-
Simmel and the Methodological Problems of Formal
SociologyAuthor(s): Talcott Parsons and William J. BuxtonSource:
The American Sociologist, Vol. 29, No. 2 (Summer, 1998), pp.
31-50Published by: SpringerStable URL:
http://www.jstor.org/stable/27698870 .Accessed: 02/07/2014
04:15
Your use of the JSTOR archive indicates your acceptance of the
Terms & Conditions of Use, available at
.http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars,
researchers, and students discover, use, and build upon a wide
range ofcontent in a trusted digital archive. We use information
technology and tools to increase productivity and facilitate new
formsof scholarship. For more information about JSTOR, please
contact [email protected].
.
Springer is collaborating with JSTOR to digitize, preserve and
extend access to The American Sociologist.
http://www.jstor.org
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
Simmel and the Methodological Problems of Formal Sociology
Talcott Parsons Edited by William J. Buxton
Introduction 1
Attempts to define the subject-matter and boundaries of
sociological science have gone somewhat out of fashion, and it is
not proposed, in the present paper, to attempt to bring such
discussions back into favor. It does, however, happen, that in
controversies over such questions, important methodological issues
some times receive a particularly pointed formulation which makes
them a favorable
point at which to enter upon the discussion of the underlying
problems. This is particularly true in the case of Simmel. Simmel's
contributions to
general social theory, as opposed to his individual essays on
particular social
forms, are relatively meager. Indeed his position in the history
of the former field rests largely on his single formula that
sociology should become the spe cialized science of "social forms,"
and on his discussion of the nature of "soci
ety" which underlies that formula. * This has sufficed to
attract a great deal of
attention to Simmel's work and even, in a modest way, to make
him the founder of a "school," the so-called "formal school" of
sociology. But the influence of the mode of analysis he advocated
has been by no means confined to people who would call themselves
explicitly formal sociologists, and discussion of some of the
methodological problems associated with Simmel's formula promises
to lead farther than either placing Simmel more accurately in the
history of social theory, or helping to arrive at a critical
judgment of the work of those who profess to follow him.
It is common knowledge that in the earlier stages of
self-consciousness of
sociology as a science, the tendency was strong to conceive it
in a synthetic or
"encyclopedic" sense, as the sys . . .
This is a transcription of the original typed manuscript along
with the handwritten changes made by Talcott Parsons. The footnotes
(indicated now with asterisks) are those of Parsons. The original
pagination has been indicated with numbers in brackets. These match
the page numbers cited by Edward Shils in his comments. Explanatory
endnotes have been added. Parsons began the article with an
alternative introduc tion, which is included here under the heading
of Introduction 1.
* See Georg Simmel, "Das Problem der Soziologie" printed as
chapter I of his Soziologie [Simmel, 1923].
Parsons 31
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
Simmel and the Methodological Problems of Formal Sociology
Simmel as a general social theorist is primarily known as the
author of a
formula for defining the scope and subject-matter of sociology,
that sociology should be the study of "social forms."
* The object of the present essay is not to
revive the discussion of the scope and limits of the science of
sociology as such. But underlying views on this subject there may
be important methodological issues of a more general nature.
Investigation of the methodological position at the basis of
Simmel's famous formula may well lead into problems of very gen
eral interest. It is as an avenue of approach to certain of these
problems that a critical analysis of the concept of social form is
here attempted.
It is common knowledge that in the earlier stages of
self-consciousness of
sociology as a science, the tendency was strong to conceive it
in a synthetic or
"encyclopedic" sense, as the sys- [2] tematic statement of all
our established
generalized knowledge of the concrete social life of man.1 This
tendency had two effects: to give an air of great pretentiousness
to the ill-established infant
discipline, which did not serve to endear it to its neighbors in
the learned world, and to raise, in an acute form, the question of
whether it had any independent contribution to make beyond that of
the older and better established fields such as economics, the many
branches of history, jurisprudence, and the others deal
ing with human social relationships. The claim only to make the
final synthesis seemed a rather thin, and scientifically dubious
basis on which to erect an
independent science. After all why could not this synthesis be
left, as it had been in the past, to the philosophy of history?
Simmel was one of the first to revolt against this encyclopedic
tendency, strongly advocating that sociology be constituted as a
special and not an ency
clopedic science. He maintained that there was no concrete class
of social phe nomena which was not already the subject of a social
science?men's economic
life, their religion, art, law, government, etc. Hence the only
place for sociology lay, in his opinion, not in the discovery of a
new class of phenomena hitherto
neglected, but in a new analytical point of view according to
which the same concrete phenomena which were the subject-matters of
these other social sci
ences, had not yet been studied. It is in this context that he
formulated his famous distinction between social "form" and
"content."
[3] Simmel couches his discussion primarily in terms of the
concept of social
relationships which is one of several possible ways of looking
at human social life. Concrete social relationships are, of course,
unified, integral phenomena.
They may, however, by a kind of abstraction, be analyzed into
two classes of
elements. The first consists of all or some of those qualities
which may be
predicated of the individuals who participate in the
relationships, but which
may be treated as analytically independent of the relationship
itself. Simmel does not attempt to give any systematic account of
these elements, but puts forward
* Developed mainly in his essay "Das Problem der Soziologie"
which was reprinted as Chapter I of his
Soziologie [Simmel 1923].
32 The American Sociologist/Summer 1998
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
a few suggestions. He uses four different terms in the
discussion: impulses (Triebe), ends (Zwecke), interests
(Interessen) and, on one occasion, motives (Motive). There is an
implied classification in the use of interests as the more general
term and impulses and ends as particular classes of interests.
When, however, inter ests are pursued within a social relationship,
as is usually empirically the case, there is an additional,
analytically separable element which he calls the "mode"
(Art) or "form" of interaction (Wechselwirkung). It is the
existence of such forms of interaction which makes both of discrete
individuals, and of their dis crete motives a "society." It would
appear that form in this sense is precisely that which makes the
difference between attributes of an analytically separated
individual, and the concrete social entity constituted by a
plurality of such indi viduals functioning in mutual interrelations
with one another.
*
[4] Simmel's use of this starting point as a basis for a
scientifically important line of analysis rests on three
interrelated empirical theses. The first is that the
conceptualization of the traditional social sciences is in fact
couched in terms of classes of interests in the foregoing sense,
which are analytically separable from the forms of the social
relationships in which he stands. The second thesis holds that
form, seen in relation to the diversity of classes of interest,
** constitutes a
common element running through all of their social
relationships. The third maintains that form and content are
independently variable, that the "same" content is manifested in
different forms, while the "same" form is similarly to be found in
connection with different contents. Thus he says that an
economic
interest, the interest in maximization of wealth, may be pursued
in a competi tive form, by engaging in a competitive struggle
against others motivated by the same interest, or may take the form
of combination in the pursuit of the com
mon goal, as in monopolistic combinations and trusts. On the
other hand, the form of "competition" may involve economic
interests, or it may involve erotic inter ests as in rivalry for
the favor of a woman, or any one of a variety of other types.
