Idealization and the Aims of Science

By Angela Potochnik

Science is the study of our world, as it is in its messy reality. Nonetheless, science requires idealization to function—if we are to attempt to understand the world, we have to find ways to reduce its complexity.
 
Idealization and the Aims of Science shows just how crucial idealization is to science and why it matters. Beginning with the acknowledgment of our status as limited human agents trying to make sense of an exceedingly complex world, Angela Potochnik moves on to explain how science aims to depict and make use of causal patterns—a project that makes essential use of idealization. She offers case studies from a number of branches of science to demonstrate the ubiquity of idealization, shows how causal patterns are used to develop scientific explanations, and describes how the necessarily imperfect connection between science and truth leads to researchers’ values influencing their findings. The resulting book is a tour de force, a synthesis of the study of idealization that also offers countless new insights and avenues for future exploration.

• Language: English
• Publisher: Open Road Integrated Media
• Release date: Nov 17, 2017
• ISBN: 9780226507194

Book preview

The University of Chicago Press, Chicago 60637

The University of Chicago Press, Ltd., London

© 2017 by The University of Chicago

All rights reserved. No part of this book may be used or reproduced in any manner whatsoever without written permission, except in the case of brief quotations in critical articles and reviews. For more information, contact the University of Chicago Press, 1427 E. 60th St., Chicago, IL 60637.

Published 2017

Printed in the United States of America

26 25 24 23 22 21 20 19 18 17      1 2 3 4 5

ISBN-13: 978-0-226-50705-7 (cloth)

ISBN-13: 978-0-226-50719-4 (e-book)

DOI: 10.7208/chicago/9780226507194.001.0001

Library of Congress Cataloging-in-Publication Data

Names: Potochnik, Angela, author.

Title: Idealization and the aims of science / Angela Potochnik.

Description: Chicago : The University of Chicago Press, 2017. | Includes bibliographical references and index.

Identifiers: LCCN 2017028476 | ISBN 9780226507057 (cloth : alk. paper) | ISBN 9780226507194 (e-book)

Subjects: LCSH: Science—Philosophy. | Idealism.

Classification: LCC Q175.P88155 2017 | DDC 501—dc23

LC record available at https://lccn.loc.gov/2017028476

This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

Idealization and the Aims of Science

ANGELA POTOCHNIK

The University of Chicago Press

Chicago and London

For Mabel and Amelia, the bookends of this book

Contents

Preface

1. Introduction: Doing Science in a Complex World

1.1. Science by Humans

1.2. Science in a Complex World

1.3. The Payoff: Idealizations and Many Aims

2. Complex Causality and Simplified Representation

2.1. Causal Patterns in the Face of Complexity

2.1.1. Causal Patterns

2.1.2. Causal Complexity

2.2. Simplification by Idealization

2.2.1. Reasons to Idealize

2.2.2. Idealizations’ Representational Role

2.2.3. Rampant and Unchecked Idealization

3. The Diversity of Scientific Projects

3.1. Broad Patterns: Modeling Cooperation

3.2. A Specific Phenomenon: Variation in Human Aggression

3.3. Predictions and Idealizations in the Physical Sciences

3.4. Surveying the Diversity

4. Science Isn’t after the Truth

4.1. The Aims of Science

4.1.1. Understanding as Science’s Epistemic Aim

4.1.2. Separate Pursuit of Science’s Aims

4.2. Understanding, Truth, and Knowledge

4.2.1. The Nature of Scientific Understanding

4.2.2. The Role of Truth and Scientific Knowledge

5. Causal Pattern Explanations

5.1. Explanation, Communication, and Understanding

5.2. An Account of Scientific Explanation

5.2.1. The Scope of Causal Patterns

5.2.2. The Crucial Role of the Audience

5.2.3. Adequate Explanations

6. Levels and Fields of Science

6.1. Levels in Philosophy and Science

6.2. Going without Levels

6.2.1. Against Hierarchy

6.2.2. Prizing Apart Forms of Stratification

6.3. The Fields of Science and How They Relate

7. Scientific Pluralism and Its Limits

7.1. The Entrenchment of Social Values

7.2. How Science Doesn’t Inform Metaphysics

7.3. Scientific Progress

Acknowledgments

List of Figures

List of Tables

Notes

References

Index

Preface

Physicists sometimes assume that surfaces are frictionless planes. Biologists sometimes assume that populations of organisms are infinite in size. And economists sometimes assume that humans are perfectly rational agents. None of these things is true; they are all idealizations. Idealizations are assumptions made without regard for whether they are true and often with full knowledge that they are false. Idealizations of all kinds pervade science, and it’s uncommon for scientists to try to replace them with more accurate assumptions. On the face of it, this is a puzzle. Why do scientists deliberately maintain falsehoods in their theories and models? What do idealizations contribute to science?

