Abstracts of talks to be presented

Chuan Fei Chin (University of Oxford & NUS)

Pictures and Parallels: How Ethics and Epistemology Mix in Pain Science

Can a foetus feel pain? (1) I will describe this controversy among American and British pain scientists. Their neural, behavioural and physiological arguments have led to an impasse, where both sides disagree over how to interpret the same data. (2) Despite this disagreement, both sides share the same picture of their problem. This picture is a pre-theoretical model about what the problem is and how it should be solved. Yet I shall argue that this picture is partly responsible for the scientific impasse. The problem of foetal pain cannot be solved as long as we accept the picture’s assumptions. (3) I propose an alternative way of picturing the problem. This picture makes better sense of the impasse and suggests how to avoid it. It also explains how medical and popular models of the foetus can play a role in solving the problem. I conclude by considering how such models, which act as theoretical parallels, are created and chosen – and what this says about the science of pain.

Jeremy Chong (National University of Singapore):

A Model of Modelling

There is no perfect correlation between a given physical situation and the equations that are used to represent them. Therefore, equations involve mathematical idealization. Christopher Pincock has suggested a two-stage process for model building that involves matching the physical situation exactly using a ‘matching model’, and then transforming the matching model into a form which we can more easily manipulate – an ‘equation model’, which resembles the models that physicists normally work with. I show how Pincock’s two-stage account, slightly altered, can clarify the roles of abstraction and idealization in the modelling process.

Catelijne Coopmans (National University of Singapore):

Computer Modeling as Organizational Enactment: Tools, Expertise and Boundaries in a Drug Discovery Context

This paper presents a perspective on computer modeling in organizational contexts which highlights the discursive work involved in making it viable and effective. Building on previous characterizations of computer modeling as the use of a 'tool' and a form of 'expertise', the paper attempts a sideways move to show how these two faces of modeling come alive in the context of drug discovery. The study shows that the drawing and redrawing of boundaries between tools and expertise is an important way in which the impact and effectiveness of this competency is ensured in practice. As the rise of expert competencies based on modeling, simulation and complex data mining is likely to generate similar dynamics in other settings, an agnostic view to the boundedness of technology provides an important addition to analytical repertoires for studying the role of models in the interaction between organizational communities.

Axel Gelfert (National University of Singapore):
Denotation,IRepresentation, and theIMinimalitylConstraint

The present paper investigates the way mathematical models represent reality. It takes as its starting point two recent accounts of mathematical models, proposed by Pincock (2005) and Hughes (1997), respectively. According to Pincock, representational success depends on the existence of an acceptable transformation between an adequate matching model (which, though it need not itself be mathematically tractable, purports to represent all the relevant details of the target phenomenon) and a corresponding derivation model, which lends itself to analytical or numerical evaluation. By contrast, Hughes regards the representational capacity of theoretical models as due to the interplay between three aspects – characterisable, in short, as denotation, demonstration, and interpretation – of the same mathematical structure. However, as I shall argue, both accounts fall short of providing an adequate answer to how mathematical models succeed in representing reality. This calls for a more comprehensive account that accommodates both theoretical challenges in one and the same framework. Such an account should also be able to explain why some mathematical models are more successful than others when it comes to their application across a wide range of different phenomena, rather than being tailored to just one class of phenomena. This is of special significance in the case of ‘minimal models’, which, essentially, provide ‘caricature[s] of the physics of the phenomenon in question’ (Batterman 2002). In the present paper, I argue that this minimality constraint, together with a proper appreciation of the role of notation and mathematical formalism, suggests a way how one might combine the intuitions instantiated by Pincock’s matching account and Hughes’s DDI account, respectively, thereby leading to a more coherent and comprehensive account of how mathematical models represent.

Mohd Hazim Shah (University of Malaya, Kuala Lumpur):
Models, Paradigms and Scientific Realism

The use of models to represent phenomena in science has been a subject of discussion in the history and philosophy of science, especially with the works of Mary Hesse which was considered to represent a conception of science which mediates between the Logical Positivists and Thomas Kuhn.  Recently, models as scientific representation has been discussed by philosophers of science such as Nersessian, Giere and Bas Van Fraassen. In this paper I will provide a brief overview of the various conceptions of models that have been presented, especially in relation to the question of models as representation in science. Apart from the cognitive and heuristic value of models, models also have a ‘symbolic’ or ‘iconic’ value which will be discussed in relation to the distinction between sense and referent, Kuhn’s notion of paradigm, and Holton’s concept of themata.  Models construed as ‘exemplars’ by Kuhn, and themata by Holton, not only have a cognitive and informative aspect, but also a psychological and symbolic aspect, as in the case of geocentric vs heliocentric models in astronomy, or field theory vs action-at-a-distance. This indicates that the way our mind organizes knowledge in the form of models, not only has to do with the structure of the world, but also with the structure of our mind broadly conceived, which includes memories, associations etc. Thus this paper will use ‘models’ as a focal point with which to tie various issues in the history and philosophy of science such as scientific realism, intelligibility of nature, cognition and symbolism.

