| |
Merton’s
Critique of Parsons
Robert K. Merton (1910-2003), had studied
under Parsons at Harvard University, where he
received his doctorate in sociology in 1936.
He was later appointed chairman of the
department of sociology at Columbia University.
His dissertation, Science, Technology, and Society in Seventeenth-Century England,
marked the beginning of his career-long interest in
sociology of science.
His papers in sociology of science written
and published between 1935 and 1972 are reprinted in
his Sociology of Science: Theoretical and Empirical Investigations
(1973). While
Merton's interest in science is noteworthy, his
views in sociology of science are beyond the scope
of this history.
Here the focus of interest is Merton's Social
Theory and Social Structure (1949, 1968), where
he departs from Parsons' Romanticism with his own
rendering of the functionalist type of explanation
for sociology, and develops his own concept of
scientific sociological theory.
He believes that functional analysis is the
most promising yet the least codified of
contemporary orientations to problems of
sociological interpretation.
He does not claim to have invented this type
of sociological explanation, and he offers several
examples of it in the literature of sociology; he
says that his major concern in this book is its
"codification" by developing a
"paradigm" for it.
He notes that sociologists often use the term
"function" as it is used in mathematics to
describe interdependence, but he is not thereby
proposing a mathematical type of sociological
theory. In
fact he explicitly states that his purpose is to
codify the procedures of qualitative analysis in
sociology.
Merton says that the concept of social
function refers to observable objective consequences
and not to subjective dispositions such as aims,
motives, or purposes, and that the consequences of
interest are those for the larger structures in
which the functions are contained. The concept of function involves the standpoint of the
observer and not necessarily that of the
participant.
He says that failure to distinguish between
the objective sociological consequence and the
subjective disposition inevitably leads to
confusion. This
is because the subjective disposition may but need
not coincide with the objective consequence, since
the two may vary independently.
This concept of functional analysis occasions
Merton's distinction between "manifest"
function and "latent" function.
Manifest functions are those that have
objective consequences contributing to the
adjustment and adaptation of the social system, and
which are intended and recognized by the
participants in the social system.
Correlatively latent functions are defined as
those objective consequences contributing to the
adjustment or adaptation of the social system, and
which are not intended or recognized by the
participants in the social system.
As an example Merton says that criminal
punishment has manifest consequences for the
criminal and latent functions for the community.
Merton’s distinction is clearly valid, and
has been recognized by others independently. For example William H. McNeill, who is not a sociologist but
a historian of medicine, illustrates what
sociologists would call “latent functions” in
his Plagues
and People (1977), a historical study in
epidemiology. McNeill
writes that a recent large-scale outbreak of bubonic
plague, also known in earlier Europe as the “Black
Death”, occurred in Manchuria in 1911.
Investigators discovered that the disease had
been contracted from marmots, which are large
burrowing rodents with skins that commanded a good
price on the international fur market.
The indigenous nomad tribesmen of the steppe
region, where these animals live, had mythic
explanations to justify epidemiologically sound
rules for dealing with the risk of bubonic infection
from the marmots.
The tribesmen believed that departed
ancestors might be reincarnated as marmots.
Trapping was taboo; a marmot could only be
shot, and an animal that moved sluggishly was
untouchable. And
if the marmot colony showed signs of sickness,
custom required that human community to strike its
tents and move away to avoid misfortune. Such
customary practices and proscriptions reduced the
possibility of human infection with plague to minor
proportions. But
in 1911 inexpert Chinese emigrants, who knew nothing
of the tribesmen’s “superstitions”, hunted the
marmot for their furs, trapping both sick and
healthy animals indiscriminately.
The result was that plague broke out among
the Chinese and then spread along the newly
constructed railroad lines of Manchuria.
In this case the manifest function, as least
to the nomads, is the proper treatment of possible
reincarnated ancestors, while the latent function is
a hygienic hunting practice that protected the
hunter from a serious contagion.
Merton describes heuristic purposes for his
distinction between manifest and latent functions.
The distinction not only precludes confusion
between motive and function, which he notes may be
unrelated to each other, but it also aids the
sociological interpretation of many social
practices, that are regarded by observers as merely
ignorant "superstitions", yet still
persist even though their manifest purposes are
clearly not achieved.
And it also directs the sociologist's
inquiries beyond the manifest or intended aspects of
behavior to discover its generally unrecognized
consequences. Merton
thus affirms that the discovery of latent functions
represents significant increments in sociological
knowledge, because they represent greater departures
from "commonsense" knowledge about social
life. This
is more philosophically sophisticated than the verstehen
requirement that hypotheses "make sense.”
