Modeling Practices

This essay investigates Newton's thoughts on laws of nature. First, we argue that Newton’s experimental-philosophical methodology of hypotheses and queries exhibits a modal-exploratory nature. Second, this modal-exploratory character has immediate consequences for his metaphysics of laws of nature. In the introduction, we explain the main characteristics of modal-exploratory modelling as a modelling technique involving targetless models (i.e. non-existing entities; cf. Massimi 2019, 2022). In the second section, we show that Newton’s modelling of the cause of universal gravitation via his earlier (...) (presented in his optical correspondences from 1675) and later aether models (presented in the Opticks of 1704 and 1717) are paradigmatic cases of modal-exploratory modelling. In the third and final section, we show further that also the context of Newton's metaphysics of laws/active principles from “Query 31” of the second edition of the Opticks (1717) can be seen as a case of modal-exploratory modelling. Recent commentators (e.g., Biener & Schliesser 2017, Henry 2020, Hazelwood 2023, Walsh 2017) remarked that Newton’s definition of active principles as laws, causes, natures, or dispositions gives rise to various conceptual tensions. However, we aim to show that the tension inherent in his conception of laws is merely apparent since he is wrestling with multiple possibilities of making a theory of laws work. According to our modal-exploratory reading, Newton attempts to find out a plausible contender for what laws of nature may be. Relative to our reading, his definition of active principles in terms of laws and dispositions should not be taken at face value, but seen as a “never-at-rest” exploratory activity. So, we will reach the conclusion that Newton's modelling procedure is an exploratory one, being situated at the intersection of philosophical- experimental methodology and his metaphysics of laws. (shrink)

Quantum mechanics lacks a formal account of the observer, leaving the measurement postulate structurally incomplete. I introduce a minimal operator chain: J, A, C, L, R, that formalizes recognition, articulation, collapse, and observation. Inserting these operators into the standard measurement rule yields a complete and stable measurement structure without altering quantum predictions. A spin‑measurement example and a reconstruction of Wigner’s friend demonstrate that paradoxes dissolve when collapse is explicitly observer‑indexed. -/- Authored by Eloy Escagedo Gutierrez as part of The Universal (...) Principle of Collapse (UPC) Research Project. (shrink)

This paper was written for and presented at a symposium on Multiple Realizability at the Central Division of the APA in 2022. It's in somewhat rough shape, especially the later parts. I hope to be in a position soon to post a revised and more carefully worked out version. The basic argument of the first half is this: Realization of the interesting sort (and thus MR of the interesting sort) requires tidy separation of levels (with realizers being at a lower (...) level than that which they realize). The success of mixed-scale, mixed-resource (i.e., multi-grain) modeling strongly suggests that such a tidy separation of levels is not in the cards. Thus, MR is (likely) not a widespread phenomenon in our universe. The second half of the paper develops a relation of "multiple explainability" that is meant to play the role, in many contexts, of MR, while being metaphysically and methodologically messier than MR. (shrink)

This paper constitutes a radical departure from the existing philosophical literature on models, modeling-practices, and model-based science. I argue that the various entities and practices called 'models' and 'modeling-practices' are too diverse, too context-sensitive, and serve too many scientific purposes and roles, as to allow for a general philosophical analysis. From this recognition an alternative view emerges that I shall dub model anarchism.

In its Sixth Assessment Report, the Intergovernmental Panel on Climate Change (IPCC) moved from a climate model-based to a climate model-supported assessment of equilibrium climate sensitivity (ECS). Unlike all previous reports, climate model information is no longer used directly for estimates of ECS. This article offers a measurement view on the ECS assessment that allows us to evaluate this shift in practice in terms of its practical and epistemic consequences. In particular, I argue for two main conclusions involved in this (...) move to a climate model-supported assessment. First, the change in using climate models as direct sources of evidence to a supporting role involves a kind of repurposing of climate models as epistemic tools. This change should be evaluated relative to the researchers’ aims, which include more accurately assessing ECS. Second, and more generally, I suggest that this shift in practice shares a kinship to calibration activities aimed at increasing the reliability of a measurement procedure. (shrink)

In this paper, I shall provide a novel response to the argument from context-sensitivity against truth-conditional semantics. It is often argued that the contextual influences on truth-conditions outstrip the resources of standard truth-conditional accounts, and so truth-conditional semantics rests on a mistake. The argument assumes that truth-conditional semantics is legitimate if and only if natural language sentences have truth-conditions. I shall argue that this assumption is mistaken. Truth-conditional analyses should be viewed as idealised approximations of the complexities of natural language (...) meaning. From this perspective, disparity between the scientific model and its real-world target is to be expected. I elaborate on what such an approach to semantics would look like. (shrink)

