almost impossible to discover a person who does not "belong"—taking into account the intermittent character of his participation—at the same time to many organizations. . . .
It is evident from the foregoing that if persons are to be included within the concept "organization," its general significance will be quite limited. The bases or terms upon which persons are included will be highly variable—so much so that even within very restricted fields, such as a particular industry, "organizations" will mean a wide variety of entities. Hence, here again as when we included a part of the physical environment within the definition, the inclusion of persons may be most useful in particular instances, but of limited value for general purposes.
It nevertheless remains to consider whether it would actually be useful to adopt a definition from which persons as well as physical and social environments are excluded as components. If this is done, an organization is defined as a system of consciously coordinated personal activities or forces. It is apparent that all the variations found in concrete cooperative systems that are due to physical and social environments, and those that are due to persons or to the bases upon which persons contribute to such systems, are by this definition relegated to the position of external facts and factors,3 and that the organization as then isolated is an aspect of cooperative systems which is common to all of them.
Organization will then mean a similar thing, whether applied to a military, a religious, an academic, a manufacturing, or a fraternal cooperation, though the physical environment, the social environment, the number and kinds of persons, and the bases of their relation to the organization will be widely different. These aspects of cooperation then become external to organization as defined, though components of the cooperative system as a whole. Moreover, the definition is similarly applicable to settings radically different from those now obtaining, for example, to cooperation under feudal conditions. Such a definition will be of restricted usefulness with reference to any particular cooperative situation, being only one element of such a situation, except as by its adoption we are enabled to arrive at general principles which may be usefully applied in the understanding of specific situations.
It is the central hypothesis of this book that the most useful concept for the analysis of experience of cooperative systems is embodied in the definition of a formal organization as a system of consciously coordinated
3 That is, external to the organization but not external to the related cooperative system. It is to be borne in mind that we are dealing with two systems: (1) an inclusive cooperative system, the components of which are persons, physical systems,
social systems, and organizations; and (2) organizations, which are parts of cooperative systems and consist entirely of coordinated human activities.
activities or forces of two or more persons. In any concrete situation in which there is cooperation, several different systems will be components. Some of these will be physical, some biological, some psychological, etc., but the element common to all which binds all these other systems into the total concrete cooperative situation is that of organization as defined. If this hypothesis proves satisfactory it will be because (1) an organization, as defined, is a concept valid through a wide range of concrete situations with relatively few variables, which can be effectively investigated; and (2) the relations between this conceptual scheme and other systems can be effectively and usefully formulated. The final test of this conceptual scheme is whether its use will make possible a more effective conscious promotion and manipulation of cooperation among men; that is, whether in practice it can increase the predictive capacity of competent men in this field. It is the assumption upon which this essay is developed that such a concept is implicit in the behavior of leaders and administrators, explaining uniformities observed in their conduct in widely different cooperative enterprises, and that its explicit formulation and development will permit a useful translation of experience in different fields into common terms.
v.
— ii.
■< o
FOUNDATIONS OF THE THEORY OF ORGANIZATION
Ptiilip Selzniclc
Trades unions, governments, business corporations, political parties, and the like are formal structures in the sense that they represent rationally ordered instruments for the achievement of stated goals. "Organization," we are told, "is the arrangement of personnel for facilitating the accomplishment of some agreed purpose through the allocation of functions and
Reprinted in part from American Sociological Review, 13 (1948), 25-35, by permission of the author and the publisher, The American Sociological Association.
-ff
responsibilities."1 Or, defined more generally, formal organization is "a system of consciously coordinated activities or forces of two or more persons."2 Viewed in this light, formal organization is the structural expression of rational action. The mobilization of technical and managerial skills requires a pattern of coordination, a systematic ordering of positions and duties which defines a chain of command and makes possible the administrative integration of specialized functions. In this context delegation is the primordial organizational act, a precarious venture which requires the continuous elaboration of formal mechanisms of coordination and control. The security of all participants, and of the system as a whole, generates a persistent pressure for the institutionalization of relationships, which are thus removed from the uncertainties of individual fealty or sentiment. Moreover, it is necessary for the relations within the structure to be determined in such a way that individuals will be interchangeable and the organization will thus be free of dependence upon personal qualities.3 In this way, the formal structure becomes subject to calculable manipulation, an instrument of rational action.