Furthermore he notes that the concept of
latent function has significance for social policy
or social "engineering.”
He sets forth a basic theorem, which may be
called Merton's theorem of social engineering; it
says that any attempt to eliminate an existing
social structure without providing adequate
alternative structures for fulfilling the functions
previously fulfilled by the abolished organization
is doomed to failure.
More generally Merton's theorem says that to
seek social change without due recognition of the
latent functions performed by the social
organization undergoing change, is to indulge in
social ritual rather than social engineering.
Like Habermas’ discussion of irrational
purposeful action, Merton's thesis of latent
functions reveals the inadequacy of the Parsonian
Romantic concept of theory based on motivational
analyses, but Merton furthermore recognized that a
new concept of sociological theory is needed,
although he does not adopt the Positivist's complete
rejection of Romanticism.
In his discussion of functionalism he says
that in preparation for developing a functionalist
explanation a fully circumstantial account of
meanings, i.e. the cognitive and affective
significance attached to a behavior pattern, goes
far toward suggesting appropriate lines for a
functional analysis.
Had he been less sympathetic to the
Romantics, he might have followed through to the
conclusion that the distinction between manifest and
latent functions contributes nothing to the
explanatory value of the functionalist explanation,
since its explanatory value consists not in a
functional factor being either manifest or latent
but in its being consequential for other factors to
be explained. And
this implies that the manifest-latent distinction is
informative only for Romantics, who need to be told
that motivational analysis is not adequate for
explanation in social science, except as one among
many possible heuristic devices for developing
functionalist hypotheses.
Merton's attack on Parsonian sociology is not
a frontal assault on Romanticism, but is part of his
own agenda for sociological research.
The attack is directed explicitly at the
all-inclusive type of system building practiced by
many sociologists including notably Parsons.
His principal objection to these
all-inclusive systems is that they are too vague to
be tested empirically, and he refers to them as
general orientations toward sociological analysis
rather than "theories.”
The agenda that he advocates for future
research in sociology is the development of what he
calls "theories of the middle range",
theories that he says are somewhere between minor
but necessary empirical generalizations or working
hypotheses on the one hand and the Parsonian-like
all-inclusive systems on the other.
Unlike the Romantics, who define theory in
terms of the semantics of a vocabulary referring to
subjective meanings and motives of social actors,
Merton defines theory in terms of its logical
structure. He
explicitly defines "theory" for both
natural and social sciences as a logically
interconnected set of propositions from which
empirical generalizations can be derived.
In another statement he says theory is a set
of assumptions from which empirical generalizations
are derived. And
referencing Lundberg's "Concept of Law in the
Social Sciences" he says a scientific law is a
statement of invariance that has been derived from a
theory. He
distinguishes theory from the empirical
generalization saying that the latter is an isolated
proposition summarizing observed uniformities of
relationships between two or more variables.
In the history of science there have been
significant single-equation theories, such as
Newton's theory of gravitation.
But Merton does not state explicitly whether
or not he intends by his definition to exclude from
the domain of theory language the single-equation
theories that are found in many sciences.
Referencing Whorf, Merton notes that the
empirical researcher's perceptions are fixed by his
conceptual apparatus, and that the researcher will
draw different consequences for empirical research
as his conceptual framework changes.
However, Merton does not seem to recognize
that this control of language over perception
undermines his distinction between theory and
empirical generalization, since this semantical
control operates by the linguistic context of
empirical generalizations, which means that
empirical generalizations are never isolated.
His distinction is therefore unsustainable.
Had he approached this problem by an analysis
with an adequate and contemporary philosophy of
language, he might have seen that his distinction
incurs the same difficulty that both the Romantics
and the Positivists encounter, when they purport to
distinguish theory from a semantically isolated
observation language.
The semantics of observational description is
not isolated from that of theory; semantics, logical
syntax, and belief are interdependent.
The only sustainable basis for distinguishing
theory from nontheory language is the pragmatics of
language, the functions it performs in basic
research. As
it happens, Merton comments on the functions of
theory for empirical research.
But his comments presume his distinction
between theory and empirical generalizations, and
are not definitive of a distinction between theory
and nontheory language. Furthermore his list of functions are not applicable to the
modern quantum theory, and more generally are not
sufficiently universal in the practice of scientific
research to serve as defining characteristics of
theory language.
On the contemporary Pragmatist philosophy of
science the only characteristic that distinguishes
theory from nontheory language is that the former is
proposed for testing, while the latter is presumed
for testing.
