^Acaderrty of Management Review, 1984, Vot. 9, No. I, 104-113.

Structural Sources of Intraorganizational Power: A Theoretical Synthesis^ W. GRAHAM ASTLEY PARAMJIT S. SACHDEVA University of Pennsylvania Research on organizational power has become overly fragmented; a multitude of different sources of power have been identified. Consequently, an overall picture of intraorganizational power relations is difficult to obtain. In this paper, three structural sources of power—hierarchical authority, resource control, and network centrality—are integrated in a theoretical synthesis. The complex nature of intraorganizational power is revealed in the interactions, tensions, and conflicts among the three sources of power.

Despite this controversial and problematic status of the concept of power, its use is retained in this paper as a general orienting concept possessing at least one basic connotation: power is viewed as a capacity of social actors to overcome resistance on the part of other social actors in order to achieve desired objectives or results (Dahl, 1957). Beyond this general notion, however, a more precise definition of the concept is not offered here because, following Lukes (1974), power is regarded as being what Gallie (1955-1956) has termed an "essentially contested concept." That is, power is viewed as one of those concepts that "inevitably involves endless disputes about (its) proper use on the part of (its) users" (Gallie, 1955, p. 169). Part of the problem here is that power takes on a very precise meaning only when the analysis is applied to a particular situation, and such application depends on the purposes of the particular analyst concerned. As Pfeffer (1981) states, power is "context specific." Given this, what the present paper hopes to offer is a context-specific analysis of power that is particularly applicable to social relationships within formal organizations. With this same general aim in mind, some social scientists have begun to focus on a number of dimensions of power that are grounded in specific social and organizational contexts. Thus, power has been

Every social act is an exercise of power, every social relationship is a power equation, and every social group or system is an organization of power. Accordingly, it is possible to transpose any system of social relationships into terms of potential or active power. (Hawley, 1963, p. 422).

As the above indicates, it is possible to interpret every instance of interaction and every social relationship as involving an exercise of power, because actors clearly affect one another all the time they are interacting. But this very pervasiveness tends to make the concept of power elusive and redundant, for it begins to have no meaning apart from the ideas of social interaction and organization. The problems associated with defining power consequently are well known (Wrong, 1968) and have led March (1966) to conclude pessimistically that the concept is of little analytical utility, because the apparent ease with which it can be applied to multiple different types of empirical phenomena tends to make statements about power rather empty or tautological. *The preparation of this manuscript was supported by funds from the Center for the Study of Organizational Innovation, University of Pennsylvania. The authors would like to thank Charles Fombrun and Andrew Van de Ven for their helpful comments on an earlier draft of this paper, which was presented at the 41st Annual Meeting of the Academy of Management, Organization and Management Theory Division, San Diego, 1981.

104

tions and have overlooked the potential tension between different facets of power (Bacharach & Lawler, 1980, p. 31).

dissected into different "bases" (French & Raven, 1959; Peabody, 1962); "kinds" (Etzioni, 1961); "types (Olsen, 1978); "sources" (Bacharach & Lawler, 1980; Pfeffer, 1981); and so on. Although moving in a direction that lends the concept greater empirical precision, these authors have introduced another problem. The proliferation of multiple subtypes of power has tended to stimulate research efforts dealing primarily with just one aspect of the . overall concept in isolation from its other aspects, with the result that researchers, unfortunately, often have lost sight of "power" as a global phenomenon. For example, Weber's (1947) infiuential analysis of organizational power tended to focus on its formal aspects, tying it more or less exclusively to hierarchical authority relations. More contemporary accounts, on the other hand, tend to pay rather scant attention to the nature and source of efficacy of hierarchical authority (Mintzberg, 1983; Pfeffer, 1981), or they ignore its existence altogether (Hickson, Hinings, Lee, Schneck, & Pennings, 1971). Thus, power often is analyzed as if it is grounded solely in dependence relationships (Emerson, 1962). It is treated as an emergent phenomenon that "has been typically formulated as a variance from formal structure" (Clegg & Dunkerly, 1980, p. 458). Moreover, two major variants of nonformal, or emergent, power theories have come to the fore in contemporary theory. One explains power by reference to an actor's capacity to control the resources on which others depend (Pfeffer & Salancik, 1978). The other explains power by reference to an actor's locational centrality within a network of workfiow relations (Tichy & Fombrun, 1979). At least three major approaches to the analysis of organizational power remain. Although they deal with highly interrelated phenomena, they have developed mostly independently of one another.