Michael Heidelberger (University of Tübingen):

Applying Models: The Historical Case of Fluid Dynamics

The talk takes as its starting point the ‘applicational turn’ of modern fluid dynamics as it set in at the beginning of the 20th century with Ludwig Prandtl’s concept of the boundary layer. It seeks to show that there is much more to applying a theory in a highly mathematical discipline like fluid dynamics than deriving a special case from a general explanatory theory under particular antecedent conditions. In Prandtl’s case, the decisive move was to introduce a model that provided a physical/causal conception of viscous flow, which facilitated an approximate solution to the original equations and led to many applications. After an account of Prandtl’s achievement, the paper discusses the role of the physical model and its experimental and mathematical significance. It is shown that the mathematical simplification provided by the physical model greatly expanded the explanatory capacity of the theory which the Navier–Stokes equations alone could not provide.

Tarja Knuuttila (University of Helsinki):

How to Approach Model-Based Scientific Practice

One of the central themes in the recent discussion on models has certainly been representation. Modeling and representation have been brought together through the double move of assuming, on the one hand that models are representations, and, on the other hand that models give us knowledge because they represent their supposed real target systems. Consequently, the question has been: In virtue of what do models represent – given that they usually misrepresent (involving idealizations, approximations, fictions, simplifications etc.)? In doing so the basic unit of analysis from which models have been approached has been the model-target dyad. More often than not, it has been posited that there is a special relationship between the model and a target, and the philosophical task is to present an analysis of this relation. In my presentation I will review this discussion and argue that, rather paradoxically, the very interest on models and representation has shunned the actual means of representing. As against the representational view on models I argue that a broader view into modeling should be adopted if we are to better understand how models and simulations give us knowledge. Instead of fixing our philosophical gaze on the relationship of a single model and its putative target system, we should ask how models are constructed and what is involved in the specific scientific practice called modeling. I suggest that, taken together, the recent proposals concerning models as independent entities (Morrison and Morgan 1999, Knuuttila 2005, Weisberg 2006), the role of computational templates in computational science (Humphreys 2002, 2004) and the strategy of model-based science (Godfrey-Smith 2006) provide a fresh perspective on these questions. I will discuss how these proposals relate to each other and how they contribute to our understanding of modeling, representing and their epistemic functioning. 

Wang-Yen Lee (National University of Singapore):
The Metaphysics of Modality and the Coherence of Scientific Models

The use of two strictly logically incompatible models in science has been defended by arguing that those models are not to be taken as literally true in all aspects. In this paper I shall show that the prima facie incoherence (i.e. logical incompatibility) between two scientific models disappears if we take into account Kripke's notion of necessity and rigid designators. I then compare the respective merits of the non-literal approach and my approach proposed in this paper to deal with the issue of the prima facie incoherence of scientific models.

Demetris Portides (University of Cyprus):
Abstraction in Scientific Modelling: The Process of Representation

In this paper I explicate how the process of abstraction operates in the construction of scientific models. I also explore the converse process of de-idealization, or concretization, in an attempt to explicate how it operates in the construction of both theory-driven and phenomenological models. I contrast my analysis to the structuralist construal of de-idealization, and thus to the idea that models represent by isomorphisms or partial-isomorphisms, by virtue of the fact that they are mathematical structures. I also argue that the model of concretization I propose is compatible with the conception of scientific models as partially-autonomous entities that result from a complex amalgamation of theory together with conceptual ingredients deriving from auxiliaries and semi-empirical results, that cannot be clearly distinguished within the representational device, i.e. the model.

John Matthewson (Australian National University, Canberra)
Tradeoffs in Model Building: a More Target-Oriented Approach

In his 1966 paper The Strategy of Model Building in Population Biology, Richard Levins puts forward the claim that the models used in population biology are limited in significant ways. Most importantly, he asserts that no single model can be maximally realistic, precise and general at the same time. If true, this means that certain tradeoffs between these three desiderata are necessary when constructing models in population biology. For example, a theorist will have to sacrifice generality if they wish to construct a realistic and precise model of a population. Alternatively, someone modeling the very same population might forfeit high levels of precision in order to maximise realism and generality. Levins’ paper has generated various responses from philosophers of science. For example, Steven Orzack and Elliot Sober have criticised his claims on formal grounds, while Jay Odenbaugh has defended Levins from a more pragmatic perspective. Finally, Michael Weisberg has argued that if we modify Levins’ position somewhat, it can be shown that tradeoffs between modeling desiderata do in fact exist for purely formal reasons. Unfortunately, most of this discussion considers only the models themselves, paying little or no attention to the properties of the systems being modeled. I will discuss and expand on this literature, arguing that it is no coincidence that these insights regarding tradeoffs in model building came from a population biologist, rather than a scientist working in another discipline.