But as we inspect these formal structures we begin to see that they never succeed in conquering the non-rational dimensions of organizational behavior. The latter remain at once indispensable to the continued existence of the system of coordination and at the same time the source of friction, dilemma, doubt, and ruin. This fundamental paradox arises from the fact that rational action systems are inescapably imbedded in an institutional matrix, in two significant senses: (1) the action system—or the formal structure of delegation and control which is its organizational expression—is itself only an aspect of a concrete social structure made up of individuals who may interact as wholes, not simply in terms of their formal roles within the system; (2) the formal system, and the social structure within which it finds concrete existence, are alike subject to the pressure of an institutional environment to which some over-all adjustment must be made. The formal administrative design can never adequately or fully reflect the concrete organization to which it refers, for the obvious reason that no abstract plan or pattern can—or may, if it is to be useful—exhaustively describe an empirical totality. At the same time, that which is not included in the abstract design  (as reflected, for ex-
1 John M. Gaus, "A Theory of Organization in Public Administration," in The Frontiers of Public Administration (Chicago: University of Chicago Press, 1936), p. 66.
2 Chester I. Barnard, The Functions of the Executive (Cambridge, Mass.: Harvard University Press, Í938), p. 73.                   ,
3 Cf. Talcott Parsons' generalization (after Max Weber) of the "law of the increasing rationality of action systems," in The Structure of Social Action (New York: McGraw-Hill, 1937),. p. 752.
a£
ample, in a staff~and-line organization chart) is vitally relevant to the maintenance and development of the formal system itself.
Organization may be viewed from two standpoints which are analytically distinct but which are empirically united in a context of reciprocal consequences. On the one hand, any concrete organizational system is an economy; at the same time, it is an adaptive social structure. Considered as an economy, organization is a system of relationships which define the availability of scarce resources and which may be manipulated in terms of efficiency and effectiveness. It is the economic aspect of organization which commands the attention of management technicians and, for the most part, students of public as well as private administration.4 Such problems as the span of executive control, the role of staff or auxiliary agencies, the relation of headquarters to field offices, and the relative merits of single or multiple executive boards are typical concerns of the science of administration. The coordinative scalar, and functional principles, as elements of the theory of organization, are products of the attempt to explicate the most general features of organization as a "technical problem" or, in our terms, as an economy.
Organization as an economy is, however, necessarily conditioned by the organic states of the concrete structure, outside of the systematics of delegation and control. This becomes especially evident as the attention of leadership is directed toward such problems as the legitimacy of authority and the dynamics of persuasion. It is recognized implicitly in action and explicitly in the work*sof a number of students that the possibility of manipulating the system of coordination depends on the extent to which that system is operating within an environment of effective inducement to individual participants and of conditions in which the stability of authority is assured. This is in a sense the fundamental thesis of Barnard's remarkable study, The Functions of the Executive. It is also the underlying hypothesis which makes it possible for Urwick to suggest that "proper" or formal channels in fact function to "confirm and record" decisions arrived at by more personal means.n We meet it again in the concept of administration as a process of education, in which the winning of consent and support is conceived to be a basic function of leadership.0 In short,
4 See Luther Guliek and Lydall Urwick (editors), Papers on the Science of Administration (New York: Institute of Public Administration, Columbia University, 1937); Lydall Urwick, The Elements of Administration (New York: Harper & Row, 1943); James D. Mooney and Alan C. Reiley, The Principles of Organization (New York: Harper & Row, 1939); H. S. Dennison, Organization Engineering (New York: McGraw-Hill, 1931).
6 Urwick, The Elements of Administration, op. cit., p. 47.
6 See Gaus, op. cit. Studies of the problem of morale are instances of the same orientation, having received considerable impetus in recent years from the work of the Harvard Business School group.
it is recognized that control and consent cannot be divorced even within formally authoritarian structures.
