to the way RG generates explanations of physical phenomena. I explore these differences and show why RG methods should be considered not just calculational tools but the basis for a physical understanding of complex systems in terms of structural properties and relations.

Renormalization group (RG) methods are an established strategy to explain how it is possible that microscopically different systems exhibit virtually the same macrobehavior when undergoing phase transitions. I argue—in agreement with Robert Batterman—that RG explanations are noncausal explanations. However, Batterman misidentifies the reason that RG explanations are noncausal: it is not the case that an explanation is noncausal if it ignores causal details. I propose an alternative argument, according to which RG explanations are noncausal explanations because their explanatory power is due to the application of mathematical operations, which do not serve the purpose of representing causal relations.

I show that the central notion of Constructive Empiricism, empirical adequacy, can be expressed syntactically and specifically in the Received View of the logical empiricists. The formalization shows that the Received View is superior to Constructive Empiricism in the treatment of theories involving constants or functions from observable to unobservable objects. It also suggests a formalization of ‘full empirical informativeness’ in Constructive Empiricism.

We describe a simple, flexible exercise that can be implemented in the philosophy of science classroom: students are asked to determine the contents of a closed container without opening it. This exercise has revealed itself as a useful platform from which to examine a wide range of issues in the philosophy of science and may, we suggest, even help us think about improving the public understanding of science.