Fictions in Science

Philosophical Essays on Modeling and Idealization

Mauricio Suárez

• Publisher: Routledge
• Year: 2008
• ISBN:
  9781135854713
• Edition: 1
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

Science is popularly understood as being an ideal of impartial algorithmic objectivity that provides us with a realistic description of the world down to the last detail. The essays collected in this book—written by some of the leading experts in the field—challenge this popular image right at its heart, taking as their starting point that science trades not only in truth, but in fiction, too.

With case studies that range from physics to economics and to biology, Fictions in Science reveals that fictions are as ubiquitous in scientific narratives and practice as they are in any other human endeavor, including literature and art. Of course scientific activity, most prominently in the formal sciences, employs logically precise algorithmic thinking. However, the key to the predictive and technological success of the empirical sciences might well lie elsewhere—perhaps even in scientists' extraordinary creative imagination instead. As these essays demonstrate, within the bounds of what is empirically possible, a scientist's capacity for invention and creative thinking matches that of any writer or artist.

Part I
Introduction

1
Fictions in Scientific Practice

Mauricio Suárez

1.1 FICTIONALISM IN THE PHILOSOPHY OF SCIENCE

This volume collects thirteen essays by prominent contemporary philosophers of science on the role that fictions and fiction-making play in the practice of theorizing and model-building in science. The topic itself is not new in philosophy of science, because fictionalism has a long history in the discipline that goes back at least to the writings of Hans Vaihinger in the early years of the 20th century. Although Vaihinger’s work was very popular in his own time, it fell into a kind of oblivion for years afterward, coinciding with the rise of logical positivism. The topic made a return to the philosophy of science agenda about 15 years ago, partly in the wake of debates about Bas Van Fraassen’s constructive empiricism, which can be construed as a kind of fictionalism about theoretical entities,¹ and partly as a result of the increased attention paid by philosophers to models and modeling in the sciences. However, within philosophy of science the discussion over fictions has so far been rather fragmentary, with references to fictionalism typically playing some minor rhetorical role in the realism–antirealism debate, and appeals to fictions and fictive entities appearing occasionally in diverse case studies devoted to modeling. This might be contrasted with the very extensive treatment of fictions in the philosophy of language and aesthetics literature over at least the last two decades.²

The present volume is intended to help to redress the balance. It is the first volume that treats fictions from a philosophy of science perspective.³ It is also the first collection of papers entirely devoted to the topic of fictions and fictionalizing in scientific practice.⁴ It represents the most recent and up-to-date thinking on the topic over a range of scientific disciplines in contemporary philosophy of science. The authors reflect upon the role and function of fictionalizing in the process of building models for natural and social systems. Hence the emphasis is markedly on case studies rather than general philosophical theory, because even the most general philosophical essays refer extensively to case studies from diverse sciences. Thus although the focus is often on the physical sciences, there are also essays on biology and economics, and references to case studies in other areas of science appear throughout the essays.

1.2 FINE’S VAIHINGER

Although philosophers of science have paid scant attention to fictions, there have been some exceptions. In 1993 Arthur Fine published a seminal article entitled “Fictionalism” where he attempted to bring Vaihinger’s philosophy of ‘as if’ back to the forefront of philosophical debate. Besides promoting a reevaluation of Vaihinger’s work, this essay for the first time makes an explicit connection between fictionalism and the modeling literature in the philosophy of science of the last two decades or so. Fine’s essay is already cited in anthologies as a classic “revival” piece, and it is the standard con-temporary reference on Vaihinger among both philosophers of science and metaphysicians. And because most of the chapters in this volume refer to this essay, often quoting it as a source of inspiration, it is here reprinted as the starting chapter of the book. It appropriately sets the tone and the terms for most of the other essays in the book.

