Costantino Cipolla & Guido Giarelli. Encyclopedia of Sociolog. 2nd edition, Volume 2, Macmillan Reference USA, 2001.

The term epistemology is used with two separate meanings according to different cultural traditions. In English-speaking countries, epistemology denotes the philosphical theory of knowledge in general: in this sense, it includes themes and problems such as the question of the possibility of valid knowledge, the analysis of the nature of such validity, the foundation of knowledge on reason or on experience and the senses, the analysis of different types of knowledge, and the limits of knowledge. In continental Europe, the above issues are considered part of the field of the more general discipline of gnoseology (gnoséologie in French, gnoseologia in Italian and Spanish) or theory of knowledge (Erkenntnistheorie in German), whereas the aim of epistemological inquiry is restricted to scientific knowledge.

In this second sense, with particular respect to social sciences (and sociology, in particular) the fundamental epistemological question becomes: “Is it possible to acquire any valid knowledge of human social reality? And, if so, by what means?” As these questions show, epistemological issues are inescapably interconnected with methodological problems; however, they cannot be reduced to simple technical procedures and their validity, as a long empiricist tradition among sociologists has tried to do. A full epistemological awareness, from a sociological point of view, should cope with at least four main issues:

  1. Is the nature of the object of social sciences (i.e., social reality) fundamentally different from that of the object of natural sciences (i.e., natural reality)?
  2. Consequently, what is the most appropriate gnoseological procedure with which to study and understand social reality?
  3. Are we sure that the particular knowledge we get by studying a particular social reality can be generalized?
  4. What kind of causality can we postulate between social events, if any?

Historically, the various sociological traditions or schools have answered these four questions differently. We shall trace them by following the three main epistemological debates that took place between three pairs of opposing schools of sociological and epistemological thinking: positivism versus historicism, logical empiricism versus dialectical theory, and realism versus constructivism.

Positivism Versus Historicism

At the beginning of the nineteenth century, the discipline of sociology originated as the positivistic science of human society and aligned itself against metaphysical and philosophical speculation about social life. This, according to Auguste Comte, the founder of sociology, implied the straight application of the scientific method used in physics and the other natural sciences to the analysis of human society. His social physics (Comte 1830-42) is the expression of the new culture of the Industrial Revolution and of its practical ambition of a totally planned social life. Comte’s positivistic doctrine is organic and progressive and holds a naive faith in the possibility of automatically translating scientific knowledge of the laws of society into a new harmonious social order.

The uncritical assumptions of Comte as to the reliability and universal applicability of scientific method were at least partially tempered by John Stuart Mill, who considered illiberal and doctrinaire Comte’s idea that, once sociological laws are established by the same scientific method of natural science, they can no longer be questioned. Mill objected that scientific knowledge never provides absolute certainties, and he organically settled the British empiricist tradition in his A System of Logic, Ratiocinative and Inductive (1834). Starting from the epistemological positions of David Hume, Mill considered knowledge as fundamentally grounded on human experience, and the induction as its proper method. The possibility of inductive generalization is based on the idea of regularity and uniformity of human nature, ordinated by laws; and even social events are correlated in the form of empirical laws, which we can understand using the scientific method, considered as a formal logical structure independent from human subjectivity.

The individualistic and liberal features (based on utilitarian ethics) of British positivism were particularly evident in Herbert Spencer, who opposed to Comte’s organicistic view of society the greater importance of the particular with respect to the organic whole. His central focus on the concept of evolution supplied him with a synthetic perspective with which to study reality as a whole, using the analogy of biological and inorganic evolution to consider social evolution (1876-1896). However, the most complete development of a positivistic epistemology in sociological thought is certainly contained in Emile Durkheim’s work Les régles de la méthode sociologique (1895). What he terms faits sociaux (social facts) can be characterized according to four criteria: two of the criteria are related to the subjects observed, whereas the other two are concerned with the sociologist observing them.

