This article concerns the diachronic rationality norms for comparative confidence judgments, that is, judgments of the form “I am at least as confident in $p$ as I am in $q$.” Specifically, it identifies, characterizes, and evaluates an intuitively compelling learning rule called comparative conditionalization that specifies how agents should revise their comparative confidence judgments in the face of novel evidence.

I develop a novel account of how nonepistemic aims and values can appropriately influence scientific investigation. At its heart is a process of epistemic projection in which a nonepistemic aim or value is mapped to an epistemic research problem that aligns with that aim or value. Choices in research are then justified as a means of solving that research problem. This epistemic projection approach makes research responsive to nonepistemic aims and values yet remains consistent with the value-free ideal; it could be acceptable to parties on both sides of the values-in-science debate. It also promises to be useful in practice.

In a conventional experiment, scientists typically aim to learn about target systems by manipulating source systems of the same material type. In an analogue quantum simulation, by contrast, scientists typically aim to learn about target quantum systems of one material type via an experiment on a source quantum system of a different material type. In this article, we argue that such inferences can be justified by reference to source and target quantum systems being of the same empirical type. We illustrate this novel experimental practice of wavefunction engineering with reference to the example of Bose–Hubbard systems.

Teleosemantic theories aim to naturalize mental representation through the use of functions, typically based on past selection processes. However, the historical dependence of these theories has faced severe criticism, leading some philosophers to develop ahistorical alternatives. This article presents a new dilemma for all ahistorical teleosemantic theories, focusing in particular on the theories proposed by Timothy Schroeder and Bence Nanay. These theories require certain dispositions in the producers or consumers of mental representations, but the appeal to dispositions puts the proponents in an undesirable position: mental content is either overly dependent on current circumstances or ultimately dependent on historical factors.

Some scientific models and some claims about model–target relations are fruitfully diagnosed as dogwhistles. Dogwhistles, broadly speaking, are speech acts that send different, conflicting, and often differentially inflammatory messages to listeners. I distinguish two ways in which scientific models can be dogwhistles: representational dogwhistling and fit-for-purpose dogwhistling. I illustrate both kinds of dogwhistling using an example from computational social science, the diversity trumps ability theorem. I argue that dogwhistling threatens the objectivity of science, and I propose some ameliorative strategies.

Our thinking often uses rich memories of particular past events. Yet frequently we would do better to use other forms of memory. I show that existing accounts of the function of episodic memory cannot account for such cases, then develop an account which can. Roughly: rich representations of particular past events are required for Unrestricted Learning, learning which is not limited in how much of the world’s complexity it can capture; and episodic memory’s selection for Unrestricted Learning could explain its ubiquitous (and often inappropriate) use for other tasks. This proposal suggests many avenues for further empirical and computational research.

This article argues against general claims for the epistemic superiority of experiment over observation. It does so by dissociating the benefits traditionally attributed to experiment from physical manipulation. In place of manipulation, we argue that other features of research methods do confer epistemic advantages in comparison to methods in which they are diminished. These features better track the epistemic successes and failures of scientific research, crosscut the observation/experiment distinction, and nevertheless explain why manipulative experiments are successful when they are.

In light of recent criticisms by Woodward (2017) and Rescorla (2018), we examine the relationship between mechanistic explanation and phenomenological laws. We disambiguate several uses of the phrase “phenomenological law” and show how a mechanistic theory of explanation sorts them into those that are and are not explanatory. We also distinguish the problem of phenomenological laws from arguments about the explanatory power of purely phenomenal models, showing that Woodward and Rescorla conflate these problems. Finally, we argue that the temptation to pit mechanistic and interventionist theories of explanation against one another occludes important and scientifically relevant research questions.

The central role of such epistemic concepts as theory, explanation, model, or mechanism is rarely questioned in philosophy of science. Yet, what is their actual use in the practice of science? Here, we deploy text-mining methods to investigate the usage of 61 epistemic notions in a corpus of full-text articles from the biological and biomedical sciences (N = 73,771). The influence of disciplinary context is also examined by splitting the corpus into subdisciplinary clusters. The results reveal the intricate semantic networks that these concepts actually form in the scientific discourse, not always following our intuitions, at least in some parts of science.

Modal relationism is the view that our best physical theories can dispense with substantival space or spacetime in favor of possible configurations of particles. Kenneth Manders argues that the substantivalist conception is equivalent to this Leibnizian conception of space. To do so, Manders provides a translation $f$ from the Newtonian theory ${T_N}$ into the Leibnizian modal relationist account ${T_L}$. In this article, we show that the translation does not establish equivalence because there is no translation $g:{T_L} \mapsto {T_N}$ that preserves theoremhood. This seems to show that the modal relationist theory ${T_L}$ is less parsimonious than its substantivalist rival.

We model the impact of different incentives on journal behavior in undertaking peer review. Under one scheme, the journal aims to publish the highest-quality papers; under the second, the journal aims to maintain a high rejection rate. Under both schemes, journals prefer to set very high standards for acceptance despite allowing significant error in peer review. Under the second scheme, however, in order to encourage more submissions of mediocre papers, the journal is incentivized to make its editorial process less accurate. This leads to both worse peer review and lower-quality articles being published.

Behavioral welfare economics usually aims at mere means paternalism, helping agents better pursue their own goals. This paper discusses one initially promising way to inform policies addressed at agents who violate expected utility theory (EUT), namely what I call “CPT debiasing.” I argue that this approach is problematic even if we grant the normative authority of EUT, the descriptive adequacy of CPT (cumulative prospect theory), and the general acceptability of means paternalism. First, it is doubtful whether the CPT utility function measures what its proponents intend. Second, by imposing risk neutrality on agents, the approach involves a more problematic paternalism.

Recent work on the status of astrophysical modeling in the wake of quantum gravity indicates that a “fauxrizon” (a portmanteau of faux horizon), such as is relevant to understanding astrophysical black holes according to the fuzzball proposal within string theory, might ultimately solve the familiar black hole evaporation paradox. I clarify, with general upshots for the foundations of quantum gravity research, some of what this suggestion would amount to: identification of intertheoretic constraints on global spacetime structure in (observer-relative) semiclassical models of fuzzballs.

I discuss a contemporary solution to the dynamic problem of old evidence (POE), as proposed by Sprenger. Sprenger’s solution combines the Garber–Jeffrey–Niiniluoto (GJN) approach with Howson’s suggestion of counterfactually removing the old evidence from scientists’ belief systems. I argue that in the dynamic POE, the challenge is to explain how an insight under beliefs in which the old evidence E is known increased the credence of a scientific hypothesis. Therefore, Sprenger’s counterfactual solution, in which E has been artificially removed, does not resolve the problem. I consider several potential responses.