The indivisibility of control and consent makes it necessary to view formal organizations as cooperative systems, widening the frame of reference   of those  concerned  with  the  manipulation  of  organizational  resources.   At   the  point   of   action,   of  executive   decision,   the   economic aspect of organization provides  inadequate tools  for control over the concrete structure. This idea may be readily grasped if attention is directed to the role of the individual within the organizational economy. From the standpoint of organization as a formal system, persons are viewed functionally, in respect to their roles, as participants in assigned segments of the cooperative system. But in fact individuals have a propensity to resist depersonalization, to spill over the boundaries of their segmentary roles, to participate as wholes. The formal systems (at an extreme, the disposition of "rifles" at a military perimeter) cannot take account of the deviations thus introduced, and consequently break down as instruments of control when relied upon alone. The whole individual raises new problems for the organization, partly because of the needs of his own personality, partly because he brings with him a set of established habits as well, perhaps, as commitments to special groups outside of the organization.
Unfortunately for the adequacy of formal systems of coordination, the needs of individuals do not permit a single-minded attention to the stated goals of the system within which they have been assigned. The hazard inherent in the act of delegation derives essentially from this fact. Delegation is an organizational act, having to do with formal assignments of functions and powers. Theoretically, these assignments are made to roles or official positions, not to individuals as such. In fact, however, delegation necessarily involves concrete individuals who have interests and  goals   which  do   not  always   coincide   with  the  goals   of  the   formal
system. As a consequence, individual personalities may offer resistance to the demands made upon them by the official conditions of delegation. These resistances are not accounted for within the categories of coordination and delegation, so that when they occur they must be considered as unpredictable and accidental. Observations of this type of situation within formal structures are sufficiently commonplace. A familiar example is that of delegation to a subordinate who is also required to train his own replacement. The subordinate may resist this demand in order to maintain unique access to the "mysteries" of the job, and thus insure his indis-pensability to the organization.
In large organizations, deviations from the formal system tend to become institutionalized, so that "unwritten laws" and informal associations are established. Institutionalization removes such deviations from
the realm of personality differences, transforming them into a persistent structural aspect of formal organizations.7 These institutionalized rules and modes of informal cooperation are normally attempts by participants in the formal organization to control the group relations which form the environment of organizational decisions. The informal patterns (such as cliques) arise spontaneously, are based on personal relationships, and are usually directed to the control of some specific situation. They may be generated anywhere within a hierarchy, often with deleterious consequences for the formal goals of the organization, but they may also function io widen the available resources of executive control and thus contribute to rather than hinder the achievement of the stated objectives of the organization. The deviations tend to force a shift away from the purely formal system as the effective determinant of behavior to (1) a condition in which informal patterns buttress the formal, as through the manipulation of sentiment within the organization in favor of established authority; or (2) a condition wherein the informal controls effect a consistent modification of formal goals, as in the case of some bureaucratic patterns.8 This trend will eventually result in the formalization of erstwhile informal activities, with the cycle of deviation and transformation beginning again on a new level.
The relevance of informal structures to organizational analysis underlines the significance of conceiving of formal organizations as cooperative systems. When the totality of interacting groups and individuals becomes the object of inquiry, the latter is not restricted by formal, legal, or procedural dimensions. The state of the system emerges a$ g significant point of analysis, as when an internal situation charged with"1 conflict qualifies and informs actions ostensibly determined by formal relations and objectives. A proper understanding of the organizational process must make it possible to interpret changes in the formal system—new appointments or rules or reorganizations—in their relation to the informal and unavowed ties of friendship, class loyalty, power cliques, or external commitment. This is what it means "to know the score."
The fact that the involvement of individuals as whole personalities tends to limit the adequacy of formal systems of coordination does not mean that organizational characteristics  are those of individuals.  The
7 The creation of informal structures within various types of organizations has received explicit recognition in recent years. See F. J. Roethlisberger and W. J. Dickson, Management and the Worker (Cambridge, Mass.: Harvard University Press, 1941),
p. 524; also Barnard, op. cit., c. ix; and Wilbert E. Moore, Industrial Relations and the Social Order (New York: Macmillan, 1946), chap. xv.
