Download a .pdf version here.
In alphabetical order by first listed author.
A B C D E G H I J K L M P R S T V W
Claes Andersson – ‘Inverse Ontomimetic Simulation: A Window on Complex Systems’
Abstract: The present paper introduces "ontomimetic simulation" and argues that this class of models has enabled the investigation of hypotheses about complex systems in new ways that have epistemological relevance. Ontomimetic simulation can be differentiated from other types of modeling by its reliance on causal similarity in addition to representation. Phenomena are modeled not directly but via mimesis of the ontology (i.e. the "underlying physics", microlevel etc.) of systems and a subsequent animation of the resulting model ontology as a dynamical system. While the ontology is clearly used for computing system states, what is epistemologically important is that it is viewed as a hypothesis about the makeup of the studied system. This type of simulation, where model ontologies are used as hypotheses, is here called inverse ontomimetic simulation since it reverses the typical informational path from the target to the model system. It links experimental and analytical techniques in being explicitly dynamical while at the same time capable of abstraction. Inverse ontomimetic simulation is argued to have a great impact on science and to be the tool for hypothesis-testing that has made systematic theory development for complex systems possible.
Eckhart Arnold – ‘Tools or Toys On Specific Challenges for the Epistemology of Mathematical Models and Computer Simulations in the Field of Social Sciences’
Abstract: While mathematical models are a well established tool in most of the natural sciences and their epistemological status as a link between theory and reality is fairly well understood, there exists a considerable unclarity regarding their epistemological status in the of the social sciences. In my talk I argue that this results from specific challenges that mathematical models face in the social sciences. Because of these challenges an epistemology of models that is formed on the role model of physics may not be appropriate for the social sciences. I discuss these challenges and point out their epistemological consequences, the most important of which are that greater emphasis must be placed on empirical validation than on theoretical validation and that the relevance of purely theoretical simulations is strongly limited.
Anouk Barberousse and Cyrille Imbert – ‘From Statistical Physics to Computer Science and Return: How Models are Exported and Come Back’
Abstract: It has been shown in recent years that statistical physics is not only useful in physics, in biology, in economics, but also in computer science. Phase transitions have been identified in an important class of computationally complex problems, namely the class of non-deterministic polynomial time or NP-complete problems. In this paper, we propose an analysis of this surprising phenomenon in which paradigmatically abstract and formal properties are investigated by means of models coming from the study of physical systems. How should this situation be described? We propose four options and discuss them in turn, focusing on the following questions: What is exported, precisely, when statistical mechanics, as a theory, is used in the investigation of computationally complex problems? To what extent are results obtained in physics re-used in the study of computationally complex problems? Are they only heuristic guides?
Agnes Bolinska – ‘Scientific Representation and Contessa's Interpretational Account’
Abstract: Gabriele Contessa (2007) argues that a scientific model represents an aspect of a system in the world because it allows a user to perform surrogative reasoning—reasoning to conclusions about this target system from premises in the model. That is, a scientific model is an epistemic representation of its target, like maps, diagrams, pictures and other objects that permit their users to draw inferences about the targets they represent. Contessa develops an interpretational conception of epistemic representation, according to which a vehicle represents a certain target in virtue of its user adopting an interpretation of the vehicle in terms of this target. He argues that interpretation is both necessary and sufficient for epistemic representation and thus, in turn, for scientific representation. In this paper, I will argue that Contessa’s account does not adequately characterize scientific representation. Regarding interpretation as a sufficient condition for representation is out of line with scientific practice and reduces the concept of scientific representation to a more complex kind of denotation.
Michael Borgida, Michael Cuffaro and Ryan Muldoon – ‘Employing Computer Simulations in Developing Theories of Liberal Toleration’
Abstract: Our project employs computer simulations to investigate individual motivations for adopting tolerant political attitudes. We use computer simulations where political theorists have traditionally used thought experiments. Despite their contribution to the history of political philosophy, thought experiments suffer from several inherent weaknesses - weaknesses that are ameliorated through the use of computer simulations. Using computer simulations forces us to uncover otherwise hidden assumptions and allows us to explore complexities that fall beyond the boundaries of thought experiments. Computer simulations therefore, strengthen the foundation of our political theory and point towards a novel account of individual motivations for tolerant political attitudes.
Thomas Boyer – ‘Imaginary Time Path Integral: Is There a Novel Status for Statistical Simulations?’
