Thomas Nickles. New Dictionary of the History of Ideas. Editor: Maryanne Cline Horowitz, Volume 5, Charles Scribner’s Sons, 2005.
The history of positivism falls into two nearly independent stages: nineteenth-century French and twentieth-century Germanic, which became the logical positivism or logical empiricism of the Vienna Circle that, in turn, enjoyed an American phase. In the postmodern world, “positivist” is often a term of abuse, but historical research now contests the received characterization.
In a broad sense, positivism is the philosophical expression of scientism, the view that empirical science is the primary cultural institution, the only one that produces clear, objective, reliable knowledge claims about nature and society that accumulate over time and thereby the only enterprise that escapes the contingencies of history. For positivists, that reliability is proportional to the proximity of claims to observed facts—the empirical basis of knowledge. Every substantive claim not tested by experience is sheer human fabrication. Positivists claim that they alone take fully into account the special nature and historical importance of science, with its actual and potential contribution to human life and culture. They reverse the traditional intellectual priority of science and philosophy (epistemology): philosophy is no longer prior to science but becomes the interpreter of and commentator on science. As W. V. Quine once quipped, “Philosophy of science is philosophy enough” (1976, p. 155).
Positivists aim to carry on the social mission of the scientific Enlightenment. The sciences, including the new human sciences, are to be unified under one method, usually with physics as the model. The positivists’ insistence that the hardheaded, allegedly value-free methods of the natural sciences (Naturwissenschaften) be extended to the human sciences or humanities (Geisteswissenschaften) has provoked charges of cultural imperialism from those defending historical, hermeneutical-interpretive, religious, or aesthetic modes of understanding (Verstehen).
In the broad sense, Karl Popper and even Quine are positivists, despite their trenchant critiques of the logical empiricists of the Vienna Circle (especially Rudolf Carnap), who achieved cultural authority in the twentieth century and with whom “positivism” in a narrow sense is now identified.
The Nineteenth Century: Comte To Mach
Although it owes something to the British empiricism of David Hume and to later radical empiricists such as John Stuart Mill and Alexander Bain, positivism as a movement developed on the Continent in France and later in central Europe, especially Vienna and Berlin. We can recognize positivist strains in the French Académie des Sciences around 1800, but it was the sociologist and philosopher August Comte who, in the 1830s, founded positivism as a distinctive movement, gave it its name, and also named the new science of social physics “sociology.” The conjunction was not accidental. For him, sociology was the final science, crowning the hierarchy of sciences, employing the same lawful methods as all positive sciences, and making possible a mature scientific philosophy. Comte is most famous for his law of three stages, which claims that civilization (and every field of knowledge) passes from a naïve, animistic, theological stage, through a more abstract, metaphysical-philosophical stage, to a final, scientific or “positive” stage. In the French tradition of Descartes and the encyclopédistes of the Enlightenment, Comtean positivism was an entire cultural system designed to fill the vacuum left by the French Revolution, which had swept away the religious and metaphysical ancien régime. Comtean positivism became an evangelical movement, with scientific humanism as the new religion and Comte himself as the high priest.
The law of three stages implied the need to demarcate science from other endeavors. Comte’s criterion was that scientific claims are predictive, which excluded not only metaphysics but also unstructured accumulations of singular facts. Positive science aims at lawful generalizations expressing invariable succession and resemblance, including laws of history and society—previously considered the domain of free human action and thus outside the scope of science. Positive science is cumulative and hence progressive. For Comte, something is “positive” insofar as it is precise, certain, useful, an organic organizing tendency for society, and relational rather than absolute. This last contrast means that Comte’s science seeks lawful correlations among phenomena rather than essences or underlying causes (the postulation of which smacks of metaphysics). It sticks to the observable surface of the world. “No proposition that is not finally reducible to the enunciation of a fact, particular or general, can offer any real and intelligible meaning” (vol. 3, p. 358). For Comte, explanation has the same form as prediction, namely subsumption of a fact under a general regularity rather than as the effect of a cause. Yet Comte also embraced the newly popular method of hypothesis against the old empiricist requirement that laws be induced from prior facts. All of these tenets except the strange Comtean religion are characteristic of later forms of positivism, although rarely via Comte’s influence. The great French sociologist Emile Durkheim did acknowledge a large debt to him.
Positivist strains are also evident in German scientific thinking in the decades before and after 1900, but it was Ernst Mach, physicist, historian, and philosopher of science, who made Vienna a center of positivist thinking. Positivists typically minimize the gap between appearance and underlying reality, at least knowable reality. Mach rejected atomic theory as empirically meaningless metaphysical speculation, at best of heuristic value; and his emphasis on the economy of thought led him to view scientific laws as rationally organized summaries of facts. Unlike the later positivists, he worked seriously on history of science (especially mechanics) and wrote on the processes of problem solving and discovery.