The main key to the problems discussed in this paper is the
status of the first thesis. The other two are, as they stand,
entirely acceptable, the only question being that of their
implications for the methodology of social theory. It will be
necessary, before approaching the problems involved in the
first, to attempt a clarification beyond that provided by Simmel,
of the sense in which form is a common element, and of the
independent variability of form and content.
[5] It may be noted in advance, however, that Simmel fails to
give us even the outline of a thorough critical analysis of the
prevailing conceptualization of
what he calls the "social sciences." Above all, he merely
asserts and does not
attempt to justify in terms of the history and current
literature of those sciences, that all of them lie on the same
methodological level. There can be little doubt of the
inevitability, and for certain purposes, usefulness of the
classification of human interests in terms of such categories as
economic, political, religious,
* This is evidently similar to Durkheim's concept of "society"
as resulting from a synthesis of individuals. See E. Durkheim,
"Representations, individuelles et representations collectives," in
Sociologie et Philosophie [Durkheim, 1924] and the author's
Structure of Social Action, chapter IX [Parsons, 1937]. ** I.e., in
the sense of attributes of the individual formulated in subjective
terms. For a definition of "subjec tive categories" in this sense
see Structure of Social Action [Parsons 1937], p. 46.
Parsons 33
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
erotic, aesthetic, etc. Furthermore there seems to be a
corresponding classifica tion of types of concrete activity in
which the respective classes of interests are
predominant. It does not, however, follow that the problem of
the relation of scientific conceptualization to the classification
of types of concrete social phe nomena, whether formulated in terms
of interests, or of classes of acts, can be
disposed of by the simple assertion that there is an economic
theory which deals with economic interests and acts, a political
theory which deals with political interests and acts, a theory of
religion, etc. On the contrary it is quite certain that the
relation of scientific theory to concrete reality in these fields
is far more
complex than this formula would indicate. This is evident from
one fundamental
fact; that there is no actual uniformity in the theory of the
different social sciences to correspond to the symmetry of Simmel's
scheme. Economic theory has become a highly developed analytical
system of a peculiar sort, unique in the social field. On the other
hand whatever may go by the name of a "theory of art" is
methodologically quite different from economic theory. It will,
how
ever, be more fruitful to return to some of these issues after a
discussion of the
implications of the first two theses.
[6] As Simmel defines the concept it is a sheer truism that
every social
relationship has a "form" and that more complex forms are an
inherent feature of every system of social relationships. In this
sense his second thesis is beyond dispute. It is necessary only to
guard him from one misinterpretation. When it is said that form is
a common element of the relationships of all the different
parts of society this is not to be taken to mean that all the
similarities between
any two such fields are on the level of form, and all the
differences on the level of content. Form is not the generic term
of which the classes of interests con stitute the particular
species.
* The relation is quite different.
Form is rather a common element in the sense that every system
of social
relationships has form. But this proposition, which is a truism,
has no implica tions for the relations of the particular kinds of
form to be found in any two concrete relationships. This
distinction of two senses in which form may be
thought of as a common element gives a valuable clue to the more
general meaning of the concept form. For both in his discussions of
social relationships as such, and of the analogy he dwells on
constantly, that of geometrical form, Simmel apparently means what
is often referred to as "structure." A brief discus sion of the
meaning of the concept structure will do much to clarify the impli
cations of Simmel's procedure.
One of the most deeply imbedded modes of thinking about
"reality" conceives
"phenomena" as consisting of "parts" or "units" which stand in a
system of
intelligible relations to each other. A plurality of units in
determinate mutual relations is often said to constitute a
"system." The relations between the units are then spoken of as the
"structure" of the system. [7] A unit in these terms is the
referent of a combination of empirically descriptive propositions
which have
* Both Sorokin [1928: 399-400] and Abel [1929: 28-29] attribute
this view to Simmel. He did not sufficiently guard himself against
this misinterpretation.
34 The American Sociologist/Summer 1998
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
a unitary existential reference, that is are descriptive of the
"same" thing. To say that John Jones is five feet eight inches
tall, weighs 150 pounds, has brown hair and brown eyes, constitutes
a partial description of one kind of unit, a "man." But to say John
Jones is five feet eight, George Smith weighs 175, Frank Brown has
blue eyes, and James Johnson has red hair, obviously does not
constitute even a partial description of a single person.
The conception of a system of units standing in structural
relations to each other implies a further conceptual element than
that of unit and of relations; it
implies what may be called a "frame of reference" in terms of
which the system is described. Thus in the classical mechanics a
system consists of "particles" (i.e., "things" describable in terms
of a certain combination of "properties") existing "in" and related
to one another in space and time. The concepts of
particle, space and time, are not parts of the description of
any concrete system, but the frame of reference in terms of which
any system of one class, a "me chanical" system, must be
described.
There is a further fundamental conception which seems to be
inseparably connected with those of units and of structural
relations, that of "process" or
change. Systems do not, for the most part, remain static, but
processes go on in
them; a description which was accurate at one time ceases to be
so at a later one. The foregoing considerations help to specify the
meaning of process. A
process of change happens to "something," that is to a unit, or
a combination of such units in a system. The change must be
described as involving the prop erties of one or more units in the
system, or one or more structural relations
within the system, or some combin- [8] ation of both. Finally
the only meaning ful process is one which is capable of description
in terms of the frame of reference being employed at a given time.
Thus in terms of mechanics, to say that the velocity of a given
particle doubled in a given time is a meaningful description of
process, but to say that the particle changed from a velocity of
1000 ft. per second to a state of repression of the antagonism to
the father, is not. Both descriptive propositions may be meaningful
as applied to the same concrete object, e.g., a "man" but even
though the first is applicable at an earlier
time, the second at a later, knowing these facts does not
entitle us to say that the man has undergone a change, that a
process has taken place, because the two statements are not
commensurable in terms of the same frame of reference.
This fundamental schema of thought would seem to be involved in
our think
ing about empirical reality generally, rather than merely in
connection with the
"physical world." A related, if not identical schema is involved
in our thinking about non-empirical, "ideal" objects such as
systems of logically interrelated
propositions, or artistic compositions. A symphony has its
"themes" or other units which have certain structural relations.
But the very ubiquity of the schema should make us cautious in its
use, for in different uses and contexts it may have a very
different significance. Indeed before we can say in what sense
empirical knowledge of causal relations is derivable by its use, it
is essential to be aware
of a certain fundamental relativity in the status of the
different categories of the
schema.
