Knowledge and naturalism DALLAS WILLARD
Naturalism: A Critical Analysis presents a rigorous analysis and critique of the major varieties of contemporary philosophical naturalism. It advocates the thesis that contemporay naturalism should be abandoned in light of the serious difficulties raised against it. The contributors draw on a wide range of topics including: epistemology, philosophy of science, value theory to basic analytic ontology, philosophy of mind and agency, and natural theology.
For some time now, philosophical naturalism has played a dominant intellectual and sociological role in academe. Currently, throughout a large sector of professional philosophy, philosophical naturalism determines the way philosophy is to be done, what counts as good methodology, and what is taken to be a worthwhile topic for study. This book provides a thorough and searching critique of philosophical naturalism which is squarely within the analytic tradition and which avails itself of the new and exciting literature in philosophy of religion that has burst onto the scene in the last decade.
Naturalism: A Critical Analysis accomplishes three things. First, it argues that naturalism fails to deal adequately with a number of desiderata; second, it shows that to the degree that naturalism is a consistent paradigm and is ‘located’ within the physicalist story, it should be a form of strict physicalism; and third, it provides an account of the contemporary resurgence of philosophical theism by advertising to evidentiary considerations in the natural world itself (e.g. Big Bang cosmology) which serve as signposts of transcendence.
William Lane Craig is Research Professor of Philosophy and J. P. Moreland is Professor of Philosophy both at Biola University, La Mirada, California.
Routledge Studies in Twentieth-century Philosophy 1 The Story of Analytic Philosophy
Plot and heroes
Edited by Anat Biletzki and Anat Matar
2 Donald Davidson
Truth, meaning and knowledge
Edited by Urszula M. Zeglén
3 Philosophy and Ordinary Language
The bent and genuis of our tongue
Oswald Hanfling
4 The Subject in Question
Sartre’s critique of Husserl in The Transcendence of the Ego Stephen Priest
5 Naturalism
A critical analysis
Edited by William Lane Craig and J. P. Moreland
Naturalism
A critical analysis
Edited by
William Lane Craig and
J. P. Moreland
London and New York
by Routledge
11 New Fetter Lane, London EC4P 4EE
Simultaneously published in the USA and Canada
by Routledge
29 West 35th Street, New York, NY 10001
Routledge is an imprint of the Taylor & Francis Group This edition published in the Taylor & Francis e-Library, 2001.
Editorial selection and material © 2000 Wiliam Lane Craig and J. P. Moreland Individual chapters © individual contributors
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library Library of Congress Cataloguing in Publication Data Naturalism: a critical analysis/edited by William Lane Craig and J. P. Moreland.
p. cm.
Includes bibliographical references and index.
1. Naturalism. I. Craig, William Lane. II. Moreland, James Porter, 1948–
B828.2.N367 2000
146—dc21
00–025324
ISBN 0–415–23524–3 (Print Edition)
ISBN 0-203-18613-3 Master e-book ISBN
ISBN 0-203-18736-9 (Glassbook Format)
List of figures
vii
List of contributors
ix
Preface
xi
PART 1
Epistemology
1
1 Farewell to philosophical naturalism
3
PAUL K. MOSER AND DAVID YANDELL
2 Knowledge and naturalism
24
DALLAS WILLARD
3 The incompatibility of naturalism and scientific
realism
49
ROBERT C. KOONS
PART 2
Ontology
65
4 Naturalism and the ontological status of properties 67
J.P. MORELAND
5 Naturalism and material objects
110
MICHAEL REA
6 Naturalism and the mind
133
CHARLES TALIAFERRO
7 Naturalism and libertarian agency
156
STEWART GOETZ
PART 3
Value theory
187
8 Naturalism and morality
189
JOHN E. HARE
PART 4
Natural theology
213
9 Naturalism and cosmology
215
WILLIAM LANE CRAIG
10 Naturalism and design
253
WILLIAM DEMBSKI
Index
281
9.1
Conical representation of Standard model space-time
218
9.2
Steady State model
219
9.3
Oscillating model
221
9.4
Oscillating model with entropy increase
222
9.5
Vacuum Fluctuation models
224
9.6
Chaotic Inflationary model
225
9.7
Quantum Gravity model
227
9.8
Contrast between the universe as we know it (assumed for convenience to be closed) with a more probable universe 230
9.9
A self-creating universe
233
William Lane Craig is Research Professor of Philosophy at Talbot School of Theology, Biola University, La Mirada, California. He is the author of a dozen books, including The Kalam Cosmological Argument and Theism, Atheism, and Big Bang Cosmology (with Quentin Smith), as well as over a hundred articles in professional journals such as The Journal of Philosophy, Philosophy, Philosophical Studies, Australasian Journal of Philosophy, The British Journal for the Philosophy of Science, Erkenntnis, International Studies in the Philosophy of Science and Philosophia naturalis.
William Dembski is a Fellow of Discovery Institute’s Center for the Renewal of Science and Culture, an Associate Research Professor at Baylor University, and director of Baylor’s Michael Polanyi Center, a research group focusing on complexity, information, design, and the conceptual foundations of science. He has authored and edited several books including The Design Inference (Cambridge University Press, 1998). He has also authored over forty articles in such varied journals as Journal of Theoretical Probability, Nous and Scottish Journal of Theology.
Stewart Goetz is Associate Professor of Philosophy at Ursinus College in Collegeville, PA. He specializes in the philosophy of action and the philosophy of mind, and has authored a number of articles on issues in these areas in journals such as American Philosophical Quarterly, Faith and Philosophy, Mind, Philosophical Studies and Philosophy and Phenomenological Research.
John E. Hare is Professor of Philosophy at Calvin College in Grand Rapids, Michigan.
He has taught at Lehigh University and was a visiting professor at the University of Michigan. His articles have appeared in a number of journals including Ancient Philosophy, International Journal of Applied Ethics, and Phronesis. He is also the author of The Moral Gap (Clarendon, 1996).
Robert C. Koons is an Associate Professor of Philosophy at the University of Texas at Austin. He is the author of Paradoxes of Belief and Strategic Rationality (Cambridge University Press, 1992), “A New Look at the Cosmological Argument”, American
Philosophical Quarterly (1997), and Realism Regained: An Exact Theory of Causation, Teleology and the Mind (Oxford University Press, 2000).
J. P. Moreland is professor of philosophy at Talbot School of Theology, Biola University, La Mirada, California. He has authored, co-authored, or edited twelve books including Does God Exist? (Prometheus, 1993). He has also written over forty articles in journals that include Philosophy and Phenomenological Research, American Philosophical Quarterly, Southern Journal of Philosophy, MetaPhilosophy, Religious Studies and Australasian Journal of Philosophy.
Paul K. Moser is Professor and Chairperson of Philosophy at Loyola University of Chicago. He is the author of Philosophy after Objectivity (Oxford University Press, 1993) and Knowledge and Evidence (Cambridge University Press, 1989), and co-author of The Theory of Knowledge (Oxford University Press, 1998). He is also General Editor of the book series, Routledge Contemporary Introductions to Philosophy.
Michael Rea received his Ph.D. in 1996 from the University of Notre Dame. He is currently Assistant Professor of Philosophy at the University of Delaware. His articles on the metaphysics of material objects have appeared in The Philosophical Review, Philosophy and Phenomenological Research, Philosophical Studies and American Philosophical Quarterly. He is the editor of Material Constitution: A Reader (Lanham, MD: Rowman & Littlefield, 1997).
Charles Taliaferro is Professor of Philosophy at St Olaf College, Northfield, Minnesota.
He is the author of numerous articles in professional journals. He is also the author of Consciousness and the Mind of God (Cambridge, 1994), Contemporary Philosophy of Religion (Blackwell, 1998), and co-editor of A Companion to Philosophy of Religion (co-edited with Philip L. Quinn) (Blackwell, 1997).
Dallas Willard is Professor of Philosophy at the University of Southern California, where he has been since 1965. He began his teaching career at the University of Wisconsin (Madison), and has held visiting appointments at the University of California at Los Angeles and the University of Colorado. He is the author of Logic and the Objectivity of Knowledge and of numerous articles, including, recently,
“Knowledge,” in The Cambridge Companion to Husserl, “Who Needs Brentano? The Wasteland of Philosophy Without Its Past,” The Brentano Puzzle, and “Degradation of Logical Form,” in Axiomathes (1997).
David Yandell is Associate Professor of Philosophy at Loyola University of Chicago.
He is the author of a number of recent articles in the history of philosophy which have appeared in such journals as British Journal for the History of Philosophy and History of Philosophy Quarterly.
Throughout much of contemporary intellectual culture, the world-view of naturalism is the accepted creed. David Papineau boldly proclaims that “. . . nearly everybody nowadays wants to be a ‘naturalist’ . . .”1 Along similar lines, Jaegwon Kim states that most philosophers today have accepted naturalism – and more specifically, physicalism, as the best contemporary version of naturalism – as a basic commitment. Such a commitment, says Kim, involves recognizing that naturalism “is imperialistic; it demands ‘full coverage’ . . . and exacts a terribly high ontological price.”2 This volume aims to challenge this prevailing world-view by making clear the price which must be paid to accept it.
Both in its various historical and contemporary manifestations, it would be hard to state naturalism to everyone’s satisfaction. Thus, we have left to each of the contributors to the present volume the task of specifying the specific version of naturalism within his purview. Still, a cluster of positions and a set of philosophical commitments may fairly be associated with contemporary philosophical naturalism (hereafter, simply naturalism) as it has come to be widely understood by its friends and foes in the current philosophical setting.
Minimally, as a negative thesis naturalism implies that theism is false. Roughly, scientific naturalism is the view that the spatio-temporal universe of entities studied by the physical sciences is all there is. Positively, naturalism usually includes (1) different aspects of a naturalist epistemic attitude (for example, a rejection of so-called “first philosophy” along with an acceptance of either weak or strong scientism); (2) a Grand Story which amounts to an etiological account of how all entities whatsoever have come to be told in terms of an event causal story described in natural scientific terms with a central role given to the atomic theory of matter and evolutionary biology; and (3) a general ontology in which the only entities allowed are those that bear a relevant similarity to those thought to characterize a completed form of physics.
For most naturalists, the ordering of these three ingredients is important. Frequently, the naturalist epistemic attitude serves as justification for the naturalist etiol-
ogy which, in turn, helps to justify the naturalist’s ontological commitment. Moreover, naturalism seems to require a coherence among the postulates of these three different areas of the naturalistic turn. For example, there should be a coherence among third-person scientific ways of knowing, a physical, evolutionary account of how our sensory and cognitive processes came to be, and an ontological analysis of those processes themselves. Any entities that are taken to exist should bear a relevant similarity to entities that characterize our best physical theories, their coming-to-be should be intelligible in light of the naturalist causal story, and they should be knowable by scientific means.