There is a question of whether de-idealization is needed for justified use of -- for 'checking' -- idealizations. We argue that the standard philosophical account of de-idealization has become too idealized, but that this does not preclude the possibility of justificatory practices which show how models can be used to make inferences about the world. In turn, motivated by examples in physics, we provide a more expansive and practice-driven account of de-idealization by relaxing the standards for closeness to more realistic (...) theoretical items, identifying at least three kinds of procedures for de-idealization: intra-model, inter-model, and measurement de-idealizations. These examples highlight how idealizations can be -- and indeed have been -- scrutinized within physics without appealing to the philosopher's idealized notion of de-idealization. (shrink)

In contrast to the syntactic conception of scientific theories, the semantic conception holds that theories are not statements about the world, but families of models. Recent debates have tended to blur the differences between these two views. Practice-oriented philosophers of science have also challenged both views on the ground that models are central to science, but autonomous from theories. However, they have not proposed any alternative. This article is an attempt to sharpen the challenges faced by the semantic view in (...) order to pave the way for a truly pragmatic conception of theories. I examine different ways in which a focus on models could make the semantic view different from the syntactic view. In the end, they share the same resources to individuate theories, they account just as well for different layers of interpretation, and they face the same issues regarding the relation between formal representation and informal experience. I argue that the only relevant contrast lies not between linguistic and non-linguistic representation, but between descriptive and schematic representation, but this ultimately favours a pragmatic view by implying a role for context in theoretical interpretation. I sketch an account of schematic representation that could serve as a basis for a pragmatic conception of theories. (shrink)

The central distinction—between evaluative and generative metabolic considerations—appears less theoretically secure than Haueis and Colaço suggest. First, the generative examples may ultimately reduce to evaluative reasoning applied earlier in the modeling process. Second, the principal generative examples do not clearly show metabolic considerations functioning within a cognitive model.

Feynman diagrams are used to calculate scattering amplitudes in quantum field theory, where they simplify the derivation of individual terms in the corresponding perturbation series. Considered mathematical tools with an approximative character, the received view in the philosophy of physics denies that individual diagrams can represent physical processes. A different story, however, can be observed in physics practice. From education to high-profile research publications, Feynman diagrams are used in connection with particle phenomena without any reference to perturbative calculations. In the (...) first part of the paper, I argue that this illuminates an additional use of Feynman diagrams that is not calculatory but representational. It is not a possible translation into mathematical terms that prompts this practice but rather the epistemic insights into the target phenomenon that the diagrams provide. Based on this practical use, I intend to push back against the received view. In the second part of the paper, I conceptualize the representative use of Feynman diagrams as models that provide modal understanding of their associated target phenomena. The set of Feynman diagrams corresponding to an interaction is taken as a possibility space whose dependency relations can be analysed, allowing an agent to grasp possible target behaviour, leading to understanding. In clearly separating the diagrams from perturbative calculations for their use as a model, the concerns that hinder a representative reading can be resolved. (shrink)

When it originated in the late 19th century, psychometrics was a field with both a scientific and a social mission: psychometrics provided new methods for research into individual differences, and at the same time, these psychometric instruments were considered a means to create a new social order. In contrast, contemporary psychometrics - due to its highly technical nature and its limited involvement in substantive psychological research - has created the impression of being a value-free discipline. In this article, we develop (...) a contrasting characterization of contemporary psychometrics as a value-laden discipline. We expose four such values: that individual differences are quantitative (rather than qualitative), that measurement should be objective in a specific sense, that test items should be fair, and that utility of a model is more important than its truth. Our goal is not to criticize psychometrics for employing these values, but rather to bring them into the open and to show that these values are not inevitable and are in need of systematic evaluation. (shrink)

