KEGAN, Daniel Laurence, TRUST, OPENNESS, AND ORGANIZATIONAL DEVELOPMENT: SHORT-TERM RELATIONSHIPS IN RESEARCH AND DEVELOPMENT LABORATORIES AND A DESIGN FOR INVESTIGATING LONG-TERM EFFECTS. Northwestern University, Ph.D., 1971 Social Psychology Copyright © 1971 DANIEL L KEGAN, ALL RIGHTS RESERVED Reproduced by University Microfilms with permission of the copyright owner. Further reproduction prohibited without permission. A DISSERTATION?SUBMITTED TO THE GRADUATE SCHOOL IN PARTIAL FULFILLMENT OF THE REQUIREMENTS for the degree DOCTOR OF PHILOSOPHY Field of Industrial Engineering and Management Sciences Evanston, Illinois June 1971 Text from OCR scan. -14- 2.01-01 CHAPTER TWO: Theory 2.01 Chapter overview?This chapter presents a theory of trust and openness in organizational behavior, together with an eclectic sampling of the research literature to support the theory. In general, the literature is not discussed exhaustively. The literatures of several disciplines (including experimental, social, and humanistic psychologies, organization theory, sociology, political science, anthropology, and engineering) are brought to bear on the problem of trust in organizations; but a complete review of all literatures is?beyond the scope of this dissertation. However, a comprehensive review of the relevant research literature is made.?The specific items chosen for citation are those with which the author has come in contact and found relevant. Naturally, this has involved some idiosyncratic elements. Many of the items cited are considered to generally reflect the literature of that area. Those that are more provincial, however, still provide an initial guide and entering wedge into that body of research. The research on trust and organizational behavior has been exhaustively studied. Related areas, such as psychotherapy, laboratory games, and international relations, are mentioned, but are not extensively considered. This study addresses two general questions from a humanistic viewpoint:?a) what contributes to "effective" behavior in organizations, and b) how can such effective behavior be facilitated. Measuring organizational effectiveness has been a difficult problem for the general organizational theorist; the problem is much more severe for the theorist examining research and development laboratories. Economic results are only the indirect products of an R&D lab; the direct product is information. An underlying thesis of this study, then,?is that attention to information creation and exchange--to communication--?can relate directly to effectiveness. Appropriate trust and openness are believed to improve creativity and communication, especially in dynamic environments, Trust may be conceived as a preconscious condition permitting one to enter a situation with minimal defensiveness. Its behavioral manifestation involves opening oneself up to -15- 2.01-02 a situation, exposing oneself to the remaining unpredictibility of the situation. Two aspects of trust and two aspects of openness may be differentiated. There?is trust of oneself and there is trust of "the other." Also, openness may be with respect to receiving information or with respect to giving information. Trust of self facilitates an individual's reality-testing, adaptation, and learning. In addition, appropriate trust of one's coworkers and openness with them facilitates their adaptation, learning, and improved effectiveness. It is improved information exchange that connects trust with effectiveness (see Figure 1). A simplified sketch of this chain of relationships may be made. Trust of self permits openness to and awareness of the organizational environment--its goals, resources, and constraints. That is, nondefensive perception allows the individual to obtain a more valid conception of "reality"; and a more valid "reality" permits an individual to function more fully and more effectively. Nondefensive communications help one's coworkers receive clearer information, which in turn can lead to a more valid "reality" and to more effective functioning on their part. Nondefensive and open communication and behavior also facilitate collaborative rather than competitive behaviors on the part of others. This is especially significant for sequential decision processes, such as are typical of many management and research programs. Competitive strategies have a tendency to result in suboptimal solutions, while collaborative strategies?may evoke synergistic solutions. Such collaborative strategies may include participative decision-making-- a mode of organizational operations in which decisions are arrived at by the persons who are affected, or feel they are affected, by them. ' However, participative decision-making violates some traditional bureaucratic norms, especially those concerned with organizational control. Under various collaborative strategies, organizational control is not abandoned, but is made more specified Of the three aspects of control--objectives, processes, and feedback--retention of some control aspects by the organizational hierarchy and distribution of some to the organizational membership may facilitate creative and effective organizational functioning. In summary, trust, nondefensive perceptions, open communication, and appropriate collaboration and control can contribute to effective behavior in organizations. However, there are individual and organizational barriers to trust, openness, and collaboration--and therefore to improved effectiveness. One possible -16- 2.01-03 cause of such barriers is poor learning. Many individuals in contemporary America have learned that it is dangerous to be open with others. Moreover, they have learned to mistrust themselves--to mistrust their perceptions,?their feelings, and their competences. Such mistrust and defensiveness tend toward a closed, non-adapting system. It is unlikely that such an organizational climate will change without some external stimulation. Organizational development (OD) is one form of 11external11 stimulation.?OD is an educational strategy employing experienced-based behavior in order?to achieve a self-renewing organization. OD generally stresses the value--?for both the individual and for his work effectiveness in the organization--?of confrontation, choice, and collaboration. These behaviors require some trust of self and of one1s coworkers. OD seeks to facilitate the development of?trust and openness, so that the organizational members may remove barriers which prevent the release of human potential within the organization. This chapter--especially sections 2.04 through 2.09--covers a broad range of literatures. The above synopsis may aid the reader in relating the various sections of this chapter to its general theme~ an integration and synthesis of the literatures dealing with trust, openness, and change in organizations. The sections of the chapter may be partitioned into four main parts. First, sections 2.02 and 2.03 (together with the theoretical discussion of control in section 2.09) present a framework,focused on change processes, for the analysis of individual and organizational behavior. This framework borrows from role theory, information processing and decision theories, and humanistic psychology. Second, sections 2.04 through 2.09 relate significant variables to the problem of trust in organizations. Included are considerations of effectiveness at?the individual, group, and organizational levels; trust and openness; perception and reality; communication; decision-making; and creativity and control. The third part of the chapter concerns extensions of the basic theory of trust. The evolution of the newly coalescing sub-discipline of organizational development (OD) and some strategic issues in OD are considered in section 2.10. As a further extension, section 2.11 discusses some of the environmental and cultural parameters which might affect the basic theory. In the fourth part of the chapter, section 2.12, the formal operational hypotheses of this research study) together with several secondary and tertiary hypotheses are presented. -17- 2.02-01á 2.02 A Model of Man. A conception of maná In the present existential world, there are a few absolutes, Anthropology and utopian literature inform us that there are many ways of organizing society, yet at least three conditions must hold in any human culture. These are a) the birth from a woman of man as a baby, b) the ultimate death of each man, and c) man's dislike of pain and like of pleasure.1 With the~e three elements and utilizing psychological learning theory and the theory of the simulation of cognitive processes, a rudimentary model of man may be constructed. Man may be conceived as a probabilistic, information processing, open system, where some of the information (or reduction of uncertainty or of possibilities) is programmed before birth and some is programmed during life. A key element in this model is man's desire for survival. 2 Given that man is alive, he wished to continue being alive. Man's dislike of pain and his like of pleasure provide principal additional driving mechanisms for his learning. Man's knowledge of his mortality and his ignorance of the time and manner of his death provide him with both a central certainty and an uncertainty. Through his conception, man acquires the genetic programming that will influence part of his life-learning experience, Through his experiences with those who care for him as a baby and child, man learns additional basic programs (cf. Erickson, 1963).?As a result of these genetic and early developmental experiences, programs are created that may be viewed as needs (cf. Etzioni, 1968). These 1 It is conceivable that at some future time man may be able to grow men from testtubes and may be able to prevent death. However, these developments are not likely to occur for a considerable time, if at all. If the nature of birth and death were thus changed, however, the entire nature of man would be considerably altered. This section concerns man as he is and as he has been known for:the past several millennium. It is recognized that each of these three conditions, especially the third, could be made more rigorous. However, some point 'of departure must be assumed as "good enough." 