It may be noted here by way of a postscript
to this discussion of Merton, that some economists
also recognize what Merton calls "latent
functions", even if the economists have no
particular name for it.
1976 Nobel laureate economist Milton
Friedman's "Methodology of Positive
Economics" (1952), reprinted in his Essays
in Positive Economics (1953), is one of the more
popular methodological papers written by an
economist for economists in the post-World War II
era. A
contemporary philosopher of science would likely
view this paper as an effort to de-Romanticize
neoclassical economics.
Although this paper sets forth a somewhat
naive semantical thesis, its semantical metatheory
is more sophisticated than the neo-Positivist view
in Friedman’s Theory
of the Consumption Function, and his phrase
"positive economics" here does not mean
Positivist economics.
Like the Pragmatists, Friedman says that the
only relevant test of the validity of a hypothesis
is comparison of its predictions with experience; he
thus accepts no ontological criteria in his view of
scientific criticism, including the Romantics'
mentalistic criteria involving descriptions of
motivations. He
explicitly rejects objections to the rationality
postulates or to any other assumptions employed by
economic theory, including the objections of the
Institutionalist economists, when they are not based
on the predictive performance of the theory. For example he notes that businessmen do not actually
calculate marginal cost or marginal revenues and
solve a system of simultaneous equations as do
economists, and that businessmen seldom do as they
report when asked about the factors affecting their
decisions. But
Friedman says that businessmen must act as if they have compared marginal costs and marginal revenues,
because they will not remain in business if their
behavior is not consistent with the theory of
rational and informed maximization of returns.
In philosophers’ terms, this means the
economist is not a Romantic examining what the
entrepreneur thinks, but rather is a Pragmatist
examining the consequences of what he does.
Or, in Merton's terms: it is the functional
consequences that are relevant, and the motives are
latently functional when their unintended
consequence is satisfaction of the marginalist
conditions set forth in neoclassical economics.
Lundberg’s
Positivist Sociology
Parsonian Romanticism has not been without
its critics. Not
surprisingly the science that was founded by the
founder of Positivism, namely Auguste Comte, has
offered new Positivist critics to oppose Parson’s
latter-day variant of Romanticism.
The principal protagonist in this critical
role, who was contemporary to Parsons, was George
Lundberg (1895-1966).
As it happens, Lundberg's criticisms did not
effectively persuade American sociologists, and
post-World War II sociology took the Parsonian path. Nonetheless a brief rendering of Lundberg's criticism will
describe the philosophy which for many years
American academic sociologists viewed as their
principal philosophical alternative to Parsons.
Lundberg traces his philosophical heritage to
Comte. In
his "Contemporary Positivism in Sociology"
in American
Sociological Review (1939) Lundberg gives three
quotations from Comte's Positivist
Philosophy, that he says suggest the principal
survivals and modifications of Comte's work that may
be regarded as contemporary Positivism in sociology. The first quotation is a statement of the principal aim of
science, according to which the business of science
is to analyze accurately the circumstances of
phenomena, to connect them in invariable natural
laws according to the relation of succession and
resemblance, and to reduce such laws to the smallest
possible number.
The second quotation sets forth a secondary
aim of science, namely to review existing sciences
to show that they have a unity of method and a
homogeneity of doctrine.
The third quotation affirms the importance of
observation and rejects the view that the sciences
of human behavior should attempt to study facts of
inner experience.
Lundberg finds himself at variance with
Parsons, and he quotes anti-Positivist comments from
a lengthy footnote in Parsons' Structure
of Social Action, in which Parsons states that
all Positivisms are untenable for both empirical and
methodological reasons.
Lundberg wrote several methodological works.
His principal philosophical work is a
monograph of about one-hundred fifty pages titled Foundations of Sociology (1939), which includes his views set forth
in a previous papers including one titled
"Concept of Law in the Social Sciences"
published in Philosophy
of Science (1938).
The 1964 edition of the Foundations
monograph contains an "Epilogue" as a new
chapter, in which Lundberg maintains that the
Parsonian approach to sociology is converging toward
the Positivist view.
In 1929 he wrote Social Research: A Study in Methods of Gathering Data, which he
extensively revised in 1942.
In 1947 he wrote Can
Science Save Us? and in 1953 he co-authored Sociology,
a textbook in seven parts with a methodological
discussion constituting the first part, and that
went through four editions.
Lundberg was very impressed by the successes
of natural science especially in comparison to
sociology, and he stated that the history of science
consists largely of the account of the gradual
expansion of the realms of the natural and physical
at the expense of the mental and the spiritual.
