I
Power and Influence: An Introduction
James T. Tedeschi and Thomas V. Bonoma
In his analysis of the structure and development of scientific disciplines, Kuhn (1962) has postulated that it is possible to subdivide the development and growth of any body of scientific knowledge into two characteristic phases: the preparadigm and paradigm stages. The general trend of investigation and research in the former phase is characterized by ânearly random activity,â emphasizing rival but diverse and unorganized subfields of inquiry, and a lack of theoretical or empirical direction. The paradigm stage is said to occur when the accumulation of knowledge is reorganized logically and theoretically to define a coherent, integrated new approach to the field as (1) âunprecedented enough to attract and maintain an enduring group of adherentsâ who are convinced of the significance of this ânew look,â but at the same time (2) âsufficiently open-ended to leave all sorts of problems for the redefined group of practitioners to solveâ (p. 10). It is the paradigm state, then, which provides for âparticular coherent traditions of scientific research.â
The âstate of the artâ of social psychology is quite evidently in the preparadigm phase of development. Social psychological theory as such does not exist; rather, there are a multitude of competing âlocalâ conceptualizations each of which provides some semblance of construct-data correspondence within its subfield of inquiry. The use of common sense language and trial-and-error forays into research areas suggested by disconnected, ambiguous hypotheses have produced an anarchy of concepts and empirical data. What seems to be necessary is the kind of revolution that Cubism brought to the arts. We need to fractionate and break apart present forms to discover their fundamental properties, and then restructure them so that we can see them more profoundly and in their proper relationships.
The present chapter attempts to break down and restructure much of social psychology by analyzing the fundamental concepts of power and influence. A minimum set of influence relationships are proposed which we think exhaust the kinds of dyadic interactions persons can experience, and which may serve as the basis of a novel and coherent perspective of social psychology. An attempt will be made to relate the major research interests of contemporary social psychology to one aspect or another of the social influence processes. The intended result is an overview of the writings in this book. No attempt will be made to fill in the theoretical details, or with few exceptions, to evaluate the theoretical assumptions by an examination of empirical data. In short, we hope to provide a unified perspective of the problems of social psychology and a foundation from which a paradigm may eventually be developed.
Macht
The concept of power, or Macht, has generated a great deal of controversy among social scientists of all disciplines. One reason is that the concept includes so much of the content of interest to each discipline. A number of writers have concluded that the study of power and influence may well be coextensive with their particular field of interest. Lasswell (1966), for example, asserts that the study of politics is equivalent to the study of influence and the influential. Karl Deutsch (1966) has defined political science as the study of how compliance is obtained. Bazelon (1965) has demonstrated that economics can be conceived of as a system of threats and promises. And Homans (1958), drawing from the principles of elementary economics, views sociology as based on processes of social exchange.
One possible reason why the concept of power encompasses so much aggregate data and so many dissimilar disciplines is that it has developed from a series of intuitive analyses spanning several centuries, and with each new analyst criticizing, revising, and extending the previous analystâs position so that one more crucial class of events could be added in order to gain greater scope and utility. Such defining, refining, and redefining has been particularly intense during the last two decades, during which both Grand Theorists (Dahl, 1957; Parsons, 1967) and mathematical modelers (Harsanyi, 1962a; Shapley & Shubik, 1954) have attempted classification and summation. Often these theoretical efforts have fallen prey to the nominalistic error, which assumes that once a set of events has been named, they have been explained. Also, there is the danger of assuming that the greater the number of events which can be subsumed under the rubric âpower,â the greater the theoretical achievement. Perhaps the generic relationship between the words âpowerâ and âenergyâ has given the social scientist the illusion that power has the same theoretical status in social science as does energy in physics.
The definitions of power to be examined include the notions of power as interpersonal causation and as coercive influence. Neither of these definitions has produced a scientific theory of power, but explication does help provide an analytical breakdown of the problems encountered in the construction of a scientific theory.
Power as Interpersonal Causation
Morgenthau (1960), in presenting a realist theory of international relations, relies upon a concept of power to analyze the problems of diplomacy, war, and other political processes. He defines power as âanything that establishes and maintains the control or man over man. Thus, power covers all social relationships which serve that end, from physical violence to the most subtle psychological ties by which one mind controls anotherâ (p. 9). Unfortunately, Morgenthauâs far-reaching definition raises difficulties in specifying a form of interpersonal interaction which is not power.
Robert Dahl (1957), another political scientist, has defined power in terms of interpersonal causationâthe powerful person, P, has power over the weak person, W, to the extent that P causes W to do something that W would not otherwise do. More formally, Dahlâs definition states that the amount of power P has over W with respect to response x is a function of (1) the probability that, when P does y, W does x, minus (2) the probability that, when P does not do y, W does x (Riker, 1964). Dahlâs definition requires that P make some âpower attemptâ and that as a result of this attempt, P changes Wâs behavior. Dahlâs own insistence that the exercise of power requires that P manifest an intention to influence W is softened somewhat by his suggestion that if W imputes an intention to P, and W shapes his behavior to meet the intentions he imputes to P, then power has been manifested. Empirically, it has been demonstrated that the attribution by W that P intends to influence him makes a considerable difference in the outcome of the interaction (Walster & Festinger, 1962).