2 Since this desire can be empirically demonstrated, its origin is not here at issue. However, two influences may be hypothesized: a) instinct?and b) learned anticipation of future rewards or pleasures. Kelley (1958: 49) notes that any person is motivated, motivated if for no other reason than?that he is alive. -18- 2.02-02 needs may be described as physical', belongingness, self-esteem, and self- actualization.3 "The universally observable means by which individuals in?all societies reach for the satisfaction of these needs are identity and role. Identity answers the question, 'Who am I?' and role answers the question, 'What am I doing?'" (Hsu, 1963: 160). Complete answers to both of these questions involve both thought and feeling. Most information-processing models of man have been principally concerned with only an aspe~t of the means of satisfing these needs. Their principal concern is with the doing; with thinking and problem-solving.4 This limitation may have been due to the developmental association with students of artificial intelligence,5 it may reflect characteristic American concern with means rather than ends, or it may be due to other causes. For whatever reasons, it has resulted in incomplete and imbalanced models of man. An accurate model of man needs to include not only his cognitive abilities, but his affective activities as well. However, this does not require a?complete restructuring of older cognitive theories of man, for affective elements can be incorporated into information-processing models (Simon, 1967). Insofar as man is guided by identity and role in the cognitive and affective satisfaction of his needs, a consideration of these factors is necessary to understand man. Of course, neither identity nor role is fixed, but reciprocally evolve as the individual tries to satisfy his needs, i.e., as the individual lives. Thus, a complete conception of man needs to recognize man's multiple needs, his affective as well as his cognitive activities, his means of satisfing these needs (including answering both who he is and what he does-- identity and role), and the cumulative process of growth and development, both for the individual and for his culture. 3 The nature versus nurture issue need not be decided here. The point?is that all men (with the possible exception of "wolf children") do have both these experiences and these needs. Davies (1963: 9) does not consider Maslow's (1943) safety need as a basic need, but as a means of achieving other basic needs. 4 Gullahorn and Gullahorn's (1963) HOMUNCULUS does include two emotions, anger and guilt, in its simulation of human interaction--a simulation based on Homan's model. 5 "The attempt to discover an efficient way for the computer to do the task should be called a problem in 'artificial intelligence,' whereas the attempt to imitate the human being should be called a problem in 'simulation'" (Miller, Gaianter, & Pribram, 1960: 54). -19- 2.02-03 Contemporary role theory. The concept of "role" has been seen as extremely promising for a synthesis of psychological and sociological theories. However, although rich in empirical referents, role theory lacks much of the interrelational calculus and rules of correspondence linking theory to the data that a true scientific theory has (Deutsch &Krauss,- 1965: 6, 173). Sociologists have been able to use role theory in their descriptive studies, but the psychologist has found he cannot readily use contemporary role theory to predict behavior. A principle reason for this inability to predict behavior is that role theory focuses too narrowly on behavior itself. A role is often defined as?the behavioral expectancies usually associated with given positions and?evoked by the appropriate symbols, where positions are aggregations of functions standardized within a social system and "cued" by some set of conventionalized symbols (Hollander &Hunt, 1963: 256). Current usage differentiates three aspects of the concept role (Deutsch &Krauss, 1965: 175). The system of expectations which exist in the social world surrounding the occupant of a position--expectations regarding his behavior toward occupants?of some other position--may be termed the Prescribed role. The-specific expectations the occupant of a position perceives as applicable to his own behavior when he interacts with the occupants of some other position may be termed the subjective role. Finally, the specific overt behaviors of the occupant of a position when he interacts with the occupants of some other position may be termed the enacted role. The author believes that it is this focus on behavior that has hindered the development of role theory. This difficulty is evident in Merton1 s theoriesá 'of 'reference groups and of anomie (Deutsch & Krauss, 1965: 190-203). For example, in his discussion of anomie, Merton (1938) describes five possible modes of adjustment by individuals to their society. However, the prediction of which mode will be chosen by any particular individual cannot be made?using Merton's article alone. He merely states that the mode of adjustment chosen "will be determined by the particular personality, and thus, the particular cultural background" (ibid., 678). Role conflict is another aspect of descriptive theory lacking predictive elements. Often the description of the presence of role conflict in a given social situation is the endpoint of a social analysis.?Even when possible methods of resolving such conflict are discussed (e.g. Katz &Kahn, 1966), predictive models of when which method will be chosen are rare in the traditional role theory literature. -20- 2. 02-04 The problem is that traditional social role theory often has approached society as static, and thus has approached roles as static and fixed. (A notable exception is Becker's Outsiders, 1963, with its dynamic concept of career.) Such traditional role theory has focused on specific institutions (e.g., business organizations, the family, educational systems), and thus often loses sight of the integration of society within the individual. These neglects parallel the neglects of cognitive simulation: neglect of the cumulative process of growth and development, of the function of emotions, and of the multiplicity of human needs served at any single time. The place of the emotions in social role theory may not be readily apparent. In its broadest psychological meaning, the term "emotion" refers to a stirred-up state of the organism, reflected in a) emotional experience, b) emotional behavior, and c) physiological changes in the body (Krech & Crutchfield, 1958: 230; see note 11 of this section on the terms "affect" and "emotion"). Illustrating the necessity to distinguish between emotional behavior (including both gross overt behavior and the detailed physiological changes) and emotional experience, Goldstein (1968: 24) approaches emotions as intervening variables. In his discussion of an information-processing model, Simon (1967: 36) describes an interruption mechanism in man's control program that has the properties usually ascribed to emotion. It is this interruption mechanism that allows man to pay attention to his multiplicity of needs and permits him to alter his goals and activities. A weakness of contemporary role theory has been described as too narrow?a focus on overt behavior. Social role theory describes the interplay of roles?in society, but plays scant attention to the interplay of roles within the individual.6 Animal psychologists have studied dynamic phenomena such as?learning, but too often, in generalizing their results, readers forget the?limitation that th~ rat was placed in the given maze. Child development psychologists have studied why and how the child learns what he learns, but the interest of some scholars often ends with the end of childhood. The ego-psychologist such as Erickson, have extended the developmental theory of man, but their?theory usually lacks the specific detailed.data (such as that used by other psychologists and by sociologists) that .permit prediction. Role theory can 6 This critique of the various theories and disciplines is, of course, quite generalized and perhaps somewhat overbroad. However, the central thrust of the disciplines and theories is the point of issue, not the entire range of activities put under each field's umbrella. -21- 2.02-05 become a central link between social and individual theories. However, to do this, roles must be conceived in their complete context, not abstracted and isolated from reality. A reformulation of role theory. Reality is the complex interaction of a universe full of various elements. Unto this complexity man imposes order.7 This imposition is primarily an epistemological one (Maruyama, 1965), reflected both in perception (e.g. Campbell, 1958; Kohler, 1947) and in linguistics (e.g. the Whorfian hypothesis, Whorf, 1956). Most individuals in contemporary Western culture live in a classificational universe inherited from Aristotle (Maruyama, 1965). This hierarchical universe consists of substances organized into mutually exclusive categories, subcategories, and supercategories.8 The classificational universe of Western Man is well described by an enumeration of states. In turn, possible behaviors of man can also be described by an enumeration of alternative states. An objective of descriptive (as constrasted with normative) decision theory is to describe the behavior of man?by describing the alternatives he will choose. Role theory describes the behavioral expectancies of individuals. Role and decision theories may be explicitly related by rephrasing role theory as the description of the behavioral alternatives that the individual selects to engage in.9 By considering the selection mechanism as well as the alternatives, role theory may overcome its inability to predict individual behavior outside of restricted social settings, i.e., to predict when, which role will be chosen. Important elements in the decision-making process are a) the objective?state of the universe and its component parts, b) the individual's preconscious10 perceptions of the universe (which for most Western individuals results in state-descriptions or alternatives), and d) the individual's selection mechanism. 7 The theological aspects of order will not be discussed here. Cf. Hora (1962: 31), "man does not produce truth, he manifests it and is defined by it." 8 The relational universe is not hierarchical and is event-oriented. Instead of the classificational preoccupation with causality, in the relational universe the mutuality of interactions predominate. This relational universe was characteristic of pre-communist China (cf. Needham, 1954, 1956). 