In this book, I motivate a strong view of idealizations’ centrality to science, and I reconsider the aims of science in light of that centrality. On the account I develop, science does not pursue truth directly but instead aims to support human cognitive and practical ends. Those are projects to which idealizations can directly contribute in a number of ways.

The first three chapters are used to develop my account of idealization’s central role in science. In Chapter 1, I discuss how science is shaped by its human practitioners and by the world’s complexity. Together, these two ideas inspire a view of science as the search for causal patterns, a search that invariably relies heavily on idealizations. Idealizations contribute to science in a variety of ways, including by playing a positive representational role. These ideas are developed in Chapter 2. In Chapter 3, I detail a few case studies that demonstrate the ubiquity of idealization in science, as well as the wide range of purposes it serves.

The last four chapters explore the implications of this account of idealization for central philosophical debates about the aims of science. Chapter 4 motivates the idea that the epistemic aim of science is not truth but human understanding. Understanding is a cognitive achievement, and unlike truth, it can be directly furthered by idealizations. In Chapter 5, I develop an account of scientific explanation that does justice to how the production of understanding depends on human cognizers. Then, in Chapter 6, I challenge classic conceptions of scientific levels of organization, and I develop an opposed view that better accords with idealized representation across all fields of science. Finally, Chapter 7 shows how my account of idealization and the aims of science expands the influence of human characteristics and values on science’s aims and products while also constraining scientific and metaphysical pluralism.

This is ultimately a book about how science is influenced by its creators, limited human beings finding our way in a complex world. The science we create liberally employs falsehoods. As a result, scientific knowledge does not in general consist of truths about the phenomena in our world but is more partial and indirect in its relationship to the world, mediated by human concerns. The fields of science do not reflect compositional relationships found in nature but are instead a haphazard division of labor according to researchers’ interests. Any successful account of scientific explanation must be shaped by the cognitive requirements for human understanding. And, finally, human aims and values permeate scientific methods and products. Yet none of the views I develop in this book is intended to undermine the scientific enterprise. Quite to the contrary, on my interpretation of what science is designed to accomplish, the success of science becomes all the more obvious.

This book is primarily written for philosophers of science and students of philosophy of science. But I hope it is also of interest to theoretically minded scientists and anyone else who thinks hard about the nature of science. I have tried to address the topics in ways that do not assume familiarity with the specific philosophical debates to which they relate. The ideas put forward in this book have significance for classic issues in philosophy of science, but I believe they also have significance beyond that, including practical implications for scientific practice.

1

Introduction: Doing Science in a Complex World

All scientific research has been accomplished by human agents, for human ends. Each scientist is an individual human being, with a unique combination of characteristics, concerns, and background. The science these individuals pursue is aimed to provide greater human understanding of the world we inhabit and of ourselves and to further human projects of construction, manipulation, and control. Increasingly, philosophers of science have focused on accounting for scientific practice as it actually occurs: the often-messy and never-completed project of limited human beings, pursued for human ends. This stands in contrast to philosophical approaches to science popular in the past that attempted to transcend the messiness of science and the limitations of its current practitioners. One of those approaches is rationally reconstructing science to demonstrate its logical or epistemic basis, setting aside historical contingencies and any other deviations from the logically and epistemically ideal. Rudolf Carnap (1928) famously employed this approach in Der logische Aufbau der Welt. A second philosophical approach that attempts to transcend science’s imperfections is, instead of accounting for today’s actual science, aiming to predict what a future, more perfect science will look like. This strategy was explicitly employed by Paul Oppenheim and Hilary Putnam (1958), who developed a working hypothesis that all of science will ultimately be grounded explicitly in microphysics.¹

Along with philosophers’ increasing focus on actual scientific practice, there is also greater attention paid to how complexity influences science. Quite many scientists and philosophers of science have been impressed by the complexity of the world we inhabit and investigate, and appreciation for this complexity increasingly shapes scientific approaches as well as philosophical accounts of science. Phenomena occur in seemingly endless variety and permutations.² Simple accounts of these phenomena generally meet with only limited success, and approaches to studying complex systems have proliferated across many fields of science. Philosophers have variously called this world disordered, complex, dappled, and unsimple (Dupré 1993; Bechtel and Richardson 1993; Cartwright 1999; Wimsatt 2007; Mitchell 2012b). These philosophers and others have attempted to articulate the implications of this complexity for our theories of science and metaphysics, with wide-ranging results. Our best scientific laws may, strictly speaking, be false (Cartwright 1983). Our categories, including scientific categories, may not carve nature at its joints (Dupré 1993). Phenomena may not actually be law governed or predictable (Cartwright 1999). Scientific practices may vary widely, consisting mainly of fallible, heuristic procedures (Wimsatt 2007). The influence of different causes on a phenomenon may not be separable, even in principle (Mitchell 2012b).