Parsons 35
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
[9] Every system has a determinate structure. This is true
whether the sys tem is "static," that is when no changes
significant to the observer are going on, or it is "dynamic," where
the observer's interest is centered on changes over
time, which may, of course, involve structural changes. * But it
must not be
forgotten that the frame of reference is one of the fundamental
conceptual elements of the schema of thought into which the concept
of structure here considered fits. Any given structure is relative
to the frame of reference in terms of which the relevant systems
are described. A structure is a set of determinate relations
between a plurality of a certain kind of units, described in terms
of a certain frame of reference. On any given level of analysis a
structure may be conceived as static without process, or certain
processes may be conceived as
going on "within" an unchanging structure, or finally, the
structure itself may be conceived as undergoing a process of
change. Thus the structure of a table in
ordinary common-sense terms is constant, and for most purposes
we are not interested in any processes which may be occurring
within it. It is usually described as a static system. The plumbing
system of a house, however, is a different kind of system. There is
a constant physical structure of the pipes and
appliances, but within this structure an important process goes
on, the continu ous or intermittent flow of water. Finally a
developing embryo, from fertilized ovum to mature organism, is
still a different kind of system in that, from the
point of view which interests biologists, it is undergoing a
continuous process of structural change.
But most authorities would be agreed that in relation to all
these types of
systems there is a certain fundamental relativity in the use of
the concept struc ture as a set of constant relations even at any
given time. There is a sense in
which the static aspect of all structures is an illusion. What
is described as a
"rigid" structure consists [10] "really" only in certain
uniformities of process. Thus a candle flame has a specific "form,"
a structure, but here even for com mon sense the burning of the
candle is a process, within the form there is intense movement of
hot gases, and rapid chemical change. But the physicist
will tell us the same is true of the cool wax, or of the static
table. These
seemingly so solid and substantial structures are only
relatively uniform and
stable resultants of very complex processes. If this is true of
the systems which common sense treats as static, it is doubly so
for that of the developing embryo.
But pointing out this underlying aspect of process does not
dispose of the
relevance of the concept of structure. When the physicist
attempts to describe
the processes going on within the table he does so partly in
terms of a deeper
lying structure, that of the physical systems of atoms and
electrons which make
up the table. What has happened is that the shift from
consideration of structure in the common sense description of the
table, to that of process in the physicist's
* It should be remembered that whether a system is static or
dynamic is not a question of whether process "really" goes on in
the system, or not, but of whether the attention of the observer is
focussed on the
processes, or on the structural uniformities of the system.
Thus, in the example used below, to common sense a table is "solid"
and unchanging over a short period (unless for instance it catches
on fire) but to a physicist the same object is a dynamic system of
rapidly moving atoms or electrons.
36 The American Sociologist/Summer 1998
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
description, involves a shift in frames of reference. There is
no inherent meth
odological reason to set a limit to the possible number of such
steps. A table which appears so simple to common sense may be
interpreted as a very com
plex system involving a highly dynamic process by the physicist.
But his most
elementary units, the electron, proton etc., and elementary
structures like the
atom, may in turn prove to be equally complex when analyzed in
terms of a still more elementary frame of reference.
When dealing with a concrete type of system, such as a system of
social
relationships, it is hence not enough to inquire what is the
structure of the
system. Different levels of structural categories will be
relevant according to the frame of reference in terms of which the
system is described and analyzed, and
according to the methodological use to which the structural
categories are put. [11] This brings us to some of the implications
of Simmel's third thesis, that form and content are independently
variable. It is in this connection that the
problem arises as to the sense in which a study of social
"forms" can constitute the analytical framework of a theoretical
science. Simmel apparently considers this question only on one
level which may, for present purposes, be called the
concretely descriptive. The form or structure of a system of
social relationships is a "descriptive aspect" of such a system. It
consists in certain facts about the system
which can be stated in relative isolation from other facts, that
is, in the present case, in isolation from the classificatory
character of the "interests" which are involved in the system. On
the level on which Simmel claims independent vari
ability there can be no doubt that this variability is a fact.
Furthermore there can be no doubt that this independent variability
under
certain conditions, * possesses some causal significance. For in
the logic of cau
sation, or if that term is objectionable, "determination," there
is no other test of causal significance than the independent
variation which can be demonstrated
by comparison of different cases involving the same value of one
variable in different combinations with the values of others, by
experiment under con trolled conditions, or by what is usually
known as "comparative method." What constitutes variables in a
logically acceptable sense cannot be discussed in detail here, but
there does not seem to be any obvious reason why form and content
on the one hand, or various sub-categories of each, should not
consti tute logically satisfactory variables in the most formal
sense. The problem then, is not whether it is possible,
scientifically, to accomplish "anything" by Simmel's procedure, but
whether it is more fruitful than other method
ological possibilities. [12] At this point the previously
introduced considerations of the relativity of the concept
structure, become relevant to our problem. It may be asked whether
in the sciences which have achieved the highest development of
analytical theory the structure of the concrete systems they deal
with in the most concrete descrip tive level, is usually treated as
one, or a system of variables. The answer is
uniformly no. The description of the structure, as also of the
units related in it,
* See below for a further discussion of this problem.
Parsons 37
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
is treated as the statement of the problem to be tackled by
theoretical analy sis, not as providing the conceptual materials of
the analysis itself. Put some
what differently, concrete structures are generally treated, not
as the values of
variables, but as the resultants of the interaction of more than
one variable. A
good example is the way in which the structure of a waterfall
would be treated in elementary physics. It, like any other object
of analysis, has a "form," which is perfectly capable of
geometrical description as involving for instance, an arc of a
certain type of parabola. But it would never even occur to a
physicist to
explain any aspect of the flow of water over a fall as the
result, in any part, of the arc of a parabola considered as a
causal factor. He would treat as "causes," on the one hand a
deeper-lying structure, that of the contours of the riverbed,
including the fact that they were rigid enough not to erode
significantly over a short period. This structure would be treated
as a set of "constants" of the
problem. The other main constants would be the relevant
properties of water, its specific gravity and viscosity. Then as
the values of variables would be treated the volume and rate of
flow of water, from which, in combination with the
constants, the concrete form of the fall, including the formula
for the particular parabola in question, would [13] be derived.
This does not mean that it would not be possible for physicists to
ascertain the parabolas of waterfalls by direct
observation, and use these formulas as the values of a basic
variable in the
explanation of what happens. It is only that by that method it
would not be
possible, with the same degree of simplicity, to derive
theoretical formulas of a
comparable degree of generality, applying not only to
waterfalls, but to a large variety of other physical phenomena.