9 In this essay, the term "choice" is used to imply conscious selection, while "selection" implies an elimination of alternatives by any means, conscious, unconscious, or purely physical. For example, a stone rolling downhill?"selects" the path it will take, but an individual may "choose" which way he?will walk to work. 10 Kubie (1965: 118f) describes the preconscious as "a continuous and remarkably rapid stream of afferent, integrative, creative, and efferent activity, mediated by coded cues and signals, but without fully developed symbolic representation." Besides the effects of real-time stimuli from external -22- 2.02-06 While a predictive role theory may be constructed using the decision or behavioral role as the unit of analysis, a more integrated theory results from considering the decision-premise as the unit of analysis (Simon, 1957: xxx). Thus, Simon defines a role as "a specification of some, but not all, of the premises that enter into an individual's decisions." A decision premise, then, may be considered as a conditional instruction in a program. That is, if various other premises are held (if the situation is appropriately cued), then the individual will put a certain operating program in control (will enact a certain role). A reconsideration áof role theory to include decision premises, behavioral alternatives, and a selection mechanism, as well as multiple needs, development, and cognition and emotion may now be made (see Figure 2). Beginning with an?individual interacting with his social environment, as long as he perceives his environment and himself as behaving lawfully according to the program then in control of the individual, he will continue interacting according to that role. However, this control program may be interrupted in several ways. The environment may change. Other people may have altered their behavior, so that it is no longer compatible with the individual's expectations. Or the spacial-temporal location of the interaction may have changed from that which the individual's control program de~s appropriate. Alternatively, the operating program ~ay have achieved its goal, in terms of aspiration level, satisficing, impatience, or discouragement (Simon, 1967: 32), or the ranking of the individual's multiple needs may have changed. Finally, the individual __________ 10, continued stimuli-objects, the preconscious also consists of active residual imprints from prior experiences. "Conscious symbolic processes sample all of this activity: that is, the incoming signals, the preconscious processing, and the outgoing products. By means of conscious symbols we generalize from individual bits of data, assemble them into patterns, and make abstractions from them by scanning, ordering, and arranging temporal and spatial sequences, and so forth¥¥¥¥Relatively speaking, the processes of conscious symbolic sampling are slow and limited, and provide only weighted fragments of preconscious activity, to make up what is erroneously referred to as 'conscious thought.'" "Unconscious processes constitute a distorted and even more fragmentary symbolic sample of conscious samples. Note in this connection that it is never the symbolic image, thought, or act which is unconscious, but its relation to that which it is supposed to represent, but masks instead." -23- Figure 2 Figure 2. Model of Individual Behavior, Its Interruption, and Its Modification. 2.1: Preconscious structure. i= Focal individual j= Physical and social environment k= Target entity (individual or group) t= Time No 2.15: Search for a new program. 2.14: Substitute new program. Yes Is here another adequate program? 2.12: Reintrepret. Yes 2.11: Can behavior or expectation be reintrepreted? Are Expectations adequately met? Yes 2.10: Interrupt program. No -24- á2.02-07 may become emotional, and this may interrupt the control program.11?When an operating program, or enacted role, is interrupted, another program must be selected. The substituted program may be the interrupted program, with the conflicting elements reintrepreted. For example, an individual may decide that the lack of a greeting from a colleague at a convention is not an intentional slight, but is due to difficult perception across a crowded hall. Or the substituted program may be a new program. In this case, the individual must select one appropriate program, or enacted subjective role, from his repertoire of programs. Under this latter model?of "rational" behavior, the individual will select that program, or role,?which best satisfies the conditional instructions, or decision premises,?of that program, or role. That is, the individual will compare his state description of the universe to his repertoire of alternative behavior programs, or roles, to find the role which seems most appropriate.12 Since there are likely to be some conflicting elements in his detailed perception of the universe, the individual will employ a probabilistic weighting mechanism to arrive at a unified conception (cf. the preconscious and the conscious, note 10). This analytic model of man provides a framework within which many possibilities exist. But to approach the human quality of man, the model needs to be fleshed out. Humanistic psychology and its writers (Allport, 1955; Bronowski, 1965; Bugental, 1965; Gunther, 1968; Laing, 1959, 1967; Maslow, 1962; Otto &Mann, 1968; Perls, 1969; Rogers, 1961; Schutz, 1967; Shostrom, 1968)?do just that. Humanistic psychology sees man as an organic, whole person, who exists.?Life is often seen as a process of discovery: discovery of the self in experience and of the outer world through experience. Rogers (1961: 115f) postulates 11 In this dissertation, the term "emotion" is used to refer to the common usage--a moderately high level of affect, readily observable to others--and the' term "affect" is used to refer to that constant cosymptom of life. There is a wide range of possible causal factors of emotion, Simon (1967: 35) suggests physiologicEJ.J. needs, loud cognitive associations, and uncertain environmental events. ::~1 ácontrast to the previously mentioned interrupting mechanisms, emotion belongs to a different dimension. Emotion is not so much an interrupter as a result of an interruption. It may be a cosymptom of an interruption, it is not the cause of the interruption. 12 "A person chooses for himself that alternative in a dichotomized construct through which he anticipates the greater possibility for extension and definition of his system" (Kelly, 1958: 59). -25- 2.02-08 four characteristics of the~fully functioning person (a term he prefers to "healthy" person). The fully functioning person is open to experience, trusts his own organism, has an internal locus of evaluation, and becomes more a process than a product. Moreover, humanistic psychology is optimistic. It sees a whole man as living his life, fulfilling his potential, experiencing joy (Schutz, 1967). It recognizes cultural, institutional, and personal restrictions on the individual, but maintains a "spirit of inquiry" concerning the basic nature of man, his limitations, and his possibilities.13 The Association of Humanistic Psychology (undated leaflet) defines the field as centering on a "commitment to the fundamental uniqueness and importance of human life" including 11the convictions that all human beings?are basically creative and that intentionality and values are basic determinants of human action." Although discussing theater, Spolin's (1963: 3) comments on experience and learning are consistent with this humanistic spirit: We learn through experience and experiencing, and no one teaches anyone anything. This is as true for the infant moving from kicking to crawling to walking as it is for the scientist with his equations. If the environment permits it, anyone can learn whatever he chooses to learn; and if the individual permits it, the environment will teach him everything it has to teach. Humanistic psychology places great stress on "trust" and "openness." (These words are frequently encountered in reading the humanistic psychology literature; they are relatively infrequent in the literature of sociology and the other psychologies.) A synthesis of the humanistic model of man may be constructed, focusing on these variables. The major dependent variable is man's effectiveness: how well he is achieving his potential, how fully functioning he is. If the individual trusts himself enough to be sufficiently open to the experience of his organism, which in turn reflects experience of the world, he will discover whatever guidance and control is necessary for his healthy growth. Openness to others promotes individual well-being ,(Jourard, 1964), while hiding one's 13 The theories of humanistic psychology generally assume, at least implid.tly, a physiologically "sound," "healthy," or "normal" person. This may reflect a concern primarily with the generally healthy middle-class. However, the humanists do recognize the "mind-body" interaction. For example, Maslow (1943) postulates the physical needs as of crucial primary importance. Returning to the "optimistic" quality of humanistic psychology, Wilson (1967) reviews some of theácorrelates, facilitators, and hinderences of avowed happiness. Fellows (1966) discusses happiness from a humanistic psychology viewpoint. -26- 2.02-09 feelings or trying to "play the part" of another's expectations hinder authentic behavior (Brumbaugh, 1968). This is a self-centered model of man, but not in a pejorative sense. Man is a social being; his individual organism naturally reflects concerns of other members of his species. It would only be by keeping closed to part of his experience that man could be, in the pejorative sense, selfish. As many organization theorists have noted (e.g. Argyris, 1969; McGregor, 1957; Knowles & Saxberg, 1967), one's assumptions of the nature of man and the world influence how he perceives, feels, thinks, and acts. In the following sections, the humanistic-role theoretic-information processing model of man presented in this section will be elaborated and compared with general research findings in the literature. -27- 2.03-01 2.03 Individual and Organizational Change Growth and change. Both the individual man and the social organizations he creates are open systems, and are therefore subject to factors of growth and of change. While it is possible to use these two words to refer to many differing concepts, this paper shall use them in a specific, if not atypical, manner. To begin with, the individual will be the entity under discussion; the parallel case for organizations readily follows. Man may be conceived as an open, information-processing system, with a monitor or control program, an interrupting mechanism, and many operating programs (see section 2.02). The overriding goal of this program is to live and to satisfy physical, belongingness, self-esteem, and self-actualization needs.1 The operating programs may be seen as hierarchical sets of?subroutines (cf. the plans, images, and TOTE units of Miller, Galanter, & Pribram, 1960). Using this conceptual schema, growth and decay may be?related to program and subroutine elaboration within the existing framework?of programs. Change, on the other hand, would relate to the development of parallel or alternative programs, resulting in a new overall program structure. Aging (either growth or decay) is usually "caused" by internal factors, change by external factors. Since human organizations may\be defined as open systems of roles?