These two observations—that science is ultimately the project of limited human beings and that the world we inhabit is incredibly complex—together constitute the starting point of my investigation in this book. Both are familiar ideas in today’s philosophy of science. Nonetheless, tracing out their full implications leads to surprising conclusions, conclusions that conflict with a variety of widely held philosophical ideas about science. Most basically, a science practiced by limited human beings in a complex world results in widespread idealization. Idealizations are assumptions made without regard for whether they are true, generally with full knowledge that they are false. Classic examples are the assumption of a frictionless plane in physics and the assumption of perfectly rational agents in economics. Despite their falsity, idealizations appear in most every scientific project and product, for a range of purposes, and they are not eliminated or even controlled for in the ways we might expect. This widespread idealization outstrips what most philosophers are willing to accept. Accordingly, the full scope of the use of idealizations in science has wide-ranging implications for our best theories of science. In this book, I investigate the implications of idealization for what science shows us about the world, levels of organization and the relationship among fields of science, the nature of scientific explanations, the role of human values in science, and even the very aims of science. In this chapter, I begin by considering in greater depth these observations of science as a human project and of the complexity of the world.

1.1   Science by Humans

In The Descent of Man, Charles Darwin (1871) observed of animals that the males are almost always the wooers and that the female, on the other hand, with the rarest exceptions, is less eager than the male…she is coy. This idea, that in most species males are aggressive and sexually promiscuous while females are passive and sexually selective, has been widely held by biologists, from Darwin up to today. It is one of the primary ideas of what is called sexual selection theory, one part of evolutionary theory. Male elephant seals fight each other for dominance over harems of females. Peacocks display their colorful trains in the hopes that a peahen will choose to mate with them instead of other peacocks.

FIGURE 1.1. These are Bateman’s original figures, depicting the findings of his experiments for the relationship between the number of matings and reproductive success. The lines compare the data for male fruit flies (solid lines) and female fruit flies (dashed lines). Figures (a) and (b) result from two different series of experiments (Bateman 1948, Fig. 1). Reprinted by permission from Macmillan Publishers Ltd: Heredity, copyright 1948.

A. J. Bateman (1948) provided an explanation for this phenomenon, based on his research on Drosophila (fruit flies). According to what has been dubbed Bateman’s principle, males stand to produce more offspring as a result of more matings, whereas in most species, females gain few if any additional offspring from multiple matings. If this is true, then males who mate many times are evolutionarily advantaged over other males, and male promiscuity and aggression in securing mates are expected to evolve, but the same is not true for females. Figure 1.1 shows Bateman’s representation of the data from his fruit fly research. He ran two experiments, and in both, the number of offspring increased more dramatically for male fruit flies who mated more often than they did for female fruit flies who mated more often. Bateman says,

This would explain why…there is nearly always a combination of an undiscriminating eagerness in the males and a discriminating passivity in the females. Even in a derived monogamous species (e.g. man) this sex difference might be expected to persist as a rule. (365)

And so, biologists widely accept that most male animals, but not female animals, have evolved to be promiscuous and aggressive and that Bateman’s principle explains why: it is evolutionary advantageous for males, but not females, to mate many times.

However, it has been argued that this view of the natural world, based on Darwin’s ideas from the late nineteenth century, is also infused with Victorian moral sentiment. Jonathan Knight (2002), for example, says that Bateman’s work has been used to extend dated preconceptions about human sexual behaviour to the entire animal kingdom, sometimes to the detriment of scientific knowledge (254). Knight shows that, while many biologists still endorse Bateman’s work as basically sound, most also emphasize that the situation is often much more complicated than Bateman recognized. He notes one biologist’s analysis that the reason Bateman’s observation became ‘Bateman’s principle’ is that it appealed to people’s intuition about the behaviour of individuals (256). Figure 1.2 is a much more recent representation of Bateman’s principle by Krasnec et al. (2012). Comparison between this figure and Bateman’s original figure shows how Bateman’s principle has taken on a life of its own. First, the relationship depicted in this treatment is more sharply defined than what Bateman presented based on his data. Second, this treatment posits a relationship to overall fitness, which is a stronger claim than Bateman’s finding regarding fertility, just one component of fitness. And so, we find both a codification of Bateman’s principle and, increasingly, the recognition that this simple principle does not reflect the full complexity of sexual behavior in animals. A few biologists even reject Bateman’s principle entirely, along with Darwin’s original observation of aggressive males and coy females, as grounded entirely in scientists’ values and expectations instead of sound evidential reasoning (Roughgarden 2004, 2009; Gowaty et al. 2012).

