I describe the key features of pure process realism-realism about the processes that are identified by experimental dynamics structured by scientific models-showing that the view meets criteria for scientific realism. I argue that process realism resolves many of the worries of the antirealist, including the problems of idealization, underdetermination, contextuality, multiplicity, and the pessimistic meta-induction. I show this resolution in the context of a contentious model from physics: the Bohr model of the atom. I then generalize from this discussion to (...) a collection of upshots and constraints on process realism as a view, in order to further distinguish it from orthodox, substance-paradigmatic- or thing-realism. Therefore, pure process realism is shown to be a form of realism compatible with empiricism. (shrink)

The importance of epistemic values in science is universally recognized, whereas the role of non-epistemic values is sometimes considered disputable. It has often been argued that non-epistemic values are more relevant in applied sciences, where the goals are often practical and not merely scientific. In this paper, we present a case study concerning earthquake engineering. So far, the philosophical literature has considered various branches of engineering, but very rarely earthquake engineering. We claim that the assessment of seismic hazard models is (...) sensitive to both epistemic and non-epistemic values. In particular, we argue that the selection and evaluation of these models are justified by epistemic values, even if they may be contingently influenced by non-epistemic values. By contrast, the aggregation of different models into an ensemble is justified by non-epistemic values, even if epistemic values may play an instrumental role in the attainment of these non-epistemic values. A careful consideration of the different epistemic and non-epistemic values at play in the choice of seismic hazard models is thus practically important when alternative models are available and there is uncertainty in the scientific community about which model should be used. (shrink)

Tracking technology has been heralded as transformative for animal ecology. In this paper I examine what changes are taking place, showing how current animal movement research is a field ripe for philosophical investigation. I focus first on how the devices alter the limitations and biases of traditional field observation, making observation of animal movement and behaviour possible in more detail, for more varied species, and under a broader variety of conditions, as well as restricting the influence of human presence and (...) observer bias. I reconstruct these as shifts in scope, objectivity, accuracy and fruitfulness. The second transformation is slightly less obvious but equally significant for animal ecology. Tracking devices generate complex data that demands both statistical and biological expertise, which has led to increasingly frequent and intensive collaborations between statisticians and biologists. Based on interviews, I examine how researchers in these interdisciplinary collaborations negotiate the collection, analysis and interpretation of movement data, integrating research interests, methodological constraints, previous field observations, and background theory. Tracking technology is therefore also shifting which disciplinary considerations are brought to bear on research into animal movement and behaviour and how this research is conducted. (shrink)

According to an adequacy-for-purpose view, models should be assessed with respect to their adequacy or fitness for particular purposes. Such a view has been advocated by scientists and philosophers...

Supervised machine learning has found its way into ever more areas of scientific inquiry, where the outcomes of supervised machine learning applications are almost universally classified as predictions. I argue that what researchers often present as a mere terminological particularity of the field involves the consequential transformation of tasks as diverse as classification, measurement, or image segmentation into prediction problems. Focusing on the case of machine-learning enabled poverty prediction, I explore how reframing a measurement problem as a prediction task alters (...) the primary epistemic aim of the application. Instead of measuring a property, machine learning developers conceive of their models as predicting a given measurement of this property. I argue that this _predictive reframing_ common to supervised machine learning applications is epistemically and ethically problematic, as it allows developers to externalize concerns critical to the epistemic validity and ethical implications of their model’s inferences. I further hold that the predictive reframing is not a necessary feature of supervised machine learning by offering an alternative conception of machine learning models as measurement models. An interpretation of supervised machine learning applications to measurement tasks as _automatically-calibrated model-based measurements_ internalizes questions of construct validity and ethical desirability critical to the measurement problem these applications are intended to and presented as solving. Thereby, this paper introduces an initial framework for exploring technical, historical, and philosophical research at the intersection of measurement and machine learning. (shrink)

This paper provides the first systematic epistemological account of simulated data in empirical science. We focus on the epistemic issues modelers face when they generate simulated data to solve problems with empirical datasets, research tools, or experiments. We argue that for simulated data to count as epistemically reliable, a simulation model does not have to mimic its target. Instead, some models take empirical data as a target, and simulated data may successfully mimic such a target even if the model does (...) not. We show how to distinguish between simulated and empirical data, and we also offer a definition of simulation that can accommodate Monte Carlo models. We shed light on the epistemology of simulated data by providing a taxonomy of four different mimicking relations that differ concerning the nature of the relation or relata. We illustrate mimicking relations with examples from different sciences. Our main claim is that the epistemic evaluation of simulated data should start with recognizing the diversity of mimicking relations rather than presuming that only one relation existed. (shrink)

Over the last twenty-five years, climate scientists working on the attribution of climate change to humans have developed increasingly sophisticated statistical models in a process that can be understood as a kind of calibration: the gradual changes to the statistical models employed in attribution studies served as iterative revisions to a measurement(-like) procedure motivated primarily by the aim of neutralizing particularly troublesome sources of error or uncertainty. This practice is in keeping with recent work on the evaluation of models more (...) generally that views models as tools for particular tasks: what drives the process is the desire for models that provide more reliable grounds for inference rather than accuracy to the underlying mechanisms of data-generation. (shrink)

