Modelling and representing: An artefactual approach to model-based representation

The recent discussion on scientific representation has focused on models and their relationship to the real world. It has been assumed that models give us knowledge because they represent their supposed real target systems. However, here agreement among philosophers of science has tended to end as they have presented widely different views on how representation should be understood. I will argue that the traditional representational approach is too limiting as regards the epistemic value of modelling given the focus on the relationship between a single model and its supposed target system, and the neglect of the actual representational means with which scientists construct models. I therefore suggest an alternative account of models as epistemic tools. This amounts to regarding them as concrete artefacts that are built by specific representational means and are constrained by their design in such a way that they facilitate the study of certain scientific questions, and learning from them by means of construction and manipulation.     

Model-as-replica,model-as-instrument: Representational power and contextual versatility in animal models

Existing scholarship on animal models tends to foreground either of the two major roles research organisms play in different epistemic contexts, treating their representational and instrumental roles separately. Based on an empirical case study, this article explores the changing relationship between the two epistemic roles of a research organism over the span of a decade, while the organism was used to achieve various knowledge ends. This rat model was originally intended as a replica of human susceptibility to cardiac arrest. In a fortunate stroke of serendipity, however, the experimenters detected the way mother-infant interactions regulated the pups’ resting cardiac rate. This intriguing outcome thus became the model’s new representational target and began driving the development of an experimental system. Henceforth, the model acquired an instrumental function, serving to detect and measure system-specific differences. Its subsequent development involved creating stimulus-response measures to explain and theorize those differences. It was this instrumental use of the model that pushed the experimenters into unchartered territory and conferred to the model an ability to adapt to varied epistemic contexts. Despite the prominence of this instrumental role, however, the model’s representational power continued to guide research. The model’s representational target was widened beyond heart rate to reflect other functional phenomena, such as behavioral activity and sleep/wake rhythm. The rat model was thus transformed from an experimental organism designed to instantiate cardiac regulation to a model organism taken to represent the development of a whole, intact animal under the regulatory influence of maternal care. This article examines this multifaceted transformation within the context of the salient shifts in modeling practice and variations in the model’s representational power. It thus explores how the relationship between the representational and instrumental uses of the model changed with respect to the varying exigencies of the investigative context, foregrounding its contextual versatility.     

Models as interpreters (with a case study from pain science)

Most philosophical accounts of scientific models assume that models represent some aspect, or some theory, of reality. They also assume that interpretation plays only a supporting role. This paper challenges both assumptions. It proposes that models can be used in science to interpret reality. (a) I distinguish these interpretative models from representational ones. They find new meanings in a target system’s behaviour, rather than fit its parts together. They are built through idealisation, abstraction and recontextualisation. (b) To show how interpretative models work, I offer a case study on the scientific controversy over foetal pain. It highlights how pain scientists use conflicting models to interpret the human foetus and its behaviour, and thereby to support opposing claims about whether the foetus can feel pain. (c) I raise a sceptical worry and a methodological challenge for interpretative models. To address the latter, I use my case study to compare how interpretative and representational models ought to be evaluated.     

Understanding climate phenomena with data-driven models

In climate science, climate models are one of the main tools for understanding phenomena. Here, we develop a framework to assess the fitness of a climate model for providing understanding. The framework is based on three dimensions: representational accuracy, representational depth, and graspability. We show that this framework does justice to the intuition that classical process-based climate models give understanding of phenomena. While simple climate models are characterized by a larger graspability, state-of-the-art models have a higher representational accuracy and representational depth. We then compare the fitness-for-providing understanding of process-based to data-driven models that are built with machine learning. We show that at first glance, data-driven models seem either unnecessary or inadequate for understanding. However, a case study from atmospheric research demonstrates that this is a false dilemma. Data-driven models can be useful tools for understanding, specifically for phenomena for which scientists can argue from the coherence of the models with background knowledge to their representational accuracy and for which the model complexity can be reduced such that they are graspable to a satisfactory extent.     