(Katz &Kahn, 1966: 172), and since roles may be directly related to programs, it follows that the above conceptions of growth and change for an individual also apply to human organizations. This description of growth and change will not result in two completely disjoint and non-overlapping subsets. However, it should prove useful in this discussion of individual and organizational change. Individual growth and change. In the process of becoming an adult, a child learns many new programs. Most of this learning is a growth (with?some decay) process. Change phenomena may be important in determining the child's basic personality structure. However, some change phenomena viewed from an isolated individual vantage, may be seen as growth phenomena from a cultural viewpoint. Thus, for example, Erickson (1963) sees the adolescent's identity problem as but one of eight developmental stages of man. The 1 Note that this goal is coordinate: to live and to satisfy the needs. By making these two aspects conjunctive rather than hierarchical, self- destruction programs, such.as suicide, may more readily be understood. -28- 2.03-02 individual's culture will provide a set of roles which may resolve the conflicts in each of these stages. Whether the individual chooses an accepted or disapproved established role, or whether he innovates a new role, will be determined largely by idiosyncratic factors, i.e., the specific interplay of growth and change. While the separation between growrth and change for childhood has considerable overlap, the distinction is much clearer for the adult. By this time, his basic personality is established and he has found his major social roles. Thus, most of the program alteration in the adult is due to program elaboration, rather than alternative program creation. There are several reasons for this consolidation in the adult. First, the adult has passed through almost of his developmental stages (six or seven of the eight, depending on whether the adult has completed his childrearing). Second, the adult's cognitive field has become more differentiated and its boundaries more rigid (Thompson, 1962: 19). That is, later events are experienced through the framework developed in earlier times. While it is possible for this framework to change, for the adult it usually grows.?Finally, where there is a cognitive choice, the adult may prefer to remain?within his established framework, thus conserving his "investment" (Homans, 1961). There are, however, ways in which the adult may change his programs. These closely parallel the interruption mechanism used to select new roles (see section 2.02). When the individual can not understand (i.e., does not know?how he is expected t~ respond to2) major segments of his life-space, or universe, he will search for a means of understanding (for an appropriate subjective role). If the individual is able to reintrepret the situation so that he finds current subjective roles (programs) that enable him to understand the situation, no change in the individual will take place. However, if the intrepretation of the situation can be maintained ambiguous for a period, then the individual may develop new roles (programs) before he forces the situation into a definition appropriate to his old repertoire of roles. In this latter case, change will have occurred. 2The existential expectation of "letting be" is a possible, though not popular, legitimate role. Ignoring an aspect of the universe, or "shrinking" regions of the lifespace, is another non-change solution; it is used more frequently. -29- 2.03-03 Men generally have an intolerance to ambiguity concerning major segments of their universe, and therefore they try to find appropriate subjective roles as soon as possible.3 Therefore, to permit change, strong forces must be established to prevent premature interpretation of the situation (premature selection of a program or role). These forces are the same as those which interrupt an enacted role, except that they are much more extreme. The environment may change. But the behavior of a few other people, while sufficient to interrupt a role, usually will not be sufficient to induce individual change.4 A change of expectations in the entire social world,?i.e., in the prescribed role system, on the other hand, may well induce individual change. While small spacial-temporal shifts may not cause individual change, larger shifts may leave the individual with confused expectations.5 Alternatively, a major goal may be achieved. A man may make his million, and then wonder how to enjoy it. Or a Negro may become discouraged and impatient with the slow progress of equal opportunity statutes, and seek new ways to achieve his objectives. Finally, the individual may become emotional.6 Many programs to change individual behavior, such as psychotherapy, sensitivity training, or "the week in the country," are structured around a situation which is new and which inhibits intrepretation according to old categories (cf. Lewin's concepts of freezing, unfreezing, and refreezing, 1947). By removing or reducing the counterforces, they provide the individual an opportunity to try new behaviors and to make new interpretations of his world. 3 There are some times when men seek ambiguity. However, these rarely refer to major segments of their life-space. Work and family are examples of major segments; an amusement park would be a minor segment. 4 Of course, if those few others occupy extremely central positions in the individual's life-space, this may be strong enough a force to induce change. 5 Small and large are relative matters. For a provincial peasant without access to mass communications, a move to a city may cause dislocation. A cosmopolitan business executive may feel comfortable flying around the world, yet he may feel in an ambiguous situation if visiting an urban ghetto for a day. 6 Again, note that emotion is a cosymptom, not a cause, of interruption. In this connection, another set of cosymptoms may be mentioned. Neurotic and psychotic actions may be viewed as behavior under incomplete or incorrect interpretations of the universe. -30- 2.03-04 Rational, reasonable, and real maná Planned change programs often?encounter resistance. One common reason for this is that the planners use an inappropriate model: they often assume their organization's members arc rational beings (in the strict economic model sense). Much more often, a "reasonable" or "real" man model would be more appropriate.7 Rational Man, the child of economics, has all the relevant information which could be needed in making his decisions, has a strictly increasing monotonic utility curve, can perform any needed calculations, can make infinitely fine discriminations, and always acts so as to optimize his utility. Typically, subjective variables do not enter into Rational Man's preference function. Administrative Man operates by the principle of efficiency. This principle is a definition of what is meant by "good" or "correct" administrative behavior. It does not tell how accomplishments are to be maximized, but merely states that this maximization is the aim of administrative activity ¥¥¥ (Simon, 1957: 39, italics in original). March and Simon (1958: 205) discuss how external economics change Administrative Man's goal from "optimizing" to "satisficing." Reasonable Man is a legal concept. Reasonable Man is somewhat?similar to Administrative Man, but the lineages differ. The Reasonable?Man has general acquired behavioral dispositions, but no idiosyncratic?acquired behavioral dispositions (Campbell, 1963). The legal "reasonable man" test is: would any other person occupying the same role, and thus with the same general acquired behavioral dispositions, and with the identical information at the moment, arrive at the same conclusion, as did the person under test. In contrast to Rational and Administrative Man, Reasonable Man does include subjective variables in his utility function. But Reasonable Man is not yet real. 7 To maintain the aliteration, what is often termed "Economic Man" is here termed "Rational Man" (cf.V/.Edwards, 1954; Simon, 1957: 75f, on rationality). Charles W. N, Thompson has helped clarify the concept of Reasonable Man for the author. -31- 2.03-05 It is Real Man's idiosyncratic acquired behavioral dispositions that make him real. While these idiosyncracies are entirely lawful considering the individual's entire past history of reinforcement (and perhaps some genetics), it is they which make prediction of individual behavior difficult. Four types of expected outcomes to planned change efforts may be differentiated. The first type involves those programs which, if implemented, would cause benefits to each individual member of the organization, as well as to the organization as an entity in itself. ("Benefit" and "harm" are discussed below.) The second would be those programs which, if implemented, would cause a benefit to the organization, but would cause some harm to?one or more individual organization members. These two types of programs are the principle ones involved in planned change. Third, an implemented program may benefit one or more individuals, but harm the organization; such an effort would be deemed unacceptable to the organization. However, an individual with high power may choose such a program for his own benefit, in spite of the consequent harm to the organization. The fourth type of change effort would be one which, if implemented, would cause harm both to the organization and to each individual. Such a program would be unacceptable to the organization and to any non-malevolent individual. If all change efforts were perceived as the first type, there would be little problem with implementation. Also, if all proposed programs were of the fourth type, the organization would have little reason to continue to support this group of change agents. However, it is with solutions of the second and third types that implementation difficulties arise. In a simple sense, benefit and harm to the organization may be?considered to be the increase or the decrease respectively of the long-run assets of the organization, provided that action which benefits the organization in the long-run may not, in the short-run, cause the organization to cease to survive as a viable entity. The assets of the organization include readily measured cash assets, as well as less readily measured good will (in the strict commercial sense), employee expertise, and employee willingness to produce. For Rational Man, benefit and harm would have definitions parallel to?that for the organization. Reasonable Man would consider not only objective variables, but also general subjective variables. Thus, feelings of economic security, pride in accomplishment, a sense of participation in the decisions of -32- 2. 03-06 the organization, and time for leisure all may be considered benefits for the individual. (Reasonable Man assumes some given cultural context; contemporary America is here taken to be the cultural context.) Although Reasonable Man considers subjective as well as objective assets, any other Reasonable Man would agree with the classification of benefits and harms. However, one Real Man may consider as a benefit that which another would call a harm. Thus, "benefit" and "harm" are defined by the individual's perceptions of the proposed change effort. These evaluations will be?consistent across individuals i.n similar roles for Reasonable Man, but they may differ for various Real Men. Some of the differences between Reasonable and Real Man evaluations may be eliminated by specifying Reasonable Man's role in greater detail (e.g., a marketing manager or a manufacturing manager rather?than just a manager); however, some idiosyncracies will also remain with?Real Man. It is primarily because of the differences between Reasonable Man and Real Man that implementation problems in organizations arise. If planned change efforts were all of the first type (individual and organizational benefit), there should be few problems with implementation. However, a change program may change from the first type to the second type (organizational benefit, some individual harm) by a change in the model used?to measure individual benefits. A first type of solution under' a Reasonable?Man model may become a second type of solution under a Real Man model. That is, some people, under some circumstances, do not behave "reasonably." If the "non-reasonable" man is in a powerful position (see Cartwright, 1965, for a