Abstract
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.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
In the next section, I provide a further taxonomy between kinds of model and template transfer.
A quick aside: a growing number of philosophers of science endorse the view that there are a plurality of epistemic goods in scientific modeling besides accurate representation or even successful surrogative function: prediction, validation of other models, exploratory function, checking a dataset, etc. I too endorse this view! (This view goes under a few names, with subtle differences between each articulation, notably the “artefactual” view of Knuuttila (2011) and the “adequacy-for-purpose” view of Alexandrova (2010) and Parker (2020).) My usage of phrases like “representational work” or “representation of” in this paper are not intended to imply any further commitments about the nature of scientific modeling or what counts as successful scientific modeling. Nor do any of my remarks in this paper commit me to a view about the nature of the scientific representation relation or the ontology of scientific models. Thanks to an anonymous referee for pressing me to clarify this point.
There also exist one-dimensional and multi-dimensional (> 2) Ising models, but our discussion need not involve these in any detail.
A referee asks about the possibility that by using an in vitro sample, early studies (such as Schneidman et al., 2006) could have removed higher-order correlations between neurons (i.e., > 2 neurons). However, the reproducibility of these early results in in vivo conditions suggests that no such knock-on effect took place. More importantly, quite a few of these studies end up concluding that higher-order correlations between neurons do not contribute significantly to observed neuronal activity (Tang et al., 2008, p516, Yu et al., 2008, p2894, Ohiorhenuan et al., 2010, p619).
A referee flags another possible worry about the sample size of some of the earlier studies: 10 neurons is not a lot. However, later experiments showed that the results were robust across a number of larger sample sizes, as well as being reproducible in a number of conditions (early results were in vitro, but later results were in vivo; early results focused only on retinal tissue, but later results reproduced the model’s success in cortical tissue as well).
Some readers, especially philosophers of cognitive science or neuroscience, might be aware that there are previous attempts to model certain cognitive processes on Ising models in artificial neural networks. I think that these models of “Hopfield networks” (sometimes also called ‘spin glass networks’) are importantly distinct from the kinds of applications of the Ising model discussed here, primarily because they seem to be aimed at providing a different kind of scientific explanation. More on this in Sect. 4.3.1, below.
There are some differences between Maxwell’s and von Helmholtz’s use of the method of physical analogy. I have chosen here to focus on Maxwell’s contribution, because of its convergences with observations and because it is more influential in the history of electromagnetic theory. Alisa Bokulich (2015) provides a detailed discussion of the differences between Maxwell’s and von Helmholtz’s contributions.
Cf. Nersessian (2002a, p. 147).
This is what drives the well-known divide et impera strategies for defending selective realism from the pessimistic induction (Psillos 1999). To be sure, these strategies have faced much-discussed objections in the context of how well they answer historical challenges like the pessimistic induction and P. Kyle Stanford’s “new induction.” But the strategy at least prima facie applies to the current cases of model transfer, and these well-known historical objections would not be well-placed in the current situation because the current puzzling cases do not concern cases of theory change in the sciences.
While Salmon’s endorsement of the ontic conception is often mischaracterized as being a view according to which all explanations are causal, clearly he does not think so.7 In another illustrative passage, he writes: “Explanations tell us how the world works. [Mechanistic] explanations are local in the sense that they show us how particular occurrences come about; they explain particular phenomena in terms of collections of particular causal processes and interactions—or, perhaps, in terms of non-causal mechanisms, if there are such things” (1990, p184).
Woodward does not himself endorse this particular version of the ontic conception, although he does endorse in broader strokes the ontic conception writ large: “The theory I am proposing is thus what Salmon (1984) calls a realist or “ontic” theory of explanation, rather than an epistemic or logic‐based model, in the sense that it insists that one cannot read off just from the logical or deductive relations between a putative explanans and explanandum whether the former genuinely explains the latter” (2003, p202).
There has been some controversy lately on whether the ontic conception is, as I have stated it, the view that explanations must represent the ontological dependence-structure of the explanandum (see, e.g., Woodward (2003), Povich (2018), Craver and Povich 2017), or as the view that explanations are “full-bodied things in the world” (see for instance Illari 2013 and Wright 2015 for criticism of this view). I accept the first formulation and also broadly sympathize with criticisms of the second formulation, but these details are unimportant here.