The first criterion considers the social facts as a reality external to the individual conscience: social institutions have their own life independent from individual life. The second criterion is that this external, objective nature of social facts consequently confers upon them a normative, coercive power over the individual: these social facts impose themselves on him, even without his will. Morals, public opinion, law, customs are all examples of this. These two features of social facts require, on the part of the social scientist observing them, the observance of two further qualifying criteria. Firstly, they should be considered as “things,” that is, they should be studied as an external, objective reality, separate from the individual consciousness. Secondly, consistent with their nature, social facts can be explained only by other social facts: the nature of social causality is specific, and can be reduced neither to psychological causes of individual behavior, nor to biological causality, as Spencer’s evolutionism suggested. The emerging nature of social phenomena from the level of psychological and biological phenomena grounds their autonomous, distinct reality: this, in turn, constitutes the specific field of sociology and of the other social sciences. To explain social phenomena, sociology should adopt a unilinear concept of causation: the same effect always corresponds to the same cause. Any plurality of causes, according to Durkheim, involves the impossibility of sorting a scientific principle of causality.

In sum, we can say that the classic positivistic paradigm of sociology is characterized by the recognition of the specific nature of social facts as emerging from the other spheres of reality. However, this implies an idea of sociology as a naturalistic science of society, using the same methods already applied with success in natural sciences (methodological monism), and grounded on experience as perceived by the senses and generalized on the basis of universal human nature (induction). Finally, a proper explanation of social facts should take into account only one cause for each effect (mono-causal determinism).

A serious challenge to this epistemological view of sociology came from the German Historical School: the debate about the method (methodenstreit) which took place in Germany during the last decades of the nineteenth century is the first example of epistemological debate in sociological history. It started with Wilhelm Dilthey, the major representative of German historicism, who argued against Comte and Mill’s proposal of introducing the methods of physics into the “moral sciences.” What he termed “spiritual sciences” (Geisteswissenschaften) were of a radically different type from the natural sciences and therefore they could not share the same method (Dilthey 1833). In fact, their objects were states of mind, spiritual experiences that could be apprehended only by means of an “empathic understanding” (Verstehen): “we explain nature, we understand psychic life” (ibid.). In the first phase of Dilthey’s thought, this understanding, on which the sociohistorical nature of the spiritual sciences is grounded, was not considered to be mediated by sense perception, but rather as producing a direct and immediate intuitive knowledge-by-acquaintance. In the second phase of his thought (1905), however, he believed that psychic life was not immediately understandable, but would require the hermeneutic interpretation of its objective displays in cultural life. In any case, his distinction between the aim of natural sciences (to explain erklären data by means of external senses) and the scope of human sciences (to understand verstehen through an intrapsychic experience) was posited as a long-lasting distinction in the social sciences.

Wizhelm Windelband and Heinrich Rickert, two other adherents of German historicism, criticized Dilthey’s distinction between nature and spirit and the related sciences, since they considered human knowledge as always a spiritual activity, without regard to its object. In its place, they proposed a distinction based on the form and on the different type of methods used: whereas the idiographic method is a description of singular events, the nomothetic method is concerned with the inquiry of regularities and general laws (Windelband 1894). The first type of procedure is typical of—though not exclusive to—the sciences of the particular, the sciences of culture that, being historical, had to interpret and understand the individual character of the historical event; the second type of method is more typical of the sciences of the general, natural sciences that are aimed at establishing general laws. Knowledge is always a simplification of reality, which is a heterogeneous continuum (eterogenes Kontinuum); knowledge can either proceed by making a homogeneous Kontinuum, as the natural sciences do, or by sectionalizing a portion of heterogeneous Diskretum, such as the sciences of culture do. Rickert (1896) further developed the necessity of referring to values as a foundation for sociohistorical knowledge: the sciences of culture (Kulturwissenschaft) are not idiographic disciplines only, they should also refer to value relevences (Wertbeziehung) if they want to understand the actual meaning of sociohistorical events.

The most significant attempt to reconcile the two antitheses (understanding/explanation and idiographic/nomothetic) has been carried out by Max Weber, the founder of German sociology, who was profoundly influenced by German historicism, although critical of its idealistic orientation. In his methodological essays (1904-17) and in subsequent works, Weber undertakes an inquiry into the nature and validity of methods used in the social sciences and presents a general epistemological framework for his interpretive sociology (Verstehende Soziologie). He argued in particular for the necessity of when investigating social actions, 1) resorting to an interpretive understanding, which is not separated by causal explanation, 2) while avoiding value judgements, establishing value relevances (Wertbeziehungen) of the subject matter as criteria for its cultural importance and scientific pertinence, and 3) preserving the social and cultural uniqueness of the historical event, using the idealtype methodology. These three points constitute Weber’s main elaboration of the epistemological tenets of German historicism.

