Clint Hurshman

This paper presents a novel theory of artifact functions, drawing from persistence-based accounts of social functions, according to which the function of an artifact consists in those of its effects that contribute to the persistence of its kind. First, the paper argues that artifact functions have an underacknowledged “interventionist task”: functional ascriptions have implications for the ways that users have reason to use technologies, and how they have reason to intervene when technologies have undesired effects. Then, it argues that the ways that users have reason to use technologies and intervene are informed by the present persistence conditions of artifact kinds. It is therefore useful to incorporate these persistence conditions into the artifacts’ functions. By focusing on the social and political embedding of artifacts, the theory also allows for fruitful connections between the theory of artifacts, the political philosophy of technology, and the literature on functions in the philosophy of social science.

This paper proposes a reevaluation of assessment methods in philosophy higher education, advocating for a shift away from traditional written assessments towards oral evaluation. Drawing attention to the rising ethical concerns surrounding large language models (LLMs), we argue that a renewed focus on oral skills within philosophical pedagogy is both imperative and underexplored. This paper offers a case for redirecting attention to the neglected realm of oral evaluation, asserting that it holds significant promise for fostering students with some of our traditional academic values that we want to maintain. We identify implications of this shift in emphasis which situates our discipline to contribute positively to solving some of the most pressing socio-political issues. Additionally, our proposal aims to demonstrate how philosophy can solidify its relevancy to the twenty-first century student and academy more broadly.

The functions of technical artifacts are closely associated with design. Increasingly, however, we depend on technologies that are not designed: algorithms produced using machine learning (ML). Machine learning uses automated optimization processes to produce algorithms that are often opaque even to developers. I argue that these opaque ML models cannot be ascribed functions on the leading design-based account, the ICE theory of Houkes and Vermaas (Technical functions: On the use and design of artefacts, Springer, 2010). Specifically, I argue that the form of rational justification that the ICE theory provides for the uses of artifacts in typical cases cannot be provided for opaque algorithms, since developers lack the necessary “support beliefs” about how the algorithm works. This does not mean that opaque ML models have no proper functions at all; rather, it lends support to non-intentional theories of technical functions, modeled after function theories in the sciences. However, these theories are meant to perform different work than the ICE theory. As a result, the kind of justification they provide for artifact use is weaker than that provided by the ICE theory and fails to justify many ongoing uses of algorithms. The paper therefore has two upshots. First, opaque ML models lack intentional technical functions. Second, while they may have functions on alternative theories, in many cases these functional ascriptions still leave the use of algorithms unjustified. Rather than extensions of our agency, opaque algorithms are a kind of found object with unknown properties that may or may not be suited to the purposes to which we put them.

Joshua Christie, Carl Brusse, Pierrick Bourrat, Peter Takacs, and Paul Griffiths argue that selected-effects (SE) functions generally fail to causally explain traits because they omit some explanatorily essential information. Heterogeneous environments, bet-hedging strategies, and frequency-dependence all produce selection dynamics that are explanatorily important but that are left out when we focus exclusively on the conditions under which a given trait was adaptive. Thus, they argue, the SE theory gives inadequate explanations since it only picks out a limited set of explanatorily relevant factors. Some other authors in this volume have responded by trying to clarify the explanatory goals of SE functions in order to show that their omission of these phenomena is unproblematic. I take a different approach, arguing that in some important cases—including, perhaps, the case of signalling—environmental heterogeneity and coevolution can be explanatorily relevant because they are some of the functional effects of traits. I agree with the authors that these dynamics are important for explaining adaptations; however, I suggest that they need not be left out by SE explanations, because they can be among the effects for which traits are selected. Niche construction theory offers a useful framework for understanding these dynamics in this way.