To remedy this situation, this paper analyzes intraorganizational power as the joint product of three sources of power: the official authority vested in hierarchical positions (hierarchical authority); the capacity for obtaining resources from the environment and controlling the supply of these resources to others through processes of exchange (resource control); and the centrality of location within the organization's network of workfiow linkages (network centrality). Each of these is described as a "structural" source of power in the sense that it refiects the properties of the organization as a social system rather than being a product of the psychology or social psychology of particular individuals within the organization. Recent research is synthesized, and tentative hypotheses regarding the intersection of the three power sources are examined. The objective is to construct an overall view of intraorganizational power that reconstitutes power as a global phenomenon and still takes into account its analytical dissection into discrete and qualitatively different subtypes.

Three Sources of Power Hierarchical Authority

So much emphasis has been laid on Emerson's (1962) explanation of power as a dependence-generated phenomenon that power and dependence are treated almost as synonyms, and it has become difficult to talk of power without implying that it is grounded in dependence. Focusing on the dependencies that are generated by task allocations in the division of labor, many contemporary theorists have thereby emphasized "functional" sources of power at the expense of "formal" sources of power (Peabody, 1962). However, the popularity of power-dependence theory should not allow one to overlook the long-standing central importance of hierarchical authority, a concept that is not based primarily on the analysis of dependence (Blau & Schoenherr, 1971; Fayol, 1949; Gouldner, 1954; Gulick & Urwick, 1937; Mooney & Reiley, 1939; Pugh, Hickson, Hinings, & Turner, 1968). The interpretation of power in terms of authority relations, on which all of these latter studies focus, is grounded in the conception of power as something that inheres in official positions. Power is viewed as the product of formal decree. Subordinates obey superiors not so much because they are dependent on

In consequence, research on intraorganizational power has become overly fragmented. Too often the differentiated aspects of power are analyzed as if they are detached from each other and unrelated. Except for a few rare exceptions—for example, Bacharach and Aiken, (1976); Drake and Mitchell (1977); and Grimes (1978)— little attempt has been made at specifying the conditions under which one source of power becomes more salient than another. And little attention has been focused on the identification of interactions between sources of power. Comparative organizational studies... have failed to recognize the multidirectional nature of power rela105

tion must cope; consequently, the capacity for dealing with such uncertainty can be regarded as a valuable resource (Pfeffer, 1981). It is for this reason that Pondy (1977) views the resource-dependence perspective and Hickson et al.'s (1971) "uncertainty coping" perspective as basically variants on each other. In this respect, for example. Woodward (1965), Lawrence and Lorsch (1967), and Hambrick (1981) detailed how different industry environments present organizations with different sources of critical uncertainty, so that particular subunits become dominant within particular industries, as the task of scanning the environment tends to shift from the domain of one functional area to that of another.

the latter, but because they believe that the latter have a right to exercise power by virtue of their position. Sheer incumbency in office provides sufficient grounds for the superior to expect obedience. The superior's right to issue commands is enacted by a legitimized procedure that is considered by both superior and subordinate as correct. Thus, authority requires a degree of obedience as an end in itself; obedience is due a superior more because of institutionalized privilege than because it is instrumental in fulfilling the dependent subordinate's needs. Formal prerogative lies at the heart of the hierarchical differentiation of power. Resource Control

Network Centrality

The notion that power is derived from the ability to control the supply of resources to others has been emphasized in the work of Yuchtman and Seashore (1967), White (1974), Mindlin and Aldrich (1975), and Pfeffer and Salancik (1978). Nord (1980) has pointed out the respects in which these "resourcedependence perspectives" rely heavily on underlying "exchange theory" concepts (Blau, 1964; Emerson, 1962; Homans, 1961; Thibaut & Kelley, 1959). Thus, organizational subunits supply resources to others in exchange for a return of resources upon which they are dependent; and asymmetry in the dependencies that underlie such exchanges explains asymmetry in power between the actors involved. The perspective outlined here links this approach to the notion that resources ultimately are derived from the environment. It contends that organizations are open social systems that require a supply of resources from the environment in order to sustain their operations. The resources that are obtained in transactions with the environment vary in terms of how difficult they are to secure and in terms of how critical they are to the organization's functioning. Those organizational actors that obtain the most critical and difficult to secure resources consequently acquire power because of the dependencies that are generated. For example, Salancik and Pfeffer (1974) revealed that the ability of a university department to secure outside grants and contracts was related to its power; and Aldrich and Herker (1977), Whetten (1978), and Spekman (1979) each have suggested that boundary spanning subunits derive power from their capacity to obtain information directly from the environment. Further, the environment can be viewed as a major source of uncertainty with which the organiza-