lengthier treatment of power), change programs of the third type (some individual benefit, organizational harm) may be involved. The problem, then, for a group seeking to institute change in the organization, is to determine which powerful individuals will not behave as Reasonable Men for which specific change programs. From these considerations, four propositions concerning "real man" may be constructed. a) If an individual evaluates a proposed change effort as being of type one (individual and organizational benefit) he will support it and aid in its implementation. b) If an individual evaluates a proposed change effort as being of type four (individual and organizational harm) he will not support it and will attempt to hinder its implementation. c) Predictions of individual behavior for proposed change efforts of types two or three will depend on the relative evaluations of individual and organizational effect, on the individual's degree of "altruism," -33- 2.03-07 on the individual's risk propensity, and on the individual's temporal "discount factor"--a compensation factor for the time difference between benefits, harms, and the present. (These factors may be related to Figure 4.) d) The net effect (either benefit or harm) of any proposed project solution must be greater than an individual's threshold for action to cause him to engage in any action: either aiding or hindering implementation. The laws governing this threshold are similar to response threshold laws of psychology (Campbell, n.d.; Hill, 1963). While these propositions are rather simple, they may be seen as directly related to many more distal propositions and problems of organizational change. Organizational change. Although organizations may be defined as open systems of roles, the most obvious facet of organizations is that it is men who occupy the positions in the role system. Since organizations are systems or roles, to change the organization it is necessary to change the roles. But since men are involved in the role system, it is necessary, in turn, also to change the men. (Such a change may involve replacement or modification of the men.) The central difficulty in organizational change is to isolate the role system from forces which would prevent or inhibit its change. There are two principle ways of inhibiting counterchange forces. One?is to change the entire role system of the organization at the same time. The other is to change aspects of the role system sequentially and to give continued support to the changed portions of the role system while other portions of the role system remain unchanged. A key in either method of change is to find the leading subsystem in the organization. 8 8 Katz &Kahn (1966: 63) define a leading subsystem as a subsystem?"whose output exerts the greatest influence on the inputs of other component" subsystems. To make the concept more generally applicable, this author prefers to define the leading subsystem as that subsys.tem whose output?exerts the greatest influence on the "characteristic equation" (G) of other component subsystems. (See Figure 6 and section 2.09 for a more detailed discussion of these terms.) Thus, in the Katz &Kahn formulation, subsystems have differing inputs, while in this formulation, all subsystems have the same potential distal input. However, the proximal input (I) to subsystems may differ because of different filtering proceduresá (G'). The distal input (U) is the energy from the objective universe. This proximal input (I) is the same as Katz &Kahn's input. The distinction is useful in discussing subsystem change. In this formulation, a change?in the characteristic equation may result in a different output (O) and/or different proximal inputs. Using the Katz &Kahn formulation, it would be more awkard to explain differences in input and output by the single application of the leading subsystern concept. In effect, this formulation -34- 2.03-08 For change to begin to occur, the current operating programs must be interrupted and the selection of a replacement operating program must be temporarily somewhat uncertain. If the changed subsystem is not a leading subsystem, the other subsystems in the organization are likely to further transform the changed subsystem so that the organization as a systematic whole remains unchanged. However, if the changed subsystem is a leading subsystem, the organization as a whole may be altered. Major changes can usually be attributed to one of two interrupting mechanisms--either to internal system strain or to changed inputs (Katz & Kahn, 1966: 446). Either of these causes may result in further change, either planned or unplanned (Rubenstein & Haberstroh, 1966: 287ff). Of the many attempts at planned organizational change, two methods have been fairly successful: a) direct systematic change and b) continued inhouse sensitivity training of and organizational development with a leading subsystem,?usually top mangement (Katz &Kahn, 1966: chap. 13; Greiner, 1967). Many aspects of sensitivity training indicate that it has potential for systematic organizational change. In the training group (T-group), the individual is confronted with a novel situation, one in which he cannot immediately find an appropriate subjective role. He may find that most of the roles he does enact lead to unanticipated consequences, and hence will also be evaluated as inappropriate. These are conditions which can readily lead to a search for new, appropriate roles. Since all members of the group are in a similar, searching situation, appropriate prescribed roles may be developed. These prescribed roles may be translated with minimal distortion into subjective roles. In turn, because of the opportunity for practice, new enacted roles may result. The high affect (or emotional) atmosphere serves both to prevent the recursion to old roles and to support the members while searching and learning their new, more effective roles. When the members of the T-group leave the training sessions and interact with other members of the organization, they will be under pressure to behave according to their old roles. To be effective, the sensitivity training __________ , continued. focuses on the transformational properties of the subsystem, i.e., the characteristic equation, while Katz &Kahn focus on the environment of the subsystem, its input and output. A further distinction may be made between Aleading subsystem and THE leading subsystem. THE leading subsystem exerts the greatest influence, A leading subsystem exerts a "large" influence on other subsystems. -35- 2.03-09 sessions must give the trained individual support outside of the T-group and must enable him to persuade (with both cognitive and affective elements) his coworkers to accept his new behavior and to alter their own expectations and behaviors, i.e., their subjective and enacted roles. If the subsystem (or subsystems) chosen for sensitivity training is a leading subsystem and if the training sessions do result in sufficient individual change, then the organization may be changed in this chain-reaction manner. Whether the change is as planned, and whether a more effective organization results, are additional questions. If the planners (this may be an external change agent or the T-group itself) have a fairly valid understanding of the structure of their organization and its environment, and of the forces within and without the organization, then they should be able to develop plans for change which will result in being implemented. This is not to say that no difficulties will result, but that the difficulties will either have been forseen or may be overcome within the framework of the change plan.9 The question of the effectiveness of the changed organization is answered by similar conditions, summarized in the phrase, "the wisdom of the planners." The chief difficulty in change by sensitivity training is ensuring that?a leading subsystem experiences the training. If the leading subsystems of the organization are hostile or indifferent to the planned change (or to the new role system), the organization has little chance of changing. Direct systematic change avoids this problem by starting with the cooperation of the leading subsystem (top management) as a given. Thereafter, its problems are the same as for progressive change led by sensitivity trained members: other individuals must be persuaded (in both cognative and affective terms) to accept new roles and these new roles must be supported while the residual forces of the old role system are extinguished. The central question of planned organizational change, then, is the location and conversion of the leading subsystems of the organization. In a well managed organization, the leading subsystem will soon become aware of 9 An alternative way of describing the results of effective sensitivity training programs is to say that they result in higher interpersonal competence. The greater the interpersonal competence (Argyris, 1965: 4), the greater is?a) the awareness of relevant problems, b) the ability to solve problems in such a way that they remain solved, and c) the probability that the problem-solving process involved has not been harmed or negatively influenced. -36- 2.03-10 changed environmental conditions or of internal stresses and will seek appropriate strategies to ensure the continued effectiveness of the organization. Depending on the specific conditions, these strategies may or may not involve sensitivity training or alterations in the organization's role system. In a less well managed organization, the leading subsystem may remain unaware of the need to seek new strategies. Under these conditions, it is much harder for the subsystem that is aware of the need for organizational change to induce such change. This aware subsystem could try to directly change the organization even though it is not a leading subsystem. In this case, much effort would have to be expended by the subsystem for rather small amounts of net organizational change. Alternatively, the aware subsystem could work to make the leading subsystems of the organization aware of the need for change, of the change needed, and of effective strategies for achieving such organizational change. Other, less adaptive possibilities also exist. The aware subsystem?could decide that it is not worth trying to influence the larger organization, and could "live" with the old system. Finally, the members of the aware subsystem may decide to leave that organization for another. This may result in the leading subsystems becoming aware that a problem exists, or it .may only result in the individuals moving to a new system. (This latter possibility is non-adaptive from the organization's viewpoint; it may be quite adaptive from the individual's vantage.) The critical questions of organizational change, then, are a) does a need for change exist; b) which are the leading subsystems; c) are the leading subsystems of the organization aware of this need; d) do the leading subsystems have appropriate strategies for implementing effective changes; and d) are there other subsystems, either within the organization or in its environment, which:are aware of the need for change, have appropriate implementing strategies, desire to influence the organization, and have the ability to effectively influence the organization. In the well managed organization, the high technical and interpersonal competence of the leading subsystems should yield effective and adaptive organizations. In less effective organizations, even if there are subsystems that are aware of problems (or of the need for change), the forces against change and the desire ("is it worth it?") of the aware subsystem for organizational change must be considered. -37- 2. 