The paper thus shows that scientific rationality under finite conditions consists not only in stabilization, critique, and exploration, but equally in the controlled capacity to transform existing model orders. It begins from the observation that scientific rationality cannot be adequately described if one considers only stabilization, falsification, search, or complete model replacement. Between the provisional retention of a model and its complete replacement lies a field of regulated transformations, which the present contribution conceptualizes and systematizes as revision. Revision denotes the (...) rule-governed transformation of a model or model nexus under conditions of rising friction, declining validity, increasing costs, or intensifying domain conflicts. The paper clarifies the structural position of revision within Epistemics and distinguishes it from adjacent operations such as falsification, search, and mere stabilization. On that basis, it develops a typology of local, structural, domain-related, and global revision. It also identifies typical triggers of revision, formulates criteria of rational revision, and analyzes central pathologies, including omitted, delayed, cosmetic, and disproportional revision, as well as ontologization that blocks revision. The central thesis is that revision constitutes the mediating transformative operation between friction, falsification, and search. Friction signals strain, falsification marks loss of validity, search opens the space of possible alternatives, and revision organizes the actual transformation of the model. The paper thus shows that scientific rationality under finite conditions consists not only in stabilization, critique, and exploration, but equally in the controlled capacity to transition existing model orders. (shrink)

Computer simulations are commonly employed when researchers work with analytically intractable or practically unsolvable mathematical modeling equations. In such cases, scientists seem to deal with two different but interrelated kinds of models: a mathematical model and a simulation model. This raises at least two philosophically interesting questions. First, does one or the other model figure centrally in the activity of generating an explanation in such situations? And second, what could an account of explanation involving both mathematical models and simulation models (...) look like? I will argue that, in a large set of cases, the simulation model serves the central role in explanatory discovery. On this basis, I will then present a counterfactual account of simulation model-induced explanation. I claim that on this approach, we often find that the simulation model possesses an explanatory autonomy from its underlying mathematical models and conclude by relating this notion to extant views on the autonomous role of scientific models. (shrink)

Scientists often disagree on the best theory to describe a scientific event. While such debates are a natural part of healthy scientific discourse, the timeframe for scientists to converge on an ideal method may not always align with real-life knowledge dynamics. In this article, I use an event from the history of chemistry as inspiration to develop Agent-Based Models of epistemic networks, exploring method selection within a scientific community. These models reveal several situations where incorrect, simpler methods can persist, even (...) when substantial evidence supports a more complex method. This becomes particularly evident when different evidence-sharing timeframes are analyzed. The network structure connecting the scientists plays a crucial role in determining how and when convergence on the correct method is achieved, guided by real-world evidence. This framework provides a foundation for further exploration of scientists’ behavior in past and future discoveries, as well as how agents internalize scientific information. (shrink)

Scientists and philosophers of science often draw methodological lessons from successful theories to justify methods of theory construction and to guide research programs. This paper proposes an epistemic framework for this practice, articulated in terms of the notion of meta-induction. By analogy to Goodman's `New Riddle of Induction', it introduces the concept of projectability of meta-inductive arguments, and demonstrates its significance in any account of meta-inductive reasoning. Likewise to scientific induction, meta-induction is shown to be constrained by naturalist epistemology in (...) determining which classifications of theoretical methods are `projectable'. Projectability judgments in meta-inductive arguments help ensure that when taking lessons from past theorizing (either as a part of scientific discussions or in a philosophical reflections about science), they would be based on the way evidence was exploited in theory construction or on theoretically plausible hypothesis about the natural world, rather than on human-made concepts. This framework emphasizes the primacy of empirically guided patterns of conjecture over purely formal considerations, enabling epistemic evaluation of research projects, even when theorizing is primarily grounded on abstract theoretical methods and at stages when direct evidence is unavailable. The paper discusses possible implications on philosophical debates concerning theory assessment in fundamental physics, particularly regarding meta-empirical justification, as well as discussions on model transfer in the special sciences. (shrink)

In this Open-Access-book, the author investigates the value dimension of Integrated Assessment Models and their application to questions of feasibility. Integrated models provide a quantified representation of the interaction between the socio-economic system with the climate and serve as a pivotal tool at the intersection of climate science, policymakers, and society. This book critically examines how IAMs approach the concept of feasibility. It unpacks the value assumptions embedded within integrated modeling, critiques the implicit normativity of these models, and proposes principles (...) for responsibly managing value judgments in scientific advice. Arguing for a thick conception of feasibility—one that integrates key normative dimensions—it highlights the gap between conceptual discussions of feasibility and the scientific practices that inform feasibility assessments. Addressed to both philosophers and integrated modelers, this book sheds light on the implicit values shaping our collective visions for the future. -/- . (shrink)