For our purposes, it is important to say a bit more about naturalist ontological commitments. A good place to start is with what Frank Jackson calls the location problem.3 According to Jackson, given that naturalists are committed to a fairly widely accepted physical story about how things came to be and what they are, the location problem is the task of locating or finding a place for some entity (for example, semantic contents, mind, agency) in that story. Without passing judgment on Jackson’s own analysis of the location problem, we maintain for at least two reasons that the naturalist ontology should, in some sense, be located in light of the naturalist etiology and epistemic attitude. First, if this is not done, then the naturalist ontology may well amount to a mere assertion and a controversial one at that, given the proliferation in recent years of robust alternatives to naturalism, e.g. theism. Second, the naturalist claims to have epistemic, explanatory, and methodological superiority on his or her side, especially when naturalist positions are understood to have the authority of science behind them. In order to justify this assertion of superiority, the naturalist ontology should be located in the naturalist etiology and epistemology.
Not all naturalists agree with this, though most would accept a location requirement of some sort, not altogether unlike the one mentioned above. More importantly, however, not all naturalists will agree regarding the resulting naturalist ontology that is so located. Here we may offer a pair of useful, if not exhaustive, distinctions regarding a naturalist ontology – global vs. local naturalism and weak vs. strong naturalism. Roughly, global naturalism is the view that the spatio-temporal universe of natural entities studied by science is all there is. Global naturalists (for example, Wilfrid Sellars, D. M. Armstrong) reject abstract objects of any kind, including sets, numbers, propositions, and properties/relations in general taken in the traditional sense as non-spatio-temporal abstract entities. Local naturalists (for example, Jeffrey Poland, Collin McGinn) either accept or are indifferent towards abstract objects but they insist that the spatio-temporal universe consists only of entities studied by the natural sciences. Local naturalists reject Cartesian souls, Aristotelian entelechies, and so forth.
We can also distinguish strong and weak naturalists. Strong naturalists (for example, David Papineau) accept strict physicalism for the natural world, while weak
naturalists (for example, John Searle) embrace various emergent entities. According to strong naturalists, naturalism constrains properties and relations just as much as it does particulars, and all entities whatsoever must be susceptible to exhaustive description in the language of an ideal physics (supplemented, perhaps, by chemistry and biology). What is clearly eschewed by strong naturalists is the need for psychological categories. Such categories may be useful for a number of purposes, but they do not carve the world at the joints. By contrast, weak naturalists have offered various forms of supervenience in order to cash out the proper requirements of naturalism.
Whatever the form, the goal is to make supervenient entities dependent upon a subvenient physical base.
For someone to be a consistent naturalist of whatever form, he or she must hold that the entity which emerges or comes-to-be (as naturalists understand that notion) (1) fits solely within a naturalist ontology (that is, it bears a relevant resemblance to paradigm case natural entities of physics in its being and behavior), and (2) can, at least in principle, be given a naturalist etiological explanation in keeping with the naturalist epistemic attitude and Grand Story if it is to be explained at all. The naturalist ontology must be selected in such a way that it will cohere well with the naturalist epistemology and Grand Story in order for the naturalist to be justified in asserting the scientific explanatory superiority of naturalism over rival world-views.
In light of these considerations, critics and advocates of naturalism disagree about what ontology best expresses the naturalist world-view. Naturalist critic Reinhardt Grossmann has argued that the naturalist ontology should be taken to imply a rejection of abstract objects (such as sets, properties, propositions). Many naturalists accept this implication and either adopt some version of nominalism (for example, Hartry Field) or try to work out an acceptable form of naturalism that includes a realist construal of properties without taking them as abstract objects (for example, D. M. Armstrong). Likewise, many naturalists (for example, David Papineau) take naturalism most reasonably to require strict physicalism (all individuals, events, properties, relations, and so on are completely physical entities), though others (for example, John Searle) accept the existence of irreducible, genuinely mental, emergent or supervenient properties or events. Virtually all naturalists (for example, Daniel Dennett, John Bishop) eschew libertarian agency though naturalist William Rowe and anti-naturalist Tim O’Connor believe that emergent libertarian agents are consistent with an acceptable form of naturalism. The vast majority of naturalists eschew literal, irreducible teleology and substance dualism. Naturalists also have difficulty allowing for the existence of a number of features of the moral point of view, and they employ various strategies to handle problems in value theory.
All the contributors to this volume reject naturalism. Accordingly, we have a three-fold purpose for the essays that follow.
First, the volume contains arguments that, if successful, justify the claim that naturalism fails to deal adequately with a number of desiderata (for example, its fails to provide an adequate account of acts of knowing, and it similarly fails to offer an adequate ground for or analysis of the preconditions necessary for scientific realism).
Second, we want to show that to the degree that naturalism is a consistent paradigm and its ontology is “located” within the physicalist story, to that same degree it should be a form of global, strong naturalism, as Papineau and others have correctly seen. Therefore, it should include rejection of abstract objects, irreducible mental entities, libertarian agency, and first philosophy. Those versions of naturalism that countenance abstract objects and knowledge thereof, supervenient mental entities, emergent libertarian agents, and so forth do so at the price of appearing ad hoc, of giving up the claim that science and science alone is adequate to explain (or is, superior to non-naturalistic attempts to explain) everything in one’s ontology, of weakening and trivializing what it means to “locate” these entities, in contrast with naturalist successes in locating paradigm case macro-entities such as solidity.
As an illustration of this last point, there is a serious debate about naturalist employments of supervenience on at least two fronts: 1) Some question whether the supervenience relation itself (or states of affairs that contain supervenient entities) is an existent which a naturalist should countenance; (2) Others charge that supervenience is just a label for the problem naturalists must solve and not a solution.
It seems that naturalism faces a dilemma: either naturalism involves an epistemic attitude and etiology that express strong versions of scientism, in which case naturalism suffers from some obvious defects (no account of proper functioning, denial of consciousness) or else it must weaken its ontology to adopt certain entities (abstract objects, mental properties), in which case it loses the unity of science and its right to claim a strong naturalist epistemic attitude, explanatory hegemony, and an adequate etiological account of the coming-to-be of everything. This dilemma is part of the widely discussed explanatory gap or location problem for certain entities in the naturalist ontology.
Third, we want to provide a sampling of contemporary work in the burgeoning field of philosophy of religion which is pertinent to the debate over naturalism. In a recent retrospect of the twentieth century, Tyler Burge has remarked that “the central event” in philosophy during the last half century has been “the downfall of positivism and the re-opening of discussion of virtually all the traditional problems in philosophy.”4 Nowhere has this been more evident than in the field of philosophy of religion. Whereas philosophers of religion – what few of them there were – struggled during the fifties and sixties to defend the very meaningfulness of theological discourse in the face of positivist critiques, today’s philosophical theologians find themselves in the midst of a renaissance of interest in religious perspectives on epistemol-
ogy, metaphysics, and science. All of the traditional arguments of natural theology, such as the ontological, cosmological, teleological and axiological arguments for God’s existence, are matters of renewed discussion, along with theistic approaches to typical problems of philosophy normally pursued with a naturalistic methodology.
We include in the present volume articles which advert to evidentiary considerations in the natural world itself, for example, contemporary Big Bang cosmology, indica-tions of intelligent design discernible in the bio-world, and the apprehension of value through moral experience, which serve, as it were, as “signposts of transcendence”
pointing beyond the natural world to its transcendent ground.
Finally, we wish to express our indebtedness to Lisa Vasquez, Nate King, Mercy Hung, Mike Austin, Richard Graziano, and Robert Garcia for their tireless efforts in preparing the typescript for publication.
Notes
1 David Papineau, Philosophical Naturalism (Oxford: Blackwell Publishers, 1993), 1.
2 Jaegwon Kim, “Mental Causation and Two Conceptions of Mental Properties,”
unpublished paper delivered at the American Philosophical Association Eastern Division Meeting, Atlanta, Georgia, December 27–30, 1993, pp. 22–23.
3 Frank Jackson, From Metaphysics to Ethics (Oxford: Clarendon Press, 1998), 1–5.
4 Tyler Burge, “Philosophy of Language and Mind,” Philosophical Review 101 (1992): 49.
Epistemology
naturalism1
Paul K. Moser and David Yandell
Many philosophers have held, in the tradition of Plato and Aristotle, that there are uniquely philosophical, non-empirical methods of inquiry and that there are things whose investigation is reliably conducted via such methods. Rejection of any such methods or any such objects other than those available to sense experience or scientific methods traces back to the monism of the presocratic philosophers. Call any position entailing such a rejection “philosophical naturalism.” (At times we will simply call such a position “naturalism,” for short.)
Philosophical monism in general, including idealism as well as materialism, demands a single standard in metaphysics or epistemology, contrary to pluralism in metaphysics or epistemology. Explanation in philosophy and science is inherently unifying, subsuming a multiplicity of phenomena under classificatory unity – for example, under a unifying cause in the case of causal explanation. Explanation contributes unity and thus organization to what may otherwise appear as mere diversity. Proponents of monism, however, must attend to the risk of neglecting genuine data and truths resistant to a monistic explanatory scheme. What monism gains by unification of multiplicity in data may be lost by neglect of genuine recalcitrant data. Explanatory unity may be a virtue, but it will be virtuous only if pertinent truths and data are not excluded for the sake of theoretical simplicity. Since naturalism is, as we will see, monistic in its implications for metaphysics or epistemology, it faces the threat of neglecting genuine recalcitrant data.
Contemporary motivation for naturalism derives largely from the success of the empirical sciences in representing their disciplinary phenomena as uniformly natural with respect to their being explained (in a sense of “natural” to be clarified below).
Philosophical dispute concerning naturalism ranges over many issues, but at bottom it concerns the nature of philosophy itself. A basic question is whether there is a legitimate form of philosophical procedure, often called “first philosophy,” that has ontological authority but employs methods “prior to” or at least not based on sense experience or the empirical sciences. In particular, can a philosopher operating
4 Paul K. Moser and David Yandell without reliance on sensation or the empirical sciences legitimately engage in inquiry that posits real objects or at least yields genuine truths? Naturalists say no; antinaturalists, yes.
We will pursue not the history of disputes over first philosophy but rather the tenability of naturalism regarding first philosophy. In particular, we seek answers to the following questions: What motivates and supposedly warrants methodological and ontological naturalism? Does naturalism commit one to methodological or ontic presuppositions inconsistent with naturalism? If the methodology and ontology of naturalism aren’t grounded in traditional first philosophy, can they be grounded at all, and if so, how? By attending to these questions and proposing a dilemma for naturalism, we’ll contend that naturalism regarding first philosophy is unwarranted at best and self-refuting at worst. We will conclude, therefore, that naturalism is rationally unacceptable.
1. Naturalism: ontological and methodological
Ontological naturalism takes various forms. We will understand such naturalism in terms of this core view: every real entity either consists of or is somehow ontically grounded in the objects countenanced by the hypothetically completed empirical sciences (that is, in the objects of a natural ontology). As a general position about what actually exists, ontological naturalism is a metaphysical view, and not (just) an epistemological view about how we know or just a semantic view about the meaning of terms.
Ontological naturalists typically, but not always, endorse some kind of materialism (or, physicalism) about reality. So, two prominent construals of “material” are noteworthy. First, something is material (or, physical), according to some philosophers, if and only if it is extended in space. This approach portrays materialism as the view that everything that actually exists is extended in space, and thus that nothing non-spatial exists. This portrayal of materialism seems straightforward, but the relevant notion of “spatial extension” evidently depends on the very notion of
“material” in need of elucidation. If there is such dependence, conceptual circularity will trouble the proposed characterization of materialism.
