The COVID-19 pandemic and global climate change crisis remind us that widespread trust in the products of the scientific enterprise is vital to the health and safety of the global community. Insofar as appropriate responses to these (and other) crises require us to trust that enterprise, cultivating a healthier trust relationship between science and the public may be considered as a collective public good. While it might appear that scientists can contribute to this good by taking more initiative to communicate their work to public audiences, we raise a concern about unintended consequences of an individualistic approach to such communication.

Transparency has been proposed as a way of handling value-ladenness in machine learning (ML). This article highlights limits to this strategy. I distinguish three kinds of transparency: epistemic transparency, retrospective value transparency, and prospective value transparency. This corresponds to different approaches to transparency in ML, including so-called explainable artificial intelligence and governance based on disclosing information about the design process. I discuss three sources of value-ladenness in ML—problem formulation, inductive risk, and specification gaming—and argue that retrospective value transparency is only well-suited for dealing with the first, while the third raises serious challenges even for prospective value transparency.

Under what conditions does machine learning (ML) model opacity inhibit the possibility of explaining and understanding phenomena? In this article, I argue that nonepistemic values give shape to the ML opacity problem even if we keep researcher interests fixed. Treating ML models as an instance of doing model-based science to explain and understand phenomena reveals that there is (i) an external opacity problem, where the presence of inductive risk imposes higher standards on externally validating models, and (ii) an internal opacity problem, where greater inductive risk demands a higher level of transparency regarding the inferences the model makes.

A recent controversy about neural networks allegedly capable of detecting a person’s sexual orientation raises the question of whether all research on homosexuality should be permitted. This paper considers two arguments for limits to such research, and concludes that there are good reasons to limit at least the dissemination of applied research on the etiology of homosexuality. The paper then briefly sketches how this could work, and looks at three objections against these limitations.

This article analyzes the scientific debate surrounding the appropriateness of pharmacological interventions aimed at increasing female desire. It has three objectives: (1) to outline the nonepistemic values that underlie scientific concepts of desire; (2) to conduct an analysis of scientific misuses when justifying a drive-based model of desire and pharmacological intervention; and (3) to show how in this and similar cases of vague constructs we can demonstrate the role of nonepistemic values and assess their influence.

The standard view about sex differences in sexual desire is that males are lusty and loose, while females are cool and coy. This is widely believed and is a core premise of some scientific programs like evolutionary psychology. But is it true? A mountain of evidence seems to support the standard view. Yet, this evidence is shot through with methodological and philosophical problems. Developments in the study of sexual desire suggest that some of these problems can be resolved, and when they are, the standard view looks, at best, to be an exaggeration.

Following recent arguments that cultural practices in wild animal populations have important conservation implications, we argue that recognizing captive animals as cultural has important welfare implications. Having a culture is of deep importance for cultural animals, wherever they live. Without understanding the cultural capacities of captive animals, we will be left with a deeply impoverished view of what they need to flourish. Best practices for welfare should therefore require concern for animals’ cultural needs, but the relationship between culture and welfare is also extremely complex, requiring us to rethink standard assumptions about what constitutes and contributes to welfare.

Definitions of animal welfare often invoke consciousness or sentience. Marian Stamp Dawkins has argued that to define animal welfare this way is a mistake. In Dawkins’s alternative view, an animal with good welfare is one that is healthy and “has what it wants.” The dispute highlights a source of strain on the concept of animal welfare: consciousness-involving definitions are better able to capture the normative significance of welfare, whereas consciousness-free definitions facilitate the validation of welfare indicators. I reflect on how the field should respond to this strain, ultimately recommending against splitting the concept and in favor of consciousness-involving definitions.

We show that as game theory was transferred from mathematical oncology to experimental cancer biology, a new mode of inquiry was created. Modeling was replaced by measuring. The game measured by a game assay can serve as a bridge that allows knowledge to flow backward from target (cancer research) to source (game theory). Our finding suggests that the conformist and creative (Houkes and Zwart 2019) types of transfer need to be augmented. We conclude by introducing the expansive and transformative types to get a four-tier typology of knowledge transfer.

Any given ecosystem includes macro- and micro-organismal ecology. I examine the concept of stability and its meaning in these two size scales. Historically, the stability concept follows species population dynamics and equilibrium. In micro-ecology, however, classification is rare and, in the majority of cases, not possible. Therefore, the microbial taxonomic composition is regarded in a nonspecific way, as opposed to animals and plant communities. Such different consideration leads to the emphasis on functional stability in micro-ecology versus the structural in macro-ecology. I discuss the implication of these differences in meanings and the possibilities of integrating them.