Abstract: I propose a case-study of a generic tool used in contemporary physics: the imaginary time path integral (ITPI). It allows to compute indirectly, for example, the energy eigenvalues for a quantum system by means of measures made on the simulation of a dual physical system, suitably chosen so that a formal correspondence is possible with an imaginary time. The aim of this paper is to characterize precisely the novel use of statistical simulations allowed by a ITPI formulation. I argue for some interesting features on the conceptual ground; on the practical one, judging the efficiency of the tool requires several distinctions.
Otávio Bueno – ‘Computer Simulations: An Inferential Conception’
Abstract: In this paper, I offer an inferential conception of computer simulations, emphasizing the role that simulations play as inferential devices to represent empirical phenomena. Three steps are involved in a simulation: an immersion step (from aspects of the empirical set up to the simulated model), a derivation step (that yields the relevant results), and an interpretation and correction step (that interprets the results in light of the empirical set up). After developing the view, I present a case study on the simulation of the current flow between silicon atoms and buckyballs, and argue that the inferential conception accommodates the integration of empirical and theoretical data that is involved.
Julia Bursten – ‘The Role of Modern Valence Bond Models in Explaining Chemical Bonding’
Abstract: I explore the explanatory benefits of the modern valence bond models of chemical bonding and argue that philosophical analysis of different bonding models can inform both theoretical chemistry and contemporary theories of explanation. By highlighting particular explanatory virtues possessed by modern valence bond models in contrast with those possessed by molecular orbital models, an alternative and incompatible type of bonding model, I demonstrate how both types of models inform chemical research in differently useful ways. This demonstration leads to a discussion of the need for a plurality of models to explain bonding phenomena. During this discussion, I apply the conclusions of recent pluralist accounts of philosophical explanation, drawing heavily on Sandra Mitchell’s and Kyle Stanford’s recent work.
Gabriele Contessa – ‘Scientific Realism: Between Theories and Models’
Abstract: This paper focuses on some recent arguments that suggest that, on an indirect view of the relation between theories and world, traditional scientific realist arguments from the empirical success of a certain theory to its approximate truth do not support a robust form of scientific realism. In it, I argue that the realist can acknowledge the a crucial role of models in applying our theories to the world while claiming that theories describe the world directly and that the empirical success of a theory at a more fundamental level provides us with a special kind of evidence for its truth.
Christian Dieckhoff – ‘The Construction of Energy Scenarios – A Reconstruction of Modeling Practices and Mindsets in the Field of Energy Economic Eystems Analysis’
Abstract: In the field of energy economic systems analysis the application of computer models is a major element of research practice and policy advice. The models are used to produce quantified descriptions of the development of the energy system – mostly in form of scenarios. In my paper I will present first empirical results of my Ph-D project. Based on descriptions of scientists working in this field, I investigate the individual modeling and scenario building practices and the underlying theories, concepts and beliefs. The aim is to clarify the claims to validity raised with energy scenarios and the way the models are used to fulfill them.
Brian Epstein and Patrick Forber – ‘The Perils of Tweaking: When can Macrodata be Used to Set Parameters in a Microfoundational Simulation?’
Abstract: We consider a common but potentially problematic technique in constructing microfoundational simulations: the use of macrodata to set or tweak the parameters of microentities. This technique is regularly employed both in the initial construction of simulations and in the course of refining simulations over multiple iterations. We aim to identify when it is advisable to tune microparameters with macrodata and when it is illicit. We argue that this technique faces two very different kinds of risks: (1) the risk of overfitting, and (2) the risk of smuggled or covert macroproperties. We propose heuristics for simulation construction.
Axel Gelfert – ‘Strategies of Model-Building in Condensed-Matter Physics’
Abstract: In the philosophical discussion about models so far, attention to condensed matter physics has focused on either the use of mathematical models (as given by a specific Hamiltonian in terms of the operator formalism of quantum many-body physics), or on the limitations of such models in terms of accuracy and tractability. For example, it has been argued that if we are looking for accurate quantum mechanical predictions about, say, superconductivity, instead of constructing a model Hamiltonian one should look to phenomenological approaches, that is an ‘ad hoc combination of considerations from thermodynamics, electromagnetism, and quantum mechanics itself’ (Cartwright 1999). The present paper argues that, instead of construing the practice of modelling in condensed matter physics as involving a choice between either model construction or phenomenological approaches, the very process of model construction needs to be characterised in terms of trade-offs and competing considerations. Even at the level of fundamental theory, the range of in principle theoretically acceptable Hamiltonians is vastly underdetermined, and developing a many-body model that combines (however limited) theoretical and computational tractability with even a modicum of explanatory success and empirical adequacy, is generally considered to be a rare achievement. In developing this line of argument, the present paper attempts to make space for a sustained discussion of model construction as opposed to the much-discussed problems of representational success and empirical accuracy, both of which have dominated discussions of models in this area of science.