Modeling cuts across sundry scientific practices, contributing to theorizing, experimentation, prediction, measurement, scientific instrumentation, and science education. Beyond the sciences, modeling plays a crucial role in citizen engagement with science and public policy decision-making. It plays a major role in the efforts to address the huge challenges of the 21st century, including but not limited to climate change, shortage of natural resources, loss of biodiversity, and economic forecasting in increasingly unforeseeable situations. The diversity of scientific models is astounding; side-by-side mathematical (...) and scale models, technological advances such as rapidly expanding big data, computational and synthetic approaches, and generative AI are pushing modeling toward new frontiers, redefining the epistemic agency between humans and scientific instruments. A discussion of what we can achieve through modeling, and how we should manage model-based practices, is critical for ensuring a good and responsible use of this epistemic resource. The chapters of The Routledge Handbook of Philosophy of Scientific Modeling, written by experts in various areas of the philosophy of science, seek to provide enduring philosophical insights and useful analyses for understanding modeling in its multiplicity. -/- . (shrink)

Models are used to explore possibilities across all scientific fields. Climate models simulate the potential future climatic conditions under various emissions scenarios, macroeconomic models investigate the implications of various fiscal and monetary policy initiatives, and infectious diseases models study the spread of viral diseases under a range of conditions. Such modeling approaches have not gone ignored by philosophers of science, but they have only recently started to explicitly address modeling the possible. So far, the discussion has been spread across a (...) variety of more or less isolated pockets of debate in the philosophy of science. Modeling the Possible: Perspectives from Philosophy of Science draws together these studies, focusing specifically on how various modeling practices probe possibilities and justify claims concerning them. -/- The volume is divided into three sections, plus an introductory chapter. The introductory chapter provides a state-of-the-art survey of the discussions of modeling possibilities within the philosophy of science, as well as an introduction to the book’s main themes and individual papers. The three sections focus on different kinds of possibility concepts, possibility spaces, and how-possibly modeling in practical situations. The chapters contained in this volume address conceptual and theoretical issues while also presenting case studies from various scientific domains: physics, evolutionary and synthetic biology, network science, climate science, economics, and epidemiology. -/- Essential reading for philosophers of science, epistemologists, and modelers in various scientific disciplines, Modeling the Possible is also suitable for anyone interested in model-based scientific inferences, their validity, and the policy conclusions derived from them. (shrink)

Scientific thinking is one of the most creative expressions of human cognition. This paper discusses my research contributions to the cognitive science of science. I have advanced the position that data on the cognitive practices of scientists drawn from extensive research into archival records of historical science or collected in extended ethnographic studies of contemporary science can provide valuable insight into the nature of scientific cognition and its relation to cognition in ordinary contexts. I focus on contributions of my research (...) on analogy, model-based reasoning, and conceptual change and on how scientists enhance their natural cognitive capacities by creating modeling environments that integrate cognitive, social, material, and cultural resources. I provide an outline of my trajectory from a physicist to a philosopher of science to a hybrid cognitive scientist in my quest to understand the nature of scientific thinking. (shrink)

The idea of using lattice methods to provide a mathematically well-defined formulation of realistic effective quantum field theories (QFTs) and clarify their physical content has gained traction in the last decades. In this paper, I argue that this strategy faces a two-sided obstacle: realistic lattice QFTs are (i) too different from their effective continuum counterparts even at low energies to serve as their foundational proxies and (ii) far from reproducing all of their empirical and explanatory successes to replace them altogether. (...) I briefly conclude with some lessons for the foundations of QFT. (shrink)

This chapter examines modal modelling in climate science. It considers two related topics. The first is the use of climate models to attribute extreme weather events to climate change. The second is the interpretation and use of collections of climate models. Each topic is the subject of a current debate within climate science and philosophy of science, and each has an important modal component. The debates are similar in that each involves a contrast between probabilistic and non-probabilistic methods. In each (...) case, the chapter examines the kind of modal claims scientists make and how they are justified. As climate change is an important practical problem, each debate has a link to policy decisions, and so the chapter ends with a discussion of how modal modelling, and the possibilities it identifies, can support decision-making. (shrink)

Several philosophers of science have drawn attention to a number of modeling practices where scientific models primarily contribute modal information. Examples now abound, and, recently, there have also been some preliminary attempts to address questions of under what conditions, and by virtue of what, models can perform this modal epistemic function. This paper sets out to constructively review those attempts through a prism of the more general literature on the epistemology of modality. One aim of this exercise is to expose (...) the underlying justificatory strategies of the modal modeling accounts – which are strikingly convergent with those already identified by modal epistemologists. Another aim is to highlight some challenges familiar to modal epistemology and discuss the form they take in the context of modal modeling as well as the resources that modal modeling accounts have for dealing with them. This is all done with an eye to pointing out ways forward in further developing the work on the epistemology of modal modeling. (shrink)