In the context of astrophysical modeling at the solar system scale, we investigate the modalities implied by taking into account different levels of detail at which phenomena can be considered. In particular, by framing the analysis in terms of the how-possibly/how-actually distinction, we address the debated question as to whether the degree of plausibility is tightly linked to the degree of detail. On the grounds of concrete examples, we argue that, also in the astrophysical context examined, this is not necessarily (...) the case. (shrink)

While the predominant focus of the philosophical literature on scientific modeling has been on single-scale models, most systems in nature exhibit complex multiscale behavior, requiring new modeling methods. This challenge of modeling phenomena across a vast range of spatial and temporal scales has been called the tyranny of scales problem. Drawing on research in the geosciences, I synthesize and analyze a number of strategies for taming this tyranny in the context of conceptual, physical, and mathematical modeling. This includes several strategies (...) that can be deployed in physical modeling, even when strict dynamical scaling fails. In all cases, I argue that having an adequate conceptual model—given both the nature of the system and the particular purpose of the model—is essential. I draw a distinction between depiction and representation, and use this research in the geosciences to advance a number of debates in the philosophy of modeling. (shrink)

Estimating whether the Earth’s biota is in the middle of a crisis relies heavily on comparisons between present and past data about biodiversity or biodiversity surrogates. Although the past is a crucial source of information to assess the severity of the current biodiversity crisis, substantive conceptual and methodological questions remain about how paleodiversity and biodiversity are to be properly compared. I argue that to justify claims of a current biodiversity crisis is harder than it appears. More precisely, I claim that (...) paleodiversity measurements are incommensurable with contemporary measurements, given the different ways that biodiversity is conceptualized and operationalized. I conclude by proposing three possible ways of overcoming this incommensurability problem. (shrink)

In both scientific and popular circles it is often said that we are in the midst of a sixth mass extinction. Although the urgency of our present environmental crises is not in doubt, such claims of a present mass extinction are highly controversial scientifically. Our aims are, first, to get to the bottom of this scientific debate by shedding philosophical light on the many conceptual and methodological challenges involved in answering this scientific question, and, second, to offer new philosophical perspectives (...) on what the value of asking this question has been — and whether that value persists today. We show that the conceptual challenges in defining ‘mass extinction’, uncertainties in past and present diversity assessments, and data incommensurabilities undermine a straightforward answer to the question of whether we are in, or entering, a sixth mass extinction today. More broadly we argue that an excessive focus on the mass extinction framing can be misleading for present conservation efforts and may lead us to miss out on the many other valuable insights that Earth’s deep time can offer in guiding our future. (shrink)

Many decades ago Patrick Suppes argued rather convincingly that theoretical hypotheses are not confronted with the direct, raw results of an experiment, rather, they are typically compared with models of data. What exactly is a data model however? And how do the interactions of particles at the subatomic scale give rise to the huge volumes of data that are then moulded into a polished data model? The aim of this paper is to answer these questions by presenting a detailed case (...) study of the construction of data models at the LHCb for testing Lepton Flavour Universality in rare decays of B-mesons. The close examination of the scientific practice at the LHCb leads to the following four main conclusions: raw data in their pure form are practically useless for the comparison of experimental results with theory, and processed data are in some cases epistemically more reliable, real and simulated data are involved in the co-production of the final data model and cannot be easily distinguished, theory-ladenness emerges at three different levels depending on the scope and the purpose for which background theory guides the overall experimental process and the overall process of acquiring and analysing data in high energy physics is too complicated to be fully captured by a generic methodological description of the experimental practice. (shrink)

Biodiversity science is in a pivotal period when diverse groups of actors—including researchers, businesses, national governments, and Indigenous Peoples—are negotiating wide-ranging norms for governing and managing biodiversity data in digital repositories. These repositories, often called biodiversity data portals, are a type of organization for which governance can address or perpetuate the colonial history of biodiversity science and current inequities. Researchers and Indigenous Peoples are developing and implementing new strategies to examine and change assumptions about which agents should count as salient (...) participants in scientific projects, especially in projects that build and manage large digital data portals. Two notable efforts are the FAIR (Findable, Accessible, Interoperable, Reusable) and CARE (Collective benefit, Authority, Responsibility, Ethics) Principles for scientific data management and governance. To characterize how these principles influence the governance of biodiversity data portals, we develop an account of fit-for-use data that makes explicit its social as well as technical conditions in relation to agents and purposes. The FAIR Principles, already widely adopted by biodiversity researchers, prioritize machine agents and efficient computation, while the CARE Principles prioritize Indigenous Peoples and their data sovereignty. Both illustrate the potency of an emerging general strategy by which groups of actors craft and implement governance principles for data fitness-of-use to change assumptions about who are salient participants in data science. (shrink)