Peter Gildenhuys and Balazs Gyenis – ‘Classical Population Genetics and the Semantic Approach to Scientific Theories’
Abstract: While a number of writers have argued that the semantic approach to scientific theories provides a better way to understand evolutionary theory, and in particular classical population genetics, than does the syntactic view (Beatty 1982, 1987; Thompson 1983, 1987, 1989; Lloyd 1983, 1988), none of these writers has actually tried to present more than small fragments of classical population genetics in terms of the semantic approach. It is our aim to fill the lacuna here: Using the semantic approach to scientific theories, we present (as far as it is possible to do so) the most well-known model in classical population genetics, the discrete generation Wright-Fisher model.
Matthias Heymann – ‘Constructing Evidence and Trust: How did Climate Scientist's Confidence in their Models and Simulations Emerge?’
Abstract: Computer simulation was adopted quite smoothly in the atmospheric sciences. Climate simulation is a case in point. By 1979 climate scientists had reached agreement that climate warming was likely to occur in future decades, even though empirical evidence of warming was lacking at that time. Climate models and simulation helped to construct the evidence of warming. But how and why did trust in climate models and simulation emerge? The paper will attempt to make a first step in exploring the emergence and consolidation of trust by analyzing publications and documentations of climate scientists.
Rafaela Hillerbrand – ‘Epistemology of Computer Simulations. The Case of High Energy Physics’
Abstract: In recent years, computational sciences such as computational field theory supplemented theoretical and experimental investigations in many scientific fields. Often, there is a seemingly fruitful overlap between theory, experiment, and numerics. This paper aims at investigating the latter and discusses possible problems when computer simulation and laboratory experiment are intertwined. As regards their epistemic claims, three types of computer simulations are distinguished. The simulations and experiments within high energy physics as pursued at the Large Hadron Collider (LHC) at CERN serve as a study case.
Giora Hon and Bernard R. Goldstein – 'The Concept of Model in Victorian Physics: Revisiting the Case of Maxwell’
Abstract: In current scientific discourse the use of the term, model, abounds. However, this term was only used occasionally in technical discourse in early modern times, and it was not linked to a methodology until the late nineteenth century. As philosophers of science, we ask, What was a scientific model in nineteenth-century physics, and did Maxwell, the foremost contributor to the study of electrodynamics, appeal to it? In addressing these questions, we adhere consistently to a historiography which remains faithful to the primary sources as much as possible. Our methodology is intended to yield historically informed philosophy of science.
Lara Huber and Lara Kutschenko – ‘Mutant Mice: Experimental Organisms as Materialised Models’
Abstract: Standardised experimental organisms are broadly used in biomedical research. They are regarded as both scientific objects and epistemological tools. As such they are conceptualised as “animal models”. In taking biomedical research on Alzheimer’s disease as an example we will address epistemological challenges that arise out of the specificities of animal models: We will present how the functionality of a given experimental design shapes the concept of animal-based modelling. These issues will be discussed against the background of the very materiality of models in biomedical research.
Vincent Israel-Jost – ‘The Role of Simulations in Observation’
Abstract: While philosophers have seriously started to study models and simulations in relation with various categories of scientific practice (explanation, experimentation, prediction, etc.), very little has been said regarding the role of simulations in observation. In this paper, I argue that the most common characterizations of simulations, namely that they involve no direct physical interaction and rely on dynamic models, are unsatisfactory to capture any difference between simulations and observation. The analysis must be pushed further, I suggest, by looking at the domain of underlying hypotheses used in simulations and in observation.