One could speculate that when intentions are falsely attributed to P by W, different consequences will follow from the interaction than if P actually possessed the intentions imputed to him. False or nonveridical attributions may lead to communication problems which would not result if each party was clear about the otherâs intentions (White, 1965). Analysis of power relationships by Dahlâs requirements, then, would require that a causal connection be discovered between Pâs behavior and Wâs behavior, even when it is the latterâs misattributions which cause the change in Wâs behavior (Nagel, 1968). The essential point to be made is that while Dahlâs definition of power represents a delimitation of Morgenthauâs all-pervasive construct of control of âman over man,â the inclusion by Dahl of the requirement that P intend to influence W introduces the novel but equally complex problem of determining the attributions and intentions of the participants involved before the presence or absence of power can be determined.
A number of social scientists have noted that the assertion âP has power over Wâ may be synonymous with âPâs behavior causes Wâs behaviorâ (for a detailed treatment relating power to causation, see Cartwright, 1959; Dahl, 1968; Heider, 1958; Simon, 1957). As with Morgenthauâs concept of âcontrol,â the concept of âcauseâ includes at least all of social psychology within its domain, and does not differentiate subsets of independent variables as functionally related to subsets of dependent events. The experience of psychology has been that such all-pervasive sovereign concepts are not only vague, but are immune to empirical evaluation (Allport, 1954). It is a truism to comment that men partially control or determine each otherâs behavior, but it does not further scientific understanding to label all such interpersonal causation as power. Thus, if the concept of power is to be rescued as a scientific construct, the referent events must be carefully delimited.
Power as Coercion
There has been a growing consensus among social scientists which accepts a series of qualifying conditions regarding those types of interpersonal relationships which may be considered power-relevant. Weber (1947), for example, was careful to distinguish power from other forms of social control in which cooperation is crucial. For Weber, power (Macht) is the probability that one actor in a social relationship will be in a position to carry out his own will despite resistance, regardless of the basis on which this probability rests. Hence, in the exercise of power, resistance and conflict in one degree or another are considered as common or probable. In this sense, Weberâs conception of power implies a process of overpowering. Bendix (1962) has perceptively observed that Weberâs concept of power is very similar to Clausewitzâ (1962) definition of war.
Harsanyi (1962a, 1962b) has applied Weberâs ideas to bargaining situations. Harsanyi views most power as bilateral in nature where each person has some control over the behavior of the other. The amount of Pâs power in P and Wâs joint policy with respect to some controversial issue X is defined as the probability of Pâs being able to get the joint policy Xp adopted when P favors this policy and W favors a different policy Xw. Harsanyi further specifies that power relations become relevant in a social group when two or more individuals have conflicting preferences, and a decision must be made as to whose preferences will prevail. Thus, power is considered to be relevant only where social conflict exists between P and W. Parenthetically, Chein (1967) poses an interesting paradox by combining Dahlâs notion that where there is no intent there is no power with Harsanyiâs and Weberâs contention that, where there is no resistance there is no power. The Paradox of Unconditional Omnipotence states that any person who is omnipotent must at the same time be powerless, for he can have neither occasion to want anything nor any resistance to overcome!
Both Weber and Harsanyi agree that power is relevant only under conditions of social conflict. Accepting this primary restriction, Bachrach and Baratz (1963) have limited the domain of power to an event set of even smaller size. They argue that there are three relational characteristics associated with power: power exists (1) when there is a conflict of interests or values between two or more persons or groups; (2) only if W actually complies with Pâs demands; and (3) only if P can threaten to invoke sanctions. Thus, power is the process of affecting the policies of others with the help of threatened deprivations for nonconformity with the policies recommended.
The insistence that power should only be equated with successful coercion leads Bachrach and Baratz to provide some hypotheses about the conditions of such success. The threat of sanction, though necessary, is not considered a sufficient condition of power. It is insufficient because the availability of sanction endows P with power over W only when certain conditions are met: (1) W must be aware of what P wants, presumably through clear communications between P and W; (2) the threatened punishment must be perceived as costly by W; (3) the cost of Wâs nonconformity to Pâs wishes should be greater than the costs of conformity; and (4) W must believe that P will probably punish nonconformity. These propositions are clearly testable, and, in fact, have been generally supported by the evidence (Tedeschi, 1970). At the same time, it is equally clear that these propositions are much too simple to explain the entire domain of power even if the limited definition provided by Bachrach and Baratz is adopted.
The actual application of sanctions for nonconformity is considered by Bachrach and Baratz to be a property of force, rather than power, a differentiation with which a number of other theorists concur (e.g., Gamson, 1968; Lasswell & Kaplan, 1950; Parsons, 1963, 1967). Force is considered to be associated with manipulation rather than with power, though potential force is the basis for the successful exercise of power. Further, Bachrach and Baratz assert that the absence of any threat of sanctions coupled with an attempt to control or change the behavior of W is what differentiates power from persuasion. Since threats imply the intent by P to procure compliant behavior from W, Bachrach and Baratz apparently agree with Dahl that without source intent there is no power relationship.
The implications for both P and W when P employs force are quite different from the employment of other forms of influence. A personâs scope of decision making is radically curtailed under force. Once the bullet or missile has been fired, the target person is no longer in the position of choosing between compliance or noncompliance. As Chein (1967) has pointed out, in a power relationship it is W who chooses what to do, while in a force relationship it is P. Similarly, Simmel (1950) has noted that in power relationships, the âweakâ target of influence has much more latitude of behavior than is commonly ascribed to him, for he may choose to comply or not comply with the source of influence. Both Tedeschi (1970) and Fisher (1969) have noted that, from an ahistorical viewpoint, punishment of W for noncompliance to a threat may be considered a nonrational action on the part of...