The work of complex organizations typically is segmented into discrete positions that are differentiated both vertically by hierarchical level and horizontally by division or department. These differentiated positions are reintegrated through interconnecting workflows that form a relatively stable network of patterned interactions. Tichy and Fombrun (1979) have argued that organizational power relationships can be analyzed profitably by reference to this network of interactions. To the extent that actors are located at tightly coupled interconnected nodes in the network, they gain power because their immersion in multiple interdependencies makes them functionally indispensable (Dubin, 1957; Hickson et al., 1971; Mechanic, 1962; Thompson, 1956; Woodward, 1965). By being centrally located, they act as conduits for integrating the discrete functional contributions of others who are not directly interconnected (Freeman, 1979; Freeman, Roeder, & MulhoUand, 1980). Network centrality can be regarded as an additional source of intraorganizational power over and above an actor's ability to generate dependencies through resource exchange. Such power is attached to an actor's position in the network rather than derived from a control of resources within any particular dyadic relationship. Of course, it is true that degree of access to multiple resource fiows will affect a subunit's ability to generate dependencies through dyadic resource exchange, but this overlapping does not erase the analytical distinction between the two sources of power. In this vein, Emerson (1976) refers to structural properties of location, such as network centrality, in terms of their capacity for 106

creating "structural dependence" as opposed to dependence within a two-party exchange relation. Thus power derives not only from resource dependencies within specific exchange relations, but also has a structural component, namely the position or location of the actor within the exchange network (Cook, 1977, p. 72).

Pairwise Interactions Hierarchical authority, resource control, and network centrality appear together in organizational reality in interdependent configurations. Pairwise interactions among the three power sources pinpoint, systematically, instances in which they are mutually reinforcing, in which one dominated over another, in which tension is produced between them, and so

on. Hierarchical Authority Versus Resource Control Although Weber (1947) recognized that hierarchical power was based on formal authority, he also understood that formal authority was substantially reinforced by the superior's ability to control resources upon which the subordinate depends. This is refiected in what Gouldner has described as Weber's "biforked conception of bureaucracy as based on expertise and on authoritarian discipHne" (1954, p. 219). That is, Weber asserted that bureaucracy authority was based not only on formal decree, but also on the superior's functional contributions of knowledge and expertise. One could say that Weber saw two sources of power stemming from the overall "rational-legal" legitimation of authority: authority based on incumbency in office (its "legal" component), and authority based on functional expertise (its "rational" component). Unfortunately, Weber's coupling of "rational" with "legal" implies that the institutionalization of formal authority in a hierarchical office always tends to correspond to the importance of its functional role. Much of the time, however, such correspondence will be far less than perfect. Offe (1978) characterizes this situation as "task-status discontinuous organization" in contrast to "task-status continuous organization," in which formal authority and functional expertise do correspond; and ^ contends that the former is becoming increasingly^ prevalent in contemporary society. A number of studies have, in fact, indicated that formal authority and functional expertise do not always coexist, and it is this that makes the differentiation of hierarchical authority as a discrete source 107

of power empirically interesting. For example, Lehman's (1975) findings suggest that not only is the "distribution of resource control" important, but also important is the "distribution of responsibility for decision making." Similarly, Astley (1978) and Hambrick (1981) report that hierarchicallevel accounts for large proportions of variance in decision making power after the effects of functional expertise and interdependence are held constant. Gouldner (1954) has even utilized the distinction between formal and expert power to posit the existence of two types of bureaucracy: the "representative bureaucracy," which parallels Offe's "task-status continuous organization"; and the "punishment-centered bureaucracy," which parallels Offe's "taskstatus discontinuous organization." In Gouldner's study, moreover, it is particularly interesting that the representative bureaucracy is characterized by an absence of confiict, whereas the punishment-centered bureaucracy is confiict-ridden because of the tension produced by a lack of correspondence among the different sources of power. This is the same source of tension that typically is highlighted by analyses of bureaucratic-professional confiict (Benson, 1973). Resource Control Versus Network Centrality Because the network analysis of organizations is still in its infancy (Fombrun, 1982; Tichy & Fombrun, 1979), there is little empirical research dealing with the interaction of resource control and network centrality as sources of intraorganizational power. The present authors are aware of only one study, that by Hinings, Hickson, Pennings, and Schneck (1974), that attempts to look at this interaction. Even this study did not self-consciously spell out the qualitative difference between resource control and network centrality; it preferred, instead, to refer to both of these simply as "strategic contingencies." However, the study did succeed in comparing the relative effects on subunit power of "coping with uncertainty" (which, as has been argued earlier, as a form of resource control) and "pervasiveness," defined as the extent to which a position is centrally interconnected to other positions. The authors found that possession of each source of power was necessary, but not alone sufficient, for a subunit to gain a control of "strategic contingencies" and thereby make itself the dominant subunit within the organization. Subunits that ranked high on both coping with uncertainty and pervasiveness were able to amass greater power than were subunits