What was classical genetics?

I present an account of classical genetics to challenge theory-biased approaches in the philosophy of science. Philosophers typically assume that scientific knowledge is ultimately structured by explanatory reasoning and that research programs in well-established sciences are organized around efforts to fill out a central theory and extend its explanatory range. In the case of classical genetics, philosophers assume that the knowledge was structured by T. H. Morgan’s theory of transmission and that research throughout the later 1920s, 30s, and 40s was organized around efforts to further validate, develop, and extend this theory. I show that classical genetics was structured by an integration of explanatory reasoning (associated with the transmission theory) and investigative strategies (such as the ‘genetic approach’). The investigative strategies, which have been overlooked in historical and philosophical accounts, were as important as the so-called laws of Mendelian genetics. By the later 1920s, geneticists of the Morgan school were no longer organizing research around the goal of explaining inheritance patterns; rather, they were using genetics to investigate a range of biological phenomena that extended well beyond the explanatory domain of transmission theories. Theory-biased approaches in history and philosophy of science fail to reveal the overall structure of scientific knowledge and obscure the way it functions.     

Inconsistent idealizations and inferentialism about scientific representation

Inferentialists about scientific representation hold that an apparatus's representing a target system consists in the apparatus allowing “surrogative inferences” about the target. I argue that a serious problem for inferentialism arises from the fact that many scientific theories and models contain internal inconsistencies. Inferentialism, left unamended, implies that inconsistent scientific models have unlimited representational power, since an inconsistency permits any conclusion to be inferred. I consider a number of ways that inferentialists can respond to this challenge before suggesting my own solution. I develop an analogy to exploitable glitches in a game. Even though inconsistent representational apparatuses may in some sense allow for contradictions to be generated within them, doing so violates the intended function of the apparatus's parts and hence violates representational “gameplay”.     

The realizers and vehicles of mental representation

The neural vehicles of mental representation play an explanatory role in cognitive psychology that their realizers do not. Cognitive psychology individuates neural structures as representational vehicles in terms of the specific causal properties to which cognitive mechanisms are sensitive. Explanations that appeal to properties of vehicles can capture generalisations which are not available at the level of their neural realizers. In this paper, I argue that the individuation of realizers as vehicles restricts the sorts of explanations in which they can participate. I illustrate this with reference to Rupert’s (2011) claim that representational vehicles can play an explanatory role in psychology in virtue of their quantity or proportion. I propose that such quantity-based explanatory claims can apply only to realizers and not to vehicles, in virtue of the particular causal role that vehicles play in psychological explanations.     

An appraisal of Mendeleev’s contribution to the development of the periodic table

Historians and philosophers of science generally conceptualize scientific progress to be dichotomous, viz., experimental observations lead to scientific laws, which later facilitate the elaboration of explanatory theories. There is considerable controversy in the literature with respect to Mendeleev’s contribution to the origin, nature, and development of the periodic table. The objectives of this study are to explore and reconstruct: a) periodicity in the periodic table as a function of atomic theory; b) role of predictions in scientific theories and its implications for the periodic table; and c) Mendeleev’s contribution: theory or an empirical law? The reconstruction shows that despite Mendeleev’s own ambivalence, periodicity of properties of chemical elements in the periodic table can be attributed to the atomic theory. It is argued that based on the Lakatosian framework, predictions (novel facts) play an important role in the development of scientific theories. In this context, Mendeleev’s predictions played a crucial role in the development of the periodic table. Finally, it is concluded that Mendeleev’s contribution can be considered as an “interpretative” theory which became “explanatory” after the periodic table was based on atomic numbers.     

Computational models: a modest role for content

The computational theory of mind construes the mind as an information-processor and cognitive capacities as essentially representational capacities. Proponents of the view (hereafter, ‘computationalists’) claim a central role for representational content in computational models of these capacities. In this paper I argue that the standard view of the role of representational content in computational models is mistaken; I argue that representational content is to be understood as a gloss on the computational characterization of a cognitive process.     

