EMPIRICISM.

Empiricism is a family of theories of knowledge (epistemology) claiming that all knowledge about the extant universe is based on experience, primarily on perception via the five senses. Some empiricists add introspection, a moral sense, or a special sensitivity to religious or aesthetic experience. Strong empiricists claim that all knowledge whatever derives from experience. They must show how empiricism can handle apparently a priori knowledge, including logic, mathematics, and ordinary truths such as "Bachelors are unmarried males." Empiricism also provides an account of mind, language, and learning. The traditional contrast of empiricism is with rationalism and nativism, the view that we do possess a priori knowledge, either furnished by reason alone or innate. Empiricists tend to perceptualize the mind and its operations, while rationalists tend to intellectualize it. With its down-to-earth emphasis on concrete experience and clarity, empiricism has flourished in Anglophone countries, whereas the more speculative rationalist and Kantian ideas have flourished on the Continent. This is one aspect of the divide between Continental philosophy and Anglo-American, "analytic," and "linguistic" philosophies.

In the twenty-first century nearly everyone is an empiricist in the everyday sense of taking experience seriously as a basis for knowledge claims about the natural world and human behavior, but most philosophers reject traditional, doctrinaire empiricism—the view that human sense experience provides a special connection of the knowing mind to the world and thus provides a foundation on which knowledge can build, step by step.

A Thumbnail History

In ancient times Aristotle was an empiricist relative to Plato's other-worldly rationalism. Modern empiricism began around 1600 with Francis Bacon (1561–1626), who promoted a new, experimental philosophy combining experience and reason, and with Galileo Galilei (1564–1642), who united experimental observation with a Platonic mathematical framework. Thomas Hobbes (1588–1679) further enriched early empiricist thinking, but the "big three" British empiricists were John Locke (1632–1704), George Berkeley (1685–1753), and David Hume (1711–1776). Locke first systematically expounded modern empiricism (see below). He was followed in the eighteenth century by Berkeley, notorious for his subjective idealism, the radical empiricist view that there are no material objects, that everything can be analyzed into minds and their ideas. Hume then took the further step of denying that there is even a substantial mind or ego. We introspect only a bundle of passing impressions. The mind is governed by natural laws of association, analogous to Isaac Newton's (1642–1727) law of gravitation, without needing an executive overseer. Hume also denied that inductive inference can be justified by logical argument, but he defended a wider conception of rationality (or at least sensible action) based on our natural impulses to believe and act. As he wrote in A Treatise of Human Nature (1739–1740) and its popularization, An Enquiry Concerning Human Understanding (1748), passion or custom, not reason, is "the great guide of human life."

The leading nineteenth-century empiricist was John Stuart Mill (1806–1873), who developed a full-fledged phenomenalism. Mill held that simple induction by enumeration (ravens 1, 2, 3, … n are black; therefore all ravens are black) is sufficient to support both science and mathematics: even the principles of logic and mathematics are very general empirical laws. Twentieth-century empiricists such as Bertrand Russell (1872–1970), George Edward Moore (1873–1958), and Alfred Jules Ayer (1910–1989) denied this, as did the logical empiricists of the Vienna Circle, who contended that the laws of logic and mathematics are both a priori and analytically true, that is, true by virtue of logical form and our linguistic conventions, hence completely empty of empirical content. By contrast, many twentieth-century thinkers, following Willard Van Orman Quine (1908–2000), have returned to a more naturalistic pragmatism.

The most damaging criticisms of British empiricism were leveled first by Immanuel Kant (1724–1804) and by the German and British idealists who followed him, then by Ludwig Wittgenstein (1899–1969), Gilbert Ryle (1900–1976), Quine, Wilfrid Sellars (1912–1989), Thomas Kuhn (1922–1996), and other twentieth-century figures, who attacked the entire Cartesian-Lockean conception of mind, experience, and language. Shocked by Hume's apparent skepticism about causality and Newtonian science, Kant synthesized rationalism and empiricism, while critically transcending both. The human mind itself furnishes the conceptual and rational apparatus necessary to organize our experience, as he argued in his Critique of Pure Reason (1781). "Concepts without percepts are empty; percepts without concepts are blind." Kant retained a vestige of the rationalist idea that we possess a special sort of intuition that enables us to make substantive, "synthetic" claims about the world that are nonetheless known a priori. He held that Euclidean geometry and the basic principles of Newtonian mechanics are such "synthetic a priori" truths. Post-Kantian empiricists deny this.