The apparent problem is that the notion of spatial extension is actually the notion of something being extended in physical space, or being physically extended. It is coherently conceivable that something has temporal extension, in virtue of extending over time, even though that thing is not extended in physical space. It is not self-contradictory, in other words, to hold that something is temporally extended but is not a body. The proposed characterization of materialism should thus be qualified to talk of physical space or physical extension. In that case, however, the problem of conceptual circularity is transparent. Even if there is no circularity here, the notion of spatial extension seems too close to the notion of being material to offer significant
Farewell to philosophical naturalism 5
clarification. Minimally, we need an elucidation of spatial extension, if talk of such extension is to elucidate talk of what is material.
The second prominent construal of “material” invokes the empirical sciences, as follows: the language of the empirical sciences determines what it is to be material, or physical. A predicate (such as “is a quark”) signifies a physical item, on this view, if and only if the empirical sciences rely on that predicate in their formulations. Three noteworthy problems arise. First, the empirical sciences do not have an exhaustive list of truths about the physical world and are thus now incomplete; nor is it clear that they will ever have such an exhaustive list, given the complexity of the physical world. Such incompleteness allows for predicates that the empirical sciences do not rely on, but that nonetheless pick out something in the physical world. It is thus unclear that the empirical sciences have a monopoly on predicates signifying physical items. Second, the empirical sciences use mathematical predicates that evidently do not signify physical items. The predicate “is a member of a set,” for instance, does not signify a physical item. Third, invoking “the empirical sciences” raises the problem of specifying exactly when something is an empirical science. This is no small task, given (a) the actuality of such highly theoretical empirical sciences as cosmology and particle physics and (b) the inclusion of the social sciences as empirical sciences. It would not help now to portray an empirical science as a discipline that investigates, in a special way, only the material, or physical, world, because we are after elucidation of the relevant notion of ‘material’.
We now need elucidation of what it is to be material, or physical. Is social psychology, for example, an empirical science? Social psychologists often use mentalistic vocabulary (for example, talk of beliefs, desires, intentions, fears, and hopes) that does not obviously signify material things. So, we need clarification of the sense in which social psychology is materialist, if it is an empirical science. In general, then, ontological naturalists who are materialists would do well to characterize an empirical science without relying on talk of what is material, or physical.2
Ontological naturalists who are materialists oppose ontological dualism (the view that there are two irreducible kinds of things that actually exist) as well as other forms of ontological pluralism. The most influential kind of ontological dualism comes from Descartes, who affirmed that there are mental substances (particularly, thinking individuals) as well as physical substances (particularly, matter), and that mental substances do not depend ontically on physical things.3 Proponents of Cartesian dualism must specify the sense in which mental substances are ontically independent of physical things. Cartesians allow for relations of causal dependence between the mental and the material, but they deny that mental things are part of the material realm. Being a mental thing, according to Cartesians, is different in kind from being a material thing. If there’s a coherently conceivable distinction between minds and material bodies, it follows that materialism, understood as entailing mind–body identity, is not conceptually, or analytically, true – that is, true just in virtue of the
6 Paul K. Moser and David Yandell meanings of the terms “mind” and “body.”
Any argument from a coherently conceivable distinction between mind and material body to an actual ontic distinction between the two may face trouble. An actual ontic distinction between minds and material bodies entails that actual minds aren’t material bodies. This goes beyond a distinction between concepts or definitions; it is a difference between things. Even if it is not conceptually true that minds and material bodies are identical, according to many philosophers, it may still be contingently true that minds are material bodies. An identity relation between minds and material bodies may hold as an actual matter of fact, according to these philosophers, even if it does not obtain as a matter of meaning, definition, or necessity.
Other philosophers deny that the notion of contingent identity is coherent and accept an inference from the coherent conceivability of mind–body distinction to the real distinction of mind and body. We need not endorse either position here. According to most contemporary materialists, the coherent conceivability of the falsity of materialism does not challenge the actual truth of that position. The contingent truth of materialism is adequate for most materialists.
Ontological naturalists who are materialists reject Descartes’s ontological dualism, in particular, its implication that a human mind is composed of an immaterial substance different in kind from material bodies. They also reject Descartes’s view that some mental properties are actually exemplified by certain immaterial things, things not spatially extended. Materialists are uniformly monistic in their view that all actually existing individuals are material, and in their view that only material things exemplify mental properties. These two fundamental materialist views, according to most contemporary materialists, are justifiable empirically, not a priori. The evidence for these views, in other words, must come from considerations dependent on experience. Neither reason nor definition alone can produce evidence for such views.
We should distinguish the following two views about the relation between ontological naturalism and materialism: (a) ontological naturalism is logically neutral regarding materialism, and thus is logically compatible with ontological dualism; (b) ontological naturalism logically presupposes materialism, and thus is logically incompatible with ontological dualism. Given (a), a defense of materialism on the basis of ontological naturalism should appeal to supporting evidence beyond naturalism, presumably empirical evidence. Given (b), an appeal to ontological naturalism in defense of materialism would be question-begging, as materialism would then be a logical prerequisite of ontological naturalism. Most contemporary ontological naturalists apparently prefer (a) to (b).
W. V. Quine evidently would oppose (b) as follows:
. . . nowadays the overwhelming purposes of the science game are technology and understanding. . . . The science game is not committed to the physical, whatever
Farewell to philosophical naturalism 7
that means. . . . Even telepathy and clairvoyance are scientific options, however moribund. It would take some extraordinary evidence to enliven them, but, if that were to happen, then empiricism itself – the crowning norm . . . of naturalized epistemology – would go by the board. For remember that that norm, and naturalized epistemology itself, are integral to science, and science is fallible and corrigible.4
Quine thus suggests that naturalism is not logically committed to materialism even though naturalism does give primacy to fallible science. Quine’s sense of “scientific”
in this passage invites trouble. It plainly is not restricted to empirical science in the usual sense; for it implies that empiricism could “go by the board” as the result of our discovering evidence of parapsychological phenomena, but that this would not involve moving outside of “science.”
A problem now emerges for Quinean naturalists. Quinean naturalists endorse a strong fallibilism, with good reason given the enormous shifts in the scientific conception of reality that have occurred within merely the last few generations. Just as a commitment to absolute space as the basis for material reality led many to oppose theories of space-time relativity, any a priori commitment about the nature of science could readily bring Quinean naturalists into conflict with the best contemporary science. Indeed, Quinean constraints on the ontology of the sciences are so loose that they permit the postulation of parapsychological processes, and Quinean constraints on methodology are so loose that they permit the jettisoning of empiricism. It seems, then, that the sciences are so unconstrained that they permit even non-empirical methodologies to ground belief in the immaterial. So, it is altogether unclear how non-empirical methods of the sort employed in first philosophy are to be ruled out at all by Quinean naturalism. After all, if what characterizes the sciences is the pursuit of understanding, and there is no requirement that the objects of understanding all fall within the range of sense-experience and scientific observation, then there is no result that scientific pursuit of understanding is to be constrained at all by empiricism.
The problem for Quinean naturalists stems from their reliance, for opposition to first philosophy, on a conception of science they deem “fallible and corrigible.” In particular, they offer no reason to think that science, as merely a pursuit of understanding, should not include first philosophy. Indeed, given (1) that Quinean naturalists take any pursuit of understanding as science and (2) that at least one intelligible philosophical question (say, a question about the consistency of a position or about valid inference from a position) is not empirically resolvable but is nonetheless a pursuit of understanding, an inclusion of first philosophy by science seems required. The only way out for Quinean naturalists would be to deny (2), but given their unwillingness to commit themselves, as a matter of principle, to empiricism in epistemology, we see no ground for their denying (2).
8 Paul K. Moser and David Yandell Given the internal problem of Quinean naturalism, we will relate naturalism to empirical science, in keeping with the fact that naturalists do uniformly limit good scientific explanation to empirically relevant explanation. Without digressing to the complex task of supplying a precise account of empirical relevance, we will construe
“empirical science” broadly to encompass any scientific discipline housed in a typical college of natural or social sciences. So, astronomy is in, and astrology out; psychology is in, and parapsychology out. Anthropology is in, but philosophy and theology are out, given their regular place in humanities or arts, owing to their being insufficiently empirical to be empirical sciences. Our criterion is rough, but it enables us to take steps nonetheless.
The previous core definition of ontological naturalism encompasses eliminative, reductive and non-reductive forms of ontological naturalism, which may be characterized as follows:
(i) Eliminative ontological naturalism: every real entity is capturable by the ontology of the hypothetically completed empirical sciences, and language independent of those sciences is eliminable from discourse without cognitive loss.
(ii) Non-eliminative reductive ontological naturalism: every real entity either is capturable by the ontology of the hypothetically completed empirical sciences or is reducible to something capturable by that ontology.
(iii) Non-eliminative non-reductive ontological naturalism: some real entities neither are capturable by the ontology of the hypothetically completed empirical sciences nor are reducible to anything capturable by that ontology, but all such entities supervene on entities capturable by that ontology.
Proponents of (i) have included W.V. Quine, Paul Churchland, and Daniel Dennett.5
J.J.C. Smart, among others, has suggested (ii).6 Donald Davidson and David Papineau, among many others, have suggested (iii).7
Views incompatible with positions (i)–(iii) are sometimes deemed naturalist. For example, some philosophers regard themselves as ontological naturalists even though they countenance the existence of non-physical abstract objects that neither consist of nor are somehow ontically grounded (by reduction or supervenience) in the objects of a natural ontology. They embrace ontological naturalism owing to their rejecting the existence of mental substances. Such a construal of ontological naturalism as a position in the philosophy of mind differs from the general ontological naturalism just identified. Nonetheless, many, if not most, philosophers rejecting Cartesian dualism in the philosophy of mind are also ontological naturalists in the core sense specified earlier.
Beyond ontology, the term “naturalism” picks out a range of views about the nature of legitimate inquiry that are logically independent of ontological naturalism. These views, falling under the general category of “methodological naturalism,” share this
Farewell to philosophical naturalism 9
core position: every legitimate method of acquiring knowledge consists of or is grounded in the hypothetically completed methods of the empirical sciences (that is, in natural methods). This is not ontological naturalism, as claims about methods of acquiring knowledge are not the same as claims just about what exists. In agreement with most naturalists, we understand talk of the “methods of the empirical sciences” broadly, to encompass the kind of “lay scientific” observation underlying ordinary perceptual beliefs.
Methodological naturalism manifests a kind of variety analogous to that of ontological naturalism. We may thus distinguish:
(iv) Eliminative methodological naturalism: all terms, including empirically disputed terms (for example, normative and intentional terms), employed in legitimate methods of acquiring knowledge are replaceable, without cognitive loss, by terms employed in the hypothetically completed methods of the empirical sciences.
(v) Non-eliminative reductive methodological naturalism: all terms, including empirically disputed terms, employed in legitimate methods of acquiring knowledge either are replaceable, without cognitive loss, by terms employed in the hypothetically completed methods of the empirical sciences or are reducible (by, for example, definition, mutual entailment, or entailment) to terms employed in those methods.