This example is taken from a simple field of application of a
particular theo retical system, the classical mechanics. The long
experience of the science of
physics seems to have shown that this system gives accurate
results only in so far as it is possible to conceive the concrete
systems to which it applies as
closely approaching a certain type, which has come to be called
"atomistic" or "mechanical." The relevant feature of such systems
for us is that it is possible to treat the units and their
properties as descriptively independent of their relations to other
units of the system. Thus in the Newtonian treatment of the solar
system the masses of the bodies which make it up are treated as
entirely independent of the distances from or velocities relative
to each other. It is, however, precisely this assumption which, in
relation to certain astronomical
problems, the theory of relativity has had to question. If I
understand it cor
rectly, the innovations introduced by the theory of relativity,
and the other great branch of modern physics, the quantum theory,
have in common the fact that
they no longer treat their systems as strictly atomistic in this
sense. *
If a certain modification of this type of theory has proved
necessary even in
physics, it is not surprising that currents of thought of
fundamental importance * A comparative example from the social
field is that of the way in which an economist would ordinarily
treat
a competitive market situation. Such a market has a structure
just as a waterfall has. The term competition is descriptive of an
aspect of that structure just as the parabolic curve is of the
physical structure. More over, competition, being a mode of the
interrelations of human individuals in a process of social
interaction is in Simmel's technical sense a "social form."
Competition is not, however, a variable in the system of
38 The American Sociologist/Summer 1998
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
in the biological and social [14] fields have, since conscious
reflection on such
subjects began, always refused to consider it an obligation of
scientific analysis to treat the systems they were concerned with
as strictly atomistic. The term
most generally acceptable for the property which primarily makes
the difference seems to be that such systems are to a greater or
less degree "organic."** The usual definition of an organic system
is one relative to which it is impossible to treat the properties
of the parts as independent of their mutual relations in and to the
whole. The question thus arises whether the "organicity" of a
system alters the status of structure as a variable from the case
of the atomistic system.
There is undoubtedly an important sense in which this is true.
To be sure it is never possible to understand the behavior of a
system solely in terms of the
properties of its units. There are always certain minimum
relational terms which are required, in terms of the frame of
reference, in order that the conception of an existent system
should make sense at all; what I have elsewhere called its
elementary relations. ***
These, like distance in a mechanical system, are always taken
account of in an atomistic analysis. When a system is organic to an
impor tant degree, however, the total system has descriptive
features which are not
logically required by the frame of reference as such. Since they
are properties of the total functioning system, not derivable from
those of the units, they are,
according to the definition adopted here, closely identified
with the form or structure of the system.
It is true that Simmel was self-consciously attacking certain
extreme forms of
"organic" social theory; social form to him was not a
"substance," but a mode of process. But nevertheless it may well be
suspected that in treating form as an
independent variable he was seeking a way of dealing adequately
with certain of these [15] organic features of social systems. This
may well be the main reason why he did not bring in the rather
obvious considerations just discussed as to why structural
categories on the immediately descriptive level do not have an
important place in the analytical theory of mechanics.
This presumption is greatly strengthened by an important fact of
the history of social thought, the extent to which economic theory
and the closely related "utilitarian" social theory has been
"individualistic," since one of the several connotations of this
term is atomism in the sense of this discussion, and has,
particularly in Germany in Simmel's time, been continually attacked
as such. It
may well be suspected that Simmel was seeking a way of taking
account of the
organic elements neglected by individualistic theories, without
involving the wholesale repudiation of the latter in all respects,
which has been so common in the organic traditions of social
thought. The analytical situation underlying this presumption is in
need of further clarification.
economic theory, it is a descriptive aspect of the systems to
which economic analysis is habitually applied. That a market is
competitive is not the explanation of anything, but is something
for the economist to
explain. As causal factors, on the other hand, the economist
will use the numbers of participants in the market on both sides,
the distribution of economic resources on either side, among them,
the "conditions" of demand for the particular product or service
dealt with and the supply of cost conditions. ** See Structure of
Social Action, pp 3Iff, pp 738ff. *** Ibid., p 734.
Parsons 39
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
There is a sense in which the basic explanatory schema of
economics in relation to a competitive market situation is closely
analogous to that of classical
mechanics. Such a competitive market undoubtedly has a
structure, the individu als in it stand in certain determinate
forms of relationship to one another, on the one hand as
competitors, on the other as parties to actual and potential sales.
But traditionally economists, like physicists in the case of the
waterfall, would not attempt to explain anything by the "form" of
competition. The competitive market would rather be described as a
statement of the analytical problem. Explanation however, would be
in different terms, this time of elements of the "motivation" of
individuals, the maximization of money proceeds from the trans
actions, and, on the other, hand, of certain deeper-lying
features of the situation,
partly structural, such [16] as the available quantities of the
relevant commodi ties or services, or the "costs" of producing
them, and the "conditions of demand" on the part of buyers as
ultimate consumers, or if they are not consumers, of others in the
society.
This kind of schema for the analysis of certain types of social
relationships has, in the utilitarian tradition of thought, been
broadened out to the more
general conception of "contractual relationships." As such it
has occupied a
central position in what has perhaps been the most important
theoretical tradi tion in the social field in modern times. At the
same time there has been an
important movement of thought which has shown that the elements
usually explicitly formulated in the economic analysis of
competitive markets, or in the broader account of contractual
relations, are inadequate to account for certain of the important
features of systems of such relations, above all their stability,
and elements of motivation which extend beyond the range of
"immediate eco nomic self-interest."* The essential point is that
while individualistic theories have tended to lay stress on what
may be called the "factual" conditions of contractual relations,
natural resources, the biological work-powers of men etc. on the
one hand, their biological needs on the other, the movement of
thought in question has laid stress on what may be called
"normative" conditions, such as the institutional rules of honesty
and other aspects of the property system, and of the ultimate
values directly expressed in economic activities.
Both these types of elements have in common, that for purposes
of the analy sis of the action of a particular individual, or even
of a limited sub-system of action such as a particular market-price
problem, they may, with relatively little
damage to accuracy often be neglected. Their crucial analytical
[17] importance as variables, has only become evident with the
attempt to formulate a general ized analysis of total social
systems.
Seen in this context, Simmel's attempt was certainly leading in
the right direc tion. In terms of his approach these elements,
since their importance was not
necessarily evident from the consideration of an isolated
individual, are neces
sarily closely involved in the structure of total systems.
Hence, his emphasis on
* Durkheim, 1893: bk II, ch. 7; Max Weber, The Protestant Ethic
and the Spirit of Capitolism [Weber 1930]. See also Structure of
Social Action, chap. XIII and XIV.
40 The American Sociologist/Summer 1998
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
structural categories is a way of stating the problems, which
makes it much more difficult to neglect the empirical role of the
non-individualistic elements than was the case with the utilitarian
mode of thought. In this respect it plays a role similar to that of
conceptions of organic totality in biology, from Aristotle's
entelechy down.2 But this relative usefulness, important as it
may be, should not be allowed to
distract attention from the fact that Simmel's schema obscures
certain very important analytical possibilities which other
developments of social theory, not open to the above criticisms
directed against atomism, have opened up. I should like to
develop briefly two aspects of these possibilities, both of
which rest upon an
analysis, not of form as such, but of what Simmel wished to
exclude from
sociological conceptualization, the "motives" or "interests" of
individuals. Simmel distinguished form from "content," which rather
consists of the "mo
tives" or "interests" of individuals. The latter are then
classified in relation to the differentiation of human activities,
as "economic," "religious," "political," etc. It is clear on the
one hand that what is ordinarily called the "motivation" of action
is not involved in the analytical concept of social form, on the
other that a
motive or an interest is for Simmel an integral unit which for
his purposes he does not attempt to subject to any [18] further
analysis. Simmel does not carry his discussion of the problem far
enough to commit himself, but the presump tion is that he thinks in
terms of a one-to-one correspondence between motive and act, each
act having its appropriate motive.