Nicholaos Jones – ‘What Makes Idealized Hypotheses False, and What Doesn't’
Abstract: No idealized hypothesis is false in virtue of its constituent idealizations. For example, on my view, the idealization <gas particles are noninteracting point-masses> restricts the scope of phenomena that the ideal gas law characterizes, and it does so in a way that prevents gas behaviors at high pressures from providing disconfirming evidence for the law. I shall provide evidence for a particular way of interpreting idealized hypotheses, involving a distinction between their actual and apparent content, that supports this view and is superior of competing views that invoke ceteris-paribus clauses and validity limits.
Aki Lehtinen – ‘Economic Models, Derivational Robustness and Confirmation Holism’
Abstract: Economic model building can be characterised as derivational robustness analysis. It proceeds by investigating whether a given theoretical result can be with derived with different modelling assumptions. I will argue in this paper that, although derivational robustness analysis is not an empirical method in that no new data is being collected, it may be a crucial part of an empirical investigation, and that it provides a way of dealing with Duhem-Quine problems in economic modelling. My argument is that derivational robustness analysis, if used together with empirical data, provides contrastive information on individual assumptions in models, and thereby provides a way of evaluating their truth.
Johannes Lenhard – ‘Mix until Blended: Bayesianism Processed by the Computer’
Abstract: Bayesian approaches are experiencing a steep rise in scientific practice since the early 1990ies. The paper identifies two sources for this upswing. The accumulation of massive data obtained by high-throughput devices lends itself to so-called Empirical Bayes methods and Monte-Carlo-Markov-Chain simulations make tractable conditional probabilities in complex models. Bayesianism, it is argued, undergoes a significant transformation: its success in practice comes with a loss of the subjectivist part of it – which many consider to be the epistemologically defining part.
Maël Lemoine – ‘Mosaicism and Chimerism of Models’
Abstract: Behavioral neuroscience commonly uses several models of a single disease such as depression. This multiplicity takes two forms: mosaicism and chimerism of models. The “received view”, the semantic, 'polytypic' and 'autonomist' conceptions of models do not provide a sufficient account for what they consist in. The concept of disciplinary explicative values, that is, a set of rules that define the conditions on which a set of observable facts can be taken as a satisfying explanans, is proposed instead. Several sets of rules may apply to the same facts, producing different, even conflicting explanations.
Gregory Lusk – ‘Models and Scientific Explanation: A Reply to Alisa Bokulich’
Abstract: Little research has been conducted on the role of models in scientific explanation. Recently, an account of model explanation has been developed by Alisa Bokulich. In this paper, I provide an overview of Bokulich’s account and show that it faces several problems. It cannot show how fictional models can be justified from theory, nor how they can answer counter factual questions. I attempt to expand the view so that many of these problems can be avoided. I conclude that the modified account can mark as explanatory models justified top-down from theory, but perhaps not models built bottom-up from empirical data.
Micheal McEwan – ‘Abstraction, Generality and the Theory-Model Distinction’
Abstract: In this paper I reexamine the theory-model distinction from a new perspective by considering three common views: (1) theories are more general than models, (2) theories are more abstract than models, and (3) models play a mediating role between theories and the systems they represent. Despite appearances to the contrary, I show that these three views are divergent. That is, they support radically different attitudes concerning truth, explanation and prediction. I use this example to demonstrate that a clear characterization the theory-model distinction is needed before any general attitudes towards models and theories can be established.
Steven L. Peck – ‘New Wine in Old Bottles? Novel Philosophical Problems in Representing Ecological Systems with Agent-Based Models’
Abstract: Agent based systems have become very important to understanding the complex interactions of organisms in ecological and evolutionary systems. Like the complexity found in natural systems these models allow complexity to bubble-up from lower-level scales as digital organisms allow representation at spatial and temporal multiple scales. These types of models present several problems to understanding how a simulation represents and what role it can play in scientific discourse. I explore the question, ‘What allows us to argue that these models are telling us anything about the ecological systems they are designed to represent and how can they be credentialed?‘
Isabelle Peschard – ‘Relevance For What, Relevance For Whom: Numerical Simulation and Experimental Measurements Face-to-Face’
Abstract: Situation: scientists investigate a certain phenomenon, some conducting measurements, some conducting numerical simulations. Some of the measurements conflict with the numerical results. Question: What arguments would settle the dispute in favor of one or the other? Distinguishing between supportive arguments, used to support one’s claim, and undermining arguments, used to undermine the opponent’s claims, I examine what the four types of arguments and draw some conclusions to highlight some crucial epistemological differences between two forms of investigation.