(vi) Non-eliminative non-reductive methodological naturalism: some empirically disputed terms employed in legitimate methods of acquiring knowledge neither are replaceable by terms employed in the hypothetically completed methods of the empirical sciences nor are reducible to those terms, but the referents of these terms supervene upon those of the terms employed in the hypothetically completed methods of the empirical sciences.
W.V. Quine has suggested (iv), and Jerry Fodor has suggested a version of (vi).8
Perhaps a reductionist like J.J.C. Smart would endorse (v).9 Talk of “empirically disputed” terms in (iv)–(vi) connotes terms deemed by some to be insufficiently empirical to figure in the empirical sciences. Normative terms and mentalistic terms are familiar instances of empirically disputed terms. In (iv)–(vi), the word
“knowledge” may be taken in the broadest sense, to include knowledge proper (however analyzed), justified or warranted belief, and even evidence for a belief. In other words, methodological naturalists claim that the hypothetically completed methods of the empirical sciences define or otherwise ground the only sorts of inquiry that yield knowledge, justified or warranted beliefs, and even evidence for a belief.
We do not have a simple recipe to identify the methods of the empirical sciences.
Perhaps the best we can do is to refer to those methods employed in our best physics, chemistry, biology, anthropology, sociology, and so on. The debate will turn, quite naturally, to what exactly is picked out by “and so on” or by talk of our “best” empirical
10 Paul K. Moser and David Yandell sciences. We take it for granted that the methods of the empirical sciences, whatever specifically they include, are designed to be reliable, that is, truth-conducive. We do not deny that the methods of the empirical sciences seek other cognitive virtues, such as fruitful explanation and predictive success. We simply acknowledge that those methods aim at least to be reliable, regardless of whether they actually succeed. The turbulent history of the empirical sciences illustrates that they do not always succeed.
That limitation, however, does not preclude intended reliability in the methods of the empirical sciences. Scientific methods aim for truth, even when they fail. We thus assume that every scientific method is epistemically relevant in that it aims for reliable conclusions. We turn now to some problems, including a troublesome dilemma, for naturalism.
2. A dilemma
Our dilemma will bear on positions (i)–(vi) inasmuch at it bears on the aforementioned core statements of naturalism satisfied by those positions, namely: Core ontological naturalism: every real entity either consists of or is somehow ontically grounded in the objects countenanced by the hypothetically completed empirical sciences (that is, in the objects of a natural ontology).
Core methodological naturalism: every legitimate method of acquiring knowledge consists of or is grounded in the hypothetically completed methods of the empirical sciences (that is, in natural methods).
These are core statements of ontological and methodological “scientism,” for they exalt the empirical sciences as the criterion for metaphysical and epistemological genuineness. They entail ontological and methodological monism in that they acknowledge the empirical sciences as the single standard for genuine metaphysics and epistemology. These core positions thus promise remarkable explanatory unity in metaphysics and epistemology, but is their promise trustworthy? For brevity, let us call the conjunction of these two positions Core Scientism, while allowing for talk of its ontological component and its methodological component.
Core Scientism is not itself a thesis offered by any empirical science. In particular, neither its ontological component nor its methodological component is a thesis of an empirical science. Neither component is represented in the empirical scientific work of either physics, chemistry, biology, anthropology, psychology, or any other natural or social empirical science. As a result, no research fundable by the National Science Foundation, for instance, offers Core Scientism as a scientific thesis. In contrast, the National Endowment for the Humanities would fund work centered on Core
Farewell to philosophical naturalism 11
Scientism. In addition, Core Scientism proposes a universality of scope for the empirical sciences that they themselves eschew. Individual sciences are marked out by the particular ranges of empirical data they seek to explain. Similarly, empirical science as a whole is marked out by its attempt to explain all empirical data and thus by the range of all empirical data. Given this constraint on empirical science, it would be surprising indeed if the empirical sciences had anything to say about whether nonempirical categories, the investigation of which lies outside those sciences, are empty. Sweeping metaprinciples about the nature of legitimate inquiry, particularly metaprinciples bearing on non-empirical inquiry, are not the fruits of the empirical sciences; they rather issue from philosophy, good or bad. Accordingly, Core Scientism is not the kind of scientific thesis characteristic of the empirical sciences. The empirical sciences flourish, have flourished, and will flourish without commitment to Core Scientism. In addition, antiscientism is in no way antiscience or antiscientific.
Friends of Core Scientism will remind us that it invokes not the current empirical sciences but rather the hypothetically completed empirical sciences. They will thus be unmoved by the absence of Core Scientism from the theses of the current empirical sciences. Still, the problem persists, because we have no reason whatever to hold that Core Scientism is among the theses of the hypothetically completed empirical sciences. A general problem is that predictions about what the completed sciences include are notoriously risky and arguably unreliable. The revolutionary history of the sciences offers no firm basis for reasonable confidence in such predictions. In addition, nothing in the current empirical sciences makes it likely that the completed sciences would include Core Scientism as a thesis.
The problem with Core Scientism is its monopolistic posture, a posture common among philosophical claims. It makes a claim about every real entity and every legitimate method for acquiring knowledge. The empirical sciences, for better or worse, are not thus monopolistic; nor do we have any reason to think that they will become so.
Neither individually nor collectively do they offer theses about every real entity or every legitimate epistemological method. The empirical sciences limit their theses to their proprietary domains, even if wayward scientists sometimes overextend themselves with claims about every real entity or every legitimate epistemological method.
Support for this claim comes from the fact that the empirical sciences, individually and collectively, are logically neutral on such matters as the existence of God, the reliability of certain kinds of religious experience, the objectivity of moral value, and the reality of thinking substances. Each such science thus logically permits the existence of God, the reliability of certain kinds of religious experience, the objectivity of moral value, and the reality of thinking substances. We have no reason to suppose that the hypothetically completed empirical sciences differ from the actual empirical sciences in this respect. Naturalists, at any rate, have not shown otherwise; nor has anyone else.
Proponents of Core Scientism may grant that the empirical sciences are not
12 Paul K. Moser and David Yandell logically monopolistic regarding real entities or legitimate epistemological methods, and thus that Core Scientism is not a thesis of the empirical sciences. Still, they can insist that Core Scientism is epistemically warranted by the empirical sciences. (A
‘thesis’ of the empirical sciences is, in our language, logically entailed by the empirical sciences, whereas a claim warranted by the sciences need not be thus logically entailed.) This move would turn the debate to principles of epistemic warrant appropriate to the empirical sciences. Specifically, what principles of epistemic warrant combine with the (hypothetically completed) empirical sciences to justify Core Scientism? More relevantly, are any such principles of warrant required, logically or epistemically, by the (hypothetically completed) empirical sciences themselves? No such principles seem logically required, as the (hypothetically completed) empirical sciences logically permit that Core Scientism is not warranted. Whether such principles are epistemically required depends on the epistemic principles warranted by the (hypothetically completed) empirical sciences, and the latter matter remains unsettled. We have, at any rate, no salient evidence for thinking that the (hypothetically completed) empirical sciences warrant principles of epistemic warrant that justify Core Scientism. The burden for delivering such evidence is on naturalists and remains to be discharged.
Minimally, the empirical sciences rely on abductive epistemic principles that certify inferences to a best available explanation of pertinent phenomena. The empirical sciences are, after all, in the business of best explanation. Even so, their domains of pertinent phenomena to be explained are not, individually or collectively, monopolistic in the manner required by abductive warrant for Core Scientism. For example, those domains do not, individually or collectively, preclude every kind of religious experience suitable to abductively warranted theistic belief. So, the abductive epistemic principles accompanying the empirical sciences will fall short of warranting Core Scientism, given the monopolistic assumptions of the latter. More generally, any domain of evidence outside the scope of the (hypothetically completed) empirical sciences will raise potential problems for the abductive warrant of Core Scientism, given its monopolistic assumptions.
By way of reply, some philosophers may propose that Core Scientism is constitutive of the hypothetically completed empirical sciences properly understood. The proposal suggests that the proper understanding of empirical science involves Core Scientism.
This would amount to a vindication of Core Scientism by semantical fiat: it is just part of the proper understanding of empirical science that Core Scientism is correct. This solution is, however, too easy to be correct. It offers nothing to block critics of Core Scientism from making an analogous move, whereby a position contrary to Core Scientism becomes part of the proper understanding of empirical science. For example, one might as well build a theistic design hypothesis into the “proper understanding”
of empirical science. Two, then, can play the unconvincing game of semantical fiat. If one objects that a theistic design hypothesis is insufficiently empirical to figure in the
Farewell to philosophical naturalism 13
proper understanding of empirical science, the same objection is invited by the ontological and methodological components of Core Scientism. If the proper understanding of empirical science rests on semantical fiat alone, Core Scientism will run afoul of various available “proper understandings” of empirical science.
Naturalists sympathetic to either pragmatism or some kind of antirealism may try to rescue Core Scientism via rejection of our characterization of scientific and philosophical understanding as truth-seeking. Pragmatists insist that the purpose of inquiry is to produce (theoretically) “useful” beliefs rather than to uncover truths. In addition, antirealists claim that the purpose of inquiry is to produce consensus regarding certain claims or to pursue any of a number of other non-alethic goals.
(Pragmatism and antirealism face decisive self-referential problems, but we will not digress here.) The relevant problem with pragmatism and antirealism emerges from their assumptions about “the” purpose of inquiry. We have no reason to suppose that there is such a thing as “the” purpose of inquiry or that truth is no goal of inquiry. Even if inquiry has non-alethic goals of the sort favored by pragmatists and antirealists, it can still have truth as a central goal. Like those engaged in the empirical sciences, we are concerned with the nature of legitimate inquiry in those cases where inquiry is directed to the discovery of truth. That different standards (whether naturalist or otherwise) would govern inquiry with alternative goals is no surprise, and it has no relevance to the dispute over naturalism offered as a view about the nature of legitimate truth-seeking inquiry. It may be “useful” (assuming that this term avoids alethic implications) not to posit immaterial objects or not to employ non-scientific methods, but, even if this were useful, this fact would be irrelevant to the current dispute over Core Scientism, as we have defined this position. So, we will not concern ourselves further with pragmatist or antirealist defenses of Core Scientism.
We can now construct a troublesome dilemma for Core Scientism. Either Core Scientism is not itself a thesis included in or warranted by the (hypothetically completed) empirical sciences or its warrant depends on a special reformulation of the “proper understanding” of empirical science. In the first case, Core Scientism is self-defeating; in the second case, it is merely philosophically harmless. In either case, however, it is philosophically ineffectual. The self-defeat of Core Scientism would result from its failing to be included in or approved by its own proposed single standard for methodological and ontological integrity: the (hypothetically completed) empirical sciences. By its own proffered standard, in that case, Core Scientism would suffer defeat. If, however, Core Scientism is to be rescued via semantical fiat, as a desideratum for the proper understanding of empirical science, it becomes harmlessly stipulative in nature, with no firm basis in what actually goes by the name “empirical science.” In that case, Core Scientism loses its ontological and epistemological teeth, inasmuch as it becomes divorced from the ontological and epistemological successes of the actual empirical sciences. Given this dilemma, Core Scientism becomes philosophically innocuous. We will explore the fate of naturalism
14 Paul K. Moser and David Yandell further, in light of the dilemma facing Core Scientism.
3. The logical status of naturalism
If any version of naturalism is true, it is either true necessarily or true contingently. If naturalism is true necessarily, the necessity is a product of either (a) the form of the characteristic claims, (b) the definitions of the terms employed, (c) the relations between the concepts underlying those terms, or (d) the necessary nature of reality.