But it is precisely here that the analytical possibility on
which Simmel's ap proach closes the door, opens up. There is no
better empirical justification for
assuming a one-to-one correspondence between act and motive than
there was for the logically similar assumption underlying the
so-called "unit character"
theory of inheritance in biology, *
that there was a one-to-one correspondence between the
analytical unit of heredity, the "gene," and the concrete somatic
character of the developed organism. In both cases a similar
undesirable rigidity results, on the one hand the denial of the
possibility of environmental influence on the organism, on the
other such conceptions as that of the "egoism" of the economic man,
accepted as a literal total description of the concrete
business
man's concrete motives.
The fact is that the "motives" of economic theory or of
psychological and
sociological theories should not be treated as names for
concrete entities, but as
analytical categories. In particular the classifications of
motives, such as that which Simmel employs, are not classifications
of concrete motives, but rest on
analytical distinctions. Concrete motives are seldom "pure
types" but almost
always involve some combination of the elements distinguished in
the classifica tions.
This proposition could be rigorously proved in terms of a large
number of different empirical fields, if space permitted. Here it
will have to suffice to call attention to the situation with
reference to the problem of "economic motives."
*
* On this theory, see H.S. Jenning's The Biological Basis of
Human Nature [Jennings, 1930].
Parsons 41
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
Many of the most objectionable empirical inferences from the
older economic
theory flowed precisely from this identification of the [19]
economic element of concrete motivation, the "rational pursuit of
self-interest" with the total con crete motivation of a concrete
system of action. On the one hand analysis has, I venture to say,
definitely proved, that economic motivation in this sense can not
exhaust the concrete motivation of the individuals involved in
market or other "economic" relations, but that any concrete
motivation system relevant to the analysis of such phenomena must
be far more complex. This is true in spite of the fact that for
certain limited purposes it is quite legitimate to treat only these
economic elements of motivation as variables in the analytical
problem in
hand, taking account of the others as constants, which, however,
are logically essential premises of any sound empirical
conclusions.
On the other hand, the investigations which have demonstrated
the empirical inadequacy of a conception of a system of purely
contractual relations, or of the
conception of the concrete actions of modern business men as
motivated en
tirely by the pursuit of self interest, have, in the process of
analytical develop ment through which their authors have carried
them, come to be directly inte
grated with a generalized theory of the motivation of action in
which the element of economic motivation has a crucially important
place, not however, as the concrete motive of a class of action,
but as an element standing in certain
quite definite structural relations to other, quite different
elements. In the case of Max Weber's analysis of the role of
"disinterested" elements in the spirit of
capitalism, this articulation with a general theory of
motivation has been direct. In that of Durkheim's analysis of
contractual relations, it was more indirect, but
through his conception of the role of a sense of moral
obligation in the enforce ment of institutional rules, came to be
none the less definite.
[20] It is true that this generalized theory of the motivation
of action contains structural elements of fundamental importance.
Indeed it is in relation to the structure of social systems of
action that it is by far [the] most highly developed so far. But
this is not a vindication of Simmel's analytical approach, it is
rather a
conspicuous example of what has already several times been
mentioned. Social structure described in Simmel's terms as a
complex of forms of social relationship, is broken down
analytically, is not treated as an independent variable, or system
of them, but as a resultant of the interdependence of a group of
other variables. And these variables are not elements of the forms
of relationships, but of the motives of action, the very elements
which Simmel explicitly excluded from the
analytical consideration of sociological science. If one's
approach to sociological theory is, as in the case of the present
author, through the theory of social
action, Simmel's formula, far from being acceptable, "throws out
the baby with the bath." By definition it excludes from the start
the conceptual elements which form the basis of the analytical
system of sociological theory.
* See the author's "Reflections on the 'Nature and Significance
of Economies'" Quarterly Journal of Econom ics, Vol. 48 [Parsons,
1934] and Structure of Social Action, [Parsons, 1937] chap. IV, pp.
I6lff, chap. VI pp. 264ff.
42 The American Sociologist/Summer 1998
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
Before developing the consequences of this exclusion a little
farther, I should like to say a few words about the other respect
in which Simmel's procedure shuts the door on analytical
considerations of fundamental importance to soci
ology. Structure is, as has been said, a descriptive aspect of
all empirical sys tems. But precisely because it is common to all
empirical systems, basing analy sis on the concept of structure
fails to build into the methodological basis of a
particular analytical system, certain of the elements which are
of especial im
portance to the analysis of the particular type of systems to
which it is applied. [21] The further the analytical theory of
social action has developed, the more evident it has become how
fundamental to it is one feature of action systems, what I have
elsewhere called their "normative orientation."* Normative
orientation is an
indispensable logical component of the frame of reference of
action, the means-end
schema; it is impossible to think in such terms without it.
Moreover the basic fact of normative orientation necessitates the
analytical distinction between two classes of elements of action
systems, the normative and "conditional" elements.
Now Simmel's concept of the concrete system "society" is far
from being logically independent of the frame of reference of
action, and of the means-end schema. His definition of content by
such terms as impulse, end, interest and
motive is sufficient to prove that. Indeed, "society" for him is
precisely a process of continuous mutually oriented action of
individuals in relation to one another. The essential processes are
processes of action in the technical sense of the
theory of action now under consideration. But the "line" which
Simmel's distinction between form and content draws through the
facts singles out the concrete structural forms of social
relationship for the attention of sociological theory. Concrete
structural form is, however, a category to which the whole fact of
normative orientation, and the distinction and Spannung3 between
normative and conditional elements is conceptually irrelevant. This
results in pushing the
whole problem of normative orientation over into the sphere of
"content"; it becomes a matter for the "social sciences," not for
sociology.
But this fact has the most far-reaching consequences. For in
relation to the
theory of action, it is the active orientation to normative
elements which plays the analytical role analogous [22] to that of
energy in physical theory. It is by virtue of that alone that the
analytical system becomes the theory of dynamic systems, of process
and change. But without the capability of analyzing, not
merely describing, process in the relations of the independent
variation of vari
ables, it is impossible to establish causal relationships or
"laws" in the usual
analytical scientific sense. This situation is, in my opinion,
the principal source of one of the most striking, and
scientifically unsatisfactory features of "formal"
sociology, its "static" character. It does not yield a causal
analysis of empirical phenomena, which are always in some sense
dynamic systems.
Simmel's procedure and formula have an important historical
justification in at least two respects. It has been pointed out
that in one connection, it may be
regarded as an attempt to work out a mode of analysis of social
phenomena * See Structure of Social Action, [Parsons, 1937] chap.