With regard to the first point, Weber criticized Wilhelm Wundt and George Simmel (who were influenced by Dilthey), who reduced sociohistorical knowledge to psychological understanding, since this position cannot support an objective knowledge: to properly understand a socio-historical event requires establishing some relations of cause and effect in order to test our interpretive comprehension. Therefore, Weber conceived an idea of sociology as both a generalizing, nomothetic and interpretive, idiographic science, arguing for the complementarity of interpretive understanding and causal explanation: that is, a researchers personal understanding should be balanced by empirical and statistically established regularities of scientific explanation (1913).

Secondly, Weber distinguishes value judgements, based on personal faith and belief, and fact judgements, based on science. Scientific knowledge cannot be neutral, in the sense that the values of the researcher inevitably influence his choices of relevance—the specific selection of problems and focus. However, it can be nonevaluative and objective in the sense that, once its priorities have been selected according to the researcher’s personal values, the researcher should proceed by testing empirical evidence supporting his hypotheses, thus expressing fact judgements only.

Finally, the ideal-type concept was employed by Weber to explain certain unique historical events using a model of plural causality: a singular event is always the effect of a multiplicity of historical connections. The problem becomes, in the infinite flow of events, to sort out the proper model of factors that under certain conditions and at a certain time can probably explain that effect. This also implies a probabilistic idea of causation, which gets beyond the deterministic approach typical of positivistic unilinear causation by a model of plural probabilistic-causal connections.

Logical Empiricism Versus Dialectical Theory

The epistemological legacy of German historicism—with its dualistic approach to the relationship between social and natural sciences, its preferential bias for a gnoseological model based on interpretive understanding, and its probabilistic model of multiple causality—has undoubtedly influenced a significant part of contemporary twentieth century sociology, such as the School of Chicago (Park’s contacts with Windelband and Simmel are well known) and, more strongly, the later interpretive sociology (through Weber and Simmel) and the phenomenological tradition in sociology (through Schutz). Its impact is manifest even in the antipositivistic methodological choices of many sociologists as different as Pitrim Aliksandrovič Sorokin (1959) and Charles Wright Mills (1959). Altogether, these different currents answer negatively the epistemological question about the possibility of acquiring knowledge of human social reality by means of empirical data alone. They reject the methodological unity of the natural and social sciences and do not accept the straight application of the scientific method to sociological analysis. When researchers undertake empirical research, they resort to methodological perspectives that result from the researcher’s epistemological choices: they tend to reject technical terminologies and statistical quantification, to privilege common-sense conceptualization and language, and to take the point of view of the social actor instead of that of the researcher and of the scientific community. Qualitative methodologies of this kind are also favored by those microsociologies—such as symbolic interactionism, ethnomethodology, and sociologies of daily life of different varieties—that have not been directly influenced by the idealistic and antipositivistic tenets of the German historicism. Although seldom explicitly stated, their epistemological foundations, which tacitly direct their choice of methods and tools of social research, are mostly consonant with those of interpretive and phenomenological sociologies.

The large majority of sociologists, however, followed an empirical-quantitative approach and accepted the scientific method borrowed from natural sciences as the appropriate and valid method of social sciences. The positivistic epistemological foundations of this mainstream sociology are quite evident in its most significant representatives Lazarsfeld and Lundberg, and in their utilization of the social survey as the main research tool of sociological inquiry. Lazarsfeld, in particular, although reluctant during his work at Columbia University to directly tackle epistemological issues while preferring practical empirical research, was well known for his neo-positivistic convictions (Gallino 1973:27), which were rooted in his mathematical and psychological background within the Vienna philosophical movements of the post-World War I period.