some way by the presence or absence of the third source of power. Consider, first, the relationship between resource control and hierarchical authority. Parsons (1960) and Thompson (1967) have suggested the existence of a positive relationship between these two sources of power. Differentiating among three organizational functions—"technical," "managerial," and "institutional"—Thompson (1967) states that the technical function is mostly closed to the environment (closed off in the technical core); that the institutional function is largely responsible for articulating the organization's rights to command outside resources and is thus mostly open to the environment; and that the managerial function, because it procures outside resources for internal use in the technical core, is intermediate with respect to environmental openness. At the same time, these functions are located at "three distinct levels of responsibility and control" (Thompson, 1967, p. 10), which implies a positive relationship between hierarchical authority and resource control. Despite this, however, it seems clear that some control of environmental resources typically is also available at lower hierarchical levels, in which case the two power sources would be inversely related. The empirical appearance of such a negative association appears best exemplified by the typical customer relations department of an industrial organization, which normally would have little formal authority but would perform a great deal of boundary spanning activity and be responsible for the control and mediation of resources in the form of information about customers. It is suggested here that the direction of the relationship between resource control and hierarchical level in any particular organizational situation typically is also related to a subunit's score for the third source of power: network centrality. Thus, the closer a subunit is to the apex of the hierarchy, the more likely it will be centrally located in the organization's workfiow in the sense that it typically will perform a centralized coordinating role in channeling the fiow of environmentally derived resources to multiple other subunits. In this situation its contacts with the environment tend to be critical, because it is located in a key position for safeguarding the continued functioning of units throughout the organization. It therefore is necessary to match its functional importance with formal authority to ensure that its performance is successfully undertaken. On the other hand, for subunits such as customer relations departments, which are not so centrally connected to multi-

that ranked high on only one of these factors. Moreover, a regression analysis pointed to interactive effects between the two variables, indicating that it is important to analyze the two power sources within a single theoretical framework. Network Centrality Versus Hierarchical Authority

A major limitation of the Hinings et al. (1974) study was that its sample of subunits was drawn from the same level of hierarchy, so that pervasiveness, as an operationalization of network centrality, was correspondingly allowed only to vary in a horizontal dimension, and not vertically across levels. This limitation, however, was overcome by Tichy and Fombrun's (1979) analysis of interaction and infiuence data in three organizations. These authors make a distinction between "prescribed" and "emergent" networks. Prescribed networks are operationalized as aspects of "formal structure"— they were "derived from organization charts" and were based on the hierarchical differentiation of "formal status." In other words, Tichy and Fombrun's prescribed network is a surrogate for what is referred to here as the hierarchy of authority. Emergent networks, on the other hand, were operationalized by sociometric measures of interaction generated by "work matters," and thus correspond to the network of workfiow interaction. Prescribed and emergent networks were seen to be interdependent: Metaphorically, we assume that the prescribed organization provides the pegs upon which the emergent networks hang. Variations in these pegs, therefore, alter the form of the emergent network (Tichy & Fombrun, 1979, p. 929).

However, with regard to this hypothesized correspondence between prescribed structure and emergent network, the analysis found that the three organizations studied were quite different. In two of the organizations, which were described as "mechanistic," a reasonably high degree of "fit" existed between prescribed structure and emergent network, indicating that hierarchical authority was reinforced by network centrality. On the other hand, in the third organization studied, described as "organic," the degree of "fit" was zero, indicating that the two sources of power varied independently. Again, this indicates the possibility of tension between different facets of organizational power, and it highlights the need for systematic investigation of such divergences.