This is one example of how scientific findings can be informed by human expectations. Darwin’s observations of the natural world bear a striking resemblance to the social norms of Victorian England, and Bateman’s research was later taken to underwrite a general principle in part because it accorded with researchers’ intuitions. Research informed by human expectations can still be well founded, and it can still lead to successful results. Darwin’s idea of aggressive males and coy females may mirror the expectations for men and women in Victorian England, but it is nonetheless the case that male elephant seals tend to fight one another while female elephant seals do not and that the results of those fights determine male seals’ access to mating opportunities. There are also examples of researchers’ expectations leading to straightforwardly bad science. Stephen Jay Gould (1996) shows that this is so for research purported to show a genetic basis for IQ differences between races, sexes, or social classes. As a second example, Naomi Oreskes and Erik Conway (2011) analyze how a single group of scientists misled the public about tobacco research and environmental issues ranging from acid rain to global climate change, apparently motivated by their political leanings. But even more legitimate research like Bateman’s work, as well as scientific results that are even more widely accepted, still bears the mark of human expectations, human concerns, and the limits of human observation.

FIGURE 1.2. This figure is a recent illustration of Bateman’s principle (Krasnec et al. 2012, Fig. 2), reprinted with permission from Michael D. Breed. Comparison with Figure 1.1 above shows that this presents the principle in a stronger form than warranted by Bateman’s original research. First, it represents the difference between males and females as more sharply defined than Bateman’s data allowed. Second, it presents the relationship as between number of matings and relative fitness rather than relative fertility. This is a stronger conclusion, for it assumes that additional matings have no effect on fitness other than number of offspring produced. This assumption is unlikely to hold true for many animal species.

Features of the natural world that are not anticipated might escape notice entirely. The primatologist Sarah Blaffer Hrdy (1986) shows that data running contrary to the generalization about male promiscuity and female coyness in animals were available many years before any researchers began to question that generalization. In particular, female promiscuity is quite common among primates. Hrdy argues that it was thus not newly available data but a shift in the researchers themselves that led to a recognition of the limits of the promiscuity/coyness generalization. She says,

I seriously question whether it could have been just chance or just historical sequence that caused a small group of primatologists in the 1960s, who happened to be mostly male, to focus on male-male competition and on the number of matings males obtained, while a subsequent group of researchers, including many women (beginning in the 1970s), started to shift the focus to female behaviors. (Hrdy 1986, 159)

Hrdy hypothesizes that this new focus on female behaviors led researchers to finally recognize female promiscuity in primates. This in turn inspired, for the first time, a search for explanations for female promiscuity, instead of merely dismissing those behaviors as anomalies. There is still no agreed-upon evolutionary explanation for female promiscuity. Many biologists think the existing theoretical framework is accurate but needs to be employed differently for cases like the primates (Clutton-Brock 2009). Other biologists think the whole theoretical framework inspired by Darwin’s observations of differences between male and female animals needs to be abandoned (Roughgarden 2009). For present purposes, the important point is that, in this case, different individuals with different life experiences performing the science resulted in a significantly altered focus. The changed focus in turn led to an emphasis on different types of interactions and the observation of different patterns.

In philosophy, there is increasing recognition of the myriad ways in which human expectations, concerns, and limitations influence scientific practice. Feminist philosophy of science has attended to this influence for quite some time, including especially the role of social values on the process and products of science. A primary goal of this line of research is to address the question of how social influences on science can be reconciled with scientific objectivity. The current focus in philosophy of science on scientific models, including on the use of abstractions and idealizations, is another area where the limitations of human scientists are considered to be relevant. Wimsatt (2007), for instance, explicitly addresses how idealized models are the result of science produced by limited human agents. Finally, there is a growing literature in philosophy of science that directly addresses the relationship between science and broader social concerns. Special journal issues on the topics of socially relevant and socially engaged philosophy of science (Fehr and Plaisance 2010; Potochnik 2014) address this trend.

