Empiricism

John Paley. Encyclopedia of Qualitative Research Methods. Editor: Lisa M Given. Sage Publications, 2008.

The central claim of empiricism is that experience is the foundation of knowledge and that the project of gaining access to a reality other than experience is problematic. However, like positivism, a term with which it is closely associated, empiricism has been used to designate different claims and tendencies during its long history, and the concept has evolved to such an extent that those who are now regarded as copybook empiricists—for example, the British empiricist trio of John Locke, Bishop George Berkeley, and David Hume—were strongly inclined to reject that description of themselves. Both terms suffer from radical ambiguity, for just as positivism cannot be identified with a single view uniquely defining a distinctive position, so too is empiricism extremely difficult to pin down precisely, especially as many of the ideas routinely labeled positivist could, with as much or as little justification, be equally described as empiricist. It might, therefore, be sensible to read this entry alongside the corresponding one on positivism.

The entry begins with a review of empiricist ideas in philosophy and then considers the impact of these ideas on the social sciences, specifically the way in which they are reflected in qualitative methods.

Empiricist Claims in Philosophy

Historically, there are a number of beliefs and attitudes that have been attributed to empiricist authors. Most frequently cited, perhaps, is the claim that the only source of knowledge is experience. However, there are ambiguities in this sort of formulation, and it gains in precision only as alternative views are specified and rejected. For example, granted that experience is the only source of knowledge, does an empiricist permit operations to be performed on experience—and, if so, what kind of operations? Some empiricists think that the only permissible type of operation is simple numerical induction; others, such as John Stuart Mill, reject this stringent limitation and argue for more sophisticated inductive methods (including, in Mill’s case, the methods of agreement and difference) capable of identifying causes and effects. In Mill’s terminology, it is empiricism that he was repudiating, although he took himself to belong to the “school of experience” rather than to the “school of intuition.” The irony is that, according to modern typologies, he would be classed as an empiricist par excellence.

Bacon’s Insects

Mill’s use was inherited from Francis Bacon, who compared empiricists to ants, “merely collecting and using,” and contrasted them with dogmatists or “spiders” who “spin webs out of themselves.” But according to The Philosophical Works of Francis Bacon, published in 1905, Bacon’s preferred insect was the bee, which gathers “flowers from the garden” and then “by her own powers transforms and digests them; and the real work of … [science] is similar” (p. 288). The empiricist tendency to which Bacon metaphorically objected, then, was the mere aggregation of “findings” as opposed to the kind of intellectual work that does something creative with them—although it is equally opposed to dispensing with empirical data in favor of philosophy, dogma, and religion.

Innate Ideas

Next is a claim that differentiated the 17th- and 18th-century British empiricists from the continental rationalists. Locke, for example, argued that the mind is originally a tabula rasa and that all ideas are the result of experience literally imprinting itself on this blank sheet. This is in contrast to Gottfried Leibniz, for example, who believed that the mind is more like a block of marble, with innate ideas already threaded into it and ready to be sculpted by whatever experience brings. This distinction is clearly very different from the one between spiders, ants, and bees. It refers not to any methodological alternatives but rather to a view about the nature of the mind. Moreover, it is clear that the two distinctions are independent of each other; it would not be any more inconsistent for Bacon’s ant to believe in innate ideas than for the spider to believe in tabula rasa. This observation anticipates a problem (which is typical of positivism as well as empiricism); the views associated with the label do not necessarily need to belong together. The ant, for example, is an empiricist in Bacon’s sense, but not necessarily in Locke’s sense.

Nominalism

Another aspect of Locke’s doctrine is his nominalism. This is the view that general descriptive terms refer not to real structures or qualities in the world but rather to ideas derived purely from sense experience. Locke’s claim was that we can form abstract ideas on the basis of the particular things we experience but that the words associated with these ideas do not refer to anything beyond experience itself. There is an “external” fundamental reality that underlies our experience, but it is not something to which we can have access, and the general terms we use do not apply to it. In other words, concepts reflect only the organization of experience; they do not reflect the way in which reality is “carved at the joints.” Locke, then, rejected the view that it is possible to know the “real essences” of things (although in fact there are real essences) and argued that the objects of knowledge are “nominal essences”—the various combinations of experience to which we give names. In Locke’s form of empiricism, there is an unknowable reality as well as experience. In other forms (e.g., Berkeley’s), there is no unknowable reality because the only reality is experience itself.

Against Certainty

A significant corollary of nominalism is the view that experience is always specific, never general. We apprehend particular things, not universal properties, and still less universal truths. It follows from this that any general claims we do make must, by definition, go beyond the particularity of experience and, therefore, are less than certain. So not only is it true that we can never know about an underlying reality, but also generalizations about experience cannot be guaranteed either, although they may be assessed in terms of probability or, in Locke’s terms, “likeliness to be true.” To this extent, therefore, empiricism represents a principled resistance to speculation about the real world, appeals to entities and forces that cannot be observed directly, and the assumption that some things can be known with certainty.