Here I draw on a recent movement in preclinical biomedical research – use of what are often called “dirty mice” – to challenge the standard view that less controlled experiments have diminished epistemic power. Dirty mice are research mice with diverse microbial exposures, and dirty mouse methods, I argue, are strategies for loosening the control that researchers have over their experiments. I use recent experimental results to show why, as compared to conventional mouse models, dirty mice may better support the (...) translation of findings from laboratories to human clinical contexts. I discuss how to evaluate the claim that dirty mice, as compared to conventional mice, both (a) more closely match relevant human features and (b) more accurately predict human responses (to drugs and interventions). Thus, contra the control ideal, deliberately relinquishing some experimental control improves an epistemic desideratum (in this case, translatability). This paper contributes to literature on the philosophy of experiment, and the final upshot will be a suggestion for when scientists in other fields might consider purposely designing less controlled experiments. (shrink)

Scientific fields frequently need to exchange data to advance their own inquiries. Data unification is the process of stabilizing these forms of interfield data exchange. I present an account of the epistemic structure of data unification, drawing on case studies from model-based cognitive neuroscience (MBCN). MBCN is distinctive because it shows that modeling practices play an essential role in mediating these data exchanges. Models often serve as interfield evidential integrators, and models built for this purpose have their own representational and (...) inferential functions. This form of data unification should be seen as autonomous from other forms, particularly explanatory unification. (shrink)

From Plato to the present day, political theorists have used models of idealistic societies to think about politics. How can these idealistic models inform our thinking about political life in our non-ideal world? Not, as many political theorists have hoped, by providing normative guidance -- by showing us how things should be or where we should go. Even still, we can use these models to interpret the concepts we depend on to explain and evaluate political behavior and institutions, thereby sharpening (...) our explanations of how things work and our estimations of how to make things better. Seeing idealistic models in this light reveals a promising route to a form of political inquiry that integrates our explanatory aims and our normative aspirations. -/- (I have uploaded proofs for the table of contents, chapter 1, and the bibliography. A note on appendix corrections is available under "Teaching Materials".). (shrink)

According to a widely held view, scientific modelling consists in entertaining a set of model descriptions that specify a model. Rather than studying the phenomenon of interest directly, scientists investigate the phenomenon indirectly via a model in the hope of learning about some of the phenomenon’s features. I call this view the description-driven modelling (DDM) account. I argue that although an accurate description of much of scientific research, the DDM account is found wanting as regards the mechanistic modelling found in (...) many branches of biology. By analysing research practices in cancer immunology concerning the development of mechanistic models of the process of cancer metastasis, this paper presents and argues for a complementary account of scientific modelling, herein called the experimentation-driven modelling (EDM) account. In EDM, scientists investigate a set of experimental systems and then integrate the results obtained from experiments into a mechanistic model. While EDM shares some key features with DDM, the two are epistemically very different approaches to research. (shrink)

Default logic has been a very active research topic in artificial intelligence since the early 1980s, but has not received as much attention in the philosophical literature thus far. This paper shows one way in which the technical tools of artificial intelligence can be applied in contemporary epistemology by modeling a paradigmatic case of deep disagreement using default logic. In §1 model-building viewed as a kind of philosophical progress is briefly motivated, while §2 introduces the case of deep disagreement we (...) aim to model. On the heels of this, §3 defines our formal framework, viz., a refined Horty-style default logic. §4 then uses the framework to model deep disagreement, and finally §5 provides a critical discussion of the result. (shrink)

Mitchell & Gronenborn ( 2017 ) propose that we account for the presence of multiple models of protein structure, each produced in different contexts, through the framework of integrative pluralism. I argue that two interpretations of this framework are available, neither of which captures the relationship between a model and the protein structure it represents or between multiple models of protein structure. Further, it inclines us toward concluding prematurely that models of protein structure are right in their contexts and makes (...) extrapolation of findings from one context to another seem unwarranted. Instead, protein structure determination ought to be understood as _modestly monistic_. There is one model for every protein in each physicochemical context, and models of the same protein produced in different contexts are compatible with one another. ‘Integrating’ multiple models amounts to extrapolating from one context to another; this is possible because the effect of context on protein folding is relatively weak and predictable. Modest monism better describes the practice of protein structure determination than integrative pluralism and enables greater attention to how context affects protein folding. (shrink)