Foundational Empiricism

The traditional empiricists and rationalists were foundationists in epistemology. Foundationism postulates a base set of propositions that play a distinctive epistemic role plus a superstructure (comprising the bulk of our knowledge) appropriately related to the base. The empiricists and rationalists added the constraints that the basic statements must be certain and self-justifying (self-evident to reason for the rationalists and evident to the senses for the empiricists) and that the relation of base to superstructure be one of logical inference: deductive and perhaps inductive logic must suffice to generate the superstructure from the base. The justification is one-way or "linear" in the sense that the various layers of superstructure depend only on lower layers and, hence, ultimately on the base for their justification. Euclidean geometry provides the intellectual model. In this case the inferences are strictly deductive.

Given such a Euclidean geometry–inspired model, one wants the largest possible superstructure from the narrowest and most certain possible base. Two main problems stand in the way: the base problem (whether the base itself can be adequately justified) and the superstructure problem (whether the inferential resources are sufficient to support the desired superstructure on the base). From the beginning, empiricists have addressed the second problem by restricting the super-structure to claims within reach of observation and experiment and by developing the resources of logic, probability theory, and statistical inference. The British empiricists did not fully recognize the seriousness of the first problem.

Within this foundationist framework, Locke established the overall structure of a specifically empiricist theory in An Essay Concerning Human Understanding (1690), one of the founding works of the Enlightenment:

1. All simple ideas come from experience. There are no innate ideas. Contrary to nativists such as René Descartes (1596–1650), the mind is a tabula rasa—that is, a blank slate—at birth.
2. Ideas of solidity, movement, number, and so forth, resemble features of the real world (primary qualities), whereas sensations of color, sound, taste, and so forth, do not resemble the physical powers (secondary qualities) in objects that produce these sensations in the mind. They are mind-dependent.
3. Complex ideas are compounds of simple, atomic ideas, and, so, are image-like.
4. Thus knowledge, which is the intellectual recognition of the agreement or disagreement of ideas, cannot go beyond the limits of experience. (Locke's is an "idea empiricism," but knowledge requires an operation of mind in addition to the presence of ideas.)
5. Neither can meaningful language transcend experience, since the meaning of a word is an idea in the mind. Having the appropriate idea in mind is what distinguishes a person's from a parrot's uttering, "I want a cracker."
6. We each learn our native language by attaching public noises or marks (words) to ideas. We can then communicate our ideas to others by making the appropriate noises or marks.
7. Thought is a connected sequence of ideas.
8. The immediate objects of perception and thought are ideas in the mind, which in turn represent external things and situations (doctrine of representative perception, two-object theory of cognition).
9. All existing things are concrete and particular.

Empiricists immediately encountered the superstructure problem. Locke recognized that most meaningful words are general and many are abstract (rather than proper names of concrete objects, e.g., canine versus Lassie ), so how do we get the corresponding ideas (meanings) from experience, which furnishes only particular ideas? From an image of a particular triangle, said Locke, we can abstract from its being equilateral, isosceles, or scalene, and thus construct a general idea of a triangle that is "all of these and none of these at once." Berkeley and Hume improved on this unsatisfactory solution, but to this day empiricist abstraction accounts face serious difficulties. Hume added to the superstructure problem by denying the adequacy of reason alone to produce, from particular experiences, either (a) moral judgments, about what one ought to do, or (b) inductive conclusions, such as "All ravens are black" and Newton's laws. The former is his point that one cannot deduce "ought" from "is" or value judgments from objective facts, and the latter is the aforementioned problem of induction. Meanwhile, Berkeley had challenged Locke's empirical base by rejecting his distinction between primary and secondary qualities.

The Appearance-Reality Distinction

The two problems resurrect the old difficulty of bridging the gap between appearance and reality. Seventeenth-century advocates of the new science joined Plato in sharply distinguishing the world of everyday experience from underlying reality. Empiricists, with their limited resources, have tended to stick close to the experiential surface of the world by either narrowing the gap between appearance and reality, denying the existence of an underlying reality altogether, adopting the skeptical position that we simply cannot know it, or rejecting all talk of a reality beyond experience as "metaphysical" and hence meaningless. In short, they have wavered over commitment to the reality of unobservable entities and processes.