Fine’s paper is not a piece of scholarly history but an interested commentary on a historical figure and its relevance for the contemporary debate. In other words, Fine’s Finehinger is bound to differ from the original Vaihinger, and Fine himself acknowledges some unhistorical elements (Fine, this volume, section 8). The account is arguably anachronistic in the following two senses.⁵ An earlier appeal to Vaihinger appears in Bas Van Fraassen’s celebrated book The Scientific Image (Van Fraassen, 1980, p. 35–36), where Vaihinger’s philosophy is cited in support of constructive empiricism’s agnosticism regarding theoretical entities. More specifically, Vaihinger’s fictionalism is invoked there as ammunition against the verificationist’s commitment to the full theoretical equivalence of empirically equivalent theories (roughly, on this account, if two putatively different theories have exactly the same empirical consequences then they really are just the same theory). Van Fraassen argues that by contrast the fictionalist can claim that a theory that postulates a fictional unobservable entity is distinct from its empirically equivalent non-entity-postulating rival theories—even though none of these theories commits in fact to the reality of the entity. And this is just the constructive empiricist’s response to the verificationist challenge. Thus, although Van Fraassen does not make the mistake of explicitly pinning the constructive empiricist commitments on Vaihinger, the reader of these passages might be led to suppose a strong association between Vaihinger’s fictionalism and present-day antirealism of the constructive empiricist variety.

The association finds echoes in a few passages in Fine’s piece, but it is definitely not Vaihinger’s—who could not possibly have anticipated logical positivism’s verificationism, Van Fraassen’s response to it, or the main arguments in the contemporary realism–antirealism debate. Vaihinger himself was not committed to a fundamental epistemological difference between our knowledge of the observable world and that of the unobservable world. It is even questionable whether he acknowledged the antecedent distinction between observable and unobservable entities or domains of the world.

Another anachronism is arguably Fine’s close association of Vaihinger to Neurath. This nowadays looks out of place with Vaihinger’s pronounced neo-Kantian transcendentalism.⁶ The association probably responds to the attempts in the early 1990s to re-evaluate logical positivism by emphasizing its “sociological” strand, and the then comparatively neglected figure of Neurath. Fine is here trying to distinguish Vaihinger’s project from Carnap’s and Reichenbach’s brand of “logicism,” and the most straightforward way to do this in the climate of the early 1990s is to associate Vaihinger with the rising counterbalancing figure of Neurath. Arguably, the real Vaihinger is a little less committed to the particularism defended by Fine’s Natural Ontological Attitude (NOA). After all, Vaihinger promotes a theory of scientific reasoning, and advances what he takes to be a general theory of the logic of scientific fictions. Both assumptions seem estranged from Fine’s NOA, and would have been alien to Neurath. Crucially no anticipation is to be found in Vaihinger of Neurath’s most valuable and long-lasting contribution to the theory of knowledge, namely: his holistic epistemology regarding empirical knowledge.

Regardless of any anachronisms, Fine’s article has rightly set the agenda for the discussion of fictions in contemporary philosophy of science, and most of the essays in the volume refer to it in such terms.⁷ In particular, Fine lays out the connection between fictions and the contemporary philosophical literature on modeling practice more carefully and helpfully than Vaihinger would have been able to do. For Fine the recent literature on idealization and abstraction in science backs up Vaihinger’s views, whereas all cases of scientific model-building are instances of Vaihinger’s fictionalism at work: “Preeminently the industry devoted to modeling natural phenomena, in every area of science, involves fictions in Vaihinger’s sense” (Fine, this volume, p. 34). The essays in this volume address this claim in a variety of ways, sometimes by filling in detail through case studies, sometimes by adding philosophical flesh and argument, and yet at other times by taking issue with the claim or qualifying it to some extent.

1.3 A MAP OF THE BOOK

Part II of the book (“The Nature of Fictions in Science”) contains two essays besides Fine’s devoted to the nature of fictions in science. Joseph Rouse (Chapter 3) defends the radical thesis that fictions and fictionalizing are not only present in modeling, simulation, supposition, and thought-experimentation. These are activities that we may call “representational” in a broad sense, and may seem prima facie distinct from most of the activities carried out in experimental science, which typically involve actual causal interventions on experimental systems in real laboratories. Fictions and fictionalizing have typically been thought to be confined to the realm of the “possible” within representational science. Rouse defends the view that fictionalizing also plays a role in the constitution of objects and procedures actually employed in the laboratory sciences. He appeals to case studies from genetics and thermometry in order to back up the claim that hypothetical entities and fictional assumptions are embodied in the material construction of these objects. (The same view is defended by Ankeny in Chapter 12 in relation to model organisms in biology.) The point of such entities and assumptions, according to Rouse, is to introduce alternative concepts, thus opening up a new space of reasons for the justification of scientific belief. Thus fiction-making in science can not be reduced, argues Rouse, to an exercise of expediency whereby false assumptions are employed for inferential or predictive gain. Fictional assumptions are not simply false-but-useful; they actually define both what is false and what is useful. This is a radical new proposal regarding the nature of fiction in science, going beyond the claims by most other contributors to the volume, but very much deserving to be explored.