I will focus on a specific case in fluid mechanics, but consider also some cases in cognitive psychology, like the Wason experiments.
Christopher Pincock – ‘Mathematical Contributions to Scientific Explanation’
Abstract: Recent discussions of mathematical explanations in science have claimed that these sorts of explanations have major interpretative implications. Unfortunately these arguments have not clarified the different ways in which mathematics can make a contribution to scientific explanations. I argue that at least three different sorts of contributions are possible: (i) tracking causes, (ii) isolating recurring features of a phenomenon and (iii) connecting different phenomena using mathematical analogies. Once these distinctions are in place it becomes clear that mathematics does make crucial contributions to scientific explanations, but the interpretative conclusions that some have drawn from these contributions are undermined.
Gordon Purves – ‘In Order to Form a More Perfect Fiction: An Interpretation of Imaginary Cracks as Non-Fictions’
Abstract: I examine the recent work on the philosophy of scientific fictions by Winsberg, Suarez and Teller. By considering two case studies in fracture mechanics, the strip yield model and the imaginary crack method, I argue that their reliance upon the social norms associated with an element of a model forces them to remain silent whenever those norms fail to clearly match those characteristic of fictions or non-fictions. To resolve such cases, I propose a normative expansion of the epistemology of fictions which clarifies a model's ontological commitments when the scientific community lacks any clear, shared norms of use.
Isaac Record – ‘Getting a Foot in the Door: Accepting Random Numbers in Simulations’
Abstract: Random numbers – or at least simulated random numbers – are at the heart of Monte Carlo simulations. What inferences we are licensed to make on the basis of simulation results depends on the role these random numbers play. Practitioners and philosophers argue that random numbers are self-liquidating, unbiased, stand-ins for reality, or else inessential. I will argue that none of these arguments supports more than provisional acceptance without begging the question. Nevertheless, they are the foot in the door that gives simulations the opportunity to be successful – and self-vindicating.
Yasha Rohwer – ‘Game-theoretic Models and the Evolution of Prosocial Behavior’
Abstract: Two-person games are the standard models when thinking about the evolution of prosocial behavior. Focusing on coalitions, I argue for an expansion beyond standard two-person models. I propose that any model of coalitions must 1) explain both coalition formation and maintenance and 2) accord with what is known about chimpanzee coalitions. Under these criteria, the Prisoner’s Dilemma and Stag Hunt—traditional workhorses of prosocial modeling—fall short. A better model is the three-person Odd Man Out game, which satisfies both criteria. Not all prosocial behaviors can be boiled down to general moral categories that are nicely modeled by two-person games.
Alirio Rosales – ‘Causal Narratives and Theory Structure in Population Genetics: The Case of R.A. Fisher and Sewall Wright’
Abstract: Scientific theories comprise at least certain fundamental equations and mathematical models. Recent philosophy of science has begun to include stories or narratives as another component of theoretical frameworks. Drawing on lessons drawn from physics and economics, I present a biological case study from the history of population genetics. The term causal narrative is introduced as a scenario where biological processes are represented in a series of stages and a temporal order is fixed. R.A Fisher’s “mass selection theory” and Sewall Wright’s “shifting balance theory” are re-interpreted as causal narratives that frame and constrain mathematical model building and explanation.
Jen Schellinck and Richard Webster – ‘The Scientific Power of Good Models: Unifying Hypothesis Discovery and Hypothesis Testing’
Abstract: Although scientists often accept models as a key part of science, the nature of model contributions in specific research contexts are less well understood; this can lead to a lack of model uptake in these circumstances. In this paper we discuss the ways in which well-constructed models connect diverse research activities, including hypothesis discovery and hypothesis testing. We argue that models have scientific power because they unify theoretical constructs and experimental data. They thus act as a special type of cognitive tool, increasing the ability of scientists to accumulate scientific knowledge in a given research area.
Oron Shagrir – ‘Representation and Simulation in the Brain’
Abstract: The focus of this paper is the extent at which the brain employs S-representations; “S” for simulation (Cummins, 1989). I will first argue that the notion of S-representation is deeply entrenched in current computational approaches in neuroscience. I will then turn to rebut a recent argument (Ramsey 2007) to the effect that S-representations do not play an essential explanatory role in the current computational approaches in cognitive science. My claims are exemplified through a detailed examination of the computational work on the oculomotor memory.