II pp 44 ff and Note A. Also frequently discussed
throughout the book.
Parsons 43
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
which was not subject to the criticisms directed against the
atomistic theories,
particularly those of "economic individualism," and which would
likewise avoid
the more extreme "organicism" of much of the German social
theory of his time.
But in addition to this Simmel had another very important
insight which was in
advance of much of the though of his time: This was the
realization that the
basis of much of the most fruitful scientific conceptualization
is not a naive
reproduction of the concrete objects of the external world, but
involves a high
degree of analytical abstraction. "Empiricism" in this sense was
a dominant
characteristic of the methodology of social theory in almost all
schools at the time. To have vindicated the legitimacy, even the
necessity, of analytical abstrac tion was a great service.
But from the fact that abstraction is essential to scientific
theory it does not
follow that all modes and directions of [23] abstraction are
equally fruitful for all scientific purposes. That Simmel embarked
on the path of abstraction is
greatly to be commended. But it is unfortunate that he did not
analyze the various possible distinct modes of abstraction and
their mutual relations more
thoroughly. To clarify the previous discussion it may be useful,
very schemati
cally, to distinguish three different types of abstract
concepts, of which Simmel's form is only one.
Simmel was, of course, aware that a concrete phenomenon or
system, as
described for scientific purposes, is not simply a verbal
reproduction of the external world, but involves a selective
ordering and abstraction. But this is not the present concern. Once
a system has been described, it may be analyzed on
the structural level, as discussed above, into units and their
structural relations. It is obvious that the structural relations
always constitute in a sense an abstrac
tion, because it is nonsensical to think of them as "existing"
apart from the
relata, the units whose relations they state. "Marriage" is not
a thing, which can exist concretely without reference to any human
beings; it is a mode in which concrete human beings are related to
one another.
The unit of an atomistic system is not an abstraction in quite
the same sense. The properties of a particle in classical mechanics
are thought of as quite inde
pendent of its relations to other particles. It can be taken out
of the system and still be the "same" particle. But in so far as
the system is organic, this ceases to be true, and a unit in the
sense of an entity thought of as existing apart from its relations
in a concrete system is a fiction. It might be possible
experimentally or conceptually to isolate such a unit. But it would
not be a unit of that system in the [24] sense in which an isolable
particle is a unit of mechanical systems. Simmel's discussion of
abstraction is limited to this distinction of units and structural
relations. His insistence that the unit or content, as well as form
is an abstraction, is, I think, largely a result of his realization
that social systems are
highly organic in the sense of the above discussion. There is,
however, a third type of abstract concept which plays a basic
role
in the methodology of science, and which Simmel ignores
completely. That is what may be called the "analytical element" or
variable. In content this may
overlap with either of the other two, but it need not. A
variable may be simply
44 The American Sociologist/Summer 1998
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
a "property" of a unit or system of units. One can conceive a
particle as existing independently of its relations. But one cannot
conceive "mass" as independently existent. The expression "a mass"
is, in the technical sense of mechanics, non sensical. Yet mass is
one of the basic variables of the theory of mechanics.
Whether the conceptual content of a variable coincides with what
are for some purposes unit or relational concepts is not, however,
the important point. It may be said that a system of variables
never simply describes the units and structural relations of an
empirical system on the most concrete level relevant
within the frame of reference being employed. Such description,
including the abstraction of structural relations from the
properties of units always has the function of stating a problem
for causal analysis. Such analysis involves in addi tion to
description the application to the particular facts of a system of
variables. This application always involves, explicitly or
implicitly, a comparative refer ence to parts of the range of
variation of the variables which lie beyond the values involved in
the particular description. It is only by virtue of generalized
knowledge of the relations of inter- [25] dependence, of the
variable, derived
from comparison, that it is logically possible to draw causal
conclusions. Fur
thermore, the comparison from which such generalized knowledge
is derived, cannot be limited to a descriptive comparison of
structures on this most con crete descriptive level.*
It is true that structural categories often play an important
role in causal
explanation. But it is safe to say that this is never true when
they are the most concrete descriptive structures. We may revert to
the three types of cases dis cussed early in the paper. In the case
of the common-sense description of a table there is no dynamic
problem. The conception of the table may, however, be used
negatively to explain why certain dynamic processes do not take
place, why for instance objects placed upon it do not fall to the
floor. This causal
knowledge cannot, however, be derived merely from observation of
the table. It requires in addition the application of generalized
analytical knowledge. It is safe to say that Simmel did not think
of form in relation to this kind of static case.
His thinking is more analogous to the case of the waterfall. But
the analogy of form in the sense in which he treats it as causally
significant is not the concrete form of the fall, but the structure
of the riverbed. This is causally significant to the process of
fall of the water, including the concrete form of the
fall, because it sets limits to the dynamic process of flow. If
the physicist can take certain features of the riverbed as given,
can neglect their interdependence
with the process of flow, then his dynamic problem is immensely
simplified. That is structural categories, as constant data for
dynamic problems, are always of causal significance. But these are
never the structures of systems on the most concrete level, but
[26] on a deeper one. Simmel, in failing to make this distinc tion
of levels in its application to his methodological problems,
creates a serious confusion. Descriptively his forms are the forms
of concrete social relationships. But these cannot have causal
significance in relation to processes of change in
* On all this, see Structure of Social Action, [Parsons 1937]
chap. XVI, pp. 601-624 and chap. XIX, pp. 795ff.
Parsons 45
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
social structure, which most important dynamic sociological
problems in fact involve. So, in his empirical essays, we find that
he does not in fact explain social
processes by forms. He describes the forms, and then turns to
other elements for his
explanations. An excellent example is his treatment of number as
a determinant of social relationships.4
We must not forget, then, that although structural categories
may often have causal significance, it is only as constant data in
a dynamic problem. This fact
rigidly limits the generality of any theory which depends
primarily on such
concepts. Its categories are significant only within the range
in which it is
empirically legitimate to assume relative constancy. Beyond that
range new dynamic elements must be admitted to the problem, which
takes the theory of it off the structural level. After all the
essence of theoretical advance in science is the
reaching of higher and higher levels of analytical
generalization. The use of form
concepts as the constructive principle of explanatory theory, is
not an instru ment in this process of advance, it is a statement of
the limitations on theoretical
generalization which are necessitated by the scientific
ignorance of the time. Advance comes, not through the development
of the "theory of forms," but [27] through the replacement of such
concepts by those of dynamic analytical theory. It would be
difficult to conceive a more pernicious methodological doctrine, in
its long run implications for social science, than Simmers. For, if
it were adhered to with real faithfulness, the effect would be a
fixation of generalized knowledge of social processes on the
present elementary levels, the inhibition
of all analytical progress in the theory of human social
behavior. This brings us to the question of the kind of
systematization of which Simmel's
type of social theory is capable. This may best be discussed in
terms of what has
actually happened to the theory of formal sociology. Simmel's
own case is of
great interest. In spite of being the originator of the
methodological program, he never even attempted to develop a system
of theory on its basis. His other
sociological writings consist in a series of essays on
particular social forms. These are brilliantly illuminating essays,
among the finest things to be found in
sociological literature. But even taken together they clearly do
not comprise a*
system of formal sociology. Moreover, as has already been
remarked, Simmel does not consistently follow out his
methodological program even in these dis connected studies. It is
true that he talks about social forms. He describes them and uses
them as his starting point. But he does not consistently use them
as
explanatory categories. His discussions are full of reference to
motives and the other things he has methodologically banned. The
very great fruitfulness and
originality of these essays derives, not from his methodological
use of forms as
analytical tools, but from the fact that he approaches the
analysis of social
phenomena from an unusual point of view which cuts across the
conventional
compartments of economic, political, etc. But his actual
explanatory [28] theory is not "formal," it is overwhelmingly
motivational. Only it remains on a common
sense level which, however much it may illuminate certain
hitherto obscure
empirical problems, does not contribute in any important measure
to the
development of the systematic analytical theory of human social
life.