8 For an analysis of the letter in these terms, see Philip Selznick, "An Approach to a Theory of Bureaucracy," American Sociological Review, Vol. VIII, No. 1 (February, 1943).
organic, emergent character of the formal organization considered as a cooperative system must be recognized. This means that the organization reaches decisions, takes action, and makes adjustments. Such a view raises the question of the relation between organizations and persons. The signlfleanee of theoretical emphasis upon the cooperative system as such is derived from the insight that certain actions and consequences are enjoined independently of the personality of the individuals involved. Thus, if reference is made to the "organization-paradox"—the tension created by the inhibitory consequences of certain types of informal structures within organizations—this does not mean that individuals themselves are in quandaries. It is the nature of the interacting consequences of divergent interests within the organization which creates the condition, a result which may obtain independently of the consciousness or the qualities of the individual participants. Similarly, it seems useful to insist that there are qualities and needs of leaders/iip, having to do with position and role, which are persistent despite variations in the character or personality of individual leaders themselves.
Rational action systems are characteristic of both individuals and organizations. The conscious attempt to mobilize available internal resources (e.g., self-discipline) for the achievement of a stated goal— referred to here as an economy or a formal system—is one aspect of individual psychology. But the personality considered as a dynamic system of interacting wishes, compulsions, and restraints defines a system which is at once essential and yet potentially deleterious to what may be thought of as the "economy of learning" or to individual rational action. At the same time, the individual personality is an adaptive structure, and this, too, requires a broader frame of reference for analysis than the categories of rationality. On a different level, although analogously, we have pointed to the need to consider organizations as cooperative systems and adaptive structures in order to explain the context of and deviations from the formal systems of delegation and coordination.
To recognize the sociological relevance of formal structures is not, however, to have constructed a theory of organization. It is important to set the framework of analysis, and much is accomplished along this line when, for example, the nature of authority in formal organizations is
reinterpreted to emphasize the factors of cohesion and persuasion as
against legal or coercive sources.9 This redefinition is logically the same as that which introduced the conception of the self as social. The latter helps make possible, but does not of itself fulfill, the requirements for a dynamic theory of personality.  In the same jway, the definition of authority as
9 Robert Michels, "Authority," Encyclopedia of the Social Sciences (New York: Mac-millan, 1931), pp. 319 ff.; also Barnard, op. cit., c. xü.
conditioned by sociological factors of sentiment and cohesion—or niuie generally the definition of formal organizations as cooperative systems— only sets the stage, as an initial requirement, for the formulation of a theory of organization.
STRUCTURAL-FUNCTIONAL ANALYSIS
Cooperative systems are constituted of individuals interacting as wholes in relation to a formal system of coordination. The concrete structure is therefore a resultant of the reciprocal influences of the formal and informal aspects of organization. Furthermore, this structure is itself a totality, an adaptive "organism" reacting to influences upon it from an external environment. These considerations help to define the objects of inquiry; but to progress to a system of predicates about these objects it is necessary to set forth an analytical method which seems to be fruitful and significant. The method must have a relevance to empirical materials, which is to say, it must be more specific in its reference than discussions of the logic or methodology of social science.
The organon which may be suggested as peculiarly helpful in the analysis of adaptive structures has been referred to as "structural-functional analysis."10 This method may be characterized in a sentence: Structural-functional analysis relates contemporary and variable behavior to a presumptively stable system of needs and mechanisms. This means that a given empirical" system is deemed to have basic needs, essentially related to self-maintenance; the system develops repetitive means of self-defense; and day-to-day activity is interpreted in terms' of the function served by that activity for the maintenance and defense of the system. Put thus generally, the approach is applicable on any level in which the determinate "states" of empirically isolable systems undergo self-impelled and repetitive transformations when impinged upon by external conditions. This self-impulsion suggests the relevance of the term "dynamic," which is often used in referring to physiological, psychological, or social systems to which this type of analysis has been applied.11
10 For a presentation of this approach having a more general reference than the study öf formal organizations, see Taleott Parsons, "The Present Position and Prospects of Systematic Theory in Sociology," in Georges Gurvitch and Wilbert E. Moore (eds.), Twentieth Century Sociology (New York: Philosophical Library, 1945).