The caveat here is, of course, that sometimes this requirement on accuracy must be reconciled with the presence of idealization in a model. There is a voluminous literature on reconciling the presence of idealizations in scientific models with models’ abilities to provide factive explanations. I have nothing new to say about this here.
Thanks to an anonymous referee for raising this point.
This is important to bring up because the term ‘abstraction,’ in discussions of scientific modeling, often has a precise usage as a kind of modeling procedure that is distinct from idealization but is also distinct from this kind of “move-towards-generality”. The familiar slogan often attached to this other usage of ‘abstraction’ is that idealization is distortion while abstraction is omission. In the present passage and throughout the rest of the paper, I intend by my usage of ‘abstraction’ only to capture the cognitive process where existing properties are considered at a more generic or general level, rather than to capture this more precise usage where abstraction concerns the construction of models that omit certain features of their target systems. (Of course, the nature of the relationship between abstraction-as-a-move-towards-generality and abstraction-as-omission is an open question to some extent. See Pincock (2014) and Saatsi and Jansson (2017) for some discussion of abstraction-as-generality in contexts of scientific explanation. See also Levy (2018) for some recent discussion of the “abstraction is omission” view in model construction.) Thanks to an anonymous referee for pressing me to clarify this point.
There are still more examples of this sort of view. Consider Philip Dowe’s conserved-quantity characterization of a causal process: a causal process is a “world line” of some concrete object—a four-dimensional history of the object— which manifests some conserved quantity (1992, p210). Just like on Salmon’s view, Dowe’s “world line” characterization ties the realization of a process to some particular object (or, we can easily suppose, a type of object). Douglas Ehring’s (1998) view of processes likewise holds that processes are those causal interactions in which the same property instance persists over time, and at the type-level, causal processes are individuated by the particular property-types they pass along.
References
Alexandrova, A. (2010). Adequacy for purpose: The best deal a model can get. Modern Schoolman, 87(3/4), 295–301.
Andersen, H. (2017). Patterns, information, and causation. Journal of Philosophy, 114(11), 592–622.
Arkani-Hamed, N., Hall, L. J., Kolda, C., & Murayama, H. (2000). New perspective on cosmic coincidence problems. Physical Review Letters, 85, 4434–4437.
Beggs, J. M., & Timme, N. (2012). Being critical of criticality in the brain. Frontiers in Physiology, 3, 163.
Bhogal, H. (2020). Coincidences and the grain of explanation. Philosophy and Phenomenological Research, 100(3), 677–694.
Bokulich, A. (2014). How the tiger bush got its stripes: ‘How Possibly’ vs. ‘How Actually’ model explanations. The Monist, 97(3), 321–338.
Bokulich, A. (2015). Maxwell, Helmholtz, and the unreasonable effectiveness of the method of physical analogy. Studies in History and Philosophy of Science Part A, 50, 28–37.
Cartwright, N. (1983). How the laws of physics lie. Oxford University Press.
Castellane, A., & Paternotte, C. (2018). Knowledge transfer without knowledge? The case of agentive metaphors in biology. Studies in History and Philosophy of Science Part A, 72, 49–58.
Chakrabarti, B. K., Chakraborti, A., Chakravarty, S. R., & Chatterjee, A. (2013). Econophysics of income and wealth distributions. Cambridge University Press.
Chakravartty, A. (2007). A Metaphysics for Scientific Realism. Cambridge University Press. Craver, Carl F. & Povich, Mark (2017). The directionality of distinctively mathematical explanations. Studies in History and Philosophy of Science Part A 63:31–38.
Chakravartty, A. (1998). Semirealism. Studies in History and Philosophy of Science Part A, 29(3), 391–408.
de Regt, H. (2017). Understanding scientific understanding. Oxford University Press.
Donhauser, J., & Shaw, J. (2019). Knowledge transfer in theoretical ecology: Implications for incommensurability, voluntarism, and pluralism. Studies in History and Philosophy of Science Part A, 1(77), 11–20.
Dowe, P. (1992). Wesley salmon’s process theory of causality and the conserved quantity theory. Philosophy of Science, 59(2), 195–216.
Ehring, D. (1997). Causation and persistence: A theory of causation. Oxford University Press.