The philosophical literatures on models and thought experiments have been developing exponentially, and independently, for decades. This independence is surprising, given how similar models and thought experiments are. They each have “lives of their own,” they sit between theory and experience, they are important for both pedagogy and cutting-edge science, they galvanize conceptual changes and paradigm shifts, and they involve entertaining imaginary scenarios and working out what happens. Recently, philosophers have begun to highlight these similarities. This entry aims at taking (...) the idea further, by trying to systematically identify places where insights from one literature can be taken up in the other. Along the way, important differences will also be highlighted. (shrink)

A conspicuous feature of contemporary modelling practices is the use of the same mathematical forms and modelling methods across different scientific domains. This model transfer raises many philosophical questions concerning, for example, the exact object of transfer, the relationship between the model and the target domain, the specific challenges such transfer confronts, and the ways in which model transfer relates to scientific progress. While the interest in studying model transfer has increased among philosophers of science in recent years, the phenomenon (...) has not yet received its due attention. The literature is scattered and much of it has addressed questions around the factors that might explain model transfer, such as whether the different target phenomena to which a model is applied eventually prove to be structurally and dynamically similar, or whether they have to share more fundamental ontological features that explain the interdisciplinary success of distinct models. This chapter surveys the existing literature on model transfer with the main goal of highlighting the main findings of existing discussions and pointing out open issues whose discussion would improve our understanding of this highly relevant phenomenon in science. (shrink)

I argue that dimensional analysis provides an answer to a skeptical challenge to the theory of model mediated measurement. The problem arises when considering the task of calibrating a novel measurement procedure, with greater range, to the results of a prior measurement procedure. The skeptical worry is that the agreement of the novel and prior measurement procedures in their shared range may only be apparent due to the emergence of systematic error in the exclusive range of the novel measurement procedure. (...) Alternatively: what if the two measurement procedures are not in fact measuring the same quantity? The theory of model mediated measurement can only say that we _assume_ that there is a common quantity. In contrast, I show that the satisfaction of dimensional homogeneity across the metrological extension is independent evidence for the so-called assumption. This is illustrated by the use of dimensional analysis in high pressure experiments. This results in an extension of the theory of model mediated measurement, in which a common quantity in metrological extension is no longer assumed, but hypothesized. (shrink)

The digitalization of human society continues at a relentless rate. However, to develop modern information technologies, the increasing complexity of the real-world must be modeled, suggesting the general need to reconsider how to carry out conceptual modeling. This research proposes that the often-overlooked notion of ‘‘system’’ should be a separate, and core, conceptual modeling construct and argues for incorporating it and related concepts, such as emergence, into existing approaches to conceptual modeling. The work conducts a synthesis of the ontology of (...) systems and general systems theory. These modeling foundations are then used to propose a CESM+ template for conducing systems-grounded conceptual modeling. Several new conceptual modeling notations are introduced. The systemist modeling is then applied to a case study on the development of a citizen science platform. The case demonstrates the potential contributions of the systemist approach and identifies specific implications of explicit modeling with systems for theory and practice. The paper provides recommendations for how to incorporate systems into existing projects and suggests fruitful opportunities for future conceptual modeling research. (shrink)

This paper argues that investigations into how to best make something often provide researchers with an objectual understanding of their target phenomena. This argument starts with an extended investigation into the non-representational uses of models. In particular, we identify a special sort of “design model” whose aim is to guide the production of phenomena. Clarifying how these design models are evaluated shows that they are evaluated in different ways than representational models. Once the character of design models has been fixed, (...) we argue that grasping design models can provide objectual understanding of phenomena. This argument proceeds through a critical engagement with Dellsén’s ( 2020 ) position that a grasp of a good representational model of dependencies provides objectual understanding of a phenomenon. We agree with Dellsén that this is one way to achieve understanding, but maintain that grasping a good design model is another way to achieve understanding. The paper concludes by considering some important objections to our proposal and also by noting some of the broader questions about understanding and knowledge in both science and engineering. (shrink)

