"There are few social scientists who would identify themselves as positivists in any
strict sense, and fewer still who would confine the conduct of social scientific inquiry
to the canons of strict positivism. Indeed, many positivist doctrines have been superseded
in the natural sciences. Nevertheless, while the influence of positivism in the social
sciences is oblique, there are important elements of continuity between positivism and
objectivism in the social sciences. For many social scientists the methods of the natural
sciences remain an explanatory ideal, and creating general laws of social process is
often portrayed as a reasonable goal. Perhaps the most concrete expressions of
positivism in the social sciences are the techniques of survey research.  It is a basically
positivist impulse that leads practitioners to equate social scientific research with the
effort to find systematic associations between quantified variables. The legacy of
positivism is clearly detectable in attempts to quantify 'subjective' states and to treat
them in the manner of 'variables.' For social scientists of a positivist bent, a concept
like alienation, properly operationalized, can be dealt with in the same manner as a
physicist might deal with temperature or pressure. In this perspective, the subject matter
of the social sciences is to be transformed into brute data, and then cast into the same
explanatory format as the brute data of nature.

"It is not clear that positivism in the social sciences entails behaviorism, but there is an
affinity between these two doctrines. While Émile Durkheim (1858-1917), for one, tried
to combine a strain of idealism with positivism, the behaviorism favored by the Vienna
Circle is perhaps a more natural corollary of positivism.  [The oddness of Durkheim's
attempt, similar to Auguste Comte's, to combine a sense-impression theory of empiricism
with a conception of social order as consisting largely in ideas is worth puzzling. We do
not ordinarily find brute-data empiricists using concepts akin to group mind.] This is
because ideas and meanings are not, it is claimed, apparent to the senses in the manner
of the brute data favored by positivism, and hence the effort to explain human behavior in
the same fashion as matter in motion must come to grips with the special problems
presented by the 'mental' aspects of human behavior. The presence of 'mind' in the
human studies is perhaps the strongest impetus to social scientific subjectivism.
Objectivists have responded to the problem of mind in a variety of ways:

"1 Some proponents of objectivism seem to claim that, at least within the context of
science, we must treat persons as though they did not have minds. Taking his cue from
psychological behaviorism, Georg Lundberg (Foundations of Sociology, 1939) speaks
of various mental terms, including 'will,' 'feeling,' 'ends,' motives,' and 'values,' as 'the
phlogiston of the social sciences.' He thus seems to regard mental states as mythical
entities that have no place in the lexicon of science, and portrays the growth of scientific
thought as a progressive elimination of such obscurantist terminology: 'The history of
science consists largely of the account of the gradual expansion of realms of the
"natural" and the "physical" at the expense of the "mental" and the "spiritual" ' (Ibidem).

"2 An alternative approach to the problem of mind acknowledges that persons have a
mental life but argues that mental properties can be operationalized into statements
about what subjects do. This program typically takes the form of a reductionist doctrine
that psychological statements can be translated, without loss of meaning, into statements
about behavior. This seems to be Rudolf P. Carnap's (1891-1970) approach:

          'So called psychological sentences -- whether they are concrete sentences about
          other minds, or about some past condition of one's own mind, or about the present
          condition of one's own mind, or, finally, general sentences -- are always translatable 
          into physical language.' (R. Carnap, 'Psychology in physical language,' 1959)

There are some important ambiguities, however, within the behaviorist program. The
above statement speaks to the possibility of translating psychological sentences into
sentences about behavior. But Carnap also seems to argue that psychological sentences
describe nothing but behavior: 'all sentences of psychology describe physical occurrences,
namely the physical behavior of humans and other animals.' (Ibidem)  G. Lundberg
expresses a similar ambivalence. On the one hand, Lundberg claims that 'The sociologist
must demand sensory evidences of the imaginings, thoughts, and other phenomena of
'consciousness"' (Foundations of Sociology, 1939). But he also claims, 'All these words
(honor, duty, loyalty) stand for behavior' (Ibidem). Another source of confusion in
interpreting the behaviorist program arises from the fact that sometimes the behavior
in question is events in the central nervous system, sometimes it is 'molar' behavior*,
and sometimes it is both. It is also difficult to reconcile Lundberg's claim that mental
states are mere 'phlogiston' with his attempts to operationalize them. For we do not
ordinarily try to operationalize mythical entities.