46 The American Sociologist/Summer 1998
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
A very different case is that of Leopold von Wiese. Professor
von Wiese has
explicitly accepted Simmel's formula and attempted to build a
system of formal
sociology. The result of this attempt is an elaborate
classification of possible types of social relationship. It starts
from the most general possible criterion of
distinction, whether relations are "associative" or
"dissociative," and proceeds by adding progressively more and more
criteria. Empirical material has a place only by way of
illustration. There is no attempt at explanation, only at pinning
the labels of the structural types thus constructed, to certain
empirical cases.
This is not to say that classification, of structural types as
well as other things, is not an entirely legitimate and
indispensable scientific procedure. But classifi cation has its
function mainly in two respects, on the descriptive level as
stating the problems for theoretical analysis, on the analytical,
as the mode of system atizing the values of particular variables.
In the former case a classification is
scientifically significant in direct proportion to the extent to
which it is inte
grated with a system of analytical theory. Professor von Wiese's
classification does not grow out of far-reaching empirical problems
of causation, nor is it
integrated with any analytical system. It is "purely formal." As
such it is very little more than an exercise in spinning out the
logical implications of defini tions. The further elaborated it
becomes the less useful are its creations [29] as tools of
empirical research. This formal sterility is indeed inevitable if
the for
mula of Simmel is strictly followed, as the sole methodological
program of a theoretical science. The situation becomes especially
clear through comparison of the work of von Wiese with that of Max
Weber.
The extensive system of explicit conceptualization to be found
in Weber's work is the system of ideal types, most fully formulated
in the first part of
Wirtschaft und Gesellschaft [Weber 1925].6 There is no question
but what (sic) this constitutes formal sociology in the sense of
Simmel and von Wiese. It con firms the fact that the sole possible
mode of systematization for such theory is a classification of
types. But there are, apart from the specific concepts used, two
striking differences from the system of von Wiese. In the first
place Weber's
types were predominantly constructed in the course of a
far-reaching program of empirical research, above all his
comparative studies in the sociology of
religion. Each type is a tool in the task of grappling with
basically important empirical problems, problems of causal
relationship, not of classification. Hence in origin it is not
formal in the sense of Wiese's system. The formal systematics is
secondary to the empirical uses. Weber did not start with certain
formal distinctions and spin out further distinctions from
there.
Secondly, investigation of this system shows * that its
systematic aspect is by
no means limited to the "formal" side. On the contrary the basic
logical frame work of the classification is the outline of an
analytical system of theory, on a
deeper level of analysis than the structure of social
relationships. It is, in fact, one of the most complete versions of
certain aspects of the "theory of social action" to be found in the
literature. In spite of the fact that Weber was meth
* See Structure of Social Action, [Parsons 1937] chap. XVII.
Parsons 47
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
odologically unaware of [30] certain of the vital functions of
analytical theory, in his own work he built up the outline of an
analytical system, driven, we may surmise, largely by the sheer
logic of his empirical facts. Thus his "formal" classification of
relationships is closely integrated with a conceptual system on a
deeper analytical level. It is this, combined with its close
relation to empirical
problems, which saves it from the objectionable "formalism" of
von Wiese's
system. Both these features are notably lacking in the latter.
In short the formal
part of Weber's theoretical work is scientifically fruitful
precisely because he did not limit himself to Simmel's
methodological program, but carried out formal
theorizing only in connection with dynamic problems of causation
in empirical research, and integrated his formal concepts directly
with an analytical system, the concepts of which are essential
tools dealing with the same range of empiri cal problems. What
Simmel, failing to live up to his own program, did on a
dillentantish, common-sense level, with many brilliant and
arresting insights, but no imposing structure of proof of
propositions of far-reaching importance, Weber
accomplished on a far higher level, meticulously building up a
rigorous proof of his theses, in the manner not of a brilliant
dilettante, but of a sober professional scientist. Brilliant
dilettantes have their place in science, but the progress of
science can hardly be left to their efforts alone. And overcoming
that stage necessitates a careful analytical study of the
methodological problems underly ing such proposals as that of
Simmel.
In spite of these virtues, there are a number of points at which
Weber falls short of the highest level of rigor in his proofs of
which his empirical level is
capable. Analysis of his work in these respects shows that one
main source of his [31] difficulties lies in what, in essence, is
the following of a formal procedure in Simmel's sense, where that
is not methodologically advisable. This results in a kind of
"ideal-type atomism" which unfortunately cannot be dealt with
here.*
A few of the more important points of this paper may be
reiterated in con clusion. On the level of his own explicit
definitions Simmel's categories of social form are not suitable
tools in problems of causal explanation on the theoretical level at
all, but only of description for the purposes of clear statement of
explana tory problems. The importance of this function should not
be underestimated, but by the same token, in relation to just what
functions in science particular types of concepts are important,
should be made as clear as possible. Structural
concepts, on the other hand, not on the most concretely
descriptive level, but on the next level of analysis, may have
considerable significance for explanatory purposes. But even here
they play their role in the form of constant data, not as
functionally interdependent variables of a system. The more, then,
a theoreti cal scheme hopes to attain a high level of generality as
an explanatory tool, the less can it depend, empirically, on the
descriptive facts of the structure of the
* See Structure of Social Action, [Parsons 1937] chap. XVI, pp.
607ff. It is hoped the reader will excuse the continual references
to the author's own work. The explanation lies in the fact that the
content of the
present paper was originally conceived as a chapter of that book
which, for reasons of space, it was decided to omit. It constitutes
an extension of the analytical argument of the book into a slightly
different field. Hence, references must continually be made to the
starting points of the extension, in the book itself.
Without familiarity with these, comprehension of the paper will
be difficult.