In short, the necessary truth of naturalism entails that its characteristic claims are either syntactic tautologies, definitional tautologies, conceptual tautologies, or metaphysical necessities. If the characteristic claim of a given form of naturalism does not fall under any of these categories, then it is contingently true, if true at all. We will argue that none of the forms of methodological naturalism under discussion falls under any of the categories of necessity (a)–(d), and that therefore methodological naturalism is at best contingently true. One might hold that ontological naturalism is necessarily true, but since this view is not defended by any prominent naturalist, we do not bother with it.
If methodological naturalism’s characteristic claims are contingent, then any warrant we have for these claims will presumably be empirical. Given the problems identified in the previous section for the empirical warrant of methodological naturalism, the case for claiming that Core Scientism, including its methodological component, is contingently true faces serious trouble. If, then, no version of methodological naturalism is necessarily true, the case for methodological naturalism being true at all will likewise be in trouble.
Syntactic tautology and naturalism
None of the characteristic claims of the species of naturalism (i)–(vi) seems to be a syntactic tautology; neither, as expected, does Core Scientism. The denials of (i)–(vi) and Core Scientism seem to be consistent in logical form. So, the necessitarian naturalist has the burden of showing that a formal contradiction arises from denying a particular variant of naturalism or Core Scientism. Given the difficulty of discharging this burden, no prominent contemporary naturalist has defended naturalism on this basis.
Definition and methodological naturalism
A historically popular defense of methodological naturalism claims that such
Farewell to philosophical naturalism 15
naturalism follows from definitional truths. This defense proceeds on the basis of one of these options: (1) the terms in the aforementioned principles (iv)–(vi) could be stipulated to have certain meanings; (2) the common or conventional usage of those terms could give them particular meanings, or (3) those terms could pick out concepts having certain relations that yield definitional truth (this possibility will be treated in the next subsection).
Regarding option (1), any of the characteristic claims of naturalism could be made a definitional tautology by mere stipulation. If, for example, “legitimate method of acquiring knowledge” is simply defined as “method of the empirical sciences,” then eliminative methodological naturalism’s defining claim is a definitionally necessary truth. This strategy is, however, readily disarmed. One can grant the stipulation, and then shift the terms of the debate away from the stipulated terms by asking whether there are any methods for acquiring knowledge other than by means of “legitimate”
(stipulatively defined as “empirical scientific”) methods of inquiry. One could also frustrate such a strategy by refusing to grant the stipulation. Mere insistence that we so stipulate the meanings of terms is question-begging against opponents of this sort of necessitarian naturalism (including not only antinaturalists but also those holding that naturalism is contingently true and even those holding that it is non-stipulatively necessarily true).
A naturalist might claim that, as a matter of lexicographical fact, the terms employed in some version of naturalism are so used that the characteristic claim of that version is definitionally necessarily true. Given this claim, there is some standard or convention of linguistic usage on which some version of naturalism is true solely in virtue of the way the terms expressing it are used. Whether this constitutes a legitimate form of necessity is an open question, given that such standards or conventions of usage would themselves be contingent. We will offer reasons for doubting that methodological naturalism can be justified in this way regardless of whether such justification would yield genuine necessity.
Whether there are conventions regarding linguistic usage that make a form of naturalism definitionally true is an empirical matter, and must therefore be evaluated on the basis of empirical evidence. Naturalists, however, do not typically claim just that people do regard the terms of scientific methods as (or as grounding) the terms of the only legitimate methods of acquiring knowledge. Otherwise, the truth or falsity of naturalism would be established by linguistic survey, and this project would likely produce the overwhelming result “undecided” if popular linguistic usage is our guide.
Rather, naturalists typically claim that people should adopt their naturalism. The principles of naturalism, including those of Core Scientism, typically have a normative character that would be lost if they were mere reports concerning how people happen to use certain words. In addition, the necessity and indeed the truth or falsity of naturalism would then fall prey to variations in linguistic usage. Naturalism could be necessarily true (or simply true) at one time or in one region but not in another. This,
16 Paul K. Moser and David Yandell however, would be an undesirable characteristic in any scientific or metascientific principle. The sciences seek truth, not just temporal or regional consensus.
A common claim regarding the source of normativity for naturalism is that it stems not (just) from ordinary usage of terms, but from scientists’ usage. In other words, it enjoys support from expert linguistic usage. The problem, however, is that scientific linguistic usage does not enunciate monopolistic claims regarding the nature of legitimate inquiry, as Core Scientism does – even if wayward scientists occasionally engage in exaggeration via such monopolistic claims. Given the non-monopolistic conceptions of the empirical sciences standardly employed by scientists themselves, it’s understandable that the sciences do not offer monopolistic answers to philosophical metaquestions about the nature of inquiry. If a scientist endorses Core Scientism, he or she is doing so not qua empirical scientist but rather qua metaphysician or epistemologist, in short, qua philosopher.
Philosophers claiming that the methodological component of Core Scientism is definitionally necessarily true face a dilemma: either they (a) seek to ground the definitional claim in the actual sciences by claiming that empirical truths about how scientists use methodological terms somehow (despite appearances) warrant such a metaclaim and somehow make it normative, or (b) seek to ground the claim in hypothetical future science that has expanded to cover such claims, by claiming that the methods of such science rule out traditional first philosophy as unscientific.
Neither of these possibilities is promising.
Narrowly conceived, the empirical sciences include no discipline that makes even descriptive monopolistic claims about the nature of scientific practice. More broadly conceived, to include “science studies” or sociology of science, the empirical sciences will have a much larger range of methods available but still will not make normative claims, at least not monopolistic normative claims, about the legitimacy of methods of inquiry. For such normative claims to be included within the scope of the products of scientific methods, those methods would have to be expanded to include methods that employ normative terms and underwrite monopolistic claims about inquiry. If naturalism permits such expansion, the door is thereby opened to the legitimacy of competing forms of inquiry using normative terms, including those of traditional first philosophy. (This avenue is not open to any version of eliminative methodological naturalism that rejects as unscientific the needed normative terms.) Suppose, as seems likely, that non-eliminative methodological naturalists cannot provide a standard governing the use of terms, itself grounded in scientific method, that permits the use of normative terms in metaclaims favoring naturalism while excluding as illegitimate antinaturalist conceptions of inquiry using normative terms. In that case, there will be no way that a definitional necessity can ground the rejection of antinaturalist forms of inquiry. Moreover, the normative terms employed within the needed naturalist standard, and those permitted by it, will have to be reducible to, or to refer to objects supervening on objects referred to by, the non-normative terms of
Farewell to philosophical naturalism 17
scientific inquiry. Otherwise, the retention of these terms will violate the standards of reductive or non-reductive methodological naturalism. We find no basis in the empirical sciences for the kind of standard needed by methodological naturalists.
The prospects for methodological naturalism, and thus for Core Scientism, seem bleak now. The empirical sciences, collectively as well as individually, are not in the business of making monopolistic normative claims about the legitimacy of methods of inquiry. One might seek to expand the empirical sciences to include methods for making such monopolistic normative claims. In that case, however, one would have made meta-epistemology a branch of the empirical sciences. Under such an expansion, much of what is now included in first philosophy would become “empirical science,” owing not to a revision in first philosophy but rather to a sweeping liberalization of the idea of empirical science. Antinaturalists can consistently welcome such sweeping conceptual revision, given that it poses no threat to first philosophy. Naturalists, in contrast, will have to concede that such revision removes the epistemological teeth from Core Scientism, as it then allows first philosophy to proceed apace.
Conceptual necessity and methodological naturalism
One might defend the methodological component of Core Scientism on the ground that some version of it is a necessary truth owing to relations among the concepts underlying its constitutive terms. This modal status, so the reasoning goes, might arise solely from the way we conceptualize a legitimate method of inquiry. On one proposal, accordingly, the category “legitimate method” is conceived of as being included within, if not equivalent to, the category “empirical scientific method.” On another proposal to the same end, there is a conceiving-independent relation of inclusion or equivalence between these two concepts. That is, one might hold that an objective conceptual necessity applies to the characteristic claim of some version of methodological naturalism. If, in contrast, the relation between the concepts in question is merely the product of convention or of some other conceiving-dependent process, then the option under consideration is equivalent to the possibility of definitional necessity discussed in the previous subsection.
A noteworthy consideration is that conceptual items are among the things whose reality is often called into question by methodological naturalists, particularly those of Quine’s persuasion. Such items are intentional, and the conceptual relations being considered are semantic. If semantic or intentional items are needed to account for the logical status of methodological naturalism, based on conceptual relations, then eliminative naturalism cannot be necessary or even true on this basis, since it characteristically rejects the cognitive significance of such items. If the semantic and intentional features of the relevant concepts are not conventional, then the terms
18 Paul K. Moser and David Yandell denoting them are not plausibly viewed as reducible to non-intentional, non-semantic terms. Or, at least, naturalists will then need to make a special case for their reducibility, and the prospects seem bleak in light of the history of attempts at such reduction. So, reductive methodological naturalism is a dubious candidate for conceptual necessity (except where this is merely another name for definitional necessity, discussed above).
Non-eliminative non-reductive methodological naturalism allows for the legitimacy of intentional and semantic terms, but only if the intentional entities or features picked out by those terms and the semantic relations between those features supervene on non-intentional entities or features and non-semantic relations.10 The non-eliminative non-reductive methodological naturalist faces a dilemma here.
Either (a) the range of terms whose referents supposedly supervene on the referents of terms of the empirical sciences is broad enough to include the sorts of intentional terms needed to express metaphilosophical theories about meaning and to allow for reference to non-conventional semantic relations between concepts or (b) it is not. If it is, then the same latitude that makes non-eliminative methodological naturalism a candidate for conceptual necessity will permit competing views (including many that express traditional conceptions of first philosophy) to be candidates for the same status. If, in contrast, the range of legitimate terms is not this broad, then the characteristic claim of non-eliminative methodological naturalism is excluded from the domain of conceptual necessity along with its naturalist cousins. Still, only noneliminative methodological naturalism (among forms of methodological naturalism) is a serious candidate for conceptual necessity. The problem is that the broad range of intentional and semantic terms needed as legitimate for this to be a live option, according to the view’s own standards of legitimacy, would also permit (as legitimately naturalistic candidates for conceptual necessity) competing standards of legitimate inquiry that are not naturalistic in any sense that threatens first philosophy.
Metaphysical necessity and methodological naturalism
Many proponents of traditional first philosophy have held that certain principles are, although not analytic truths or tautologies, nonetheless necessarily true, in virtue of the necessary nature of reality. These principles are said to be “metaphysically necessary.” Let us call any such principle a substantive necessary truth. Ethical principles, metaphysical principles concerning explanatory and causal sufficiency, and epistemic principles regarding the nature of evidence have been proposed as having this status.
Most naturalists deny that there are any such truths. So, it would be rare for a methodological naturalist to appeal to such truths on behalf of naturalism. Still, we will challenge this option.