Three-Way Interactions It is contended that the relationship between each of the pairs discussed above typically is mediated in 108

Figure 1 Network Centrality and Resource Control across Technical Core and Boundary Spanning Units

pie units throughout the organization, responsibility for the procurement of environmental resources can be delegated to their lower hierarchical level because the immediate effects of any unsuccessful performance are restricted to a localized area. This interpretation is consonant with Hambrick's (1981) interpretation of the finding that "general managers" are more powerful than "external relations executives." Bbth categories of managers spanned organizational boundaries. General managers performed the organization's institutional function by "formulating and implementing overall policy; overall goal setting; [and] reporting to the board of trustees" (1981, p. 273). External relations executives were concerned with "public relations, fund raising, governmental relations, etc." (1981, p. 273). General managers naturally tended to be located at higher levels than did external relations executives, but even after controlling for hierarchical level, the former were much more powerful than the latter. Hambrick suggests, therefore, that "centrality" acted as an additional critical factor because general managers, "who typically span and integrate a variety of functions, have high centrahty and, hence, high power" (1981, p. 267); and external relations executives, who "are not directly involved in the mainstream tasks, have low centrality and low power" (1981, p. 267). Consider the relationship between resource control and network centrality. This relationship also is mediated by the third source of power—in this case, hierarchical authority. At lower levels, a clear differentiation typically emerges between the technical core and boundary spanning units, which buffer the technical core from direct contact with the environment. One consequence of this differentiation is that boundary spanners, with direct access to the environment, exercise dominance in the form of resource control over the technical core positions, which are linked only indirectly to the environment. On the other hand, because boundary spanning units are located in a peripheral position, they typically have less workfiow connections with other units than does the technical core, which is buffered by all of the boundary spanning units simultaneously. Thus, at lower hierarchical levels there is likely to be an inverse relationship between resource control and network centrality, as Figure 1 suggests. At upper levels of the hierarchy, however, there is no similar differentiation between boundary spanning and technical core subunits; thus, the same inverse relationship does not hold. Part ofthe reason for this is obvious.

Decreasing Resource Control

Boundary Spanning Units

Fewer people are located at higher than at lower levels, so there is less opportunity for differentiation to emerge. However, there also are additional factors to be considered, factors that are related to the kind of work that is performed at upper and lower levels. In this respect, a consideration of the differences between administrative and nonadministrative work is helpful. It is at the lower levels of the hierarchy that organizations perform their directly productive work in the "primary workfiow" (Rice, 1958). This is where organizations introduce horizontal differentiation in order to create homogeneous duties within specialized subunits (Blau & Schoenherr, 1971). In other words, at lower hierarchical levels, the task of resource acquisition becomes specialized in boundary subunits, whereas the intricate technological interdependencies that enhance efficiency become specialized in the technical core subunits, where they can be sheltered from environmental intrusion. Specialization thus gives subunits highly specific functions and restricts their acquisition of power to a single source. Consequently, if a subunit controls resources because of its location at the boundary, it does not also gain centrality from its immersion within the technical core, and vice versa. This proposition is consonant with Thompson's contention that organizational power structures are adjusted in response to two sources of change: "internally from technological developments, and externally based on changes in task-environment elements" (1967, p. 129). Technology, in other words. 109

2(a) and Figure 2(b), represented by the arrows, would completely alter the power situation. As the source of all resources in the network. Position A clearly would have greatest power in Figure 2(a), but would be a dependent recipient of resources in Figure 2(b) and weak because of this. This seems obvious, but the important point is that the situation of Figure 2(a) would occur when Position A is located close to the apex of the hierarchy, where contact with the environment would give it control over important organizational resources. The situation of Figure 2(b) would occur when Position A is located in the heart ofthe technical core, where it would be isolated from environmental contact and thus bereft of resource control.