Perspectivism, inconsistent models, and contrastive explanation

It is widely recognized that scientific theories are often associated with strictly inconsistent models, but there is little agreement concerning the epistemic consequences. Some argue that model inconsistency supports a strong perspectivism, according to which claims serving as interpretations of models are inevitably and irreducibly perspectival. Others argue that in at least some cases, inconsistent models can be unified as approximations to a theory with which they are associated, thus undermining this kind of perspectivism. I examine the arguments for perspectivism, and contend that its strong form is defeasible in principle, not merely in special cases. The argument rests on the plausibility of scientific knowledge concerning non-perspectival, dispositional facts about modelled systems. This forms the basis of a novel suggestion regarding how to understand the knowledge these models afford, in terms of a contrastive theory of what-questions.     

Modelling and knowledge transfer in complexity science

I analyse the construction and transfer of models in complexity science. Thereby, I introduce a distinction between (i) vertical model construction, which is based on knowledge about a specific target system, (ii) horizontal model construction, which is based on the alteration of an existing model and therefore does not require any references to a specific target system; and (iii) the transfer of models, which consists of the assignment of an existing model to a new target system. I argue that, in complexity science, all three of those modelling activities take place. Furthermore, I show that these activities can be divided into two general categories: (i) the creation of a repository of models without specific target systems, which have been created by large-scale horizontal construction; and (ii) the transfer of these models to particular target systems in the natural sciences, which can also be followed by an extension of the transferred model through vertical construction of adaptions and additions to its dynamics. I then argue that this interplay of different modelling activities in complexity science provides a mechanism for the transfer of knowledge between different scientific fields. It is also crucial to the interdisciplinary nature of complexity science.     

Transfer and templates in scientific modelling

The notion of template has been advocated by Paul Humphreys and others as an illuminating unit of analysis in the philosophy of scientific modelling. Templates are supposed to have the dual functions of representing target systems and of facilitating quantitative manipulation. A resulting worry is that wide-ranging cross-disciplinary use of templates might compromise their representational function and reduce them to mere formalisms. In this paper, we argue that templates are valuable units of analysis in reconstructing cross-disciplinary modelling. Central to our discussion are the ways in which Lotka-Volterra models are used to analyse processes of technology diffusion. We illuminate both the similarities and differences between contributions to this case of cross-disciplinary modelling by reconstructing them as transfer of a template, without reducing the template to a mere formalism or a computational model. This requires differentiating the interpretation of templates from that of the models based on them. This differentiation allows us to claim that the LV models of technology diffusion that we review are the result of template transfer - conformist in some contributions, creative in others.     

The primitive ontology of quantum physics: Guidelines for an assessment of the proposals

The paper seeks to make progress from stating primitive ontology theories of quantum physics—notably Bohmian mechanics, the GRW matter density theory and the GRW flash theory—to assessing these theories. Four criteria are set out: (a) internal coherence; (b) empirical adequacy; (c) relationship to other theories; and (d) explanatory value. The paper argues that the stock objections against these theories do not withstand scrutiny. Its focus then is on their explanatory value: they pursue different strategies to ground the textbook formalism of quantum mechanics, and they develop different explanations of quantum non-locality. In conclusion, it is argued that Bohmian mechanics offers a better prospect for making quantum non-locality intelligible than the GRW matter density theory and the GRW flash theory.     

