Merton - On Sociological Theories of The Middle Range
Merton - On Sociological Theories of The Middle Range
Merton - On Sociological Theories of The Middle Range
Merton
Like so many words that are bandied about, the w o r d theory threatens to b e c o m e
meaningless. Because its referents are so diverse - including everything from minor
working hypotheses, through comprehensive but vague and unordered speculations,
to a x i o m a t i c systems o f thought - use o f the word often obscures rather than creates
understanding.
T h e term sociological theory refers to logically interconnected sets o f proposi-
tions from which empirical uniformities c a n be derived. T h r o u g h o u t we focus on
what I have called theories of the middle range: theories that lie between the minor
but necessary working hypotheses that evolve in a b u n d a n c e during day-to-day
research and the all-inclusive systematic efforts to develop a unified theory that will
explain all the observed uniformities o f social behavior, social organization, and
social change.
Middle-range theory is principally used in sociology to guide empirical inquiry. It
is intermediate to general theories o f social systems which are too remote from par-
ticular classes o f social behavior, organization, and change to a c c o u n t for w h a t is
observed and to those detailed orderly descriptions o f particulars that are not gener-
alized at all. Middle-range theory involves abstractions, o f course, but they are close
enough to observed data to be incorporated in propositions tha t permit empirical
testing. Middle-range theories deai with delimited aspects o f social p h e n o m e n a , as is
indicated by their labels. O n e speaks o f a theory o f reference groups, o f social mobil-
ity, or role-conflict and o f the formation o f social norms just as one speaks o f a theory
o f prices, a germ theory o f disease, or a kinetic theory o f gases.
T h e seminal ideas in such theories are characteristically simple: consider Gilbert
on magnetism, Boyle on atmospheric pressure, or D a r w i n on the formation o f coral
atolls. Gilbert begins with the relatively simple idea that the earth may be conceived
as a magnet; Boyle, with the simple idea that the atmosphere m a y be conceived as
a "sea o f a i r " ; D a r w i n , with the idea that one can conceive o f the atolls as upward
Robert K. Merton, "On Sociological Theories of the Middle Range," pp. 3 9 - 5 3 from Robert
K. Merton, Social Theory and Social Structure (New York: Simon 6c Schuster, The Free Press,
1949).
On Sociological Theories of the Middle Range 449
and outward growths o f coral over islands that had long since subsided into the
sea. Each of these theories provides an image that gives rise to inferences. T o take
but one case: if the atmosphere is thought of as a sea of air, then, as Pascal inferred,
there should be less air pressure on a mountain top than at its base. T h e initial idea
thus suggests specific hypotheses which are tested by seeing whether the inferences
from them are empirically confirmed. T h e idea itself is tested for its fruitfulness by
noting the range o f theoretical problems and hypotheses that allow one to identify
new characteristics o f atmospheric pressure.
In much the same fashion, the theory of reference groups and relative deprivation
starts with the simple idea, initiated by James, Baldwin, and M e a d and developed
by H y m a n and Stouffer, that people take the standards o f significant others as a basis
for self-appraisal and evaluation. Some o f the inferences drawn from this idea are at
odds with common-sense expectations based upon an unexamined set o f "self-
evident" assumptions. C o m m o n sense, for example, would suggest that the greater
the actual loss experienced by a family in a mass disaster, the more acutely it will
feel deprived. T h i s belief is based on the unexamined assumption that the magnitude
of objective loss is related linearly to the subjective appraisal o f the loss and that this
appraisal is confined to one's o w n experience. But the theory o f relative deprivation
leads to quite a different hypothesis - that self-appraisals depend upon people's c o m -
parisons o f their own situation with that of other people perceived as being c o m p a -
rable to themselves. T h i s theory therefore suggests that, under specifiable conditions,
families suffering serious losses will feel less deprived than those suffering smaller
losses if they are in situations leading them to compar e themselves t o people suffer-
ing even m o r e severe losses. F o r example, it is people in the area o f greatest impact
of a disaster w h o , though substantially deprived themselves, are m o s t apt to see
others around them w h o are even more severely deprived. Empirical inquiry sup-
ports the theory o f relative deprivation rather than the common-sense assumptions:
"the feeling o f being relatively better o/"/"than others increases with objective loss up
to the category o f highest loss" and only then declines. T h i s pattern is reinforced by
the tendency o f public communications to focus on "the most extreme sufferers
[which] tends to fix them as a reference group against which even other sufferers can
compare themselves favorably." As the inquiry develops, it is found that these pat-
terns o f self-appraisal in turn affect the distribution o f morale in the community o f
survivors and their motivation to help others. Within a particular class o f behavior,
1
therefore, the theory o f relative deprivation directs us to a set o f hypotheses that can
be empirically tested. T h e confirmed conclusion c a n then be put simply enough: when
few are hurt to much the same extent, the pain and loss o f each seems great; where
many are hurt in greatly varying degree, even fairly large losses seem small as they
are compared with far larger ones. T h e probability that comparisons will be made
is affected by the differing visibility o f losses o f greater and less extent.