In addition, Rouse’s emphasis on activity and practice ahead of ontology is nicely in line with the rest of the essays in the volume. The contributors to this volume are neither distracted nor particularly concerned with questions regarding ontology. The purpose of the book is instead to focus on relevant aspects of the activity of model building. It might seem at first sight that Chapter 4 is an exception to this rule, because Barberousse and Ludwig’s aim is to fill much-needed detail into the claim—a slogan, really—that scientific models are fictions. This identity claim about the ontology of models has been made before—for example, by Cartwright (1983, pp. 151–162)—but Barberousse and Ludwig point out that both sides of the identity need to be explicated to take it beyond a mere slogan.

Moreover, the slogan fails to capture the claims defended by the contributors to this book, which are fundamentally distinct. The essays here collected focus on the important role that fiction-making plays within the activity of modeling in science. But they remain on the whole silent on the question “what is a model?” which they tend to view as an issue of secondary importance.⁸ However Barberousse and Ludwig’s approach to this issue is of a piece with the book’s outlook, because they claim that the slogan “models are fictions” is best filled in precisely by putting activity center stage, as opposed to structure or nature, as an account of both models and fictions. In particular they follow Currie (1990) and Walton (1990) in understanding fiction as the activity of imagining and role-playing, as opposed to any putative referential relations between “representans” and presumed “representanda.” They then go on to use this view in order to provide a typology of models in terms of their functions (prospective, bridge, and what they call “anti-Duhemian” models).

In keeping with the book’s outlook, part III (“The Explanatory Power of Fictions”) swiftly moves away from issues related to the nature of fiction toward a discussion of one of the main and most controversial roles that fictional assumptions may play in modeling practice. Recent philosophy of science has emphasized the importance of explanatory virtues in the assessment and evaluation of theories and models alike. A widespread assumption is that false theories that are known to be false may never have explanatory power. Scientific realism backs this up by appeal to the idea that the explanatory power of a theory depends upon its closeness to truth. (So an inference to the most explanatory theory guarantees that the one closest to the truth is selected.⁹) The antirealist disagrees that this is a valid mode of inference (the most explanatory theory need not be closer to the truth), but still finds it difficult to articulate a notion of “explanation” that will fit this bill. So the question “do fictional assumptions have explanatory power?” goes to the heart of the realism–antirealism dispute. The three essays collected in this part of the volume answer this question differently.

First, Catherine Elgin (Chapter 5) argues for a move away from the notion of explanation and toward a more general notion of scientific understanding. Her view is that representation in science is best understood along the lines of Goodman’s notion of “exemplification” of properties. Roughly, a model A exemplifies a property B if and only if A is denotative of B and moreover A possesses the property B. (So a concrete model of a building exemplifies fragility if it denotes fragility and is fragile.) The advantage of this account, according to Elgin, is that representation does not require actual denotation, but only that the ‘representans’ be a “denoting symbol or system.” In other words, actual successful reference is less important for representation than the function of denoting. Because the representational character of models does not depend on successful denotation, a model can be representational just in virtue of purporting to denote, regardless of its actual success in doing ...

Table of contents

• Routledge Studies in the Philosophy of Science
• Contents
• Acknowledgments
• Part I Introduction
• Part II The Nature of Fictions in Science
• Part III The Explanatory Power of Fictions
• Part IV Fictions in the Physical Sciences
• Part V Fictions in the Special Sciences
• Part VI Fictions and Realism
• References
• Contributors
• Index