In fact, during the twenties and thirties of the twentieth century, a new group of positivists, the so-called “Vienna Circle,” arose in the Austrian capital, profoundly influenced by the work of the physicist Ernst Mach, the pioneers of mathematical logic Gottlob Frege, Bertrand Russell, Alfred North Whitehead, and Ludwig Wittgenstein, and by the French conventionalism of Pierre Duhem and Jules-Henri Poincaré. Mach considered science not a fact-finding but a fact-related activity: it cannot claim to discover the absolute truth about reality, since its laws are not absolute but, at best, “limitations of possibilities.” Russell, among the pioneers of mathematical logic, attempted a logical foundation of mathematics that considers the element of necessity present in both disciplines. This can be considered as the basis of the logical structure of thinking. Finally, Duhem and Poincaré suggested an idea of scientific theories as being purely hypothetical and conventional constructions of the human mind.

On these premises, the main representatives of the “Vienna Circle”—Moritz Schlick, Rudolf Carnap, and Otto Neurath—preferred the label of logical empiricism or neopositivism to distinguish themselves from classical positivism, since they discarded its sensist view—the idea that knowledge originates from experience ganied only through the senses, and from the observation and verification of that experience. They rejected direct observation of experience as the only means of hypothesis testing, and they considered verification as “testability in principle.” In this way, they accepted theoretical constructs without direct empirical referents as meaningful. They renewed empiricism on the new basis of the logical analysis of language: in fact, they argued that linguistic statements can be shown to be true or false by appeal to both logic and experience, and that both means, being factual, can be considered meaningful.

Logical empiricism still mantained some fundamental epistemological tenets of classical positivism, however, such as the idea that only empirically verifiable knowledge was meaningful, that science was a cumulative process based on induction, that the method of physics was the method of all sciences including social sciences (methodological monism), and that the discovery of natural and general laws was the fundamental aim of any science. The criticism of Karl Popper was aimed at most of these foundations and reversed them in a sort of negativism (Cipolla 1990). Since his first major publication (Popper 1934), the Austrian philosopher formulated his tenets in opposition to the main assumptions of the neopositivist: first and foremost, the induction as an appropriate method to infer general laws from the observation of a singular case. He argued that one case is enough to demonstrate the falsity of induction and that verification always depends on observational theories that are often derived from substantially the same theory from which hypotheses under test have been deducted, making their test logically inconclusive. Popper dropped the notion of verification itself in favor of that of falsification, and proposed a hypothetic-deductive method, based on the assumption that a hypothesis can be definitely considered false if it fails an adequate test, while if it is congruent with data this does not necessary mean it is true. In any case, Popper argued that a theory is maintained as true only provisionally, and it always remains a conjectural hypothesis that can be confuted in the future. In this way, scientific knowledge is never a closed, completed system, but always remains open to new possibilities. And even though Popper believed in the methodological monism—the unified theory of method—this is not absolute but conjectural, critical, and subject to falsification (1963).

A substantial refusal of methodological monism is represented by the Frankfurt School with its “critical theory of society” that advocates the necessity of the dialectical method in sociology. Its epistemological tenets are derived partly from Weber’s thought on Western reason and rationalization and partly from the young Hegel-influenced Marx of the Philosophical-Economical Manuscripts of 1844. It was also strongly influenced by the two critical Marxists Lukacs and Korsch and by the psychoanalysis of Freud. Dialectical epistemology implies:

  1. The holisticapproach, that is, the necessity of considering the totality in order to understand the parts, which are mediated by totality.
  2. The denial of the separation of history and sociological theory in order to grasp the dialectical process of change.
  3. A demystifying attitude in undertaking critical analysis of society and of the irrationality of Enlightenment reason once it has become purely instrumental and separated from its goals, that is, when it’s become a pure means of domination over men and society.

In 1961 the second most significant debate in the history of sociological epistemology took place at Tübingen between Theodor Adorno and Jurgen Habermas (two of the most prominent representatives of the Frankfurt School) on one side and Karl Popper and Hans Albert on the other (Adorno et al. 1969). The neopositivistic epistemology of the latter two was strongly questioned by the former two, who criticized its dialectic nature and who considered it a sort of logical formalism without any connection with its contents and that was incapable of understanding the social reality. The aim of sociology is to go beyond apparent phenomena, to grasp social contradictions and conflicts by interpreting society as a totality. This implies a refusal of the individualistic approach of positivism, of its monism with regard to the scientific method, of its measurement and quantification of social reality. Beyond Popper and Albert, the actual target of the Frankfurt scholars was the American school of sociology, whose positivistic analytical framework (theoretical categories and their translation into research tools) was considered an ideological reflection of the domination structures in the late capitalistic society.