middle-range theories have not been logically derived from a single all-embracing
theory o f social systems, though once developed they may be consistent with o n e .
Furthermore, each theory is more than a mere empirical generalization - an isolated
proposition summarizing observed uniformities o f relationships between t w o or
m o r e variables. A theory comprises a set o f assumptions from which empirical
generalizations have themselves been derived.
Another case o f middle-range theory in sociology may help us to identify its char-
acter and uses. T h e theory o f role-sets begins with an image o f h o w social status
2
leads directly to certain analytical problems. T h e notion of the role-set at once leads
to the inference that social structures confront men with the task o f articulating
the components o f countless role-sets - that is, the functional task o f managing
somehow to organize these so that an appreciable degree of social regularity obtains,
sufficient to enable most people most o f the time to go about their business without
becoming paralyzed by extreme conflicts in their role-sets.
If this relatively simple idea o f role-set has theoretical worth, it should generate
distinctive problems for sociological inquiry. T h e concept o f role-set does this. It
raises the general but definite problem o f identifying the social mechanisms - that
is, the social processes having designated consequences for designated parts o f the
social structure - which articulate the expectations o f those in the role-set suffi-
ciently to reduce conflicts for the occupant o f a status. It generates the further
problem o f discovering h o w these mechanisms c o m e into being, so that we can also
explain why the mechanisms do not operate effectively or fail to emerge at all in
some social systems. Finally, like the theory o f atmospheric pressure, the theory o f
role-set points directly to relevant empirical research. Monograph s on the workings
of diverse types o f formal organization have developed empirically-based theoreti-
cal extensions o f how role-sets operate in practice.
The theory o f role-sets illustrates another aspect o f sociological theories o f the
middle range. They are frequently consistent with a variety o f so-called systems o f
sociological theory. So far as one can tell, the theory o f role-sets is n o t inconsistent
with such broad theoretical orientations as M a r x i s t theory, functional analysis,
social behaviorism, Sorokin's integral sociology, or Parsons' theory o f action. T h i s
may be a horrendous observation for those o f us who have been trained to believe
that systems o f sociological thought are logically close-knit and mutually exclusive
sets o f doctrine. But in fact, as we shall note later in this introduction, comprehen-
sive sociological theories are sufficiently loose-knit, internally diversified, and mutu-
ally overlapping that a given theory of the middle range, which has a measure o f
empirical confirmation, can often be subsumed under comprehensive theories which
are themselves discrepant in certain respects.