The extent of human influences on science makes it especially important for philosophy of science to account for science as it is actually conducted. The more science is recognized to be the activity of limited human beings, the more its characteristics can be expected to diverge from an idealized rational reconstruction or a hypothesized, perfect end state. Indeed, there is no reason to think that our actual scientific practice and products bear any particularly strong resemblance to a science philosophers would hold to be ideal or to an end state that philosophers would hold to be perfect. As Catherine Elgin (2010) puts the point, Science is a human achievement—a product of human endeavor. As such, it is ineluctably connected to the ways we access the world (446). Instead of finding a way to fit our actual scientific practices into a predetermined mold based on antecedent commitments or expectations of science’s goals and end point, philosophy of science must take its lead from observations about how science produced by real human beings finds success and what kind of success this is. This is the tack I take throughout this book. Such an approach to philosophy of science is a natural outgrowth of a naturalized epistemology, where our philosophical theories of knowledge are informed by how science in fact proceeds. In this case, the idea is that our philosophical theories of science itself should reflect how science in fact proceeds.

One might have the concern that, if philosophy of science aims to account for actual scientific practice—or merely aims to account for actual scientific practice—this would threaten to eliminate any grounds for making normative claims about science. There is, after all, a traditional dichotomy with normative philosophical claims about science on one side and descriptive claims of sociology and history of science on the other.³ It may seem that providing an account of actual scientific practice lands us on the wrong side of this divide to make any normative claims about science, any prescriptions for how science should proceed. Yet we require a normative account of science in order to articulate epistemically meaningful standards for scientific practice and to distinguish between successful and unsuccessful scientific results in an epistemically relevant way. Limiting ourselves to purely descriptive claims about science would, it seems, undermine philosophy’s ability to adjudicate debates about the proper aims, scope, and success of science. It would also in effect eliminate philosophy of science as an enterprise distinct from sociology and history of science.

Happily, aiming to account for science as it is actually practiced does not threaten the possibility of providing a normative account of science. It simply must be a normative account of our actual science, as it has in fact grown up, complete with accidental features and limitations due to particular human characteristics. Such an account should be guided by exemplars that are recognized to be our best scientific practices and products, as well as by the standards scientists themselves explicitly embrace or implicitly adopt. Consider an example of a philosophical position that is at once normative and motivated by ways in which our actual scientific practice deviates from what might be taken to be the ideal. Various feminist philosophers of science have argued that our science fails to satisfy traditional conceptions of objectivity, such as immunity from the influence of scientists’ values. Some of these philosophers then use insights about this failure to help motivate an alternative conception of objectivity that they take to be consistent with actual scientific practice. Such revised accounts of objectivity are normative insofar as they outline conditions required to achieve objectivity and place value on that achievement. Consider a specific example of this. Helen Longino (1990, 2001) develops an account of how objectivity emerges from science practiced in certain social conditions, in particular, by a diverse community of scientists with sufficient critical interaction. In her view, this enables the influence of values to be controlled, even though it cannot be eliminated. This may or may not be the right account of objectivity, but it is clearly a normative account. If it is correct, this account has straightforward consequences for how our science should be practiced and what such a science can accomplish. It gives a prescription for good science. Yet subscribing to this normative position requires acknowledging the limitations of our actual science—in this case, how our science falls short of an ideal objectivity. The general point is that normative claims can be made about the actual science that we humans have developed; they simply must be nuanced enough to speak to that enterprise in particular.

There is another concern one might have about a philosophy of science that aims to account only for the actual scientific practices of limited human beings. We might in fact be stuck with our particular science, warts and all, but this does not in itself eliminate any hope of providing a philosophical account of the features any science must possess. Such an account would indicate the features that must be possessed by any empirical scientific endeavor, pursued by any beings, in order for that endeavor to be epistemically sound. I said above that there is no reason to think that our science bears a resemblance to any such ideal science, but I have not so far demonstrated that it does not. Nonetheless, I suspect that the interesting generalizations that can be made about science—including generalizations with normative force—are by and large about our human science in particular. The basis for this view is the observation of how deep and thorough the mark of human characteristics is on our science. This can be established by careful attending to our scientific practices, as well as to the epistemic value to which those practices might lay claim.