Could the methodological component of Core Scientism be a substantive
Farewell to philosophical naturalism 19
necessary truth? The claim that it is implies that (a) there is a necessary nature, or essence, of every legitimate method of acquiring knowledge and (b) this essence makes the methodological component of Core Scientism a substantive necessary truth. Suppose, then, that the methodological component of Core Scientism is the correct description of such an essence and thus is a substantive necessary truth. In that case, for the methodological component of Core Scientism to be itself a thesis or warranted product of the empirical sciences, the empirical sciences must be capable of including or warranting a substantive necessary truth of the monopolistic sort represented by Core Scientism’s methodological component. We have already noted, however, that the empirical sciences do not trade in or warrant monopolistic methodological
claims,
let
alone
substantively
necessary
monopolistic
methodological claims. Even if the empirical sciences include or warrant substantively necessary truths (and we’re not going to stick our necks out here), they don’t include or warrant monopolistic substantively necessary truths. Likely claims to empirical necessity, deriving from linguistic reference to natural kinds, are irrelevant to our point concerning the substantive necessity of Core Scientism. Linguistically grounded empirical necessities, familiar from the Putnam–Kripke–Marcus theory of linguistic reference to natural kinds, do not add up to genuine metaphysical necessity, owing to the undeniable contingency of linguistic reference.
In sum, then, first philosophy has nothing to fear from claims to the necessity of methodological naturalism. Core Scientism, we have seen, either collapses by its own standard for methodological genuineness or opens the door to the legitimacy of first philosophy. Let us turn, in conclusion, to an application of this dilemma.
4. The dilemma applied
W. V. Quine has proposed an influential version of Core Scientism implying that there is no cognitively legitimate philosophy prior to, or independent of, the empirical sciences, and thus that one should not make philosophical claims exceeding the empirical sciences.11 Epistemology, by Quine’s lights, is but a chapter in the book of empirical psychology; so epistemology offers no encouragement to first philosophy.
Quine’s own version of Core Scientism, his epistemological naturalism, is itself, however, an instance of philosophy prior to the empirical sciences. To meet this objection, Quine must show that his epistemological naturalism is a thesis or at least a warranted product of the empirical sciences. Quine does not discharge this burden, because the empirical sciences, as we have noted, do not make sweeping, monopolistic claims about the status of epistemology – even if an individual scientist makes such claims on occasion, in a moment of epistemological chauvinism. This may be an empirical truth about the empirical sciences, but it is warranted nonetheless, and it holds for the empirical sciences generally. Quine’s epistemological naturalism,
20 Paul K. Moser and David Yandell combined with his general scientism, is thus self-defeating; by its own standard, it is omitted from the domain of the cognitively legitimate. Quine and other naturalists should seek to avoid self-defeat inasmuch as the empirical sciences do and inasmuch as theoretical conflict is disadvantageous to unified scientific explanation.
One might try to defend Quine’s epistemological naturalism by offering a notion of science broader than that underwritten by the empirical sciences as standardly characterized in light of a typical college science curriculum. Such a proposal would enable Quine’s epistemological naturalism to be a scientific hypothesis in the newly proposed sense of “scientific.” This move would, however, yield a dilemma for naturalists: either (a) there will be a priori constraints on what counts as science (as actual linguistic usage of “science” would not ground this extended notion) or (b) this broader notion will be altogether ad hoc with regard to its rationale. To serve the purposes of naturalism, the new notion of science must exclude traditional epistemology and first philosophy, while including epistemological naturalism, in a way that is not ad hoc. This is no small task. In addition, this strategy for escaping self-defeat demands a hitherto unexplicated notion of science, which is likewise no small order. Neither Quine nor any other naturalist has defended such a strategy; nor have naturalists otherwise resolved the problem of self-defeat or the dilemma facing Core Scientism. At a minimum, naturalists will want to avoid a return to first philosophy, but it is not clear how they reasonably can.
An adequate epistemology, whether traditional or non-traditional, relies on epistemological principles that specify, among other things, what conditions epistemic warrant consists in. Quine’s naturalized epistemology likewise relies on epistemological principles, such as his “maxim of minimum mutilation”: a principle of conservatism implying that it is good not to alter our antecedent theory more than necessary.12 Quine holds that another epistemological principle, concerning the maximization of “simplicity,” can override his principle of conservatism; but this will not affect our discussion. We will bracket cases where the two principles conflict.
Is Quine’s epistemological principle of conservatism itself a thesis of any of the empirical sciences? Quine has not shown that it is; nor has anyone else. People engaged in the empirical sciences perhaps use this principle at times, but no empirical science endorses it or seeks to confirm it. Again, the empirical sciences, given their theoretical objectives, do not concern themselves with endorsing or assessing general epistemological principles, particularly principles that are monopolistic in the way Core Scientism is. Quine’s preferred species of epistemological naturalism thus includes a principle barred by his own scientism. In this respect, Quine’s version of epistemological naturalism is self-defeating. As a result, it fails to challenge traditional epistemology or first philosophy.
Quine’s epistemological principles evidently function as analytic truths for him, contrary to his own assumptions. The following questions invite this point: In virtue of what is Quine’s epistemological principle of conservatism true, or correct? In virtue of
Farewell to philosophical naturalism 21
what is it actually the case that warranted theory revision conforms to the maxim of minimum mutilation? What, in other words, makes Quine’s epistemological principle true rather than false? These questions highlight the importance of analyticity for epistemology, including Quine’s naturalized epistemology. They ask what makes it the case that warranted theory revision is constrained by the principle of conservatism.
Apparently, it is just part of Quine’s usage of the term “epistemic warrant” that the principle of conservatism constrains epistemic warrant. If so, it is analytically true for Quine that the principle of conservatism constrains epistemic warrant. A traditional alternative to this truth’s being analytic proposes that it is a factually substantive principle about the nature of warrant, knowable a priori. This would be in keeping with the fact that the empirical sciences do not discuss, let alone endorse, any such epistemological principle. The latter alternative of substantive apriority would, however, be altogether alien to Quine’s naturalistic project, violating both his scientism and his opposition to first philosophy.
Quine suggests that his epistemological principles find support in actual scientific practice. He claims that these principles “are maxims by which science strives for vindication in future predictions”.13 This suggests that Quine’s epistemological principles are justified by their role in actual scientific practice. Quine might propose, in addition, that his principle of conservatism is true in virtue of its role in actual scientific practice. Let us grant this proposal for the sake of argument, and ask: In virtue of what is it true that actual scientific practice constrains epistemically warranted theory revision? It appears simply to be part of Quine’s usage of “epistemic warrant” that actual scientific practice constrains epistemic warrant. In that case, it is analytically true for Quine that scientific practice constrains epistemic warrant.
Appeal to scientific practice thus does not free Quine from the importance of analyticity. Barring analyticity, we would have great difficulty answering questions of the sort just raised. Perhaps this explains why Quine has very little to say on such questions.
Quine has commented on this question: What is it that makes one complete physical theory true and another false? He answers: “I can only answer, with unhelpful realism, that it is the nature of the world. Immanent truth, à la Tarski, is the only truth I recognize.”14 Even if we accept Tarski’s disquotational approach to truth, this question persists: What makes it the case, at least for Quine, that the standard of conservatism constrains epistemic warrant? Talk of “the nature of the world” is unhelpful now, since we want to know what in particular about “the nature of the world” makes it the case, at least for Quine, that the standard of conservatism constrains epistemic warrant. The most plausible answer invokes what is analytically true for Quine regarding “epistemic warrant” (and related terms).
Quine, we have seen, uses principles regarding the nature of epistemic warrant in a way that treats them as analytic. Such principles, we have argued, are not themselves warranted by the empirical sciences; rather, one needs to use such principles to appeal
22 Paul K. Moser and David Yandell to the empirical sciences for warrant. Given this, and given that such principles are precisely of the sort characterizing first philosophy, we can plausibly understand Quine to be using analytic principles regarding warrant as an implicit first philosophy.
The lesson is thus that if the analyticity of such principles involves first philosophy, as it seemingly does, then even naturalists such as Quine are wedded to first philosophy.15
Our dilemma for Core Scientism reveals, in the end, that Core Scientism is either self-defeating or ineffectively stipulative. We can nonetheless join naturalists in holding that empirical scientific methods are typically reliable for pursuing empirical knowledge of natural objects. Endorsing the empirical sciences as typically reliable in their appropriate domains, however, does not require embracing scientism of any sort. The epistemological principles justifying appeal to empirical scientific methods cannot themselves be warranted by appeal to those methods. Given this lesson, we may say not only farewell to philosophical naturalism but also welcome to first philosophy, as a foundational discipline here to stay.
Notes
1 We thank Paul Abela, Linda Mainey and Keith Yandell for comments.
2 For other difficulties in giving an informative characterization of materialism, see Tim Crane and D.H. Mellor, “There is No Question of Physicalism,” Mind 99 (1990): 185–206.
3 For the argument that Descartes was a dualist, and not a metaphysical “trialist” as many recent commentators have claimed, see David Yandell, “Did Descartes Abandon Dualism? The Nature of the Union of Mind and Body,” British Journal for the History of Philosophy (forthcoming). For discussion of the claim that Descartes conceded that interactionist dualism involved a contradiction, see David Yandell, “What Descartes Really Told Elisabeth: Mind–Body Union as a Primitive Notion,” British Journal for the History of Philosophy 5 (1997): 249–73.
4 W. V. Quine, Pursuit of Truth (Cambridge, Mass.: Harvard University Press, 1990), 20–1.
5 See W. V. Quine, “The Scope and Language of Science,” in Quine, The Ways of Paradox (New York: Random House, 1966), 215–32; Paul Churchland, Scientific Realism and the Plasticity of Mind (Cambridge: Cambridge University Press, 1979); and Daniel Dennett, The Intentional Stance (Cambridge, Mass.: MIT Press, 1987).
6 J.J.C. Smart, “Materialism,” The Journal of Philosophy 60 (1963): 651–62.
7 See Donald Davidson, “Mental Events,” in L. Foster and J. Swanson (eds.) Experience and Theory (Amherst, Mass.: University of Massachusetts Press, 1970), 79–101; and David Papineau, Philosophical Naturalism (Oxford: Blackwell, 1993). Roy Wood Sellars seems to endorse a version of this view, resting on “levels of causation” rather than supervenience, in The Philosophy of Physical Realism (New York: Macmillan, 1932).
For representation of the most prominent variants of contemporary ontological naturalism, reductive and nonreductive, see Paul Moser and J.D. Trout (eds.), Contemporary Materialism (London: Routledge, 1995). For assessments of contemporary versions of nonreductive naturalism, including those of Davidson and Papineau, see Paul Moser, Philosophy After Objectivity (New York: Oxford University Press, 1993), chap. 5; Paul Moser, “Physicalism and Mental Causes: Contra Papineau,” Analysis 56 (1996): 263–7; and Paul Moser and J.D.
Farewell to philosophical naturalism 23
Trout, “Physicalism, Supervenience, and Dependence,” in Elias Savellos and Ümit Yalçin (eds.) Supervenience: New Essays (Cambridge: Cambridge University Press, 1995), 187–217.
8 See W. V. Quine, “The Scope and Language of Science,” in The Ways of Paradox (New York: Random House, 1966), 215–32; and Jerry Fodor, “Special Sciences,” Synthese 28 (1974): 77–
115.