Deflationary representation,inference, and practice

This paper defends the deflationary character of two recent views regarding scientific representation, namely RIG Hughes' DDI model and the inferential conception. It is first argued that these views' deflationism is akin to the homonymous position in discussions regarding the nature of truth. There, we are invited to consider the platitudes that the predicate “true” obeys at the level of practice, disregarding any deeper, or more substantive, account of its nature. More generally, for any concept X, a deflationary approach is then defined in opposition to a substantive approach, where a substantive approach to X is an analysis of X in terms of some property P, or relation R, accounting for and explaining the standard use of X. It then becomes possible to characterize a deflationary view of scientific representation in three distinct senses, namely: a “no-theory” view, a “minimalist” view, and a “use-based” view—in line with three standard deflationary responses in the philosophical literature on truth. It is then argued that both the DDI model and the inferential conception may be suitably understood in any of these three different senses. The application of these deflationary ‘hermeneutics’ moreover yields significant improvements on the DDI model, which bring it closer to the inferential conception. It is finally argued that what these approaches have in common—the key to any deflationary account of scientific representation—is the denial that scientific representation may be ultimately reduced to any substantive explanatory property of sources, or targets, or their relations.     

The explanatory role of abstraction processes in models: The case of aggregations

Though it is held that some models in science have explanatory value, there is no conclusive agreement on what provides them with this value. One common view is that models have explanatory value vis-à-vis some target systems because they are developed using an abstraction process (i.e., a process which involves omitting features). Though I think this is correct, I believe it is not the whole picture. In this paper, I argue that, in addition to the well-known process of abstraction understood as an omission of features or information, there is also a family of abstraction processes that involve aggregation of features or information and that these processes play an important role in endowing the models they are used to build with explanatory value. After offering a taxonomy of abstraction processes involving aggregation, I show by considering in detail several models drawn from different sciences that the abstraction processes involving aggregation that are used to build these models are responsible (at least partially) for their having explanatory value.     

Sydney Chapman on the Layering of the Atmosphere: Conceptual Unity and the Modelling of the Ionosphere

Sydney Chapman is unanimously considered to have played a founding role in modern geomagnetism and to have opened up new lines of research in geophysics generally. Nevertheless, Chapman's conviction regarding the synthesis of the explanatory mechanisms of the atmosphere has gone practically unnoticed in the historiography of geophysics. This paper examines Chapman's contribution to ionospheric physics. It aims to understand Chapman's theory of ionospheric layer formation, and particularly its link to his theory of ozone formation. It deals first with the traits which characterized Chapman's personality, as a way of explaining—and even perhaps justifying— his quest for the integration and synthesis of geophysical knowledge. It then analyses Chapman's model of ionospheric layers and his suggestions regarding its use as an operational tool (without ontological connotations), before continuing with his account of the formation of the ozone layer, which seemed to constitute the missing link for understanding ionosphere layer formation. The paper concludes with Chapman's attempt to reconcile geomagnetic and radio evidence.     

Explanation,understanding, and unrealistic models

How can false models be explanatory? And how can they help us to understand the way the world works? Sometimes scientists have little hope of building models that approximate the world they observe. Even in such cases, I argue, the models they build can have explanatory import. The basic idea is that scientists provide causal explanations of why the regularity entailed by an abstract and idealized model fails to obtain. They do so by relaxing some of its unrealistic assumptions. This method of ‘explanation by relaxation’ captures the explanatory import of some important models in economics. I contrast this method with the accounts that Daniel Hausman and Nancy Cartwright have provided of explanation in economics. Their accounts are unsatisfactory because they require that the economic model regularities obtain, which is rarely the case. I go on to argue that counterfactual regularities play a central role in achieving ‘understanding by relaxation.’ This has a surprising implication for the relation between explanation and understanding: Achieving scientific understanding does not require the ability to explain observed regularities.     

Periodically integrated subset autoregressions for dutch industrial production and money stock

The univariate quarterly Dutch series of industrial production and money stock are both modelled with a periodically integrated subset autoregression (PISA). This model for a non-stationary series allows the lag orders, the values of the parameters and the cyclical patterns to vary over the seasons. The PISA models are found by applying a general-to-simple specification strategy, which deals with non-stationarity and periodicity simultaneously. It is found that the two series show a common asymmetric cyclical behaviour. This paper further proposes a test for periodicity in the errors, with which it is argued that a non-periodic model for the industrial production and money stock is misspecified and that seasonal adjustment does not remove periodicity in the autocorrelation function.