Michael Stoeltzner – 'What Does a Higgs-Model Represent?’
Abstract: The recently launched Large Hadron Collider has primarily been built to find the final piece of the standard model of elementary particle physics, the Higgs particle. The Higgs mechanism is responsible for the masses of the elementary particles. This paper investigates to what extent both Higgs particle and mechanism can be integrated into the contemporary debate about models. Since the Higgs mechanism stands between the framework theory and empirical predictions, and since the Higgs sector enjoys a certain autonomy within the standard model, the “models as mediators” approach seems to recommend itself. Moreover, in virtue of the intimate connection between the Higgs mass and the other parameters of the standard model, the Higgs particle can be understood as a measuring device for some of those parameters. The model’s representative features are harder to pin down because they cannot – as physicists do – be based on certain terms in the Lagrangian that can be given a particle interpretation. Instead of quantizing a classical Higgs model, the Higgs model must be quantum from the start in order to fulfill both the function of an autonomous representation and a measuring device. This is, to my mind, an interesting lesson for the “models as mediator” approach even though the unsettled physical facts at present make it impossible to provide a definitive analysis of the Higgs model.
Ekaterina Svetlova – 'Toward an Account of Model Use'
Abstract: The use of models and causalities that are related to them has never been the focus of philosophical investigations. Though this attitude is starting to change, no explicit theory yet exists that is concerned with the practical activity of model application for building useful causal hypotheses. The paper suggests first steps towards such an account. The approach is illustrated using models from modern finance. The argumentation is based upon an empirical study.
Eran Tal – ‘Can Computer Simulations Measure? A Dilemma’
Abstract: This paper discusses the method of comparability, a central method in the evaluation of measurement accuracy. I illustrate the use of this method in the determination of the accuracy of atomic clocks. I then show that a dilemma arises when one attempts to use this method to evaluate the accuracy of computer simulations. Specifically, the method entails that the accuracy of certain computer simulations can be established by comparing computer simulations to each other, without direct reference to any measurement. I discuss the implications of this counterintuitive result.
Paul Teller – ‘Precision and Accuracy in Models’
Abstract:Models fall short of the representational ideal of exact representation by being inaccurate and/or imprecise. I will explore these notions and then argue that inaccuracy and imprecision are really two sides of the same coin. Development of these ideas will require developing the ancillary idea that any application of models must take place in some framework that is treated as if it were both completely precise and completely accurate. These considerations will have repercussions for understanding the phenomenon of vagueness in language.
Marion Vorms – ‘Models of Data and Theoretical Hypotheses: A Case Study in Classical Genetics’
Abstract: This paper aims at analyzing the interplay of computational constraints on the one hand and scientists’ commitments to theoretical hypotheses explaining the data on the other hand, and their respective influence on the choice and modification of the format of presentation of data. The examination of the debate opposing Morgan’s school and Castle, around 1919, about how to present statistical data obtained from Mendelian breeding experiments shows that the choice of a format in modeling data is driven by a prior commitment to a theoretical hypothesis explaining the data. This historical analysis could shed light on some issues concerning modeling in data driven sciences, where no general theoretical framework is available.
Andrew Wayne – ‘Idealization and Explanation: In for a Penny, in for a Pound’
Abstract: One apparent role of idealization is to underwrite scientific explanation, and therein lies a puzzle: much of what physicists take to be bona fide explanation based on idealization fails to satisfy orthodox philosophical accounts of explanation. The standard response to this puzzle is to claim significant differences between what are variously called “Galilean,” “controllable,” or “harmless” idealizations and other idealizations, and to argue that only the former underwrite bona fide explanation. This paper contends that the standard response is untenable on both points. It sketches options that remain for evaluating the explanatory role of idealizations.
Aaron Wright – ‘Narrative Simulations: Case Study and Analysis’
Abstract: Roman Frigg has recently (2009) proposed an account of models in science in analogy to accounts of literary fiction. I argue that, rather than fictions, models and simulations should be considered as narratives. Narratives are both acts and physical objects; attention to narrative allows a unified treatment of simulations as objects and in use. This is elucidated through a discussion of some simulations in contemporary experimental high energy physics. The question of how simulations can represent and explain phenomena exterior to themselves becomes empirical. This approach presents a productive opportunity to connect philosophy of science to broader science studies.