04-01 2.04 Effectiveness: Individual, Group, and Organizational Explicit goals. Much "pure" behavioral science and virtually all "applied" behavioral science is concerned with effectiveness. Inasmuch as effectiveness "may be defined as the degree of goal-achievement" (Price, 1968~~3), the problem of the definition and measurement of goals is immediately raised. Often "official goals" are chosen as the focus of inquiry, sometimes "operative goals" are also included (at 3-4). Elite decision-makers are commonly chosen as the reporters of goals. A usual statement of the goal of most formal (profit-making) organizations is to maximize, within a given time á.period, the income of the organization's members: employees, management, and owners. Sometimes a few additional goals are recognized. The "informal organization" concept often .pits the goals of the employees against the goals of the formal organization. Just because they are human, organizational members will have some goals which are not formally held by the organization (e.g., Crozier, 1964: 80_; March & Simon, 1958: 64; Parkinson, 1957; Peter & Hull_, 1969; Selznik, 1953: 105). While the perceptive manager will tolerate and even facilitate some informal organization as an adaptive response of a formalized system to unexpected contingencies and changing conditions (e.g., Downs,?1967: 60f), the "individual goals" aspect of the informal organization is?often seen as illegimate: key decision-makers often will measure organizational effectiveness against the formal organizational goal. Some theorists are able to consider parallel goals. 1 Thompson and McEwen (1958) find both competitive and cooperative strategies used by the organization in its interaction with its environment. Tom Burns (1966) recognizes three interacting social systems contained in any organization:?the working organization, the political system, and the career structure.?The working organization reflects the cooperative interest the members of the organization have in its successful achievements; the career structure reflects the individual's interest in his own advancement, often made at the expense of other organization members; and the political system reflects the interests of various active coalitions. There are strong pressures on the practitioner against considering several parallel goals. The recent emphasis on cost/benefit analysis has reinforced these 1 Serial goals may be seen as connected in a means-ends chain. Parallel goals are not directly, causally related. -38- 2.04-02 a; unidimensional metric makes such a comparison relatively easy --easy but not necessarily valid. Goals are valued states. Cultures, subcultures, and individuals differ in their value structure; they will therefore differ in some of their goals. While some absolute goals may be determined for all humans (cf. section 2.02), these goals are fairly distant (in the means-ends regression) from individuals'and organizations' immediate operative goals. Thus the theorist, systems analyst, planner, or evaluator must establish some criteria for deciding whose values or goals he is to accept as his ideological framework. There is no a priori criterion for such a selection. A Calvanist orientation may focus solely on the goals of top management, a Populist may?heed only the goals of the rank and file employees, and a Socialist may attempt to implement "to each according to his needs, from each according to his ability." Often the decision of the ideological framework to be used is implicitly decided by power considerations. The analyst immediately accepts the ideology of his client or superior. Rather than accept any such ideology before analysis or evaluation is begun, a more reasonable decision may be made by keeping the decision as to the framework to be employed initially open. Under these conditions, the analyst would attempt to ascertain the goals or values of all potentially affected individuals and groups.2 Where individual and group values do not conflict, the question of the framework to be used may be circumvented. Where interactions occur or where values differ, some ideological framework must be established. But the analysis will be more powerful and more adaptive (to such possible conditions as a change in superiors or in the distribution of power) if such framework decisions are made explicit and made after the collection of the value data, rather than made implicitly and instead of the collection of value data. 3 2 The question of determining group values is, in miniature, the problem of determining organizational values. Some weighting scheme for individual values must be determined; an ideology given a rationale for such a?particular weighting scheme. The answer. is discussed throughout this section. 3 Making goal differences explicit may increase conflict (cf. Edelman, 1964). However, this short-run dysfunction may be resolved in a longer time period. As noted in the next pages, the temporal nature of goals and consequences is an important consideration. -39- 2.04-03 Temporal goals. The explicit recognition of various goals and values is but one requirement for a valid measurement of effectiveness. Another requirement is the recognition of the variable "time." As a subproblem of effectiveness, the optimal allocation of limited resources is an appropriate consideration here. Cash-flow analysis in economics has recognized a monotonic time value of money. That is, future credits are worth less than present assets and future debits are worth more than present assets. However, this monotonic model may not be rational for organizations, and often does not produce rational or reasonable behavior for individuals. The principal limitation for organizations of this monotonic valuing of resources is that it neglects potential problems in the short run. An organization needS some minimal level of resources to survive. Thus, rather than a simple "maximize profits in the very long-run" criterion, an additional constraint, "provided that a certain minimum profit in the short-run is obtained" more realistically reflects organizational goals (cf. Pfaff &Pfaff, 1967: 283). However, even this constrained criterion cannot rationally be used to control organizational behavior, nor should one feel confident using it to predict behavior in the organization. Although an organization need not be mortal, the members of organizations, human men, are. Thus, a man's primary planning range will be limited to his lifetime.4 However, even constrained to an average life expectancy, reasonable allocation procedures must consider another aspect of time. Some resources have peak value periods. This is a non-monotonic function. For example, in?animal behavior, imprinting can be seen as highly valuing certain behaviors (e.g., movement of the mother) only during a certain period of time: before or after this period, the value of the behavior is much less. There are also contingent-controlled values. For example, a college course in physics is much more useful after a course in calculus, or a vacation may be valued more highly if a period of work precedes it. Some psychologists are recognizing that extreme "Protestant Ethic" behavior may be detremental to mental health. This Protestant Ethic doctrine demands 4 People seem to consider their children, grandchildren, and some significant others as an extention of themselves and, in a secondary manner, plan for the lives of these other entities. -40- 2.04-04 work, which is considered to be a pure cost without benefit, during an individual's life to gain his reward in the afterlife. An increasingly common ideology, however, states: Instead of the model of reality being a machine, the new model will be the dance or the symphony. If we can see reality in terms of a dance or a symphony, then we will feel in harmony with, rather than alienated from, nature. We will also feel in harmony with ourselves .... Machine ideology does not value activity for its own sake; it values activity only as a means to an end. Symphonic ideology, on the other hand, values activity for its own-sake. (Warren T. Hill. Cultural dehumanization and campus discontent. Explorations, Autumn 1965. Quoted in the Journal of Applied Behavioral Science, 3, 1, 28, 1967.) There is an interaction between the need for explicitly determining the values of potentially affected individuals and the need for considering the temporal dimension of valued states. If elite decision-makers could absolutely ensure that they and their chosen posterity would remain in their positions of power, the need for recognizing others' values would not be as great. But such ensurance is not possible. Therefore, a wise member of the elite will consider, with some weighting, the values of non-elites, so that if or when conditions change, his values will be considered by the new elite. This recognition of minority rights is one of the practical bases of democracy (cfo Kendall &Carey, 1968). While there is no guarantee that such reciprocity will occur, there does seem to be a universal norm of reciprocity (Gouldner, 1960; Romans, 1961, on distributive justice). In summary, to approach rational planning, allocation decisions, or goal-achievement measurement,, e~licit consideration should be given to the valued states of all potentially affected individuals, groups, and organizations. Further, the temporal nature of these valued states should be explicitly considered. In most organizations, there are at least three salient entities whose goals must be considered for a comprehensive consideration of effectiveness. These are the individual, his work group, and the organization. The individual. Basic individual goal~ and some derivative goals were discussed in section 2.02. Man has physical, belongingness, self-esteem, and self-actualization needs. Man lives, and directs his attention to satisfying these needs. He is a process of becoming his potential; he strives toward full functioning. -41- 2.04-05 The work group. The work group5 exhibits elements of both individual and organizational effectiveness. It is concerned with its task, and it is also concerned with the needs and goals of its individual members (Cartwright & Zander, 1968: 30lf). In an R&D laboratory, the group's task may be seen as making good decisions and providing others with information so that they can make good decisions. (The R&D laboratory is considered in greater detail?in this section under organizational effectiveness, and in section 2.09?under creativity.) Closely connected with group effectiveness is managerial effectiveness?and leadership. Much of the early literature on leadership focused on the characteristics of the good leader. For example, there has been a considerable literature on "democratic versus authoritarian" leadership styles (e.g., Anderson, 1959; Korten, 1962; White & Lippitt, 1960).6 This focus on a single factor of the leader's behavior has been broadened to also include the power?of the leader (e.g., Pelz, 1951) and the nature of the task (Fiedler, 1968). This broadening focus is reflected in Hollander and Julian's review (1969) of contemporary trends in leadership research. However, although increased attention is being paid to external factors, factors internal to the manager are still significant influences on his behavior (cf. Zaleznik, 1963). Cummings and ElSalmi (1968) have reviewed the empirical research on managerial motivation. They find two basic schools of thought, that of Herzberg and that of ~orter, buf they find that the concept of psychological and social needs underlie both. Further, they see the need for explicitly testing the assumed linkage between managerial motivation and organizational effectiveness. 