Positivism

Origins

The term was coined by Auguste Comte, but even for him it has several different connotations. It refers, in part, to a theory of history according to which every branch of knowledge passes through three stages (the theological, the metaphysical, and the positive state—when explanations by appeal to unobservable entities are finally abandoned) and which asserts that improvements in knowledge are responsible for historical progress. For Comte, positivism is also the assertion that there can be a science of society aiming at universal laws akin to those in the natural sciences; the name of a proposed secular religion, involving the worship of society, and with its own priesthood and church; and, less strangely, the label for a unity of science thesis claiming that all the sciences can be integrated into a single system. But perhaps the central thread in Comte's positivism, at least from the point of view of the subsequent history, is its empiricism, the view that the only source of knowledge [Page 647]is experience. This idea is taken from the British empiricists and leads (as it did with John Locke, Bishop Berkeley, and possibly David Hume) to the view that there can be no knowledge of any reality beyond experience. It also led Comte to acknowledge the impossibility of obtaining absolute truth. This knowledge turns out to be a perennial positivist theme and is worth noting in the light of a familiar tendency to claim that positivism involves a commitment to absolute truth as well as knowledge with certainty. At any rate, the pivotal nature of empiricist ideas in positivist thought means that positivism is, in effect, a variant of empiricism.

A project frequently associated with positivism is that of quantification; indeed, for some writers, positivism and the quantitative paradigm are more or less synonymous. The incorporation of statistics into positivist thinking is normally attributed to Émile Durkheim, who built on Comte's empiricism by combining the idea of a science of society with the tradition of social physics. This tradition had developed during the 19th century and involved the collection of statistics for largely administrative purposes. What was innovative about Durkheim's proposal was the claim that statistics could be used to construct and test social theories, not just for the purposes of administration or reform. This claim has since become entrenched in popular understandings of positivism, and as a result, Durkheim's work (particularly Suicide) is recognized as one of the classic examples of positivist sociology.

Logical Positivism

However, the most iconic version of positivism is associated with the Vienna Circle and the school of logical positivism that emerged from it along with an affiliated group in Berlin. The circle's 1929 manifesto emphasizes two fundamental commitments: to empiricism (i.e., there is knowledge only from experience) and to logical analysis, by means of which philosophical problems and paradoxes would be resolved and the structure of scientific theory made clear. It is, of course, the second of these commitments that represents logical positivism's distinctive contribution to the empiricist tradition.

Empiricism, then, was a premise of logical positivism, but there was much debate as to what counted as the experiential foundation of knowledge. For Ernst Mach, an important influence on the Vienna Circle, this foundation consisted of, quite literally, the scientist's own sense impressions, although some logical positivists held that sensations were not themselves the basis of science, but that protocol sentences—sentences recording those sensations—were. Later, however, this position was abandoned, at least by some positivists, in favor of the view that the experiential building blocks are in fact sentences that record not sensations, but the behavior of observable objects (e.g., measuring devices). In replacing sentences about sense impressions (which could be known to be true) with sentences about physical objects (which could not), this group of positivists effectively gave up the idea that knowledge could be certain and that there could be such a thing as absolute truth.

The logical analysis component of positivism has been based on developments in formal logic since the 19th century. Instead of a system of generalizations about psychological processes, logic was now seen as a formal symbolic language, empty of any empirical content that could be used to define precisely the conceptual relations between sentences. This development provided the logical positivists (or so they believed) with a means of translating theoretical sentences into sets of statements about experience and enabled them to organize the whole of scientific knowledge into an axiomatic system. These projects eventually broke down, as the positivists were the first to accept, partly because the translations were not forthcoming and partly because of the realization that no account of experience, no set of observations, can be theory-neutral. So, theoretical sentences cannot be translated into an observation language because observational terms are already theory-laden. Even so, two ideas persisted: first, that there are logical relations between theory and observation and second, that explanations consist of law-like generalizations from which the occurrence of specific events can be deduced. The latter is known as the covering law model.

Verifiability

The combination of empiricism and logical analysis leads to the principle for which logical positivism is best known: verifiability. This concept was an attempt to define a criterion capable of distinguishing between statements that are meaningful and those that are not (i.e., nonscientific, metaphysical statements). There were various formulations of this principle, but [Page 648]the basic idea was that any statement is literally meaningless whose method of verification cannot be specified in terms of experience. An associated principle, verificationism, held that the meaning of a statement just is its method of verification.

Verificationism eventually failed for the same reason that other logical positivist projects failed. Yet its position as a defining principle in the movement was strong enough to confirm the positivists in a form of antirealism. For example, if subatomic particles are not directly observable in experience and if it proves difficult or impossible to translate statements about them into the language of observation, then these statements cannot be regarded as meaningful. In which case, belief in the existence of subatomic particles is at best an optional extra. In fact, almost all the logical positivists were antirealists in this sense: They were, at the very least, noncommittal about the actual existence of unobservable entities. The same is true of their attitude toward law-like generalizations of the form all X are Y. Statements of this type cannot, strictly, be verified, as it is impossible to observe all Xs, so some logical positivists were equally skeptical about the meaningfulness of universal laws.