Logical Positivism and the Vienna Circle
The most developed form of positivism was the logical positivism or logical empiricism (LE) of the Vienna Circle. LE developed in three main phases: the first Vienna Circle from about 1907; the second Vienna Circle (the Vienna Circle proper), from the mid-1920s until about 1933; and the predominantly American emigrant phase after Hitler came to power. In all three cases the logical empiricists (LEs) were scientists, mathematicians, and scientifically trained philosophers who met to discuss substantive and methodological problems of science and society. The first circle was influenced directly by Mach and other scientists such as Heinrich Hertz, Richard Avenarius, Wilhelm Ostwald, Henri Poincaré, and Pierre Duhem, and by scientific developments such as non-Euclidean geometry, David Hilbert’s axiomatization of Euclidean geometry, and Einstein’s relativity theories. The second circle was heavily influenced additionally by Bertrand Russell and Alfred North Whitehead’s attempted reduction of mathematics to the new symbolic logic in Principia Mathematica (1910-1913), Ludwig Wittgenstein’s Tractatus (1921), Hilbert’s metamathematics, the new quantum theory, behavioral psychology, and antivitalistic progress in biology.
All three phases were also shaped by their respective social contexts. The first circle experienced the events leading to World War I and the final days of the Habsburg Empire, while the Weimar period framed the sociopolitical issues of the second circle. By contrast, the “end of ideology” characterized the American period, especially after World War II. Upon the emigration to America by members of the circle, the LE social program vanished. The American LEs presented their work as purely technical and hence politically neutral.
Many postmodern intellectuals, who think of the positivists as heavy-handed, dogmatic conservatives or as emotionless technical analysts disinterested in cultural affairs, are surprised to learn that the Vienna Circle assigned itself the urgent mission of reforming and transforming all of social and political culture by adapting to present conditions the program of the scientific Enlightenment. A major initiative was linguistic reform. The Viennese positivists’ animus against metaphysics was directed as much against obfuscatory and potentially totalitarian political discourse as it was against woolly philosophy. This is apparent at once in the manifesto of 1929, “The Scientific Conception of the World: The Vienna Circle,” by Hans Hahn, Carnap, and Otto Neurath in honor of their leader, Moritz Schlick. Modernist in outlook, the Vienna Circle celebrated the machine age and the transformative reconstruction (Aufbau) of Europe after World War I. It had close ties with a similar circle of scientific philosophers around Hans Reichenbach (Einstein’s colleague) in Berlin and with the Bauhaus school of design at Dessau, which in its own way emphasized clarity of structure shorn of all baroque, metaphysical adornment. Like the Bauhaus, the circle was international in outlook and pro–working class, and some members were politically active. Neurath was a neo-Marxist social scientist who radicalized the young Carnap, a logician. Schlick led a moderate wing.
When Schlick was assassinated in 1936, Neurath and Carnap became the leaders of the circle. It was in America that the indefatigable Neurath found a publisher for his dream project of a new, systematic encyclopedia of the sciences, but the overall project was a failure. Neurath died in 1945, and the University of Chicago Press published only twenty monographs of what was intended to be a long-term monthly subscription series. (These were later reissued as the two volumes of Foundations of the Unity of Science in 1955 and 1970.) One of the last contributions was Thomas Kuhn’s Structure of Scientific Revolutions (1962), commonly regarded as a refutation of logical empiricism. Meanwhile, Carnap, Herbert Feigl, Reichenbach, and Carl Hempel (a student of Reichenbach) headed the American phase of the movement. Ernest Nagel, although of a more Deweyan pragmatic cast, was a close associate. In America, unlike Europe, the aforementioned all had important academic positions, which they used to found the new specialty discipline of philosophy of science as well as to teach a new generation of philosophers, including Adolf Grünbaum, Wesley Salmon, and Hilary Putnam. With its rigorous formal methods, LE made the pragmatism of Charles Peirce, William James, and John Dewey seem quaintly dated and gradually displaced it as the official scientific philosophy. LE remained dominant until the 1960s and still casts a large shadow at the start of the twenty-first century.