5 Traditional organization structures have had the individual a member of only one work group, reporting to one boss. Modern structures, such as project teams and matrix organizations, may give an individual membership in several groups at one time or may rapidly change his membership over time (Davis, 1962; Goode, 1960; Haberstroh, Baring, & Mudgett, 1968; Kukjian, 1963; Lazar & Kellner, 1964; Likert, 1967; Mee, 1969). While there may, in this case, be slight parametric variations from this section's description of group effectiveness, the central thrust of it should remain valid. 6 In his study of leadership styles among English managers, Sadler (1970: 19) found the consultative style was most consistently preferred. In addition, leaders who were seen. as having distinct and identifiable styles were more effective in promoting confidence and satisfaction, whatever style they adopted, than those without a distinctive style. Sadler suggests that "one important characteristic of the successful leader is consistency of behavior, which enables subordinates to know where they stand with him and to predict his actions and reactions." -42- 2.04-06 In addition to research on the characteristics of the leader, there is a growing literature on the functions of leadership: on accomplishing "a goal through the direction of human assistants" (Prentice, 1961). For example, Nealey & Fiedler (1968) have found that there are important differences between the functions of middle managers and those of top managers; thus different individual behaviors are also required. Perhaps the findings from the research on the leader can best be summarized by Tannenbaum and Schmidt's (1958: 101) description of the successful leader as both insightful and flexible: as "keenly aware of those forces which are most relevant to his behavior at any given time ¥¥¥ /and/ able to behave appropriately in the light of these perceptions." The organization. The goals of the organization itself, discounting those directly reflecting the goals of individuals and groups, are generally rather impersonal. Since the organization needs people, satisfaction of some individual goals is an important means for the organization: a means to achieving its own goals. Bennis (1962b)derives his criteria for organizational effectiveness from considerations of individual health. His three criteria for the healthy organization are adaptability, identity, and reality-testing. While these are useful conceptual criteria, they are not easy to operationally measure. Mahoney (1967: B77) studied managers' perceptions of mid-range criteria of organizational effectiveness. These criteria were "intermediate between unique operating measures which apply only to single organizations and general criteria applicable to all organizations but which cannot be assessed except at terminal points in the organization's life." Although twenty-four dimensions were found in the analysis, the study revealed that most managers operationally place primary reliance on only a few dimensions. In a subsequent study, Mahoney and Weitzel (1969) compared managerial models of organizational effectiveness in general business organizations with those in research and development laboratories. Again they found that managers utilized only a few dimensions in their judgments, but that the dimensions utilized in R&D were different from those of general business. R&D organizational effectiveness was perceived as being strongly influenced by reliability ("meets objectives without necessity of follow-up"), cooperation ("operations scheduled and coordinated with other organizations; rarely fails to meet responsibilities"), and development ("personnel participate in training and development activities; high level of personnel competence and skill"). In their discussion, Mahoney and -43- 2.04-07 Weitzel suggest that the differences between general business and R&D criteria may be due to the different nature of the technologies employed (cf. Emery & Trist, 1960; Harvey, 1968; Hunt, 1970; Woodward, 1965), the incomplete knowledge of cause-effect relationships in R&D (cf. Thompson, 1967), or the professional orientation of researchers. R&D objectives. The general organizational objective of an industrial research and development laboratory is to maintain or increase the viability of the industrial firm. This viability objective may be broken down into survival and profitability; and further broken down into more measurable operating objectives such as dollar assets (absolute and rate/GNP rate), profit per stock share (absolute and rate/GNP rate), market share/industry, and human assets (technical and interpersonal competence). However, the R&D laboratory does not deal directly with these financial operating objectives. The R&D lab's contribution to the firm1s viability is indirect and long- range. The basic function of a R&D lab is to exploit scientific and technical knowledge to enhance the firm1s viability. (This neglects a secondary function of the lab, or of any subsystem: enhancing its own viability. Such subsystem may, at times, be counter to main system objectives.) While experience in the lab may enhance its personnel's interpersonal and technical competence, scientific and technical knowledge will not, by itself, lead to survival or profitability. A product7 must be developed and must be sold. Manufacturing departments are concerned with making a product, and marketing and sales with exchanging that product for money. R&D's function is to ensure that the firm will have designs for products that can be sold and to decrease the cost of manufacturing them. If an organization's environment were unchanging, once a firm found a profitable product it would have no need for R&D.8 However, the organization has several dynamic environments and the organization itself is not static (cfa Katz &Kahn, 1966: 20lff). Scientific knowledge grows and changes, 7 Evan an organization whose sole output is research must sell its product. It must convince someone or same group to pay for a report, a patent, or the continued operation of the lab. Whether the "sale" is made before or after a product is "completed," a sale must be made. Projects must be "sold" within the organization also (e.g., by the originator of the project idea to management), but this "selling" does not directly contribute to the profitability of the firm. 8 This assumes that the firm did not wish to make more profit by making more products, or to grow in other ways. Also, the R&D function may be served by buying it from another organization, although this may be more costly and less adaptive. -44- 2.04-08 consumers¥ preferences and competitors' practices change, and government and other agency regulations change. Given these dynamic environments, an organization's current profitability does not ensure its continued viability. The R&D lab deals with ideas, with information. There is a long and fragile connection between R&D information and final product sales and profits. Because of this delayed and indirect contribution of R&D to profit, it is difficult to use money terms in constructing operating objectives for R&D, Even with an extremely long study, R&D dollar contributions to eventual product sales?is difficult to measure, and the opportunity costs (what would have happened?if ... ?) are even harder to estimate. A more proximal measure of R&D effectiveness focuses directly on information and does not require such lengthly study, This information effectiveness measure centers around probability distributions of possible states of the universe. R&D's function is to find scientific knowledge (increase the probability of the 11truth11 of a set of possible scientific states and decrease the probability of the 11truth11 of competing scientific states) which will be useful to the organization under the environmental conditions with the greatest expected value (probability and utility). From this description of the function of the lab, two measures can be evolved: confidence and regret. In its scientific work, the personnel in the lab want to be able to rationally have significantly greater confidence in a set of hypotheses than?in rival hypotheses. Since it is people who produce, evaluate, direct, and manage the lab1s work, this objective in its operational form becomes an evaluating individual i, at time t, changing his confidence in the truth of a set of hypotheses k, as reported by j (V(i,j,k,t), where V is the degree of confidence). To the extent that Confidence (i, j, k, t) remains high for one set of js and low for rival js as t increases, the lab is producing good scientific work (assuming is who are competent to reasonably evaluate the work). But good scientific work does not necessarily promote the viability of?the organization. Such work should contribute to the eventual profitability of the firm. Again assuming reasonable evaluators, to the extent that the evaluating individual i, at time s, does not regret work, project, or decision k, done during time t,. when viewed from the objectives of entity j, that work may be considered to contribute to organizational effectiveness, provided the organization is one of the j1s. (Note that in contrast to uses of the tensor for interpersonal relationships--as in Figures 2,3, and 4--for Regret, j is taken as the other entity and k is taken as a situation. This is because the primary -45- 2.04-09 relationship is seen as between i and k, under the parametric influence of j,) Thus the laboratory will want to minimize Regret (i,j,k,t)s for all s, t, and for all relevant j. A minimal set of relevant j 1 s would be the laboratory, the individual work groups in the lab, the individual researchers, other departments of the organization, and society at large. A secondary parametric condition of j, besides objectives, may be considered for Regret. Man is not "rational," he cannot have complete information. But?he can have more or less relevant information. Thus, the base state of information,j', may be considered if additional detail is desired. j 1 may?range from considering the information those working on k (this may be noted?as k1) had at time t; considering the information that was available to k' at time