11 "Structure" refers to both the relationships within the system (formal plus informal patterns in organization) and the set of needs and modes of satisfaction which characters the given type of empirical system. As the utilization of this type of analysis proceeds, the concept oi "need" will require further clarification. In particular, the imputation of a "stable set of needs" to organizational systems must not function as a new instinct theory. At the same time, we cannot avoid using these inductions as to
It is a postulate of the structural-functional approach that the basic need of all empirical systems is the maintenance of the integrity and eentinuity of the system itself. Of course, such a postulate is primarily
useful in directing attention to a set of "derived imperatives" or needs which are sufficiently concrete to characterize the system at hand.12 It is perhaps rash to attempt a catalogue of these imperatives for formal organizations, but some suggestive formulation is needed in the interests of setting forth the type of analysis under discussion. In formal organizations, the "maintenance of the system" as a generic need may be specified in terms of the following imperatives:
1. The Security of the Organization as a Whole in Relation to Social Forces in Its Environment
This imperative requires continuous attention to the possibilities of encroachment and to the forestalling of threatened aggressions or deleterious (though perhaps unintended) consequences from the actions of others.
2.  The Stability of the Lines
of Authority and Communication
One of the persistent reference-points of administrative decision is the weighing of consequences for the continued capacity of leadership to control and to have access to the personnel or ranks.
3.  The Stability of Informal Relations within the Organization
Ties of sentiment and self-interest are evolved as unacknowledged but effective mechanisms of adjustment of individuals and sub-groups to the conditions of life within the organization. These ties represent a cementing of relationships which sustains the formal authority in day-to-day operations  and  widens  opportunities   for  effective   communication.13  Conse-
generic needs, for they help us to stake out our area of inquiry. The author is
indebted to Robert K. Merton who has, in correspondence, raised some important
objections to the use of the term "need" in this context.
12 For "derived imperative" see Bronislaw Malinowski,   The Dynamics of Culture
Change (New Haven, Conn.: Yale University Press, 1945), pp. 44 if. For the use of
"need" in place of "motive" see the same author's A Scientific Theory of Culture
(Chapel Hill, N.C.: University of North Carolina Press, 1944), pp. 89-90.
33 They may also destroy those relationships, as noted above, the need remains,
generating one of the persistent dilemmas of leadership.
quently, attempts to "upset" the informal structure, either frontally or as an indirect consequence of formal reorganization, will normally be met with considerable resistance.
4.  The Continuity of Policy and of the Sources of Its Determination
For each level within the organization, and for the organization as a whole, it is necessary that there be a sense that action taken in the light of a given policy will not be placed in continuous jeopardy. Arbitrary or unpredictable changes in policy undermine the significance of (and therefore the attention to) day-to-day action by injecting a note of capri-ciousness. At the same time, the organization will seek stable roots (or firm statutory authority or popular mandate) so that a sense of the permanency and legitimacy of its acts will be achieved.
5.  A Homogeneity of Outlook with Respect to the Meaning and Role of the Organization
The minimization of disaffection requires a unity derived from a common understanding of what the character of the organization is meant to be. When this homogeneity breaks down, as in situations of internal conflict over basic issues, the continued existence of the organization is endangered, On the other hand, one* of the signs of "healthy" organization is the ability to effectively orient new members and readily slough off^those who cannot be adapted to the established outlook.                             ~>
This catalogue of needs cannot be thought of as final, but it approximates the stable system generally characteristic of formal organizations. These imperatives are derived, in the sense that they represent the conditions for survival or self-maintenance of cooperative systems of organized action. An inspection of these needs suggests that organizational survival is intimately connected with the struggle for relative prestige, both for the organization and for elements and individuals within it. It may therefore be useful to refer to a prestige-survival motif in organizational behavior as a short-hand way of relating behavior to needs, especially when the exact nature of the needs remains in doubt. However, it must be emphasized that prestige-survival in organizations does not derive simply from like motives in individuals. Loyalty and self-sacrifice may be individual expressions of organizational or group egotism and self-consciousness.