The received view of the Vienna Circle is largely a post-Kuhnian construction that is now being contested. To be sure, the LEs wanted to make philosophy (or their replacement for it) a collective, progressive enterprise like that of the sciences, but the manifesto announced a more iconoclastic unity than was actually present. Accordingly, it was easy for opponents to miss the internal discord and tar all LEs with the same brush. Although the LEs were vehemently antimetaphysical and rejected most philosophy as a meaningless, fruitless pursuit of solutions to “pseudoproblems,” they were liberal in refusing to dogmatize about empirical questions and they viewed their group as open to discussion of all views. Another source of misunderstanding was A. J. Ayer’s inflammatory Language, Truth, and Logic (1936), the book that brought German positivism to an English-speaking audience. Ayer’s “potboiler” (as it has been called) mis-located the positivists in the British empiricist tradition.
Archival research sensitive to the intellectual and cultural milieu of central Europe later provided a major reinterpretation of the Austro-German positivist movement from Mach to Hempel. The participants came from varying academic backgrounds and life experiences and they frequently disagreed over matters of philosophical content as well as strategy and politics. They were their own most trenchant critics. For example, Kurt Gödel defended a Platonist (and hence metaphysical) ontology of mathematics. Neurath was out of sympathy with Carnap’s project to reconstruct science within a formal logical system and with Schlick’s commitment to the correspondence theory of truth. Neurath rejected the foundational, linear empiricist theory of justification, from supposedly infallible basic statements up through ever-higher levels of theory, in favor of a holistic coherence position featuring mutual support, a stance that he famously articulated in his ship metaphor: “There is no tabula rasa. We are like sailors who have to rebuild their ship on the open sea, without ever being able to dismantle it in drydock and reconstruct it from the best components” (Giere and Richardson, p. 83).
The LEs also disagreed over labels. Several members attacked “positivism,” and Reichenbach sometimes denied that he was a “logical empiricist.” (By the beginning of the twenty-first century, however, the inclusive term “logical empiricism” was commonly applied to both the Vienna and Berlin groups as well as the American contingent and was preferred to “logical positivism.”) Also, the views of the leading figures developed significantly over their lifetimes. Accordingly, a summary that is both brief and accurate is impossible.
Contrary to Ayer, the LEs had too serious an engagement with Kant to be squarely in the British empiricist tradition. They were anti-Kantian up to a point, with the political goal of displacing the neo-Kantians of the Marburg school (which included Ernst Cassirer) as the dominant school of scientific philosophy in Europe. The central problem was to retain what was correct in Kant’s critique of crude, British empiricism without commitment to Kant’s permanent categories and forms of intuition, which licensed synthetic a priori judgments. The latter are necessary truths that are knowable a priori yet make substantive statements about the universe, for example, that physical space is Euclidean and the laws of mechanics, Newtonian. Without them, Kant had said, mathematics and natural science would be impossible.
Kant had realized that sensory inputs do not automatically sort themselves into intelligible perceptions about which we can make coherent judgments. Coherent perception and thought must be actively constituted by the human mind by means of its processing rules (the categories and forms of intuition). Upon analyzing relativity theory, Reichenbach and Schlick concluded that Kant was partly right: science does need constitutive framework principles that are neither logical truths nor empirical claims subject to testing and in that sense a priori. But how, then, to avoid Kant’s commitment to a special, nonnatural intuition that yields synthetic a priori truths? Briefly, the LEs’ solution, anticipating Kuhn’s paradigms by several decades, was to disambiguate Kant’s necessary a priori from the constitutive a priori of framework principles and to regard the latter as based on human convention rather than Kantian intuition. For example, Reichenbach’s analysis of space-time theory bifurcated it into two components: a purely conventional component of “coordinating definitions” that define the meaning of measurement operations (and that we could change if it proved convenient to do so), plus a purely empirical component expressing the substantive content of the theory relative to the constitutive framework.
Stated in another way, the LEs’ problem was how to wed empiricism to logic and mathematics. As Kant had emphasized, raw experience is not the sort of thing that can enter into logical relations with statements, providing justificatory reasons or evidence. And analytic claims need their own special warrant. Carnap, the most influential LE, later widened the above approach to include logic itself. The axioms of a logical system are not self-evident to reason, he said, for there is no such thing as a special faculty of rational intuition. It is not even a question of epistemic correctness; rather, it is a question of human linguistic convention—choice of language. The choice is pragmatic, not epistemic. We may freely choose any formal system we like and explore its consequences, keeping those systems that produce the most fruitful consequences for our purposes. Thus we arrive at the mature LE view that all and only empirical statements are synthetic and all and only a priori statements are analytic (in the pragmatically grounded sense). On this view, the a priori–a posteriori distinction coincides with the analytic-synthetic distinction. There is no synthetic a priori.