affects patterns of interdependence in the technical core, and environment affects patterns of dependence at the organization's boundary. What is significant for Thompson is that these two sources of dependence are segmented "into self-sufficient clusters" so that "dependence tends to be confined to that domain" (1967, p. 129). Moreover, it is clear that Thompson here is referring to lower hierarchical levels, because it is the localization of dependence in selfsufficient clusters that permits decentralization. "Decentralization dilutes the power structure by creating more power positions but limiting the organization's dependence on each" (1967, p. 129). This further implies that if functions cannot be segmented into self-contained clusters, they will remain centralized at the top of the hierarchy. Thus, at higher levels of the hierarchy duties are general rather than specific, because administrative work cuts across specialized functions. Located at the center of allocative decision making, administrators must acquire and control important resources but simultaneously be centrally positioned to achieve effective coordination of the allocation process. Higher levels of the hierarchy therefore are characterized by a positive relationship between resource control and network centrality, rather than the negative relationship found at lower levels. Finally, consider the relationship between network centrality and hierarchical level. The mediation of this relationship by resource control is particularly interesting because it seems to qualify the widely accepted conclusions of Bavelas (1950) and Leavitt (1951) regarding centrality in interaction networks. These researchers found that when faced with the need to solve problems jointly, the structure of interaction among individuals within groups inevitably led to centrally located members acting as decision makers and peripherally located members acting merely as information-relayers. In other words, the responsibility for making decisions, which can be regarded as a progenitor to hierarchical authority, was a direct consequence of network centrality. The emergence of a positive relationship between hierarchical authority and network centrality is conditioned, however, by resource control. To illustrate why this may be the case, consider Figure 2. According to the Bavelas and Leavitt experiments. Position A scores higher in terms of network centrality than do positions B, C, D, and E and therefore should have more power in decision making. However, the change in direction of resource fiows between Figure

Conclusion The organizational literature dealing with power quite rightly has differentiated the concept into a number of analytically distinct dimensions. Power is a multifaceted phenomenon and must be analyzed as such. Unfortunately, however, a fragmentation of theory has occurred. Authors tend to act as proponents for a particular perspective on power and thereby forget the existence of other aspects of power. Consequently, much of the theoretical richness and complexity of the power phenomenon within organizations has been lost. Moreover, empirical studies of intraorganizational power are misleading if they focus on only one or another aspect of power in isolation from other aspects. This is so, not only because there are multiple sources of intraorganizational power, but also because these sources are highly interdependent. Future research efforts therefore should make systematic simultaneous reference to multiple power sources rather than focus narrowly on single dimensions of power. The analytical dissection of power into discrete sources must be accompanied by conceptual frameworks specifying the complex interdependence between these sources. Only then will the concept of power be reconstituted as a global phenomenon that is bestowed with sufficient meaning to justify the rejection of March's (1966) criticism that the concept is vacuous. Although the present paper advocates a more comprehensive treatment of intraorganizational power than is typical in the literature, it also has limitations to which future research should pay attention. The first limitation is its emphasis on structural sources of power. The structural approach taken here gives a rather static picture of power configurations and 110

Figure 2 The Importance of Resource Flows for Decision Making in Networks (a)

(b)

does not focus on the dynamic processes that produce and, over time, change these configurations. In this respect much can be gained by integrating the present structural approach with the organizational behavior literature dealing with individual and group processes, because these are relevant to power analysis. Second, research on the individual and group dynamics underlying the production of power structures also can shed light on the dialectical processes through which particular power regimes are instituted (Benson, 1977a, 1977b). The problem here is that much research on organizational power analyses established power regimes as being functionally necessary ones. For instance, the resource control perspective utilized in this paper explains the acquisition of power by reference to an individual or a unit's functional contributions to organizational goals. Units controlling important resources that are critical to organizational goals thereby acquire power. But the question that goes unanswered in such research is why certain goals come to be defined as the im-

portant ones in the first place. And are these goals defined in such a way that they systematically favor the interests of some parties and disadvantage others? An examination of the processes by which certain goals are set in place, and defined as critical and functional, permits the question: Critical and functional for whom? (Astley & Rosen, 1983). Finally, such a radical approach can be linked profitably to the extension of power analysis in another direction: to extraorganizational structures of power. Astley and Van de Ven (1983) have argued that power relations within organizations can be satisfactorily analyzed only by reference to the priorities and interests of powerful stakeholders outside the organization. Intraorganizational power relations can be seen, in part, as "epiphenomena" of forces that permeate in a broader social context. Thus it becomes necessary, following Hickson, Astley, Butler, and Wilson (1981), to link structures of intraorganizational power to the power structure of the wider environment itself.

References Astley, W. G., & Rosen, M. Power and culture in organizations: A politico-symbolic dialectic. Working Paper, Department of Management, University of Pennsylvania, 1983.

Aldrich, H., & Herker, D. Boundary spanning roles and organizational structure. Academy of Management Review. 1977, 2, 217-230.