48 The American Sociologist/Summer 1998
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
empirical systems to which it applies. To elevate the
formulation of structural
categories in this sense, as near the concrete level as
possible, and the sole
possible mode of systematization of such categories, structural
type-classifica tion, into the sole methodological program for the
development of theory in any empirical science, is (precisely) to
set rigid limits, which are in principle declared to be [32]
insurmountable, to the development of analytical explanatory theory
in the field in question. This is to cut off at the source the
growth of the very type of theory which has proved most fruitful in
the advancement of knowledge in what are almost universally
regarded as the most advanced branches of sci
ence, the physical sciences. Its inevitable effect would be
either to condemn
sociology to a rigid and sterile formalism which would cut if
off from any real
grappling with dynamic empirical problems of broad scope and
significance, or else to condemn it to a brilliant dilettantism
like that of Simmel himself which, however fruitful at certain
stages in the development of a science, can hardly be a
satisfactory basis on which to erect a program for the general
theoretical
development of the field. It is this dilemma, so strikingly
exemplified in the directions which the work of the two leading
formal sociologists so far, von
Wiese and Simmel himself, have taken, which justified the harsh
judgment, that it would, in the long run interests of sociological
theory, be difficult to conceive a more pernicious methodological
doctrine than that of Simmel, if it is taken
seriously as defining the sole acceptable program of theoretical
conceptualization for an empirical science.
The author's own conviction that the dilemma need not be
accepted has been
abundantly expressed throughout the present paper. It may,
however, be noted that this conviction is based on knowledge that
analytical theory of the type here advocated is not, in the social
field a mere program, but a reality which has
already reached a relatively high stage of development. * The
progress of socio
logical theory, as that of the other theoretical social
sciences, lies in following up those beginnings to a higher state,
not in turning away from them to follow what is, from this point of
view an enticing, but essentially fruitless program, like that of
Simmel.
At the beginning of this paper it was said that we do not
propose to revive the question of the scope of sociology. No
attempt will here be made to draw the lines between sociology and
its neighbors in the field of human affairs. Suffice it to say that
it is the author's strong conviction that the most promising
way to develop its contribution to our knowledge of man and his
social life, does not lie in making it methodologically unique, as
Simmel would have us do, but in centering it on a system of
analytical theory, which is part of a still larger system of theory
dealing with human social action as a whole. It is because
Simmel's
methodological program for sociological theory constitutes a
direct obstacle to this kind of development, that it has seemed
worth while to submit to the
profession at this time a discussion of the methodological
issues underlying it.
* That this is the case is the most important thesis of The
Structure of Social Action [Parsons 1937].
Parsons 49
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
-
Notes
1. Parsons originally discussed "encyclopedic sociology" in
relation to economics. He held that if economics were to based upon
empiricist principles it would have "no place for a logically
separate body of principles of economics." Rather, economics would
be "merely the application to a particular body of concrete
phenomena of the general principles necessary for understanding
human conduct." Hence, it would con stitute "encyclopedic
sociology": "the synthesis of all scientific theory relevant to the
concrete facts of human behavior in society." He concluded that
"economics then becomes applied sociology" (Parsons, 1937:
173).
2. This means, "a realization or actuality as opposed to a
potentiality," or in vitalist philosophy "a vital agent or force
directing growth and life."
3. Tension. 4. Simmel discusses this issue at length. See Part
II, "Quantitative Aspects of the Group," in Simmel (1950).
This material was translated from Soziologie (Simmel 1923). 5.
Parsons was undoubtedly referring to Becker's adaptation of Von
Wiese's work (Becker and Von Wiese,
1932), which he placed on a bibliography for a course that he
taught on European Sociological Theory.
References
Abel, Theodore Fred. 1929. Systematic Sociology in Germany: A
Critical Analysis of Some Attempts to Estab lish Sociology as an
Independent Science. New York: Columbia University Press.
Durkheim, Emile. 1893. De la Division du Travail Social. Paris:
F. Alean. Durkheim, Emile. 1924. Sociologie et Philosophie, with a
preface by C. Bougie. Paris: F. Alean. Jennings, Herbert S. 1930.
The Biological Basis of Human Nature. New York: W.W. Norton.
Parsons, Talcott. 1934. "Some Reflections on "The Nature and
Significance of Economics,'" Quarterly Journal
of Economics 48: 511-545. Parsons, Talcott. 1937. The Structure
of Social Action. New York: McGraw-Hill. Simmel, Georg. 1923.
Soziologie: Untersuchung ?ber die Formen der Vergesellschaftung, 3.
Auflage, M?nchen:
Duncker und Humboldt. Simmel, Georg. 1950. The Sociology of
Georg Simmel. Translated, edited, and with an introduction by Kurt
H.
Wolff. Glencoe, IL: The Free Press. Sorokin, Pitirim. 1928.
Contemporary Sociological Theories. New York: Harper and
Brothers.
Weber, Max. 1925. Grundriss der Sozial?konomik. III. Abteilung:
Wirtschaft und Gesellschaft, 2 vols., J.C.B. Mohr (P. Siebeck),
T?bingen.
-. 1930. The Protestant Ethic and the Spirit of Capitalism,
trans, by Talcott Parsons, George Allen & Un win, Ltd.,
London.
Wiese, Leopold von and Howard Becker. 1932. Systematic
Sociology: On the Basis of the Beziehungslebre and Gebildelehre of
Leopold von Wiese. Adapted and amplified by Howard Becker. New
York: John Wiley and Sons.
50 The American Sociologist/Summer 1998
This content downloaded from 119.15.93.148 on Wed, 2 Jul 2014
04:15:33 AMAll use subject to JSTOR Terms and Conditions
Article Contentsp. 31p. 32p. 33p. 34p. 35p. 36p. 37p. 38p. 39p.
40p. 41p. 42p. 43p. 44p. 45p. 46p. 47p. 48p. 49p. 50
Issue Table of ContentsThe American Sociologist, Vol. 29, No. 2
(Summer, 1998), pp. 1-107Front MatterEditor's Introduction: An
Invitation to Historical Sociology [pp. 3-3]Contested Canon: Simmel
Scholarship at Columbia and the New School [pp. 4-18]Editor's Note:
Parsons and Simmel [pp. 19-20]The "Fragment" on Simmel [From Draft
Chapter XVIII (Structure of Social Action): Georg Simmel and
Ferdinand Toennies: Social Relationships and the Elements of
Action] [pp. 21-30]Simmel and the Methodological Problems of Formal
Sociology [pp. 31-50]Comments on Parsons's "Simmel and the
Methodological Problems of Formal Sociology" [pp. 51-56]From the
"Missing Fragment" to the "Lost Manuscript": Reflections on
Parsons's Engagement with Simmel [pp. 57-76]Parsons's Second
Project: The Social System. Sources, Development and Limitations
[pp. 77-82]Social Relations Undone: Disciplinary Divergence and
Departmental Politics at Harvard, 1946-1970 [pp. 83-107]Back
Matter