Locke denied that we can know the real essences of things. Our classifications are not natural but artificial—conventions made for human convenience. Hume and the later positivists, with their verifiability theory of meaning, ruled out metaphysics as meaningless. Ernst Mach (1838–1916), the Viennese physicist and positivist, denied the existence of atoms and developed a phenomenalistic account of the world. Berkeley had denied the existence of matter with his principle, "To be is to be perceived or to perceive" (Principles of Human Knowledge, 1710). Only minds and ideas exist. Does the cat then go out of existence when it disappears beyond the sofa? No, because God (the biggest mind) still perceives it, replied Berkeley. Mill later used a logico-linguistic device to remove the need for God and thus obtain a full-fledged phenomenalism. In An Examination of Sir William Hamilton's Philosophy (1865), he attempted to reduce physical objects to "permanent possibilities of sensation," expressible by (impossibly) long series of statements about what a person would experience or would have experienced in such-and-such a situation. Russell, using the new symbolic logic to the same end, attempted to reduce mind itself to a logical construction out of experiences. He took the same line for the postulated theoretical entities of physics: "Wherever possible, logical constructions out of known objects are to be substituted for inferred entities" ("The Relation of Sense Data to Physics," in Mysticism and Logic, 1917). This was a halfway house between realism and instrumentalism or fictionalism. If electrons are logical constructions out of actual and possible laboratory operations and the resulting observations, then they are not real entities of underlying reality; but neither are they complete fictions. Rather, electron talk is a convenient, economical façon de parler.

The Twentieth Century and Beyond

Twentieth-century thinkers abandoned or at least transformed British empiricism for its failure to solve the base and super-structure problems. These developments include: (1) The linguistic turn. Linguistic philosophers speak about terms in a language rather than, vaguely, about ideas in the mind. They also employ the full power of symbolic logic or the subtle devices of ordinary language to address the twin problems of relating subjective experience to basic statements and basic statements to superstructure. (2) The holistic turn. This is a further shift from the atomism of individual ideas or terms to whole statements, representing completed thoughts, and even to entire languages and conceptual frameworks. In "Two Dogmas of Empiricism" (1951), Quine argued that neither individual terms nor even full statements (not even basic observation statements) can be directly correlated with experience. Moreover, the data of experience logically underdetermine our theoretical claims. (3) Rejection of the analytic-synthetic distinction, also by Quine. We cannot factor theories into purely empirical and purely analytic components, only the first of which is vulnerable. "No statement is immune to revision," come what may, not even the statements of mathematics—for example, it is now known that Euclidean geometry is not the only conceivable geometry and that it is not even true. Quine's work called into question not only the concept of analytic statement but also that of analysis as a philosophical method, for no one has provided an adequate analysis of analysis! (4) Rejection of the scheme versus content distinction by Donald Davidson, who proclaimed this the third and last dogma of empiricism. (5) Rejection of the correspondence theory of truth and of (6) the linear-foundational model of justification. These doctrines give way either to a weaker, nonlinear and fallibilist foundationism or to a coherence theory of justification based on the idea of a mutually supporting network of claims and practices. For some, pragmatic problem-solving supplants truth as a goal of research. (7) Anti-Kantian Kantianism. Despite the rejection of Kantian intuition and synthetic a priori claims, logical empiricists Hans Reichenbach (1891–1953) and Rudolf Carnap (1891–1970) and historian Kuhn in different ways defended the need for larger structures, at least temporarily immune to serious revision, in order to make sense of the history of science as well as individual cognition. These structures are not mere hypotheses up for testing, alongside the others, for they are constitutive of experience and of normal scientific practice, in a quasi-Kantian way. To reject them would be to throw out the baby with the bath water. (8) Rejection by Karl Popper (1902–1994) and the positivists of the traditional identification of empiricism with inductivism, the view that we must gather and classify facts prior to theorizing. They developed a sophisticated, hypothetico-deductive model of scientific research, which was in turn subjected to severe criticism.

(9) Rejection of the imagist tradition that treats cognitive states or contents as little pictures before consciousness, and of (10) "the myth of the given," by Sellars and others, the idea that subjective experience provides a special, direct, infallible, nonnatural connection of knowing mind to known world. These difficulties highlight the problem of the empirical base. Insofar as our experiential claims are certain they are not about physical reality (because we have had to retreat into the certainty of our subjective sense data of the moment), and insofar as they are about reality, they are not certain (because they are now subject to override by other observers or even by widely accepted theories). The price of relevance is fallibility. Thus accepting a basic statement is a social decision. All conceptual thinking, including perception, is mediated by language (a further phase of the linguistic turn). There is no prelinguistic cognitive (conceptual) awareness. There is no thought, no fully human perception or scientific observation, prior to language. Roughly, "language games" (Wittgenstein's term) take over the role played by Kant's categories. All inquiry is thus fallible and mediated by language and by participation in an appropriate community of inquirers. The isolated Cartesian inquirer is a myth. The result is (11) the failure of phenomenalism and sense datum theories of perception and, more generally, (12) rejection of the whole Cartesian-Lockean conception of cognition and language. This conception is based on a Cartesian dualism of mind and body and, specifically, upon the privacy, immediacy, and alleged epistemological privilege of one's current mental contents. Philosophical and psychological behaviorism provided strong arguments against the Cartesian conception even for those thinkers, such as Sellars, who went beyond behaviorism.