Gallegati, M., Keen, S., Lux, T., & Ormerod, P. (2006). Worrying trends in econo-physics. Physica A: Statistical Mechanics and Its Applications, 370(1), 1–6.
Helmholtz, H. v. (1858). On integrals of the hydrodynamic equations which express vortex-motions. Crelle's Journal für die Reine und Angewandte Mathematik, 55, 25–55. English Translation by P. Tait (1867) in Philosophical Magazine, Series 4, 33(226):485–512.
Herfeld, C., & Doehne, M. (2019). The diffusion of scientific innovations: A role typology. Studies in History and Philosophy of Science Part A, 77, 64–80.
Herfeld, C., & Lisciandra, C. (2019). Knowledge transfer and its contexts. Studies in History and Philosophy of Science Part A, 77, 1–10.
Hopfield, J. J. (1982). Neural networks and physical systems with emergent collective computational abilities. Proceedings of the National Academy of Sciences of the United States of America, 79, 2554–2558.
Humphreys, P. (2002). Computational models. Philosophy of Science, 69(S3), S1–S11.
Humphreys, P. (2004). Extending ourselves: Computational science, empiricism, and scientific method. Oxford University Press.
Humphreys, P. (2019). Knowledge transfer across scientific disciplines. Studies in History and Philosophy of Science Part A, 77, 112–9.
Illari, P. (2013). Mechanistic explanation: Integrating the ontic and epistemic. Erkenntnis, 78(2), 237–255.
Jansson, L., & Saatsi, J. (2019). Explanatory abstractions. British Journal for the Philosophy of Science, 70(3), 817–844.
Jhun, J., Palacios, P., & Weatherall, J. O. (2018). Market crashes as critical phenomena? Explanation, idealization, and universality in econophysics. Synthese, 195(10), 4477–4505.
Kitcher, P. (1993). The advancement of science: Science without legend. Oxford University Press.
Kitcher, P. (2001). Science, truth, and democracy. Oxford University Press.
Knuuttila, T. (2011). Modelling and representing: An artefactual approach to model-based representation. Studies in History and Philosophy of Science Part A, 42(2), 262–271.
Knuuttila, T., & Loettgers, A. (2014). Magnets, spins, and neurons: The dissemination of model templates across disciplines. The Monist, 97(3), 280–300.
Knuuttila, T., & Loettgers, A. (2016). Model templates within and between disciplines: From magnets to gases—And socio-economic systems. European Journal for Philosophy of Science, 6(3), 377–400.
Knuuttila, T., & Loettgers, A. (2017). Modelling as indirect representation? The Lotka–Volterra model revisited. British Journal for the Philosophy of Science, 68(4), 1007–1036.
Koskinen, R. (2017). Synthetic biology and the search for alternative genetic systems: Taking how-possibly models seriously. European Journal for Philosophy of Science, 7(3), 493–506.
Levy, A. (2018). Idealization and abstraction: Refining the distinction. Synthese, 198(Suppl 24), 5855–5872.
Lin, C.-H. (2022). Knowledge transfer, templates, and the spillovers. European Journal for Philosophy of Science, 12(1), 1–30.
Loew, C. (2019). Causes as difference-makers for processes. Philosophy and Phenomenological Research, 98(1), 89–106.
Maxwell, J. C. (1861/62). On physical lines of force.
Morrison, M. (1997). Physical models and biological contexts. Philosophy of Science, 64(4), 324.
Nersessian, N. J. (2002). Maxwell and the method of physical analogy: Model-based reasoning, generic abstraction, and conceptual change. In D. Malament (Ed.), Reading natural philosophy: Essays in the history and philosophy of science and mathematics. Open Court.
Nersessian, N. J. (2002). Abstraction via generic modeling in concept formation in science. Mind & Society, 3, 129–54.
Ohiorhenuan, I. E., Mechler, F., Purpura, K. P., et al. (2010). Sparse coding and high-order correlations in fine-scale cortical networks. Nature, 466, 617–621.
Parker, W. S. (2020). Model evaluation: An adequacy-for-purpose view. Philosophy of Science, 87(3), 457–477.
Pincock, C. (2014). Abstract explanations in science. British Journal for the Philosophy of Science, 66(4), 857–882.
Povich, M. (2018). Minimal models and the generalized ontic conception of scientific explanation. British Journal for the Philosophy of Science, 69(1), 117–137.