This reasonably u n o r t h o d o x opinion can be illustrated by reexamining the theory
of role-sets as a middle-range theory. W e depart from the traditional concept by
assuming that a single status in society involves, not a single role, but an array o f
associated roles, relating the status-occupant to diverse others. Second, we note that
this concept o f the role-set gives rise to distinctive theoretical problems, hypothe-
ses, and so to empirical inquiry. O n e basic problem is that of identifying the social
mechanisms which articulate the role-set and reduce conflicts a m o n g roles. Third,
the concept o f the role-set directs our attention to the structural problem o f identi-
fying the social arrangements which integrate as well as oppose the expectations o f
various members o f the role-set. T h e concept o f multiple roles, on the other hand,
confines our attention to a different and no doubt important issue: how do indi-
vidual occupants o f statuses happen to deal with the many and sometimes conflict-
ing demands made o f them? Fourth, the concept o f the role-set directs us to the
further question o f h o w these social mechanisms come into being; the answer to
this question enables us to account for the many concrete instances in which the
452 Robert K. Merlon
role-set operates ineffectively. (This no m o r e assumes that all social mechanisms are
functional than the theory o f biological evolution involves the c o m p a r a b l e assump-
tion that no dysfunctional developments occur.) Finally, the logic o f analysis e x h i b -
ited in this sociological theory o f the middle range is developed wholly in terms o f
the elements o f social structure rather than in terms o f providing concrete histori-
cal descriptions o f particular social systems. T h u s , middle-range theory enables us
to transcend the m o c k problem o f a theoretical conflict between the n o m o t h e t ic and
the idiothetic, between the general and the altogether particular, between general-
izing sociological theory and historicism.
F r o m all this, it is evident that according to role-set theory there is always a
potential for differing expectations a m o n g those in the role-set as to w h a t is appro-
priate conduc t for a status-occupant. T h e basic source o f this potential for conflict
- and it is important to note once again that o n this point we are at on e with such
disparate general theorists as M a r x and Spencer, Simmel, Sorokin and Parsons - is
found in the structural fact that the other members o f a role-set are apt to hold
various social positions differing from those o f the status-occupant in question. T o
the extent that members o f a role-set are diversely located in the social structure,
they are apt to have interests and sentiments, values, and m o r a l expectations , dif-
fering from those o f the status-occupant himself. T h i s , after all, is on e o f the prin-
cipal assumptions o f M a r x i s t theory as it is o f much other sociological theory: social
differentiation generates distinct interests a m o n g those variously located in the struc-
ture o f the society. F o r e x a m p l e , the members o f a school board are often in social
and economic strata that differ significantly from the stratum o f the school teacher.
T h e interests, values, and expectations o f b o a rd members are consequently apt to
differ from those o f the teacher who may-thus be subject to conflicting expectations
from these and other members o f his role-set: professional colleagues, influential
members o f the school board and, say, the Americanism C o m m i t t e e o f the Ameri-
can Legion. A n educational essential for one Is apt to be judged as an educational
frill by another, or as downright subversion, by the third. W h a t holds conspicuously
for this one status holds, in identifiable degree, for occupants o f other statuses w h o
are structurally related through their role-set to others w h o themselves occupy dif-
fering positions in society.
As a theory o f the middle range, then, the theory o f role-sets begins with a concep t
and its associated imagery and generates an array o f theoretical problems. T h u s ,
the assumed structural basis for potential disturbance o f a role-set gives rise to a
double question (which, the record shows, has n o t been raised in the absence o f the
theory): which social mechanisms, if any, operate to counterac t the theoretically
assumed instability o f role-sets and, correlatively, under which circumstances do
these social mechanisms fail to operate, with resulting inefficiency, confusion, and
conflict? Like other questions that have historically stemmed from the general ori-
entation o f functional analysis, these do not assume that role-sets invariably operate
with substantial efficiency. F o r this middle-range theory is not c o n c e r n e d with the
historical generalization that a degree o f social order or conflict prevails in society
but with the analytical problem o f identifying the social mechanisms which produce
a greater degree o f order or less conflict than would obtai n if these m e c h a n i s m s
were not called into play.
On Sociological Theories of the Middle Range 453
The quest for theories o f the middle range exacts a distinctly different commitment
from the sociologist than does the quest for an all-embracing, unified theory. T h e
pages that follow assume that this search for a total system o f sociological theory,
in which observations a b o ut every aspect o f social behavior, organization, and
change promptly find their preordained place, has the same exhilarating challenge
and the same small promise as those many all-encompassing philosophical systems
which have fallen into deserved disuse. T h e issue must be fairly joined. Some soci-
ologists still write as though they expect, here and now, formulation o f the general
sociological theory b r o ad enough to encompass the vast ranges o f precisely observed
details o f social behavior, organization, and change and fruitful enough to direct
the attention o f research workers to a flow o f problems for empirical research. T h i s
I take to be a premature and apocalyptic belief. We are not ready. N o t enough
preparatory w o r k has been done.