(13) The failure of attempts to define knowledge precisely as justified true belief, which inspired (14) externalism versus internalism in epistemology. Internalism is the Cartesian-Lockean view that a person's knowledge claims must be justified in terms of the beliefs to which that person has access. The most popular form of externalism is reliabilism. According to process reliabilists, knowledge or justification consists of true beliefs formed by a reliable process whether or not the believer has sufficient Cartesian access to that process to justify it internally. Virtue epistemology, analogous to virtue ethics, is a variant of this idea: reliable beliefs are those formed by an intellectually virtuous process. (15) Recognition of the importance of tacit versus explicit knowledge (knowledge-how vs. knowledge-that) and of embodied knowledge, for example, skilled practices that we cannot fully articulate. (16) The feminist introduction of gender variables into epistemology. (17) Competing attempts to naturalize and socialize epistemology. Increasingly, empiricist philosophers work in the cognitive sciences, although few share Quine's view that epistemology will simply become a branch of psychology. Meanwhile, sociologists of knowledge regard their sociological approach as more fundamental than psychological studies of cognition. (18) The postmodern critique of empiricism. Postmodernists, including Richard Rorty and radical feminists and sociologists, regard empiricism, epistemology in general, and, indeed, the entire Enlightenment project to replace a tradition-bound life with modern life based on empirical science as a "modern" enterprise whose time is past. It is a mistake, they say, to abstract from sociohistorical contexts with their specific power and gender relations to seek the "one true account" of the world, as if there were a determinate world out there waiting for us to provide a correct description in its own language. Rather, say the critics, the world and our modes of inquiry are all socially constructed, as is empiricism itself. It is now time to deconstruct it. These controversial oppositions have generated "the science wars."

Although philosophical thinkers have abandoned both traditional rationalism and empiricism and although Quine, Davidson, and others have rejected the "dogmas" of empiricism and hence empiricism itself as a technical philosophical doctrine, there is a wider sense in which empiricism wins. For everyone is an empiricist in regarding observation and experience as crucial to justifying claims about the world, while almost no one believes that such claims can be defended purely a priori or on the basis of some kind of nonempirical intuition. However, this is no longer an empiricist epistemology in the old sense, for gone is the idea that epistemology commands special resources that can provide external or transcendental justification for any enterprise. The sciences, for example, can only justify their claims internally, by applying further scientific tests and by their own fruits.

See also Knowledge ; Positivism ; Rationalism ; Realism .

bibliography

Alston, William P. Epistemic Justification: Essays in the Theory of Knowledge. Ithaca, N.Y.: Cornell University Press, 1989.

Barnes, Barry, David Bloor, and John Henry. Scientific Knowledge: A Sociological Analysis. Chicago: University of Chicago Press, 1996.

BonJour, Laurence. The Structure of Empirical Knowledge. Cambridge, Mass.: Harvard University Press, 1985.

Dancy, Jonathan, and Ernest Sosa, eds. A Companion to Epistemology. Oxford: Blackwell, 1992.

Goldman, Alvin I. Epistemology and Cognition. Cambridge, Mass.: Harvard University Press, 1986.

Kornblith, Hilary, ed. Naturalizing Epistemology. 2nd ed. Cambridge, Mass.: MIT Press, 1994.

Longino, Helen E.. The Fate of Knowledge. Princeton, N.J.: Princeton University Press, 2002.

Quine, Willard Van Orman. "Two Dogmas of Empiricism." Reprinted in his From a Logical Point of View: Logico-Philosophical Essays. Cambridge, Mass.: Harvard University Press, 1953.

Rorty, Richard. Philosophy and the Mirror of Nature. Princeton, N.J.: Princeton University Press, 1979.

Schmitt, Frederick F., ed. Socializing Epistemology: The Social Dimensions of Knowledge. Lanham, Md.: Rowman and Littlefield, 1994.

Van Fraassen, Bas C. The Empirical Stance. New Haven, Conn.: Yale University Press, 2002.

Woolhouse, R. S. The Empiricists. Oxford: Oxford University Press, 1988.

Thomas Nickles