"3 A more subtle response to the problem of mind acknowledges that persons have
a mental life that must be incorporated into an adequate account of action. But it is
argued that explanations of action can be cast in the same terms as the explanation
of physical events by treating motives, reasons, intentions, etc. as the causes of action.
In this way, the explanation of action is considered to follow a more or less conventional
causal format.  According to Carl Hempel and Paul Oppenheim ('Theory of scientific
explanation,' 1948), 'determining motives and beliefs have to be classified among the
antecedent conditions of a motivational explanation, and there is no formal difference
on this account between motivational and causal explanation.'  [A vast literature has
accumulated on this issue in recent years. It is difficult to say anything about this matter
without saying a great deal.  Donald Davidson's ('Actions, reasons, causes,' Journal of
Philosophy, 1963) position, that motives are very much like causes, has gained
considerable support.]

"In sum, positivism in the social sciences deals with the problem of intentional action,
which is in significant measure the paradigm expression of the problem of mind, by
denying that actions exist (i.e., behaviorism), by redescribing actions in a way that
excludes the mental component, or by assimilating the explanation of action into ordinary
causal explanation. But the influence of positivism in the social sciences has not, for the
most part, taken the form of explicit philosophical doctrines. Most typically, positivism in
the social sciences takes the form of a research methodology with certain implicit
philosophical assumptions, viz. survey researc .  It is through the use of such methods,
which aim at establishing law-like associations between brute data, rather than in the
articulation of philosophical doctrines, that the influence of positivism in the social
sciences is most evident.

"A second major implicit expression of positivism in objectivist social science, not
sufficiently dealt with in the literature, is the use of structural explanation. Because of
their doubts as to the accessibility of mind to scientific method, objectivists are generally
inclined to separate the social scientific idea of society from ideas in society. This
inclination often takes the form of an affinity for structural explanations. It is not easy to
say just what the concept of social structure means, and it has sharply different meanings
for different theorists, but generally social structure is meant to describe those features of
social living that are independent of and exert controlling influence on the beliefs and
actions of social members. The concept of social structure thus typically refers to those
features of social organization known to the scientist but likely unknown to the actor. In
this way, mental properties are either deleted from social scientific accounts or they are
portrayed as dependent variables, i.e., dependent on putatively objective features of social
structure. In describing what he considered to be a virtue of Marxism, Émile Durkheim
illustrates this objectivist affinity for explaining behavior in terms remote from common-
sense thought: 'I consider extremely fruitful this idea that social life should be explained,
not by the notions of those who participate in it, but by more profound causes which are
unperceived by consciousness.' [It is important to point out that the concept of social
structure is used very diversely. Many writers, like Talcott Parsons (1902-1979), speak of
normative structures, and hence structures are not defined independently of actors' beliefs.
The attempt to define social structure in strictly objectivist terms may even be a minority
tendency; but it is an important theoretical option for objectivist social science, and, this
concept of structure has had some very prominent advocates.]  Structural analysis is often
portrayed as the uniquely sociological expression of objectivism in the study of man. If
structural analysis can be considered to be the rock bottom of sociological explanation,
and if structural variables are 'objective', i.e., if they can be known independently of the
subjective experience of social actors, a convincing argument can be made that social
explanation is ultimately not unlike the explanation of physical nature."
-- David Rubinstein, Marx and Wittgenstein: Social Praxis and Social Explanation, 1981

*See my comment below for a definition of molar.