Where does philosophy fit into this scheme? For Carnap its task is purely analytic—Wissenschaftslogik, the logical analysis of the language of science using the tools of symbolic logic. Scientific philosophers clarify the logical structure of empirical science but do not do empirical science. Thus was born both mature analytic philosophy and philosophy of science as a specialty.
Logical Empiricist Themes—and Their Reception
What follows is a list of several interlocking theses and projects and their outcomes, several of which were controversial among the LEs themselves.
1. The verifiability theory of meaning.
A sentence is empirically meaningful if and only if it is verifiable in principle and (roughly) its meaning is given by the method of its verification. The LEs quickly rejected full verifiability in favor of weaker forms of testability. However, they were never able to formulate an adequate formal criterion of meaning or justify the equation of meaning with empirical evidence. It was this “verificationism” that backed the LEs’ anthropomorphic claim that all genuine problems are empirical and therefore humanly solvable and their dismissal of all metaphysical problems as pseudoproblems. Since competing metaphysical positions, by definition, have no empirical consequences, they cannot differ in meaning; so there can be no meaningful problem of choosing between them.
2. The attack on metaphysics as meaningless.
The LEs agreed that an enlightened society has no room for metaphysics; however, they sometimes disagreed over what counts as metaphysics.
3. A sharp fact-value distinction and emotive ethics.
Ethical and aesthetic utterances are emotional reactions. Since they are not empirically testable, they have no cognitive meaning and cannot be true or false. Nonetheless, the early LEs took a strong normative stance on social and political issues.
4. The observational-theoretical distinction.
The project to distinguish epistemically unproblematic observational terms and sentences from the theoretical ones and legitimize the latter in terms of their logical relations to the former ran into similar difficulties despite important progress such as Carnap’s treatment of dispositional terms. N. R. Hanson, Popper, Kuhn, Paul Feyerabend, and Putnam noted that scientific observational language is “theory laden” and that there is no context-free, linear gradation of theoreticity.
5. The analytic-synthetic distinction.
Quine’s influential critique of this pillar of LE (and of much else), in “Two Dogmas of Empiricism” and other writings, and his return to a pragmatic naturalism were heavy blows. The second dogma was “radical reductionism,” the idea that each sentence in isolation can be correlated with a range of experience.
6. Application of the resources of the new symbolic logic and the third dogma.
The LEs (especially Carnap) developed and applied the new symbolic logic in ingenious ways, mainly in terms of purely syntactic rules; but later critiques by Quine, Hempel, Nelson Goodman, and others showed that semantic and pragmatic considerations are unavoidable, effectively dooming Carnap’s project to produce objective, ahistorical, context-free languages of science. By the 1950s and 1960s, philosophers increasingly felt that the LEs had exhausted the resources of deductive logic without adequately capturing the richness of scientific reasoning. Static logical relations seemed especially incapable of modeling scientific theories and deep conceptual change, for example, scientific revolutions. Commitment to deductive logic by the LEs and Popper has been called the third dogma of empiricism. (Others, following Donald Davidson, give this label to the so-called scheme-content distinction.) Reichenbach all along had urged a probabilistic approach (a theme continued by his student, Salmon), although he and Carnap had developed probability theory in roughly opposite ways.
7. Logical analysis of scientific confirmation, explanation, and theory structure.
The LEs admirably articulated old and new ideas here in terms of detailed models. Their work set the standard for ongoing research in these areas. Hempel’s extension of “covering law” explanation to history, psychology, and the social sciences was especially controversial since it challenged old ideas about human freedom and spontaneity and the need for sympathetic understanding. Yet it also failed to capture the force of causal explanation.
8. The unity of science.
Pitting the sciences against the rest of culture, some LEs defended the unity of science on three fronts: conceptual, doctrinal, and methodological. Conceptual unity means that there is one universal language of science; doctrinal unity, that more complex disciplines such as biology are ultimately reducible to more basic disciplines such as chemistry and physics; methodological unity states that there is one general method of science, that all legitimate theories, all explanations, and so on possess the same logical structure. All of these projects produced interesting results, but have since been abandoned. In the antireductionist, more pragmatic atmosphere of the early twenty-first century, the emphasis is on diversity, on teasing out the differences among the various sciences rather than on trying to model all of them on physics; and physics itself turns out to be internally diverse. Most experts reject the existence of a unique scientific method as a fiction of textbooks and school administrators.