Realism Versus Constructivism

Since the 1960s, logical empiricism, under criticism from many sources, has dissolved in a plurality of postpositivistic approaches, whose common denominator is a reformulation of positivistic tradition. This allows them to be grouped under the label of scientific realism, which asserts the absolute or relative independence of the reality under scientific scrutiny from the human researcher studying it. This, in turn, is based on the fundamental distinction between the objectivity of the reality observed and the subjectivity of the scientific observer studying it. The common tenet of realisms of all types is that the observer does not belong to the reality he observes; by means of his techniques of scientific inquiry, he should avoid any involvement and influence on reality, maintaining a neutral position. In this way, the subjectivity of the observer is limited to his “discovery” of the objective reality.

One of the first attacks on the neopositivistic legacy was by Willard Van Ormand Quine (1952), who proposed a holistic view of knowledge, considered as a field of forces whose limiting conditions are constituted by experience. In turn, Mary Hesse (1974) considered scientific language as a dynamic system whose continuing growth is due to metaphorical extension of the natural language. Thomas Kuhn (1962) greatly contributed to the growing consciousness of the reality of scientific change by analyzing its actual historical development, challenging the common-sense concept of science as a purely rational enterprise. Taking scientific practice into account, he showed how it is usually ruled by a paradigm, a world view, legitimized by the scientific community, that remains dominant until a new paradigm replaces it.

Paul K. Feyerabend, with his “methodological anarchism” (1975), proposed a paradoxical and extreme view of science. In Feyerabend’s view, the scientific method remains the only rational procedure for deciding and agreeing upon which theory is more adequate to describe and explain a state of affairs in the natural as well as in the social worlds. Of course, rationality depends on common premises and procedures: the advantage of accepting them is great because it is through them that intersubjectivity is achieved. Scientific objectivity resides neither in the object of knowledge, as classical positivism maintained, nor in the subject, as idealists tend to believe, but rather in the intersubjectivity that results when researchers adopt the same procedures and accept the premises on which those procedures are based. It is the reproducibility of the methods, techniques, and tools of scientific research that secures the replicability of results. And their reproducibility is due largely to their being public procedures, easily scrutinized and reapplied. The results of the research in the natural sciences may be more objective than those in the social sciences, but this is due to the standardization and publicity of procedures in the natural sciences. There is no ground for supposing that natural sciences possess a special objective attitude, while social scientists possess a value-oriented attitude. The problem is that many research procedures in the social sciences are not reproducible, because they often reflect a private state of mind communicated through linguistic expression full of connotative meanings, often not shared by all in the research community.

The critical realism of Roy Bashkar (1975) proposes an ontology based on the distinction of three spheres: the real, the actual, and the empiric, arguing in favor of the existence of structures or hidden mechanisms that can work independently from our knowledge, but whose power can be empirically investigated both in closed and open systems. This realistic view of science is supported also by Rom Harré (1986), who stresses the role of models in the development of theory. The new realists also suggest a relational paradigm for sociology in order to overcome the traditional antinomy between structure and action by a theory of structuralization (Giddens 1984) and by a transformative model of social activity (Bhaskar 1989). According to this new paradigm, the social structure should be considered as both the omnipresent condition and the continuously reproduced outcome of the intentional human action. This view is shared by the applied rationalism of Pierre Bourdieu, who systematically applies relational concepts and methodically compares his theoretical models with the empirical material that is the result of different research methodologies (Bourdieu et al. 1973). The critical view of science taken by the new realists, joined with their transformative conception of social reality, has produced a critical naturalism that considers both the positivist and the historicist traditions as dependent on the same positivistic conception of the natural sciences. This critical naturalism holds that, in actuality, both the human social life and the natural life are liable to scientific explanation, although of a different kind (Bhaskar 1989).