Bacharach, S. B.,& Aiken, M. Structural and process constraints on influence in organizations: A level-specific analysis. Administrative Science Quarterly. 1976, 21, 623-642.

Astley, W. G. Sources of power in organizational life. Unpublished Ph.D dissertation. University of Washington, 1978. Astley, W. C , & Van de Ven, A. Central perspectives and debates in organization theory. Administrative Science Quarterly. 1983, 28, 245-273.

Bacharach, S. B., & Lawler, E. J. Power and politics in organizations, San Francisco: Jossey-Bass, 1980.

Ill

Bavelas, A. Communication patterns in task-oriented small groups. Journal of Acoustical Society of America. 1950, 22, 725-730.

Gouldner, A. Patterns in industrial bureaucracy. New York: Free Press, 1954.

Benson, J. K. The analysis of bureaucratic—professional conflict: Functional versus dialectical approaches. Sociological Quarterly, 1973, 14, 376-94.

Grimes, A. J. Authority, power, influence and social control: A theoretical synthesis. Academy of Management Review, 1978, 3. 724-735.

Benson, J. K. Innovation and crisis in organizational analysis. Sociological Quarterly. 1977a, 18, 3-16.

Gulick, L., & Urwick, L. (Eds.). Papers on the science of administration. New York: Institute of Public Administration, Columbia University, 1937.

Benson, J. K. Organizations: A dialectical view. Administrative Science Quarterly, 1977b, 22, 1-21.

Hambrick, D. C. Environment, strategy and power within top management teams. Administrative Science Quarterty. 1981, 26, 253-275.

Blau, P. M. Exchange and power in social life. New York: Wiley, 1964.

Hawley, A. H. Community power and urban renewal success. American Journal of Sociology. 1963, 68, 422-431.

Blau, P. M., & Schoenherr, P. A. The structure of organizations. New York: Basic Books, 1971.

Hickson, D. J., Astley, W. G., Butler, R. J., & Wilson, D. C. Organization as power. In L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavior (Vol 3). Greenwich, Conn.: JAI Press, 1981, 151-196.

Clegg, S., & Dunkerley, D. Organization, class and control. Boston: Routledge and Kegan Paul, 1980. Cook, K. S. Exchange and power in networks of interorganizational relations. Sociological Quarterly. 1977, 18, 62-81. Dahl, R. A., The concept of power. Behavioral Science. 1957, 2, 201-15.

Hickson, D. J., Hinings, C. R., Lee, C. A., Schneck, R. E., & Pennings, J. M. A strategic contingencies theory of intraorganizational power. Administrative Science Quarterly, 1971, 15, 216-229.

Drake, B., & Mitchell, T. The effects of vertical and horizontal power on individual motivation and satisfaction. Academy of Management Journal. 1977, 20, 573-591.

Hinings, C. R., Hickson, D. J., Pennings, J. M. & Schneck, R.E. Structural conditions of intraorganizational power. Administrative Science Quarterly, 1974, 19, 22-44.

Dubin, R. Power and union-management relations. Administrative Science Quarterly, 1957, 2, 60-81.

Homans, G. C. Social behavior: Its elementary forms. New York, Harcourt, Brace, and World, 1961.

Emerson, R. M. Power-dependence relations. American Sociological Review, 1962, 27, 31-40.

Lawrence, P. R., & Lorsch, J. W. Organization and environmen Homewood, 111.: Irwin, 1967. Leavitt, J. J. Some effects of communication pattems on group performance. Journal of Abnormal and Social Psychology, 1951, 46, 38-50.

Emerson, R. M. Social exchange theory. In A. Inkeles, J. Coleman, & N. Smelser (Eds.), Annual Review of Sociology (Vol, 2). Palo Alto, Cal.: Annual Reviews Inc., 1976. Etzioni, A. A comparative analysis of complex organizations. New York: Free Press, 1961.

Lehman, E. W. Coordinating health care: Explorations in interorganizational relations. Beverly Hills, Cal.: Sage Publications, 1975.

Fayol, H. General and industrial management (C. Storrs, Trans.). London: Pitman, 1949.

Lukes, S. Power: A radical view. New York: Macmillan, 1974.

March, J. G. The power of power. In D. Easton (Ed.), Varieties of political theory. Englewood Cliffs, N.J.: Prentice-Hall, 1966, 39-70.

Fombrun, C. J. Stategies for network research in organizations. Academy of Management Review, 1982, 7, 280-291.