Because empiricism takes all legitimate beliefs to be derived from experience, it is normally regarded as a foundationalist position. Even here, however, there is something of an ambiguity. Empiricism is quite clearly foundationalist in the sense that all knowledge must ultimately be referable to experience, but that is only one rather weak criterion. A stronger criterion is that, for a belief to count as knowledge, it must be possible to demonstrate its correctness, starting with particular experiences and deriving the belief from them according to some rule of inference. This is obviously a more ambitious project, and not all empiricists have aspired to it. Locke, for example, could not be classified as a foundationalist in the second sense because his rejection of the claim that generalizations can be known with certainty shows that he did not believe that demonstration of this kind was possible (at least for that type of belief). So the idea that empiricism is intrinsically foundationalist should be treated with caution.

Observation and Theory

Still, empiricism does require a distinctive account of theory and theoretical ideas. Scientific theories, especially in physics, seem to refer to entities that do not occur in experience (e.g., subatomic particles). So, if the empiricist believes that all knowledge is derivable from experience, she or he must explain how this is possible. A number of solutions to this problem have been proposed. The classic solution involves claiming that all theoretical concepts are reducible to (i.e., definable in terms of) observational language. Wherever a theoretical term is used, it can in principle be translated into claims about what has been, or might be, observed—even if such a translation is unlikely to be forthcoming in practice. If this solution can be made to work, theoretical concepts simply become convenient forms of shorthand, ultimately equivalent to sets of actual and possible observations.

One of the main reasons this does not work, however, is that it implies a certain type of distinction between observation and theory. Specifically, it implies that there must be experiences that can be identified and described independent of theory (because all theory is reducible to the language of observation). But the overwhelming consensus in philosophy of science during the past 40 years or so is that observational terms are “theory laden”—that it is impossible, in other words, to describe even the simplest observation without making reference to some theory. So, the classic solution, it is now universally agreed, fails.

The idea that all observation is theory laden has another apparent implication—that “plain observation” is unable to adjudicate between competing theories simply because observation statements have theory built into them and so cannot (after all) be “plain.” But if that is correct, then it would seem to follow that it is futile to attempt to determine the truth of any matter because there is no neutral, theory-independent way of adjudicating between theories. This view seems to lead to some form of relativism and has had a significant impact on methodological writing in the social sciences.

Summary

On this account, then, the claims most commonly attributed to empiricism are that experience is the only source of knowledge (although there are different views about what sorts of operation on experience are permitted, ranging from mere ant-like aggregation to intellectual procedures of varying degrees of sophistication); experience is the foundation of all knowledge (opinions differ on how a strong a claim this is); there are no innate structures in the mind; we experience particular things, not universal truths; we have no access to reality (unless we take experience to be the only reality) and cannot know things with certainty; theoretical statements are ultimately a form of shorthand, translatable into accounts of what has been, or might be, observed; and observation can hope to adjudicate between competing theories and determine which of them is more likely to be true.

Empiricism and Social Research

It is presumably true to say that the majority of social scientists are empiricists in the weakest sense of the term, which involves the claim that experience, in the form of observation, is the ultimate source of knowledge without any specific implications about what forms that observation should take or about the nature of the relation between observation and theory. But this is only to say that most social scientists are committed to empirical inquiry and that very few of them believe that significant conclusions can be drawn on the basis of pure speculation, theology, philosophy, or unsupported intuition. In a similar way, the existence (or not) of innate ideas plays virtually no part in methodological thinking (although many cognitive scientists believe that the mind has innate, or innately channeled, processing systems). So, for the purposes of this discussion, these characteristic tenets of rationalism can be left to one side.

Dust Bowl Empiricism

However, other elements of empiricist philosophy do surface in writing about sociological method. For example, Bacon’s complaint about empiricist “ants” is reflected in rather similar objections to the “mere collection” of social facts, an approach to inquiry that is likewise dismissed as empiricist. It is argued that arid accumulation of this kind is atheoretical; it offers no explanations, tests no hypotheses, solves no intellectual puzzles, suggests no interpretations, and therefore provides no real understanding of the social world. This is sometimes called “dust bowl empiricism,” a metaphor derived from the term used to describe the dry, dust storm-ridden plains of Texas, Kansas, Colorado, and Oklahoma during the 1930s. The idea is that, in the absence of theory, a heap of unconnected facts is as barren as the American Dust Bowl. (On the other hand, it could be argued equally well that in some areas of social research, such as market research and public sector statistics, empiricist fact collection is justifiable.) Although this “abstracted empiricism” is most commonly associated with quantitative methods, it can also take a qualitative form. Indeed, some phenomenological researchers make a point of being purely descriptive, collecting accounts of respondents’ experiences in a manner that is self-consciously uninformed by theoretical considerations. Arguably, this is no less “dust bowl empiricism” than the stacking up of bare statistics.