This realistic view has been seriously challenged in the 1980s and 1990s by the constructivistic movement, giving rise to the third, and still ongoing, epistemological debate in the history of sociology. Constructivism proposes the unification of the objective reality observed with the subjective reality of the observer. The rationale is that the reality investigated and the science used to investigate it have an equally subjective origin, which implies that the reality observed “depends” on the observer. This does not mean that constructivism denies the actual existence of an autonomous reality from the observer, but rather that an “objective” representation of this external reality by scientific knowledge is not possible. All that can be said about reality is inevitably a “construction” of the observer (von Foerster 1984), who, in turn, being part of this reality, is a “black box” whose internal components are unobservable. Therefore, even for the constructivists, reality is an elusive objectivity, with its autonomous existence which cannot be reduced to a simple subjective and arbitrary experience. What is “constructed” is the knowledge and not the reality (Von Glasersfeld 1987).

There are different types of constructivism. The most radical (Von Glasersfeld 1995) maintains that nothing can be said even by the observer, apart from the trivial verification that he observes. An unknowable observer faces an unknowable reality. A more moderated type of constructivism considers still possible a theory of the observer, even though it denies the same possibility for the reality observed (Maturana 1988). And finally, there is a constructivism which accepts both a theory of the observer and of the observed, considering them as mental systems (Luhmann 1990).

Towards an Integrated and Pluralistic Paradigm

As the current debate between realism and constructivism shows, the positions only appear distant. Especially when they move from the more abstract theory to the empirical research, the respective positions become more ambiguous, vague, and overlapping. Moreover, there are clear indications that the old cleavages between positivistic and interpretive sociologies may slowly fade away. Even though quantitative research techniques appear more precise and rigorous—with their standardization and reproducibility features—than qualitative methods, it is difficult to justify using quantification for all types of sociological data. Many sociologists believe that not all aspects of social phenomena could be subject to the rules of quantification. And most of them start out thinking that quantitative and qualitative methods and techniques are complementary. Yet, such an integrated and pluralistic methodology needs to be justified by an explicit epistemology. Therefore, the times seem ripe for a new epistemological foundation of an emerging integrated and pluralistic methodology, even though a few attempts have been made towards this goal.

The first attempt was the ecological paradigm of Gregory Bateson (1972). Strongly influenced by his participation in the 1940s in the interdisciplinary research group on cybernetics composed of Von Neumann, Shannon, Von Foerster, and Wiener, among others, he then organized all his subsequent work in different fields around a few central cybernetics concepts, including schismogenesis, circular communication, and feedback. On the ground of this new theoretical perspective, he built up an epistemology aimed at getting beyond the boundaries between the internal and the external components of the observer, the “mind,” conceived as a network of founding relationships. In this process, the identity of the observing subject is dissolved into its ecological environment, in the relationship between subjective meanings, action, and action objective. Even though Bateson’s epistemology precedes the debate between realism and constructivism, it has clearly inspired most of the constructivistic approach.

A second attempt, historically, can be traced in Edgar Morin’s attempt to set up a new general method (Morin 1977, 1980, 1986), which can be applied in every field of knowledge according to a methodological plurivers. In this way, he tries to overcome the shortcomings of the traditional scientific paradigm, which he considers disjointedreductive, and simplified, since it does not take into account the complexity of reality. The multidimensional character of reality is precisely the starting point of his very comprehensive attempt to found a new epistemology based on the auto-eco-organizing principle. An attempt, however, which for the most part is affected by a biologistic bias, seems quite inappropriate for the social sciences.

Finally, Costantino Cipolla has proposed a new correlational paradigm based on an epistemology of tolerance (Cipolla 1997) which prefers an “inter-” and “co-” perspective linking together and integrating the traditionally opposite epistemological poles in an attempt to reassemble them into a new pluralistic approach. Closer to the sociological tradition of the first two attempts, this new correlational paradigm takes into account the reasons of both the realists and the constructivists and focuses its attention on the connection, the relation, and the link between the objective reality and the subjective observer in order to avoid any absolutism of either of the poles. It is starting from this bivalent conception of reality—which considers reality as both autonomous in itself and “constructed” by the subject, who is in turn the outcome of the real social forces, so that each pole is considered only partially autonomous and strictly interconnected with the other—that it should be possible to re-found the sociological epistemology on the ground of the methodological research questions of the discipline, in recognition of the fundamental ambivalence of reality and of the need of an adductive procedure as a “double movement” between induction and deduction, particular and general, theoretical hypotheses and social reality beyond any self-contained reductivistic monism.