9. The fall and rise of naturalistic epistemology.
Prominent LEs followed Gottlob Frege in sharply distinguishing logic and epistemology from psychology; “psychologism” was the fallacy of confusing them. It was on this point that the positivists differed most obviously from the American pragmatists. (As usual, the most important exception was Neurath, who promoted a naturalistic epistemology.) But the LEs’ own critique of Kant’s transcendental epistemology could be viewed as a step toward a naturalistic epistemology. Quine took the next step and urged a return to a naturalistic pragmatism, contending that epistemology should become a branch of behavioral psychology. Historical case studies by Kuhn, Feyerabend, and their followers showed that the failure of LE and Popperian methodologies to fit actual history was so great as to raise the question whether anything recognizable as science could fit the old rules of method. Since (as Kant said) “ought” implies “can,” the critics thereby showed that empirical information is relevant to and can in a sense “refute” a methodology despite its normative character.
This surprising turn does fit Quine’s pronouncement that “no statement is immune to revision,” come what may—not a conventional or “analytic” statement or even a normative one. The critics increasingly perceived some problems in the philosophy of science as artifacts of the LE approach and hence as pseudoproblems with respect to real science. Kuhn famously distinguished normal science under a paradigm (a quasi-Kantian categorical scheme that made normal science possible and intelligible) from revolutionary science, neither of which fit the tenets of either LE or Popperianism. In “the battle of the big systems” (initially among the LEs, Popper, Kuhn, Feyerabend, Imre Lakatos, and Stephen Toulmin), many considered Kuhn the winner, although many philosophers severely criticized his work. And yet Carnap, who increasingly thought in terms of free, pragmatic choices among available linguistic frameworks, welcomed Kuhn’s contribution as making a similar point. The work of Kuhn and Feyerabend brought all the above-mentioned difficulties of LE to a focus, which hastened its demise as the generally accepted account of science. No similarly grand consensus has replaced it.
10. The discovery-justification distinction.
This distinction was an LE bulwark against psychologism. The basic idea is that it does not matter how or why a theory or problem solution pops into someone’s head; what matters is how the claim is tested. There is a psychology of discovery but no logic of discovery, only a logic of justification. Philosophy is concerned only with the logically reconstructed products and not the processes of science. The LEs’ applications of the discovery-justification distinction drew philosophy of science closer to philosophical problems of logic and epistemology and away from the study of science as practiced by communities of scientists. Since the new historical case studies were precisely the study of the process of investigation, they challenged these applications.
11. The emergence of science studies.
As scientific insiders, the original LEs relied on their own knowledge and intuitions about how science works (or should work) and, qua analytic philosophers, saw no need for careful empirical studies of the sciences themselves. The Kuhnian revolution changed that. Their sharpest critics noted the irony that the logical empiricists urged everyone else to be empiricists but themselves! But while post-Kuhnian philosophers began to take the history of science seriously, they did not study in detail the scientific practices of contemporary science. They thereby left an opening for the new empirical sociology of scientific knowledge that has since grown into a multidisciplinary “science studies,” often defined to exclude philosophy.
12. Realism versus social constructionism and “the science wars.”
As strong empiricists, early LEs tended to advocate a minimalist stance toward theories and the entities that they appeared to postulate. Some regarded theories as instruments for calculation rather than as attempts truly to describe underlying reality. Carnap himself used Russell’s maxim as a motto: “Wherever possible logical constructions are to be substituted for inferred entities.” This is a logical constructionist position. Most science studies practitioners are also constructionists, but social constructionists. They deny that science is a special, epistemically privileged institution and regard its results as negotiated social constructions. In reaction, many philosophers now take a “realist” position that affirms objective, scientific progress toward truth and vigorously denies the cultural relativity of scientific results. As heirs of the Enlightenment, they reject the centrifugal tendencies of postmodernism and defend the special place of the sciences in human culture. This heated debate among philosophers, science studies practitioners, culture theorists, and scientists themselves is commonly known as “the science wars.” If the postmodernist critics are right, Comte’s law of three stages stopped at least one stage too soon!
13. The linguistic turn and the rise and fall of narrowly analytic philosophy.
With G. E. Moore, Russell, and the early Wittgenstein as precursors, the LEs, and especially Carnap, were the primary developers of analytic philosophy. After World War II, a wider sort of linguistic philosophy, “ordinary language philosophy,” flourished at Oxford. Both movements construed philosophy itself as a metadiscipline concerned to analyze language rather than to address substantive questions about the world and human activity. But since the 1960s, Anglo-American philosophy has become more liberal in its interests and methods. Philosophers once again vigorously discuss the metaphysical issues rejected by the LEs as pseudoproblems, and there is even something of a rapprochement with the so-called Continental philosophy of Heidegger and his successors. Carnap dominated the American phase and the received view of LE; but at the turn of the twenty-first century, many experts were examining Neurath’s position in greater detail and finding it more defensible.