Mechanic, D. Sources of power of lower participants in complex organizations. Administrative Science Quarterly. 1962, 7, 349-364.

Freeman, L. C. Centrality in social networks: A conceptual clarification. Social Networks, 1979, 1, 215-240. Freeman, L. C, Roeder, D., & MulhoUand, R. R. Centrality in social networks: II experimental results. Social Networks. 1980, 2, 119-142.

Mindlin, S. E., & Aldrich, H. Interorganizational dependence: A review of the concept and a reexamination of the findings of the Aston group. Administrative Science Quarterly 1975 20, 382-392.

French, J. R. P., & Raven, B. The bases of social power. In D. Cartwright (Ed.), Studies in social power. Ann Arbor, Mich.: University of Michigan Press, 1959, 150-167.

Mintzberg, H. Power in and around organizations. Englewood Cliffs, N.J.: Prentice-Hall, 1983.

Gallie, W. B. Essentially contested concepts. Proceedings ofthe Aristotelian Society. 1955-1956, 56, 167-198.

Mooney, J. D., & Reiley, A. C. The principles of organization. New York: Harper, 1939.

112

Spekman, R. E. Influence and information: An exploratory investigation of the boundary role person's basis of power. Academy of Management Journal. 1979, 22, 104-117.

^. Nord, W. R. Tbe study of organizations through a resourceexchange paradigm. In J. K. Gergen, M. S. Greenberg, & R.H. Willis (Eds.), Social exchange: Advances in theory "nd research. H New York: Plenum Press, 1980, 119-139.

Thibaut, J. W., & Kelley, H. H. The social psychology of groups. New York: Wiley, 1959.

' Offe, C., Industry and inequality: The achievement principle in viork and sociat status (J. Wiekham, Trans.). New York: St. ,, Martin's Press, 1978.

Thompson, J. D. Authority and power in "identical" organizations. American Journal of Sociology, 1956, 62, 290-301.

^ Olsen, M. E. The process of social organization: Power in social systems. New York: Holt, Reinhart & Winston, 1978.

Thompson, J. D. Organizations in action. New York: McGrawHill, 1967.

fc; Parsons, T. Structure and process in modern societies. New York: ti Free Press, 1960.

Tichy, N. M., & Fombrun, C. Network analysis in organizational settings. Human Relations. 1979, 32, 11, 923-965.

Peabody, R. L. Perceptions of organizational authority. Administrative Science Quarterly. 1962, 6, 463-482.

Weber, M. The theory of social and economic organization (T. Parsons, Ed.; A. H. Henderson & T. Parson, Trans.). Glencoe, 111.: Free Press, 1947.

ilj

Pfeffer, J., & Salancik, G. R. The external control of organizations. New York: Harper & Row, 1978.

Whetten, D. A. Coping with incompatible expectations: An integrated view of role conflict. Administrative Science Quarterly. 1978, 23, 254-21].

c.

* Pfeffer, J. Power in organization. Marshfield, Mass.: Pitman, h 1981. Pondy, L. E. The other hand clapping: An information processi[; ing approach to organizational power. In T. H. Hammer & S. B. Bacharach (Eds.), Reward systems and power distribution. ' Ithaca, N.Y.: Cornell University, School of Industrial and Labor ^ Relations, 1977.

White, P. E. Resources as determinants of organizational behavior. Administrative Science Quarterly, 1974, 19, 366-379. Woodward, J. Industrial organization: Theory and practice. London: Oxford University Press, 1965. Wrong, D. Some problems in defining social power. American Journal of Sociology, 1968, 73,673-681.

Pugh, D. S., Hickson, D. J., Hinings, R. C , & Turner C. Dimensions of organization structure. Administrative Science Quarterly. 1968, 13, 65-105.

Yuchtman, E., & Seashore, S. E. A system resource approach to organizational effectiveness. American Sociologial Review. 1967, 32, 391-903.

Rice, A. K. Productivity and social organization: The Ahmedebad experiment, London: Tavistock, 1958. Salancik, G. R., & Pfeffer, J. The bases and uses of power in organizational decision making: The case of a university. Administrative Science Quarterly. 1974, 19, 453-473.

W. Graham Astley is Assistant Professor of Management in the Wharton School, University of Pennsylvania. Paramjit S, Sachdeva is a doctoral student in Social Systems Sciences in the Wharton School, University of Pennsylvania.

113