Summary

Like empiricism, then, positivism is a family of claims and concepts on which different authors have placed varying degrees of emphasis. It shares with empiricism a commitment to making experience the test of all knowledge and is skeptical about the idea of an unobservable reality that includes entities and forces not discoverable in experience, a skepticism that extends even to laws of nature. In its later forms, positivism adds to empiricism an enthusiasm for statistics—indeed, for quantification in general—and the assumption that if a statement is meaningful, then it can, by definition, be subject to scientific testing and verification (an assumption subsequently weakened or dropped). It also attempts to translate what is known into formal languages, including mathematics, and to organize scientific theory into logical structures. However, if there is an overlap with empiricism, there is also common ground with American pragmatism, which had a similar preference for experience, verifiability, antirealism, and operationalism. This common ground largely explains why the logical positivists were accorded such a favorable reception in the United States following their flight from Nazi Europe in the 1930s.

It is noticeable, though, that the resonances of positivism, as well as its variety, are now often ignored. The term frequently signifies what is regarded as an exaggerated respect for the natural sciences and is inevitably associated with quantification. Moreover, it is usually assumed that positivists believed in a determinate reality and in the possibility of a correspondence between that reality and representations of it. This image of positivism, approaching a caricature, has been boosted by postmodernism, which portrays it as a reactionary force, committed to oppressive universal truths, a chimerical objectivity, and foundational narratives. In this guise, it is a convenient foil for a great deal of recent writing on social scientific methodology.

Can Qualitative Research Be Positivist?

The study just referred to takes as its criterion for positivism an emphasis on measurement, the testing of generalizations, and the determining of relationships between variables using statistical analysis. So it [Page 649]would seem that qualitative research, by definition, cannot be positivistic. This definition does appear to be one use of the term reflected in the familiar claim that qualitative and quantitative research represent different paradigms, with the quantitative paradigm often identified with positivism. The justification for this claim, where it is not simply derived from the assumption that positivism equals numbers, is that doing quantitative research entails commitment to a particular ontology and, specifically, to a belief in a single, objective reality that can be described by universal laws. In contrast, it is suggested that qualitative researchers, almost ex officio, do not share this belief: They see the universe as inherently subjective, socially constructed, more subtle and complex than mathematics can accommodate, and comprising multiple realities. On the other hand, those who are skeptical of this position ask why the use of quantitative methods, or any other technique, should presuppose beliefs about the universe at all. They point out that tools and instruments are not usually regarded as having philosophical views built into them: using a spoon, for example, does not commit one to the claim that the world consists entirely of fluids and small particles. So why should the use of specific research methods, for specific purposes, commit one to the claim that the universe is subjective or objective, multiple or singular? The alternative is to take the distinction between positivist and nonpositivist as independent of the distinction between quantitative and qualitative, with the consequence that qualitative research can be positivist. Given this view, the various inquiry paradigms—positivism, postpositivism, critical theory, constructivism, the participatory–cooperative paradigm, and so on—can all be mapped onto qualitative research, each with its distinctive ontology, epistemology, methodology, and values and each manifested in a particular way of conducting qualitative studies.

Paradigm Tables

This mapping is evident in the tables that have become a familiar feature of qualitative methodological writing, especially by authors who draw on the work of Yvonna Lincoln and Egon Guba. Typically, the paradigms are represented in the columns while the rows represent inquiry issues on which different stands can be taken. For example, the positivist account of the nature of knowledge might be verified hypotheses established as facts or laws; the postpositivist version might be nonfalsified hypotheses that are probable facts or laws; the critical realist view might refer to structural–historical insights; the constructivist position might involve individual reconstructions coalescing around consensus; and so on.

One unfortunate feature of these tables, irrespective of the intentions of the authors, is that they imply a certain rigidity in the idea of a paradigm. It can appear, at least to the novice, that each column is a vertical tramline from which there is no prospect of escape. Acceptance of what positivism says about one of the inquiry issues irrevocably commits one to accepting what it has to say about all the others because every paradigm is in effect a package deal. Accept realism, for example, and one is thereby committed to accepting dualism, reductionism, absolute truth, certainty, correspondence, knowledge by accumulation, and an extrinsic ethic “tilting toward deception.” Once the positivist column has been entered, there seems to be no way out. It is a little like getting married and finding oneself stuck with an entire family. Yet it is clear from the history of positivism that there is no such rigid structure, that it is possible to be a positivist without being a realist and without believing in correspondence, dualism, or certainty. To this extent, paradigm tables oversimplify the philosophical issues they try to elucidate and demand that one embrace or reject in its entirety something misleadingly called positivism.

Parallels between Positivism and Qualitative Methods

In fact, there are several positivist ideas that sit comfortably with the claims of other paradigms and with the convictions of some qualitative researchers. For example, the positivist's instinct is to stick with the observable phenomena and to distrust any theory that purports to give an account of reality. This instinct is quite consonant with what qualitative methodologists recommend, whether they are grounded theorists (who say that theory must be semantically tied to data), phenomenologists (who aim at an atheoretical description of phenomena), or constructivists (who present multivocal accounts, building toward the achievement of consensus rather than a theoretical evaluation). In all these cases, there is a preference for not going too far beyond the data and for not invoking theoretical, but unobservable, social forces such as class, power, socialization, or culture. Even the multiple [Page 650]realities favored by constructivists are not too distant from the position arrived at by some logical positivists. Rudolf Carnap's mature view, for example, was that there are a number of different linguistic frameworks in terms of which the world can be described and that the choice between them is conventional and pragmatic, a matter of what is suited to a particular purpose. Consequently, all standards of correctness, validity, and truth are relativized to the rules and principles associated with whichever framework has been adopted. This view is not one that constructivists should find uncongenial.