The concept of (organizational need directs analysis to the internal
J^^7^^^_            '2Lžu
28               MODERN THEORIES OF ORGANIZATION
relevance of organizational behavior. This is especially pertinent with respect to discretionary action undertaken by agents manifestly in pursuit of formal goals. The question then becomes one of relating the specific
act of discretion to some presumptively stable organizational need. In other words, it is not simply action plainly oriented internally (such as in-service training) but also action presumably oriented externally which must be inspected for its relevance to internal conditions. This is of prime importance for the understanding of bureaucratic behavior, for it is of the essence of the latter that action formally undertaken for substantive goals be weighed and transformed in terms of its consequences for the position of the officialdom.
Formal organizations as cooperative systems on the one hand, and individual personalities on tiie other, involve structural-functional homologies, a point which may help to clarify the nature of this type of analysis. If wc say that the individual has a stable set of needs, most generally tlie need for maintaining and defending the integrity of his personality or ego; that here arc recognizable certain repetitive mechanisms which are utilized by the ego in its defense (rationalization,,projection, regression, etc.); and that overt and variable behavior may be interpreted in terms of its relation to these needs and mechanisms—on the basis of this logic we may discern the typical pattern of structural-functional analysis as set forth above. In this sense, it is possible to speak of a "Freudian model" for organizational analysis. This does not mean that the substantive insights of individual psychology may be applied to organizations, as in vulgar extrapolations from the individual ego to whole nations or (by a no less vulgar inversion) from strikes to frustrated workers. It is the logic, the type of analysis which is pertinent.
This homology is also instructive in relation to the applicability of generalizations to concrete cases. The dynamic theory of personality states a set of possible predicates about the ego and its mechanisms of defense, which inform us concerning the propensities of individual personalities under certain general circumstances. But these predicates provide only tools for the analysis of particular individuals, and each concrete case must be examined to tell which operate and in what degree. They are not primarily organs of prediction. In the same way, the predicates within the theory of organization will provide tools for the analysis of particular cases. Each organization, like each personality, represents a resultant of complex forces, an empirical entity which no single relation or
no simple formula can explain. The problem of analysis becomes that of selecting among the possible predicates set forth in the theory of organization those which illuminate our understanding of the materials at
hand.
THE THEORY OF ORGANIZATION               29
The setting of structural-functional analysis as applied to organizations requires some qualification, hokvever. Let us entertain the suggestion that the interesting problem in social science is not so much why men act the way they do as why men in certain circumstances must act the way they do. This emphasis upon constraint, if accepted, releases us from an ubiquitous attention to behavior in general, and especially from any undue fixation upon statistics. On the other hand, it has what would seem to be the salutary consequence of focusing inquiry upon certain necessary relationships of the type "if . . . then," for example: If the cultural level of the rank and file members of a formally democratic organization is below that necessary for participation in the formulation of policy, then there will be pressure upon the leaders to use the tools of demagogy.
Is such a statement universal in its applicability? Surely not in the sense that one can predict without remainder the nature of all or even most political groups in a democracy. Concrete behavior is a resultant, a complex vector, shaped by the operation of a number of such general constraints. But there is a test of general applicability: it is that of noting whether the relation made explicit must be taken into account in action. This criterion represents an empirical test of the significance of social science generalizations. If a theory is significant it will state a relation which will either (1) be taken into account as an element of achieving control; or (2) be ignored only at the risk of losing control and will evidence itself in a ramification of objective or unintended consequences.14 It is a corollary of this principle of significance that investigation must search out the underlying factors in organizational action, which requires a kind of intensive analysis of the same order as psychoanalytic probing.
A frame of reference which invites attention to the constraints upon behavior will tend to highlight tensions and dilemmas, the characteristic paradoxes generated in the course of action. The dilemma may be said to be the handmaiden of structural-functional analysis, for it introduces the concept of commitment or involvement as fundamental to organizational analysis. A dilemma in human behavior is represented by an inescapable commitment which cannot be reconciled with the needs of the organism or the social system. There are many spurious dilemmas which have to do with verbal contradictions, but inherent dilemmas to which we refer are of a more profound sort, for they reflect the basic nature of the
14 See R. M. Maclver's discussion of the "dynamic assessment" which "brings the
externa! world selectively into the subjective realm, conferring on it subjective significance for the ends of action." Social Causation (Boston: Ginn, 1942), chaps. 11, 12. The analysis of this assessment within the context of organized action yields the implicit knowledge which guides the choice among alternatives. See also Robert K. Merton, "The Unanticipated Consequences of Purposive Social Action," American Sociological Review, I, 6 (December, 1936).