9 See J.J.C. Smart, “Materialism,” The Journal of Philosophy 60 (1963): 651–62. Cf. Smart, Philosophy and Scientific Realism (London: Routledge, 1963), chap. 3.
10 For the complexities involved in the phenomenon of supervenience, see the essays in Elias Savellos and Ümit Yalçin, eds., Supervenience: New Essays (Cambridge: Cambridge University Press, 1995).
11 See Quine, “The Scope and Language,” 222; and Quine, “Things and Their Place in Theories,” in Theories and Things (Cambridge, Mass.: Harvard University Press, 1981), 1–23, especially p. 21, for an identification of the scientific and the cognitive domains.
12 See W. V. Quine, Pursuit of Truth, 15, 56.
13 Ibid., 15.
14 W. V. Quine, “Responses” in Theories and Things, pp. 173–86; the quotation is from p. 180.
15 For further support for this lesson, see Paul Moser and David Yandell, “Against Naturalizing Rationality,” ProtoSociology 8 (1996): 81–96.
Dallas Willard
There is an objective difference between one who has knowledge of something and one who does not. This is true in both the occurrent and the dispositional senses of
“knowledge” and “knows.” That is, whether or not X has knowledge on a certain point or about a certain matter – knows the English alphabet, for example, the narrative content of War and Peace, or the date of Robert Kennedy’s assassination – is not a matter of how anyone, including X himself, may think or feel about X and his conscious or other states.
Knowledge with reference to specific matters is a condition which individual human beings are or are not in. Usually this condition is found in a social context. But if only one person existed it would still be possible for that person to be knowledgeable about some, at least, of the specific matters that concerned him – for example, about reliable sources of food and water in his environment.
Can this objective difference among human beings consist in properties and relations that fit within a naturalist ontology? A long tradition of ancient and modern philosophers from Plato, Descartes and Kant to T. H. Green, Edmund Husserl and Hilary Putnam has insisted that it cannot. Others – especially those in the nineteenth and twentieth centuries who insisted upon distinguishing naturalism from materialism – have held that it could be. But with the rise and development of the mind/brain identity thesis during the last half of the twentieth century, the generous naturalism (as we shall call it) of Dewey, Santayana, Sidney Hook and others has largely disappeared in favor of a narrower naturalism more commonly and more correctly called “physicalism” (the older “materialism”). For it, all distinctively human properties are reduced to strictly physical properties of the central nervous system of the human body or to these plus characteristics of the natural and social setting – or, to nothing at all.1
In this chapter I will try to explain why narrower Naturalism or unqualified Physicalism cannot find a place for knowledge, and specifically for three of its essential components: truth, logical relations and noetic unity. At this late date it is hard to say much that will be strictly new on these matters, but, apparently, there is much that
needs to be said again. What I shall say about truth and logic is practically identical with what Frege said more than a century ago, though I hold views significantly different from his on how truth and logic fit into the full context of knowing and knowledge.2
What I shall say about noetic unity adds little to what has already been said by Kant, Lotze and Husserl.
Naturalism is . . . what?
Before the question about the narrower naturalism’s ability to accommodate knowledge within its permissible categories can be raised, one has to come to some understanding of what Naturalism and knowledge are, and this is difficult to do without begging important questions.
In traditional philosophical terms, naturalism is a form of monism. It holds, in some order of interdependence, that reality, knowledge and method each are of only one basic kind. That is, there are not two radically different kinds of reality or knowledge or method. It is fundamentally opposed to pluralism, and most importantly to dualism as traditionally understood (Plato, Descartes, Kant).
In its modern forms, naturalism further specifies its monism by reference to the empirical or the sense perceptible. The one type of reality admitted by it is that of the sense-perceptible world and its constituents. All knowledge is, for it, reducible to (or in some manner continuous with) sense perception, and all inquiry essentially involves sense perception, directly or indirectly. Currently, “the sense perceptible” is de-emphasized in favor of “the scientific” – the organization of data around empirically underdetermined hypotheses. But this is understood to constitute empirical research and, hopefully, to yield empirical or descriptive knowledge.
This leads directly into the current versions of “naturalized epistemology,” where the emphasis is entirely upon the human being as a strictly physical organism acquiring beliefs in its “natural” environment. The question as to whether naturalism can accommodate knowing and knowledge is then replaced by the question of whether the various normative issues that arise with reference to knowledge and belief formation in the course of human life and scientific endeavors can be replaced by mere descriptions of actual processes of belief formation. In the famous words of Quine:
“Epistemology [after being integrated into science through psychology] still goes on, though in a new setting and a clarified status. Epistemology, or something like it, simply falls into place as a chapter of psychology and hence of natural science. It studies a natural phenomenon, viz., a physical human subject.”3
Knowledge itself, then – and, more weakly, justified belief – is simply belief that is produced in a certain way, for example, in ways that are reliable, ways that tend to produce true beliefs in actual as well as counterfactual situations that are relevant
alternatives to the actual situation.4 Quine and others hold that epistemology can be appropriately replaced by psychology. This would yield what is now widely referred to as a “naturalized epistemology.” Hilary Putnam, Jaegwon Kim and others hold, by contrast, that the normative (non-descriptive) element cannot be eliminated from epistemology, and therefore that a naturalized epistemology is impossible.5
All of this is now very familiar. But there are problems: problems which indicate to me, as it has to others, that the issue of the naturalization of knowledge is misconceived when stated in this way. And first of all, if we are going to replace epistemology with psychology we will have to decide which psychology will do. In fact, there is no existing psychology for the Naturalistic viewpoint to turn to. There is only an idealization of some adequate theory of human behavior which, supposedly, might at some time be achieved. This will, of course, have to be a “scientific” psychology. But what would that amount to? Could it possibly be a psychology that is not essentially identified in terms of naturalism itself? Highly unlikely. But then we are running in a circle, specifying naturalism in terms of psychology, but then . . . a naturalized psychology.
At present there are many socially identifiable, institutionalized forms of
“psychology,” with associated professional organizations, funding sources and avenues of publication. Research universities in the United States rarely ever have only one academic department giving advance degrees in psychology. So, when the naturalized epistemologist speaks of replacing epistemology with psychology, which department will he go to?
And we might also have second thoughts about normativity being the issue for the naturalization of epistemology. It is, of course, an issue. It is hard to imagine that the only epistemological question to be raised with respect to belief is its occurrence –
even occurrence under certain conditions, no matter how fine-tuned those conditions may be. We would still want to know if our belief is “true,” “correct,” “right.”
Putnam asks: “Why should we expend our mental energy in convincing ourselves that we aren’t thinkers, that our thoughts aren’t really about anything, noumenal or phenomenal, that there is no sense in which any thought is right or wrong (including the thought that no thought is right or wrong) beyond being the verdict of the moment [or of some longer run], and so on? This is a self-refuting enterprise if there ever was one!”6
But let us surrender the point for a moment and grant that epistemology can be replaced by psychology, the normative by the descriptive. Will it make any difference with respect to the longstanding issues around the naturalization of knowledge? Not really. The descriptive/normative contrast has no essential bearing upon the issue of naturalism in epistemology. Getting rid of normativity will not secure the naturalization of epistemology and retaining it will not exclude naturalization. It has seemed to do so only because an important philosopher (Quine) in an influential
paper set that contrast up as the essential issue, and others, for whatever reasons, accepted his formulation as the framework of the debate.
On the one hand, a hardy naturalist such as Dewey might very well insist that there is nothing non-natural about normativity.7 To think so is only to admit another of what Dewey often called “untenable dualisms.” Norms are interwoven throughout human experience and are every bit as “natural” as anything else to be found there. And a Wittgensteinian reading of linguistic rules and forms of life could come up with much the same point. Only someone who reads “descriptive” (as opposed to normative) in a strongly materialistic or physicalistic fashion would take the opposition of normative to descriptive to be the issue of naturalism in epistemology. But many Naturalists have not taken “descriptive” in any such sense, and certainly science or experimental method does not of itself require them to do so. What, exactly, is so unnatural or non-natural about norms?
On the other hand, let us suppose we surrender norms – just banish “right” in both its adverbial and its noun forms from our treatment of knowledge and justified belief.
What is so “naturalistic” about mere description? A description will be naturalistic only if what it is about – its content, what it mentions – fits into a naturalistic ontology. The issue for naturalism certainly is not just about normativity. What, for example, would make one think that an adequate description of the formation of belief by means of a reliable process would mention only things that would fit into a naturalistic ontology?
Only, I suggest, the prior assumption that the “science” of psychology will, when fully perfected, be appropriately natural – that it will “emerge” from biology as biology from chemistry and chemistry from physics, or something like that. Or, at least, that it will mention nothing “dualistic” or “transcendental” with regard to the physical world.
And this does not even raise the issue of whether or not description as such can be understood in naturalistic terms, or whether describing itself is an entirely natural event or fact. A fully naturalized semantics is, once again, just presupposed in the move to “description” without norms. But would not the description (of reliable processes of belief formation, etc.) itself have to be “correct,” “right,” “adequate,” “justified”?
And have we then gotten rid of norms if we simply do not mention them in our descriptions of epistemic processes?
I presume that the age-old discussions in ethical theory pitting intuitionism and naturalism against each other (coming into the late twentieth century through the ethical intuitionism of G. E. Moore and developing in the subsequent critique of
“descriptivism” by ethical theorists practicing “linguistic analysis”) must have influenced Quine to frame the issue of naturalizing epistemology in the way he did.
But his was a clearly unfortunate choice so far as genuine philosophical progress is concerned.
Invoking “science”
What has just been said with reference to psychology and naturalized epistemology calls attention to broader issues bearing upon the specification of naturalism.
Methodological monism is an enduring aspect of generic naturalism, and modern naturalism is often specified simply in terms of an exclusive application of scientific method in all inquiries. But how can this method support claims about the nature of reality as whole? For example, one might state that the only realities are atoms (quarks, strings, etc.) and derivatives thereof. But how is he to support his claim? It certainly cannot be derived from any specific science (physics, chemistry on up) or from any conjunction of specific sciences. And it is not to be derived through any application of experimental techniques within any science.
The naturalist must then have recourse to that popular but philosophically suspect abstraction, “science” itself, which says even less than the individual sciences about the nature of reality as a whole, because it says nothing at all. It is not the kind of thing that can say anything, though many individuals – usually, I think, not themselves scientists, and certainly not scientists expressing truths within the competence of their profession – present themselves as speaking for science, and thus as being “scientific”
in some extended but still authoritative sense.
John Searle seems to be in this position. He speaks of “our scientific view of the world,” which, according to him, every informed person with her wits about her now believes to be true. He speaks of a view of the world which includes “all of our generally accepted theories about what sort of place the universe is and how it works.”8 “It includes,” he continues, “theories ranging from quantum mechanics and relativity theory to the plate tectonic theory of geology and the DNA theory of hereditary transmission,” etc. We might imagine a very long conjunctive sentence – containing the specific theories he has in mind as conjuncts – that would, supposedly, express the
“world view” in question.
But this will hardly do what he wants. One thing that will not show up in such a conjunctive sentence is any claim about reality as a whole or knowledge in general.