An historical sense o f the changing intellectual contexts o f sociology should be
sufficiently humbling to liberate these optimists from this extravagant hope. F o r one
thing, certain aspects o f our historical past are still too much with us. W e must
remember that early sociology grew up in an intellectual atmosphere in which vastly
comprehensive systems o f philosophy were being introduced on all sides. Any
philosopher o f the eighteenth and early nineteenth centuries worth his salt had to
develop his o w n philosophical system - o f these, K a n t , Fichte, Schelling, Hegel were
only the best known. Each system was a personal bid for the definitive overview o f
the universe o f matter, nature and m a n .
These attempts o f philosophers to create total systems became a model for the
early sociologists, and so the nineteenth century was a century o f sociological
systems. Some o f the founding fathers, like C o m t e and Spencer, were imbued with
the esprit de systeme, which was expressed in their sociologies as in the rest o f their
wider-ranging philosophies. O t h e r s, such as Gumplowicz, W a r d , and Giddings, later
tried to provide intellectual legitimacy for this still "new science o f a very ancient
subject." T h i s required that a general and definitive framework o f sociological
thought be built rather than developing special theories designed to guide the
investigation o f specific sociological problems within an evolving and provisional
framework.
Within this c o n t e x t , almost all the pioneers in sociology tried to fashion his o w n
system. T h e multiplicity o f systems, each claiming to be the genuine sociology, led
naturally enough to the formation o f schools, each with its cluster o f masters, dis-
ciples and epigoni. Sociology not only became differentiated with other disciples,
but it became internally differentiated. T h i s differentiation, however, was not in
terms o f specialization, as in the sciences, but rather, as in philosophy, in terms o f
total systems, typically held to be mutually exclusive and largely at odds. As
Bertrand Russell noted about philosophy, this total sociology did not seize "the
advantage, as compared with the [sociologies] o f the system-builders, o f being able
to tackle its problems one at a time, instead o f having to invent at one stroke a
block theory o f the whole [sociological] universe." 3
454 Robert K. Merton
Another route has been followed by sociologists in their quest to establish the
intellectual legitimacy o f their discipline: they have taken as their prototype systems
of scientific theory rather than systems o f philosophy. T h i s path t o o has sometimes
led to the attempt to create total systems o f sociology - a goal that is often based
on one or more o f three basic misconceptions about the sciences.
T h e first misinterpretation assumes that systems o f thought c a n be effectively
developed before a great mass o f basic observations has been accumulated. A c c o r d-
ing to this view, Einstein might follow hard on the heels o f Kepler, without the
intervening centuries o f investigation and systematic thought a b o ut the results o f
investigation that were needed to prepare the terrain. T h e systems o f sociology that
stem from this tacit assumption are much like those introduced by the system-
makers in medicine over a span o f 1 5 0 years: the systems o f Stahl, Boissier de
Sauvages, Broussais, J o h n Brown and Benjamin Rush. Until well into the nineteenth
century eminent personages in medicine thought it necessary t o develop a theoret-
ical system o f disease long before the antecedent empirical inquiry had been ade-
quately developed. T h e s e garden-paths have since been closed off in medicine but
this sort o f effort still turns up in sociology. It is this tendency that led the bio-
chemist and avocational sociologist, L . J . Henderson, to observe:
A difference between most system-building in the social sciences and systems of thought
and classification in the natural sciences is to be seen in their evolution. In the natural
sciences both theories and descriptive systems grow by adaptation to the increasing
knowledge and experience of the scientists. In the social sciences, systems often issue
fully formed from the mind of one man. Then they may be much discussed if they
attract attention, but progressive adaptive modification as a result of the concerted
efforts of great numbers of men is rare. 4
The greater part of physical research is devoted to the development of the various
branches in physics, in each of which the object is the theoretical understanding of
more or less restricted fields of experience, and in each of which the laws and concepts
remain as closely as possible related to experience.7
The conviction among some sociologists that we must, here and now, achieve a
grand theoretical system not only results from a misplaced comparison with the
physical sciences, it is also a response to the ambiguous position o f sociology in
contemporary society. T h e very uncertainty about whether the accumulated knowl-
edge o f sociology is adequate to meet the large demands n o w being made o f it - by
policy-makers, reformers and reactionaries, by business-men and government-men,
by college presidents and college sophomores - provokes an overly-zealous and
defensive conviction o n the part o f some sociologists that they must s o m e h o w be
equal to these demands, however premature and extravagant they m a y be.