His agenda for sociology therefore is to
realize success in sociology by imitating the
methods of the natural sciences.
The philosophical understanding of natural
science during the time of his active career was the
Positivist philosophy, which also prevailed in
academic philosophy of science at the time.
But the classical Machian Positivism
implemented in the natural sciences with its
phenomenalist ontology is not easily adapted to
behavioral and social sciences, and Lundberg
therefore developed his own Pickwickian Positivism
for sociology.
Lundberg's epistemological view has
similarities to the classical British empiricists,
Locke, Berkeley and Hume, and also to the early
Positivists such as Mach. These philosophers started with the thesis that what the
human mind knows immediately is its own ideas,
sensations, or sense impressions.
This is a subjectivist view that occasions
the question of how the human mind knows the
external or extramental real world.
One answer to this problem is the copy theory
of knowledge, according to which ideas reveal
reality, since they are copies of reality.
Another answer is that there is no external
world consisting of material substances, such that
the ideas themselves become reified, and the result
is an idealist and solipsistic thesis, such as
Berkeley's
esse est percipi, “to be is to be
perceived.” Lundberg also has a subjectivist theory of knowledge, but he
has his own solution to the problem of knowledge of
reality. Lundberg
maintains that the immediate data of all sciences
are symbols, by which he means human responses to
whatever arouses the responses.
And he also calls these responses sensory
experience. His
subjectivist philosophy of knowledge is nonrealist,
because it makes subjective experience instead of
extramental reality an object of knowledge rather
than making experience constitutive of knowledge.
He then goes on to say that the nature of
that which evoked these human responses must be
"inferred" from these immediate data which
are our sensory experience; we infer both the
existence and the characteristics of anything from
these responses.
In his Positivism there are apparently some
extramental realities beyond the phenomena.
Furthermore this "inference" of the
characteristics of reality is not a deductive
inference, but consists of operationalist
definitions. In
his discussion of measurement Lundberg says that
since Einstein, physicists have blatantly declared
that space is that which is measured by a ruler,
that time is that which is measured by a clock, and
force is that which is measured by pointers across a
dial. A
thing is that which evokes a certain type of human
response represented by measurement symbols. There is an ironic aspect to Lundberg's epistemological
subjectivism, because he uses it to refute the view
that the subject matter of social science is
subjective, arguing that distinctions between what
is subjective and what is objective is not given in
the data. Thus
objectivity is not given in things, but in those
ways of responding that can be corroborated by other
persons. He
seems unaware that corroboration to establish
objectivity or intersubjectivity is itself quite
problematic for any subjectivist philosophy of
knowledge.
The most distinctive aspect of Lundberg's
version of Positivism is his rejection of the
naturalistic philosophy of the semantics of
language. In
discussing quantification he rejects any distinction
between natural and artificial units for
measurement, and he denies that scientists measure
the behavior of some things but not the being,
quality or quantity of others.
He argues that like physicists, sociologists
must recognize that all units are artificial
linguistic constructs symbolizing human responses to
aspects of the universe relevant to particular
problems. Lundberg
also implicitly recognizes that his semantical view
affirms an ontological relativity, when he says the
human knower infers the "nature" of
phenomena from his symbolic responses.
But his ontological relativity contrasts with
that of the contemporary Pragmatists, who are
realists and who maintain that the human knower
knows reality directly and not indirectly by any
kind of inference.
Lundberg's rejection of the naturalistic
philosophy of the semantics of language absolves him
from any need to characterize the observational
basis of science.
He thus evades a difficult problem for a
social or behavioral science attempting to implement
the phenomenalist thesis of the Positivist physicist
or chemist. Human
social behavior is not easily or productively
described in terms of phenomenal shapes, colors,
sounds, or other purportedly elementary sense data.
In contrast the Vienna Circle sociologist,
Otto Neurath, was ad hoc in his attempt to accomplish the same thing, when he simply
announced that a "thing language" as
opposed to a phenomenalist language is admissible in
Logical Positivism.
More importantly Lundberg's artifactual
thesis of semantics is strategic to his agenda for
rejecting the view that sociology has a distinctive
subject matter, i.e. distinctive in its subjective
nature, since human knowledge does not immediately
apprehend the nature of things. But rejection of the naturalistic semantics undercuts
Lundberg's agenda of eliminating vocabulary
conventionally referencing subjective experience as
opposed to observably objective behavior.
His philosophy of the semantics of language
does not admit the distinction he tries to enforce
as a condition for a scientific sociology.
|
|