I thus will move forward with an investigation of our limited, human science in particular with a provisional commitment to this approach. I hope that the nature of my conclusions in this book demonstrates the value of this focus. It may be enlightening to explore the features possessed by any logically possible empirical science and to consider the relationship the features of our science bear to those. But even that project requires a nuanced understanding of the particularities of the science that has been produced by humans (Adrian Currie, in conversation). And I anticipate that many of the significant generalizations to be made about science are distinctive to our science, a science created by humans and tailored to human needs, interests, circumstances, and psychological characteristics. The degree to which other creatures—real or imagined—have a science that mirrors our own depends on the degree to which their needs, interests, circumstances, and psychological characteristics match our own.

The aim of accounting for actual scientific practice seems to be broadly accepted in contemporary philosophy of science. According to de Regt and Dieks (2005), Nowadays few philosophers of science will contest that they should take account of scientific practice, both past and present. Any general characteristic of actual scientific activity is in principle relevant to the philosophical analysis of science (139). Similarly, in the introduction to their collection, Chao et al. (2013) state their view that what really matters to philosophers of science, and what philosophical discussions should be based on, is what scientists actually do and how they do it (1). Consider just a few examples of philosophers explicitly engaging in this approach. In the preface of his book, Suppes (2002) stresses the importance that empirical details have for his account of science. Strevens (2008b) is clear that he intends to provide an account of actual explanatory practice, that is, what kinds of explanations we give and why we give them (37). Longino (2001) says that her project is to account for living science, produced by real, empirical subjects, and she adds that this requires acknowledging that scientific knowledge cannot be fully understood apart from its deployments in particular material, intellectual, and social contexts (9).

This emerging philosophical focus on engaging with science as it is actually practiced may in part be responsible for the preponderance of work in current philosophy of science that is intended to generalize only to particular fields of science or even more specific areas of scientific inquiry. For example, the question of reductionism or antireductionism has been investigated as it applies to molecular and classical genetics in particular (Kitcher 1984; Waters 1990). Explanatory strategies have been articulated that are specific to biology (Sterelny 1996) and neuroscience (Machamer et al. 2000). Hierarchical levels of organization have been assessed within physics (Rueger and McGivern 2010) and within macroecology (Potochnik and McGill 2012). Philosophers of science also increasingly engage in debates endemic to particular fields. For example, philosophers of biology have actively participated in debates among biologists about adaptationism (Orzack and Sober 1994), levels of selection (Okasha 2006), and species and phylogenetics (Hull 1978).

Projects in philosophy of science that are tailored to specific scientific fields and subfields are indisputably valuable, but they do not exhaust the possible projects in philosophy of science. One might be tempted to move from a commitment to accounting for actual scientific practice, and the observation that different fields of science are very different in their aims and methods, to the conclusion that any philosophical views about science must be specific to individual fields or subfields of science. I think this would be a mistake. I believe there are still informative generalizations to be made about all of science. These are likely not universal generalizations about methodology, aim, relationship to society, and so on. There is no single, unified scientific method, and the various aims of science include, at least, explanation, prediction, retrodiction, confirmation, generalization, and basis for policy or action. These and other variable features of science are numerous and significant. Yet I suspect that there are also informative generalizations that transcend field boundaries and that apply to a range of scientific projects in other ways quite different from one another (see Currie 2015). This too I expect to be corroborated by the views developed in this book, which are largely intended to be informative generalizations about science as a whole. I return to this point explicitly in the final chapter.

The commitment to account for actual scientific practice is unevenly pursued, even by philosophers who profess it. Here is an example. John Dupré (1993) advocates revisions to classical philosophical positions inspired by actual scientific practice, and one of the positions he criticizes is reductionism, taken as the idea that the understanding of a whole is accomplished by understanding its individual parts. In the process of developing his criticism of reductionism, Dupré declares that the scientific field of population genetics is worthless. Population genetics is a subfield of biology that investigates how evolutionary processes, including natural selection, influence the distribution of genetic variants in a population. Dupré is critical of how population genetics relates the process of natural selection to genes. This is done by estimating the effects different genetic variants have on fitness—that is, how organisms with the different genetic variants fare in survival and reproduction—and then representing these fitness effects with parameters known as genes’ selection coefficients. According to Dupré, this procedure is unacceptable, for it involves the use of misbegotten mongrel concepts of genetic fitnesses (138). He objects to considering the fitness of individual genes because this requires applying a higher-level property (fitness) to lower-level entities (genes). For this reason, in Dupré’s view, population genetics as a whole is a misguided enterprise.

Dupré (1993) rightly points out that one is not compelled to grant credence to a body of research simply because of the sociological fact that the