Most accounts of mechanism discovery have focused on mechanisms that perform the work required to produce a phenomenon. These mechanisms are often subject to regulation by control mechanisms. Using the example of the molecular motor dynein, this paper examines one process by which such control mechanisms are discovered—the process by which researchers, after identifying additional components required to produce the phenomenon but not directly involved in the work of producing that phenomenon, investigate both how these components act on the original mechanism and how they do so in response to measurements of conditions relevant to the operation of the controlled mechanism.

A minimal essentialism (“intrinsic biological essentialism”) about natural kinds is required to explain the projectability of human science terms. Human classifications that yield robust and ampliative projectable inferences refer to biological kinds. I articulate this argument with reference to an intrinsic essentialist account of homeostatic property cluster kinds. This account implies that human sciences (e.g., medicine, psychiatry) that aim to formulate predictive kind categories should classify biological kinds. Issues concerning psychiatric classification and pluralism are examined.

This paper challenges the exclusion of patients from epistemic practices in psychiatry by examining the creation and revision processes of the Diagnostic and Statistical Manual of Mental Disorders (DSM), a document produced by the American Psychiatric Association that identifies the properties of mental disorders and thereby guides research, diagnosis, treatment, and various administrative tasks. It argues there are epistemic—rather than exclusively social/political—reasons for including patients in the DSM revision process. Individuals with mental disorders are indispensable resources to enhance psychiatric epistemology, especially in the context of the crisis, controversy, and uncertainty surrounding mental health research and treatment.

Transformations in the “behavioral innovativeness” of species—broadly, the capacity to generate new or novel behaviors—have been associated with significant evolutionary shifts in cognition by both philosophers and scientists. Whilst intuitively and theoretically appealing, this assumption lacks strong empirical support. One barrier is the absence of a good measure of behavioral innovation. This paper offers a solution to this problem by breaking down innovation into its components and presenting a novel multi-dimensional framework for characterising and comparing putative cases of behavioral innovation.

A common strategy in comparative cognition is to require that one reject associative learning as an explanation for behavior before concluding that an organism is capable of causal reasoning. In this paper, I argue that standard causal-reasoning tasks can be explained by a powerful form of associative learning: unlimited associative learning (UAL). The lesson, however, is not that researchers should conduct more studies to reject UAL, but that they should instead focus on 1) enriching the cognitive hypothesis space and 2) testing a broader range of information processing patterns—errors, biases and limits, rather than successful problem solving alone.

One challenge in explaining neural evolution is the formal equivalence of different computational architectures. If a simple architecture suffices, why should more complex neural architectures evolve? The answer must involve the intense competition for resources under which brains operate. I show how recurrent neural networks can be favored when increased complexity allows for more efficient use of existing resources. Although resource constraints alone can drive a change, recurrence shifts the landscape of what is later evolvable. Hence organisms on either side of a transition boundary may have similar cognitive capacities but very different potential for evolving new capacities.

In the quest to discover the neural bases of cognition, rigorous behavioral tools are equally as important as sophisticated tools for neural intervention. This paper evaluates several episodes in the development of a novel behavioral tool for rodent cognitive testing, the rodent touchscreen operant chamber. Using conceptual tools on offer in the philosophical literature on exploratory experimentation and control, I illuminate how optimization of this behavioral tool and an understanding of the causal knowledge it may be used to generate historically has involved dynamic interplay between community-driven exploratory practices operating in parallel with causal hypothesis-driven research.

Neuroscientists have in recent years turned to building models that aim to generate predictions rather than explanations. This “predictive turn” has swept across domains including law, marketing, and neuropsychiatry. Yet the norms of prediction remain undertheorized relative to those of explanation. I examine two styles of predictive modeling and show how they exemplify the normative dynamics at work in prediction. I propose an account of how predictive models, conceived of as technological devices for aiding decision-making, can come to be adequate for purposes that are defined by both their guiding research questions and their larger social context of application.

We argue that several apparently distinct responses to the hole argument, all invoking formal or mathematical considerations, should be viewed as a unified “mathematical response.” We then consider and rebut two prominent critiques of the mathematical response before reflecting on what is ultimately at issue in this literature.