Such specific scientific theories as those just mentioned – and no matter how many of them we may list – cannot provide an ontology. They never even attempt to determine what it is to exist or what existence is, and cannot by the nature of their content provide an exhaustive list of what ultimate sorts of things there are. Their existential claims are always restricted to specific types of entities as indicated in their basic concepts.
We emphasize the point that to suppose that a given scientific theory or conjunction of such theories provides an ontology constitutes a logical mistake, a misreading of what the theories say and imply. Those theories, and the bodies of knowledge wherein they are situated, actually say nothing whatsoever about the
universe or about how it – the whole “thing” – works. This is a merely semantical point about the meaning or logical content of the claims or sentences that make up the sciences. It is to be established or refuted by examining, precisely, those claims and sentences. It turns out that they do not even mention the universe, the totality of all that exists, nor do they say anything about the boundaries of knowledge generally. Such matters simply do not fall within the purview of their methods or findings.
In support of this claim we ask: Could one possibly find the place in some comprehensive and duly accredited scientific text or treatment, or some technical paper, where it is demonstrated or necessarily assumed by the science concerned that all that exists consists of particles or fields or strings – or whatever the proper subject matter of the science is? Would Searle or anyone else be able to mention the name of the physicist who established this as an “obvious fact of physics”?9 Exactly where in the
“atomic theory of matter” is the claim about what “the universe consists entirely of” to be found?
“After all,” Searle rhetorically asks, “do we not know from the discoveries of science
[italics added] that there is really nothing in the universe but physical particles and fields of forces acting on physical particles?” The answer, contrary to his assumption, is “No, we do not.” Again, could he possibly just point out when, where, how and by whom this “discovery of science” was made. Was it made?
Also, before the philosopher can use “the discoveries of science” he must determine what “science” says. But this is to reify science, to treat it as an entity that issues “results.” Science, as already indicated, says nothing at all. Particular scientists do. Unfortunately they also make unscientific statements. How can we tell when an individual scientist is making scientific statements, and “science” is therefore speaking, and when they are not? And can a “scientific” statement be false or perhaps illicitly derived and still be scientific?
If a scientific statement can be false or based on logical errors, then a scientific statement may be less than knowledge. How, then, could it be required that we accept such statements as a basis or framework for philosophical work? History shows that statements accepted as “scientific” have been both false and based on logical errors. Is the advocate of naturalism then one who works under an authority that may be and has been wrong? He himself would rarely if ever have the competence to do the scientific work and therefore must be taking the statements of “science” on authority.
But blind authority is in fact one of the things we would expect naturalism to stand against. Historically it has done so, and that has been one of its virtues. How can it avoid resting on it, however, if what Searle says is true? And is a philosopher’s statement about science, a scientific theory or scientist to be automatically regarded as itself scientific? What can its status be?
The dilemma of naturalism
Naturalism staggers back and forth between physicalism (materialism) as a general ontology and first philosophy, and outright physics-ism or scientism (which need not take the form of physics-ism) – often, though not always, trying to derive physics-ism from scientism and then physicalism from physics-ism. This continues up to the present.
In a recent review, Patricia Kitcher chides Stephen Stich for “philosophical Puritanism” when he takes naturalism to hold that the only real entities are physical.10
Such a position apparently has now led Stich to give up naturalism “in favor of an open-ended pluralism.” Pluralism, as he takes it, is a position that counts as legitimate all properties “invoked in successful scientific theories.” But for Kitcher, it seems, such
“Pluralism,” tied to “successful science,” is just the naturalism we want. She points out how “the obvious authorities” on naturalistic epistemology (Quine, Goldman) counsel us to “make free use of empirical psychology” and to “reunite epistemology with psychology.”11 Forget physicalism, her point seems to be. A loose scientism is enough to secure Naturalism for us. Indeed, many of the “generous” naturalists of the mid-twentieth century gathered around Dewey and Sidney Hook (see note 1) identified naturalism precisely with acceptance of science and only science as the arbiter of truth and reality, and seemed, at least, to accept whatever came out of the pipe of “scientific inquiry” as knowledge and reality.
But if the points made above about science, even “successful science,” and about psychology in particular, are true, Kitcher’s advice – similar to the advice of a Dewey or Hook – simply cannot be followed. It is vacuous in practice, for there is no way of identifying and accessing the “successful science” which is proposed as defining naturalism. At most you get “science now,” which is really only “some scientist(s) now.”
And certainly no science (including psychology) that was not naturalistic in some strongly physicalistic or at least empiricist sense would be accepted as “successful” by those inclined to naturalism. Then we are back in the circle: naturalism in terms of science – but, of course, naturalistic science.
For these reasons, I take it that the appeal to science cannot serve to specify naturalism. There are, then, good reasons to be a “puritan” if you want to advocate naturalism. Naturalism has to be an honest metaphysics; and that metaphysics has to be “unqualified physicalism” as referred to above. But then a thinker who would be naturalist would feel pressure to have recourse to some specific a priori analyses to render his ontological specification of naturalism plausible. Short of that, one simply can find no reason why naturalistic monism with respect to reality, knowledge or method should be true: no reason why there should not be radically different kinds of realities with correspondingly radically different kinds of knowledge and inquiry. Why should sciences be “unified”? This lack of reason is, I think, what made A. E. Murphy
conclude long ago, in his review of Naturalism and the Human Spirit, “that the naturalists, who have so much that is good to offer, still lack and need a philosophy . .
.”.12
In addition to the difficulty of coming up with such a priori analyses, however, to turn to such inquiry as might produce them would be to break with the epistemological monism essential to naturalism and introduce something like a “first philosophy.” This would be discontinuous with the empirical methods of the sciences.
In showing its right through a priori analysis, naturalism would simply give up the game.
In specifying what naturalism is, therefore, one seems to be faced with an inescapable dilemma. Either one must turn to a priori (non-empirical) analyses to establish its monism, which will refute naturalism’s basic claim about knowledge and inquiry, or its claim will have to rest upon a vacuous appeal to “science.”
That might seem to end the discussion about naturalism as a philosophical alternative. But there may be a way to keep it going. One could retreat to a mere methodological naturalism and say that scientific method is our only hope as human beings. Whether or not we can adequately specify naturalism or know it to be true, one might say, the “scientific method” must be exclusively followed for human well-being.
Naturalism would then be a humane proposal, not a philosophical claim. The proposal would be to assume in our inquiries that only the physical (or the empirical) exists and to see if inquiry based upon that assumption is not more successful in promoting human ends than any other type of inquiry.13 Our task here would then be to show that the methodological assumption proposed contradicts what knowledge itself is, and to insist that it therefore cannot be an adequate methodological assumption, since it will not allow us to understand knowledge itself.
And knowledge is . . . what?
But now what about knowledge? I take knowledge in the dispositional sense to be identical with the capacity to represent a respective subject matter as it is, on an appropriate basis of thought and/or experience. In the occurrent sense it consists in actually representing, at a point in time, the respective subject matter as it is, on an appropriate basis of thought and/or experience. This is not intended as an analysis or definition of knowledge, but as an initial description of cases which count as knowledge or knowing.
What constitutes an “appropriate basis” will vary from subject matter to subject matter, of course, as is generally acknowledged of the corresponding methods of inquiry. It is, no doubt, impossible to define “appropriate basis” in any perfectly general way, or even to specify perfectly general necessary and sufficient conditions
for having an appropriate basis. Certainly I will make no attempt to do so here.
However, a few things may be said about the necessary conditions of knowledge and knowing, without intending to be definitional or even comprehensive. The challenge to the narrower naturalism will be to accommodate these necessary conditions. If it cannot do so, it must be false, though it might still offer itself as a heuristic principle of inquiry.
Clearly one necessary condition of knowledge, both in the occurrent and dispositional senses, is truth. This is a necessary component of representing anything as it is. But this necessary element in any sufficient condition of knowing rules out all known psychologistic or sociological analyses of knowledge, such as certainty, rational acceptability, warranted assertability, reliability of process, etc., all of which apparently could be satisfied in conjunction with representations and beliefs that are false. Could, that is, unless truth is simply defined in terms of such psychological or sociological conditions. (Of course certainty, warranted assertability and the like each have psychological and other interests in their own right.) We should also note that logical relations will be essentially involved in any case of knowing. That perhaps should be expected because of their intimate association with truth. To know one must think, in the sense of actively exploring the logical interrelationships involved in and with the respective representations. This is merely to say that the subject matter in question must be reflectively and thoroughly conceptualized, and the logical relations between relevant propositions and experiences carefully examined. Knowledge, accordingly, does not simply happen to a person. It is not a passion.
Finally, in order to know, one must have a certain broad familiarity with the subject matter itself – have observed it, tested it, informed oneself, thought it through – in ways suited to the kind of subject matter in question. How and to what extent this can and must be done will be determined by the relevant process of conceptualization of the subject matter (and of course by the subject matter itself), which will go hand in hand with observation and experimentation. All of this will produce a certain unity of consciousness that is necessary for there to be knowledge. Knowledge does not come in discrete units. Noetic unity across a complex field of consciousness within one person is required.
The absence of any reference to belief in my statements on knowledge and knowing will be immediately noticed. The absence is intended, and though rare today in discussions of knowledge it is by no means unique in the history of the theory of knowledge.14 Even such a resolute naturalist as Roy Wood Sellars specifies the nature of knowledge without reference to belief.15
Belief I understand to be some degree of readiness to act as if such and such (the content believed) were the case. Everyone concedes that one can believe where one does not know. But it is now widely assumed that you cannot know what you do not
believe. Hence the well-known analysis of knowledge as “justified, true belief.” But this seems to me, as it has to numerous others, to be a mistake. Belief is, as Hume correctly held, a passion. It is something that happens to us. Thought, observation and testing, even knowledge itself, can be sources of belief, and indeed should be. But one may actually know (dispositionally, occurrently) without believing what one knows.
Whether or not one believes what one represents truly and has an appropriate basis for so representing, depends on factors that are irrelevant to truth, understanding and evidence. It depends, one might simply say, on how rational one is. Now I do not think that this point about belief in relation to knowledge is essential to the rest of this chapter, but I mention it to indicate that the absence of any reference to belief in my general description of knowledge is not an oversight. Belief is not, I think, a necessary component of knowledge, though one would like to believe that knowledge would have some influence upon belief, and no doubt it often does.
In addition, it seems to me that specification of knowledge in terms of belief is a harmfully tendentious characterization, favoring the naturalization of knowledge.
This is because belief has an essential tie to action, and is therefore easily located in the natural world – say as a mere tendency of the physical organism to behave in certain ways. I suspect that it is the almost overwhelming empiricist – and in that sense naturalist – tendency of thought in our time that has created the general presumption that knowledge must be some kind of belief. Hence we must here at least question that presumption; and, I believe, when questioned it will not prove to be obvious or, finally, sustainable.
It also should be noticed that on the view here advanced one may know without knowing that one knows. Skeptical tendencies are often associated with the mistaken view that one has to know that one knows in order to know. And, in particular, one does not have to know that one actually has “an appropriate basis in thought and experience” in order to have one, and one does not have to know that one’s representation is true in order to know. It has to be true, indeed, but we do not have to know that it is true. We “use” it or live it; we do not “mention” or focus upon it.