Deep learning (DL) is a statistical technique for pattern classification through which AI researchers train artificial neural networks containing multiple layers that process massive amounts of data. I present a three-level account of explanation that can be reasonably expected from DL models in cognitive neuroscience and that illustrates the explanatory dynamics within a future-biased research program (Feest Philosophy of Science 84:1165–1176, 2017 ; Doerig et al. Nature Reviews: Neuroscience 24:431–450, 2023 ). By relying on the mechanistic framework (Craver Explaining the (...) brain: Mechanisms and the mosaic unity of neuroscience. Clarendon, 2007 ; Stinson Eppuor si muove: Doing history and philosophy of Science with Peter Machamer. Springer, 2017, The Routledge Handbook of the computational mind. Routledge, 2018 ), I develop an account that corresponds to the stages of mechanism discovery, i.e., our shifting epistemic position and epistemic goals, and propose a representative model for each level. Generic, theoretical DL models at _Level 1_ address the general features of a cognitive phenomenon through exploration and provide us with how-possibly explanations. On the other hand, DL models at _Level 2_ either identify the interaction between the features or represent the epistemic stage when the researcher is still unsure if the modeled features are crucial or arbitrary. These models should provide us with how-plausibly explanations. Finally, specific DL models of specific brain areas, i.e., ersatz models filled with relevant cognitive and neuroscientific details, are at _Level 3_. At this level, a researcher can advance how-actually explanations of cognitive phenomena. The main strength of this account is that it elucidates both _global explanatory dynamics_ and _local explanatory dynamics_ (cf. Dresow Synthese 199:10441–10474, 2021 ). The former occurs when the transition between levels happens in accordance with the process of obtaining more details about a particular cognitive phenomenon through multiple DL models. The latter, meanwhile, involves cases in which the transition between levels takes place within a single DL model by elucidating internal mechanisms (e.g., using Explainable AI techniques for rendering models more transparent). (shrink)

Even in areas of philosophy of science that don’t involve formal treatments of truth, one’s background view of truth still centrally shapes views on other issues. I offer an informal way to think about truth as trueing, like trueing a bicycle wheel. This holist approach to truth provides a way to discuss knowledge products like models in terms of how well-trued they are to their target. Trueing emphasizes: the process by which models are brought into true; how the idealizations in (...) models are not false but rather like spokes in appropriate tension to achieve a better-trued fit to target; and that this process is not accomplished once and done forever, but instead requires upkeep and ongoing fine-tuning. I conclude by emphasizing the social importance of being a pragmatist about truth in order to accurately answer questions about science such as, “but do we really know that…”. (shrink)

A critical review of Collin Rice's book, Leveraging Distortions: Explanation, Idealization, and Universality in Science.

Scientific research is constrained by limited resources, so it is imperative that it be conducted efficiently. This paper introduces the notion of epistemic expression, a kind of representation that expedites the solution of research problems. Epistemic expressions are representations that (i) contain information in a way that enables more reliable information to place the most stringent constraints on possible solutions and (ii) make new information readily extractible by biasing the search through that space. I illustrate these conditions using historical and (...) contemporary examples of biomolecular structure determination. Then, I argue that the notion of epistemic expression parts ways with pragmatic accounts of scientific representation and an understanding of models as artifacts, neither of which require models to accurately represent. Explicating epistemic expression thus fills a gap in our understanding of scientific practice, extending Morrison and Morgan's (1999) conception of models as investigative instruments. (shrink)

Models are indispensable tools of scientific inquiry, and one of their main uses is to improve our understanding of the phenomena they represent. How do models accomplish this? And what does this tell us about the nature of understanding? While much recent work has aimed at answering these questions, philosophers' focus has been squarely on models in empirical science. I aim to show that pure mathematics also deserves a seat at the table. I begin by presenting two cases: Cramér’s random (...) model of the prime numbers and the function field model of the integers. These cases show that mathematicians, like empirical scientists, rely on unrealistic models to gain understanding of complex phenomena. They also have important implications for some much-discussed theses about scientific understanding. First, modeling practices in mathematics confirm that one can gain understanding without obtaining an explanation. Second, these cases undermine the popular thesis that unrealistic models confer understanding by imparting counterfactual knowledge. (shrink)

It is often claimed that one can avoid the kind of underdetermination that is a typical consequence of symmetries in physics by stipulating that symmetry-related models represent the same state of affairs (Leibniz Equivalence). But recent commentators (Dasgupta 2011; Pooley 2021; Pooley and Read 2021; Teitel 2021a) have responded that claims about the representational capacities of models are irrelevant to the issue of underdetermination, which concerns possible worlds themselves. In this paper I distinguish two versions of this objection: (1) that (...) a theory’s formalism does not (fully) determine the space of physical possibilities, and (2) that the relevant notion of possibility is not physical possibility. I offer a refutation of each. (shrink)