30               MODERN THEORIES OF ORGANIZATION
empirical system in question. An economic order Ipommitted to profit as its sustaining incentive may, in Marxist terms, sow the seed of its own destruction. Again, the anguish of man, torn between finitude and pride, is not a matter of arbitrary and replaceable assumptions but is a reflection of the psychological needs of the human organism, and is concretized in his commitment (o the institutions which command his life; he is in the world and of it, inescapably involved in its goals and demands; at the same time, the needs of the spirit are compelling, proposing modes of salvation which have continuously disquieting consequences for worldly involvements. In still another context, the need of the human organism for affection and response necessitates a commitment to elements of the culture which can provide them; but the rule of the super-ego is uncertain since it cannot be completely reconciled with the need for libidinal satisfactions.
Applying this principle to organizations, we may note that there is a general source_of tension observable in the split between "the motion and the act." Flans and programs reflect the freedom of technical or ideal choice, but organized action cannot escape involvement, a commitment to personnel or institutions or procedures which effectively qualifies the initial plan. Der Mensch denkt, Gott lenkt. In organized action, this ultimate wisdom finds a temporal meaning in the recalcitrance of the tools of action. We are inescapably committed to the mediation of human structures which are at once indispensable to our goals and at the same time stand between them and ourselves. The selection of agents generates immediately a bifurcation of interest, expressed in new centers of need and power, placing effective constraints upon the arena of action, and resulting in tensions which are never completely resolved. This is part of what it means to say that there is a "logic" of action which impels us forward from one undesired position to another. Commitment to dynamic, self-activating tools is of the nature of organized action; at the same time, the need for continuity of authority, policy, and character is pressing and
requires an unceasing effort to master the instruments generated in the course of action. This generic tension is specified within the terms of each cooperative system. But for all we find a persistent relationship between need and commitment in which the latter not only qualifies the former but unites with it to produce a continuous state of tension. In this way, the notion of constraint (as reflected in tension or paradox) at once widens and more closely specifies the frame of reference for organizational analysis.
For  Malinowski,   the  core  of  functionalism  was  contained  in  the view that a cultural fact must be analyzed in its setting. Moreover, he
Hppni-eniJy   araneeiveci   csf   hip   motliod   aB   paťtin&nt   tö   tba   nftalysla   af   ell
THE THEORY OF ORGANIZATION               31
aspects of cultural systems. But there is a more specific problem, one involving a principle of selection ■ which serves to guide inquiry along significant lines. Freud conceived of the human organism as an adaptive structure, but he was not concerned with all human needs, nor with all phases of adaptation. For his system, he selected those needs whose expression is blocked in some way, so that such terms as repression, inhibition, and frustration became crucial. Ail conduct may be thought of as derived from need, and all adjustment represents the reduction of need. But not all needs are relevant to the systematics of dynamic psychology; and it is not adjustment as such but reaction to frustration which generates the characteristic modes of defensive behavior.
Organizational analysis, too, must find its selective principle; otherwise the indiscriminate attempts to relate activity functionally to needs will produce little in the way of significant theory. Such a principle might read as follows: Our frame of reference is to select out those needs which cannot be fulfilled within approved avenues of expression and thus must have recourse to such adaptive mechanisms as ideology and to the manipulation of formal processes and structures in terms of informal goals. This formulation has many difficulties, and is not presented as conclusive, but it suggests the kind of principle which is likely to separate the quick and the dead, the meaningful and the trite, in the study of cooperative systems in organized action.15
The frame of reference outlined here for the theory of organization may now be identified as involving the following major ideas: (1) the concept of organizations as cooperative systems, adaptive social structures, made up of interacting individuals, sub-groups, and informal plus formal relationships; (2) structural-functional analysis, which relates variable aspects of organization (such as goals) to stable needs and self-defensive mechanisms; (3) the concept of recalcitrance as a quality of the tools of social action, involving a break in the continuum of adjustment and defining an environment Of constraint, commitment, and tension. This frame of reference is suggested as providing a specifiable area of relations within which predicates in the theory of organization will be sought, and at the same time setting forth principles of selection and relevance in our approach to the data of organization.