Price, J. (2018). The landing zone—Ground for model transfer in chemistry. Studies in History and Philosophy of Science Part A, 77, 21–28.
Psillos, S. (1999). Scientific realism: How science tracks truth. Routledge.
Reutlinger, A., Hangleiter, D., & Hartmann, S. (2018). Understanding (with) toy models. British Journal for the Philosophy of Science, 69(4), 1069–1099.
Salmon, W. (1984). Scientific explanation and the causal structure of the world. Princeton University Press.
Salmon, W. (1990). Four decades of scientific explanation. Pittsburgh University Press.
Schneidman, E., Berry, M. J., Segev, R., & Bialek, W. (2006). Weak pairwise correlations imply strongly correlated network states in a neural population. Nature, 440(7087), 1007–1012.
Shlens, J., Field, G. D., Gauthier, J. L., et al. (2006). The structure of multi-neuron firing patterns in primate retina. Journal of Neuroscience, 26, 8254–8266.
Shlens, J., Field, G. D., Gauthier, J. L., et al. (2009). The structure of large-scale synchronized firing in primate retina. Journal of Neuroscience, 29, 5022–5031.
Sjölin Wirling, Y., & Grüne-Yanoff, T. (2021). The epistemology of modal modeling. Philosophy Compass, 16(10), e12775.
Steward, H. (2013). Processes, continuants, and individuals. In Mind 122 (487):fzt080
Strevens, M. (2007). Depth: An account of scientific explanation. Harvard University Press.
Strevens, M. (2012). Replies to Weatherson, Hall, and Lange. Philosophy and Phenomenological Research, 84(2), 492–505.
Tan, P. (2022). Two epistemological challenges regarding hypothetical modeling. Synthese, 200(6), 448.
Tang, A., Jackson, D., Hobbs, J., Chen, W., Smith, J. L., Patel, H., Prieto, A., Petrusca, D., Grivich, M. I., Sher, A., & Hottowy, P. (2008). A maximum entropy model applied to spatial and temporal correlations from cortical networks in vitro. Journal of Neuroscience., 28(2), 505–18.
Thébault, K., Bradley, S., & Reutlinger, A. (2016). Modelling inequality. British Journal for the Philosophy of Science. https://doi.org/10.1093/bjps/axw028
Verreault-Julien, P. (2019). How could models possibly provide how-possibly explanations? Studies in History and Philosophy of Science Part A, 73, 1–12.
Wright, C. (2015). The ontic conception of scientific explanation. Studies in History and Philosophy of Science Part A, 54, 20–30.
Yeh, F.-C., Tang, A., Hobbs, J., et al. (2010). Maximum entropy approaches to living neural networks. Entropy, 12, 89–106.
Yu, S., Huang, D., Singer, W., & Nikolić, D. (2008). A small world of neuronal synchrony. Oxford University Press (OUP).
Yu, S., Yang, H., Nakahara, H., et al. (2011). Higher-order interactions characterized in cortical activity. Journal of Neuroscience, 31, 17514–17526.
Zlatev, I., Wang, L., & Steinhardt, P. J. (1999). Feb). Quintessence, cosmic coincidence, and the cosmological constant. Physical Review Letters, 82, 896–899.
Zuchowski, L. (2017). Modelling and knowledge transfer in complexity science. Studies in History and Philosophy of Science Part A, 77, 120–129.
Acknowledgements
Earlier versions of this paper were presented at the University of Virginia, Texas Tech University, the College of the Holy Cross, Fordham University, Peking University, and Tsinghua University. I am grateful to participants at all these locations for their feedback. For additional discussion of the paper, I thank Harjit Bhogal, Lin Chia-Hua, Zee Perry, Nick Smyth, Karsten Stueber, and Joel Velasco. I am also very grateful to two anonymous referees at this journal. Most importantly, I am grateful to the late Paul Humphreys, whose work on model templates and influence in general are I hope evident not only in this paper but in whatever work of quality I produce in the philosophy of science.
Funding
This research received no funding.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
I have no conflict of interest to report.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Tan, P. Interdisciplinary model transfer and realism about physical analogy. Synthese 201, 65 (2023). https://doi.org/10.1007/s11229-023-04065-x
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s11229-023-04065-x