Climate science is expected to provide usable information to policy-makers, to support the resolution of climate change. The complex, multiply connected nature of climate change as a social problem is reviewed and contrasted with current modular and discipline-bounded approaches in climate science. We argue that climate science retains much of its initial “physics-first” orientation, and that it adheres to a problematic notion of objectivity as freedom from value judgments. Together, these undermine its ability to provide usable information. We develop the (...) notion of usability using work from the literature on adaptation, but our argument applies to all of climate science. We illustrate the tension between usability and the objective, physics-first orientation of climate science with an example about model development practices in climate science. For solutions, we draw on two frameworks for science which responds to societal challenges: post-normal science and mandated science. We generate five recommendations for adapting the practice of climate science, to produce more usable information and thereby respond more directly to the social challenge of climate change. These are: (1) integrated cross-disciplinarity, (2) wider involvement of stakeholders throughout the lifecycle of a climate study, (3) a new framing of the role of values in climate science, (4) new approaches to uncertainty management, and (5) new approaches to uncertainty communication. (shrink)

I examine error diagnosis (model-model disagreement) in climate model intercomparisons including its difficulties, fruitful examples, and prospects for streamlining error diagnosis. I suggest that features of climate model intercomparisons pose a more significant challenge for error diagnosis than do features of individual model construction and complexity. Such features of intercomparisons include, e.g., the number of models involved, how models from different institutions interrelate, and what scientists know about each model. By considering numerous examples in the climate modeling literature, I distill (...) general strategies (e.g., employing physical reasoning and using dimension reduction techniques) used to diagnose model error. Based on these examples, I argue that an error repertoire could be beneficial for improving error diagnosis in climate modeling, although constructing one faces several difficulties. Finally, I suggest that the practice of error diagnosis demonstrates that scientists have a tacit-yet-working understanding of their models which has been under-appreciated by some philosophers. (shrink)

Model transfer is the scientific practice of taking a model which was initially applied in one particular kind of target system in some particular scientific domain and applying it to represent a novel target system in a novel scientific domain. This paper motivates a realist interpretation of empirically successful model transfers and the implications of such an interpretation for the metaphysics of science. The paper uses two examples of empirically successful model transfer, the first of which is a strikingly successful (...) recent application of the Ising model to neurobiology, the second of which is Maxwell’s “method of physical analogy”. The paper first defends the need for a realist picture regarding the transfer of model templates and the apparent discovery of physical analogies. Then, it examines further the implications of such a picture, finding it to be quite revisionary in its conception of the nature of causal processes. (shrink)

What has fiction to do with science? At first glance, the two activities seem to have entirely different aims and products. Science aims at truth, while fiction can deviate wildly from it. Science produces theories, which we are asked to believe. Fiction produces stories, which we are asked to imagine. Given these differences, associating science and fiction might seem like a serious mistake, or even a threat to science. And yet many authors have tried to understand science by looking to (...) fiction. This article will consider three strands of thought on the topic: Hans Vaihinger’s classic work on fictions; fictionalist views, especially van Fraassen’s constructive empiricism; and recent work on models and fiction. As we will see, none of these strands pose a threat to science. Instead, each tries to understand the nature of scientific knowledge by drawing on fiction in different ways, and for different theoretical purposes. (shrink)

Biologists often study particular biological systems as models of a phenomenon of interest even if they already know that the phenomenon is produced by diverse mechanisms and hence none of those systems alone can sufficiently represent it. To understand this modeling practice, the present paper provides an account of how multiple model systems can be used to study a phenomenon that is produced by diverse mechanisms. Even if generalizability of results from a single model system is significantly limited, generalizations concerning (...) specific aspects of mechanisms often hold across certain ranges of biological systems, which enables multiple model systems to jointly represent such a phenomenon. Comparing mechanisms that operate in different biological systems as examples of the same phenomenon also facilitates characterization and investigation of individual mechanisms. I also compare my account with two existing accounts of the use of multiple model systems and argue that my account is distinct from and complementary to them. (shrink)

Previously, I (Boesch 2017) described a notion called “representational licensing”—the set of activities of scientific practice by which scientists establish the intended representational use of a vehicle. In this essay, I expand and develop this concept of representational licensing. I begin by showing how the concept is of value for both pragmatic and substantive approaches to scientific representation. Then, through the examination of a case study of the Mississippi River Basin Model, I point out and explain some of the activities (...) of representational licensing that help to establish the representational nature of this model. Throughout the exploration of the case study, I pause to identify some important lessons which apply more generally about the nature of representational licensing in science. (shrink)