It will be noted that we have set forth this frame of reference within the over-all context of social action. The significance of events may be defined by their place and operational role in a means-end scheme. If functional analysis searches out the elements important for the maintenance
IB This is not meant to deprecate the study of organizations as economies or formal systems. The latter represent an independent level, abstracted from organizational
81 rue t tiros Hb äööfisrdtlva  or niU£t[vo «ystemo  ("oľnanístna"),
32              MODERN THEORIES OF ORGANIZATION
of a given structure,  and that structure is on4 of the materials to be
manipulated in action, then that which is functional in respect to the structure is also functional in respect to the action system. This provides a ground for the significance of functionally derived theories. At the same time, relevance to control in action is the empirical test of their applicability or truth.
SUQG-EiSTIOIXS FOR
A SOCIOLOGICAL APPROACH
TO THEORY OF
ORGAJNIZATIOIVS
Talcott Parsons
For the purposes of this article the term "organization" will be used to
refer to a broad type,of collectivity which has assumed a particularly important place in modern, industrial societies—the..type to which the term "bureaucracy" is most often applied. Familiar examples are the governmental bureau or department, the business firm (especially above a certain size), the university, and the hospital. It is by now almost a commonplace that there are features common tcxall these types of organization which cut across the ordinary distinctions Between the social science disciplines. Something is lost if study of the firm is left only to economists, of governmental organizations to political^ scientists, and of schools and universities to "educationists."1
\
This is the first article of a two-part series, reprinted^ in part from Admi/rietrative Science Quarterly, 1 (1956), 63-85, by permission of the author and the publisher, Cornell University.                                                            \
1 There is already a considerable literature on organization-, which cuts across disciplinary lines. It is not the intention of this paper to attempt to review it. Three writers have been particularly important in stimulating the author's thinking in the field: Max Weber, Chester I. Barnard, and Herbert Simon. See'particularly, Weber, Theory of Social and Economic Organization (New York,  1947), ch. in; Barnard,
APPROACH TO THE THEORY OF ORGANIZATIONS                33
The study of organization in the present sense is thus only part of
the study of social structure as that term is generally used by sociologists
(or of "social organization" as ordinarily used by social anthropologists). A family is only partly an organization; most other kinship groups are even less so. The same is certainly true of local communities, regional subsocieties, and of a society as a whole conceived, for example/as a nation. On other levels, informal work groups, cliques of friends, and so on, are not in this technical sense organizations.
THE CONCEPTOF ORGANIZATION
As a formal analytical point of reference, primacy of/orientation to the attainment of a specific goal is used as the defining/characteristic of an organization which distinguishes it from other typespf social systems. This criterion has implicationsXfor both the external relations and the internal structure of the system referred to here as an organization.
The attainment of a goál^ is defined as ai relation between a system (in this case a social system)\ and the relevant parts of the external situation in which it acts or operaj.es. Trjis relation can be conceived as the maximization, relative to the relevant conditions such as costs and obstacles, of some category of output %f the system to objects or systems in the external situation. These considerations yield a further important criterion of an organization. An organization is a system which, as the attainment o£ its goal, "produce^7' an identifiable something which can be utilized in some way by another system; thai is, the OUtpUt Of the organization is, for some other/system, an input. Iru the case of an organization with economic primacy, this output may b\ a class of goods or services which are either .consumable or serve as instruments for a further phase of the production process by other organizations. In the case of a government agency the output may be a class of regulatory decisions; in that of an educational organization it may be a certain type of "trained capacity" on the/^art of the students who have been subjected to its influence. In any of these cases there must be a set of consequences of the processes which go on within the organization, which mak&a difference to the /functioning of some other subsystem of the society;\that is, without the production of certain goods the consuming unit must behave differently, i.e., suffer a "deprivation."                                    •            ^
T/ííe availability, to the unit succeeding the organization in the series, of the organization's output must be subject to some sort of terms, the^
The Functions of the Executive (Cambridge, Mass., 1938); Simon, Administrative Behavior: A Study of Decision Making Processes in Administrative Organization (New York, 1951).