Empiricism and Qualitative Methods

Fact gathering is the form of empiricism that both Bacon and Mill rejected. But if empiricism, taken less narrowly, begins with experience and derives theories and explanations inductively from data (as in Mill’s own account), then commonly adopted methods in qualitative social research can evidently be classified as empiricist. Grounded theory is an obvious example. The whole point of grounded theory is to build theory out of data, with theoretical terms defined in terms of codes emerging from the analysis of data and, therefore, semantically bound to evidence. In this respect, grounded theory is a qualitative version of operationalism, the empiricist strategy more frequently associated with quantitative methods, in which theoretical concepts are defined by how they are measured. In qualitative work, measurement is not in question; however, if the meaning of a theoretical concept is tied strictly to the procedures for analyzing data as in grounded theory, then the link between observation and theory is comparable. In a similar way, the analytic induction tradition, extending from Mill to Charles Ragin, generates explanations on the basis of purely logical relations between qualitatively defined variables without any reference to external theoretical constructs.

The contrast here is with independently defined theoretical terms and inference to the best explanation (sometimes called “abduction”). The recent interest in social mechanisms provides an excellent example of an approach—largely qualitative—that is certainly not empiricist because it postulates mechanisms (conceivably unobservable themselves) underlying observable phenomena and finds evidence for their existence, and the nature of their operation, by testing corresponding hypotheses during data collection. Although there are several variations on this theme, they all are rooted in an explicit critique of empiricism and positivism. By the same token, they are realist in orientation, taking as a premise that there is a knowable reality behind appearances (in opposition to Locke), a reality that cannot be identified just with our experience of it (in opposition to Berkeley). Qualitative methods suited to this approach have been discussed, particularly in the context of evaluation research and case studies.

Observation as Theory Laden

The view that all observation is theory laden complicates matters and has encouraged many writers to reject both empiricism and realism. Because observation, according to this view, is not neutral and cannot be described independent of prior theoretical commitments, the operationalist strategy of grounded theory is blocked (the antiempiricist argument). At the same time, however, it is impossible to adjudicate between competing realist theories given that observation cannot provide an independent court of appeal (the antirealism argument). This line of thought leads to constructivism, interpretivism, and hermeneutics because the only strategy left open is the exploration of various accounts of experience, each of which will embody different theories, different interpretations, and different preconceptions. These positions imply that the researcher’s own account is just one interpretation among others, one that cannot be shown to be more “accurate” or more “true” than any other, for this would require neutral, non-theory-laden observations to test competing accounts—and that, according to the premise, is impossible.

Or so the argument goes. But one familiar counterargument should be mentioned. This recognizes that observation is theory laden but points out that the theory being tested is not necessarily the same as the theory that is built into the observation (in fact, this is highly unlikely). The project of adjudicating between theories need not be abandoned, therefore, because there is no circularity involved.

Parallels between Empiricism and Constructivism

Despite their rejection of some empiricist claims, constructivists and interpretivists are very close to empiricism in other respects. They share with the classical empiricists a skepticism about “universal truths” and generalizations, preferring to focus on the particularities of unique situations and experiences. Like Locke, they are hostile to the idea of reality, the aspiration to certainty, and the assumption that it is ever possible to check for a “correspondence” between reality and theory. Like Berkeley, they believe that (interpreted) experience is all there is, although their idealism is more extravagant than Berkeley’s because they favor multiple realities instead of just one reality. The concepts they develop cannot be identified with Locke’s real essences but instead are nominal essences, as “constructed” as the concepts used by their research participants. Only in studies allegedly based on the phenomenology of Edmund Husserl (especially in nursing research) is there any ambiguity about whether the “essential structure” of a phenomenon is essential to somebody’s experience or is essential to something (the phenomenon) that is independent of that experience.

Conclusion

Empiricism, then, is a family of claims, with not all of them compatible with one another, and to that extent it makes little sense to either reject or embrace empiricism tout court. As with positivism, however, the assumption that it represents a single, coherent unified paradigm has taken hold over the past 40 years or so, making it more difficult to evaluate individual claims on their own merits. The unexamined view that all of these claims stand or fall together has tended to polarize methodological and epistemological discussion and creates the impression that the empiricist “package” can only be rejected in its entirety. Still, perhaps the recent reexaminations of the history of both empiricism and positivism in philosophy will loosen some of the more rigid beliefs, and prompt qualitative researchers into recognizing that they do not necessarily need to abandon every empiricist idea just because they have rejected one.