This conviction erroneously assumes that a science must be adequate to meet all
demands, intelligent or stupid, m a d e o f it. T h i s conviction is implicitly based on
the sacrilegious and masochistic assumption that one must be omniscient and
456 Robert K. Merton
M y emphasis upon the gap between the practical p r o b l e ms assigned to the soci-
ologist and the state o f his accumulated knowledge and skills does not m e a n , o f
course, that the sociologist should not seek to develop increasingly c o m p r e h e n s i v e
theory or should not w o rk on research directly relevant to urgent practica l p r o b -
lems. M o s t o f all, it does not mean that sociologists should deliberately s e e k out
the pragmatically trivial problem. Different sectors in the spectrum o f basic r e s e a r c h
and theory have different probabilities o f being germane to particular p r a c t i c a l p r o b -
lems; they have differing potentials o f relevance. But it is i m p o r t a n t to re-establish
8
science are close to the solutions o f some problems and remote fro m others. It must
be remembered that necessity is only the mother o f invention; socially a c c u m u l a t e d
knowledge is its father. Unless the two are brought together, necessity remains infer-
tile. She may o f course conceive at some future time when she is properly m a t e d .
On Sociological Theories of the Middle Range 457
But the mate requires time (and sustenance) if he is to attain the size and vigor
needed to meet the demands that will be made upon him.
This book's orientation toward the relationship o f current sociology and practi-
cal problems o f society is much the same as its orientation toward the relationship
of sociology and general sociological theory. It is a developmental orientation, rather
than one that relies on the sudden mutations of one sociologist that suddenly bring
solutions to m a j o r social problems or to a single encompassing theory. T h o u g h this
orientation makes no marvellously dramatic claims, it offers a reasonably realistic
assessment o f the current condition o f sociology and the ways in which it actually
develops.
F r o m all this it would seem reasonable to suppose that sociology will advance
insofar as its major (but not exclusive) concern is with developing theories o f the
middle range, and it will be retarded if its primary attention is focused on devel-
oping total sociological systems. So it is that in his inaugural address at the L o n d o n
School o f E c o n o m i c s , T. H . M a r s h a l l put in a plea for sociological "stepping-stones
in the middle d i s t a n c e . " O u r m a j o r task today is to develop special theories applic-
10
able to limited conceptual ranges - theories, for example, o f deviant behavior, the
unanticipated consequences o f purposive action, social perception, reference groups,
social control, the interdependence o f social institutions - rather than to seek im-
mediately the total conceptual structure that is adequate to derive these and other
theories o f the middle range.
Sociological theory, if it is to advance significantly, must proceed on these inter-
connected planes: (1} by developing special theories from which to derive hypo-
theses that can be empirically investigated and (2) by evolving, not suddenly
revealing, a progressively more general conceptual scheme that is adequate to con-
solidate groups o f special theories.
To concentrate entirely on special theories is to risk emerging with specific
hypotheses that account for limited aspects of social behavior, organization, and
change but that remain mutually inconsistent.
To concentrate entirely on a master conceptual scheme for deriving all subsidiary
theories is to risk producing twentieth-century sociological equivalents o f the large
philosophical systems o f the past, with all their varied suggestiveness, their archi-
tectonic splendor, and their scientific sterility. T h e sociological theorist who is exclu-
sively committed to the exploration o f a total system with its utmost abstractions
runs the risk that, as with m o d e r n decor, the furniture o f his mind will be bare and
uncomfortable.
The road to effective general schemes in sociology will only become clogged if,
as in the early days o f sociology, each charismatic sociologist tries to develop his
own general system o f theory. T h e persistence o f this practice can only make for
the balkanization o f sociology, with each principality governed by its own theoreti-
cal system. T h o u g h this process has periodically marked the development o f other
458 Robert K. Morton
NOTES