t; considering information that was available at time t + x; to considering no restriction of information (i.e., assuming truly rational man). If j1 is not specified, it can be assumed to be without consideration of restrictions of information, i.e., without considering the limits on rational decision-making, does i regret the decision k. Note that with these secondary variables,?the tensor becomes Regret (i,j-j1 ,k-k1 ,t)r,s. (rands are methodological subscripts. r refers to the reporter of the information, s refers to the time of the report.) The Confidence and the Regret measures can be combined in attempting to evaluate the contribution of R&D toward organizational objectives. Thus, at any time s, an evaluating individual may be more or less sure that a given past decision k, was helpful to the organization. The timing of decisions. In a completely static or completely predictable world, rational man need not worry about the speed of decision-making. In fact, rational man instantaneously makes perfect decisions. But rational man does?not exist and the organization and its environments are neither static nor completely predictable. Given dynamic conditions, a wise man should make his decisions as quickly as possible, so as to minimize the time the organization is not properly adjusted to its environments. As considered here, a wise man makes good decisions. But all decision-makers are not always wise. Therefore, sometimes it is better to wait for more information before making a decision.9 9 Tbe time of decision should not be confused with the time of the implementation of the decision. Since other decisions will be made using the first decision as a premise, it is the time of the decision, not of implementation, that is here of?prime importanceo For example, an individual may decide today to begin work on a new project next month. He may then plan to reduce work on his current project and to order equipment. If this initial decision is carried out, it will not have been implemented until next month, but it has had consequences before that t1me. -46- 2.04-10 A decision is based on a number of premises, some of which are predictions or expectations of future states. By waiting, these expectations may be tested. Neglecting the other changes in the world that have occurred, by waiting and correct1ng poor expectat1ons,10 better decisions can be made. But these other changes cannot be neglected: they may alter the validity of other premises. Thus, a basic conflict in decision-making is: should an individual make his decision now or should he wait for additional information. In industrial situations this question is of utmost importance. Competitors may steal markets with improved products or may capture markets with new products. New regulations may necessitate lengthly changes in products or procedures; changes which may be relatively inexpensive now but more costly later (e.g., changing labeling requirements or additional building regulations for a new factory). And new scientific information may enable the firm to capture or steal markets from its own competitors. If a decision is made too soon and later events provide information that indicates a counter decision must be made, decisions which used the first decision as a premise may need to be altered also. (This may be described as a situation of overcontrol.) If any of these decisions have been implemented, organizational resources will have been less than optimally used. However, if a decision is made too late, opportunities may be missed. At a minimum, additional time has been used in finally making the decision and other decisions will have been harder to make. (Other decisions will include as premises expectations concerning the eventual outcome of the first decision. Since this first decision premise is uncertain, it will cause greater uncertainty in later decisions. Decisions are made--or made easier--as uncertainties reduce.) Thus, for any decision there i~ an optimal period of time to make it; an optimal period surrounded on either side by periods of greater cost to the organization and loss of profitable opportunities. If a decision is made and later events do not prove signficant (in terms of decision sensitivity) premises false or poor, that decision was not made too soon. However, if premises are proved poor, the decision may have been made too soon. Whether it has or not depends on the cost of waiting and the benefit 10 The term "poor" rather than "wrong" is used to emphasize that expectations really result from probability distributions, rather than from discrete dichotomous categories. The distributions may be partitioned into categories, but the underlying basic mechanism is the distribution. -47- 2. 04-11 of the improved decision (some improved decisions may not be worth the cost). Since decisions always include, at least implied, premises that cannot be tested until much later times (e.g., the continued operation of a capitalistic system), a final answer on the appropriateness of a decision and its timing cannot really be made. However, good answers can be made, using measures such as Confidence and Regret, described above. Since the value of a quick decision depends on later events, it cannot easily be directly related to organizational effectiveness. Perhaps the best that can be done is to partition an organization's subsystems and its environments into those potential states which probably would require an extremely rapid decision, those probably requiring a quick decision, and those probably requiring only delayed decisions. A finer partitioning may be made by considering hours, day, and months, and by making more explicit the potential states and their implied expectations. Under this schema it may be found,?for example, that most R&D management decisions on project selection and termination should be made over a period of months, many marketing decisions should be made within weeks, and some management decisions should be made within hours. These examples are rough and speculative. A post hoc study could be made to determine the distribution of decision times for various organizational subsystems making various kinds of decisions and relating these decisions to some measure of organizational effectiveness. But this is not a major concern of this paper. Since questions concerning the speed of decision-making can be so contingent upon later conditions, it can be more useful to focus on the speed and accuracy of decision premise transmission. The phrase "speed of decision-making" would imply that the period under consideration is that between when an individual realizes (at a conscious?level for decision, preconscious also for selection) the need for a decision and his final decision. But the critical element is the time of the final decision, not the speed of the decision. (The speed of decision has some relation to the cost of the decision-making process itself, but this is not a direct relationship. Time and effort expended may be related to that cost, but not time alone.) The element of speed becomes important when considered in relation to time of realization of need of decision; thereafter the timing, rather than the speed, of the decision is the critical variable. -48- 2.04-12 After an individual realizes the need for a decision (this is itself a decision premise), his estimation of the situation will specify the time he should spend in making that decision. In turn, his estimation of the situation will depend on the premises he holds. While later additional information may alter his specification of the time he should spend in making the decision (this may be a conditional as well as absolute time: after Mr. A makes his decision; when the experiment shows a significant result; or a total time of four hours by next month), the final decision, the time spent in making that decision, and the time of the final decision are all decided on the basis of decision premises. Thus, it would seem that an individual should receive as many decision premises as soon as possible. But this may cause premature decisions and overcontrol. As discussed above, not all expectations (or premises, which are expectations deemded to be relevant) are valid or confirmed. The problem is somewhat resolved by realizing that an individual does not start every decision de novo. An individual holds a number of expectations, or decision premises, as the result of his prior experiences in life, and specifically, as part of his role (cf. section 2.02). These underlying decision premises guide the individual in deciding how to allocate his?time, as well as in his final decision~ In a sense, the individual is continually making the study mentioned?above, partitioning his lifespace into rapidly changing and fairly static probability distributions. A competent individual's lifespace (expectations) will fairly well represent reality (later confirmation of the expectation);?less competent individuals will more often find their expectations unsupported.11 Thus, the relation between a decision and its organizational effectiveness is mediated by two factors: the individual's general expectations and the premises specific to the decision. The question of decision effectiveness then reduces to: what is the relation between the flow of premises specific to a decision and organizational effectiveness. It may be hypothesized that: assuming technically and interpersonally competent personnel, the more speedily and completely decision premises flow in the organization, the more effective the organization will be. By assuming competent personnel, a fairly accurate set of general expectations are likewise assumed, so that specific decision premises may be properly 11 An individual may be competent in one set of situations, and not in another. He may be technically competent when dealing in organic chemistry, but not so competent in other scientific fields, and fairly incompetent in interpersonal relations. -49- 2.04-13 evaluated. Premises must be fairly completely, as well as speedily, communicated. This completeness and explicitness will aid in the evaluation of the premise and its relation to urgency of action and final decision. For example, a decision premise that "the market will buy a million units of product x if we can distribute them starting next month" may be incomplete unless accompanied by such often implicit premises as "if they can be sold for under a dollar each" and "this estimate was made by Mr. B, who is known to be notoriously optimistic." If the complete premise is not given (or able to be completed by general experience, e.g., knowledge of Mr. B1s optimism), a poor decision may be made--in this case, rushing into research and production. Since the decision-makers are assumed to have a fairly accurate set of general expectations, they should be able to estimate the danger of overcontrol-- the chief danger of having information too soon--and usually avoid it. But?this assumption may be questioned as assuming away the problem. It doesn't however, for most organizational operations assume competent personnel,?personnel that are able to fulfill their roles. In the event that parts of the organization know that other parts are deficient in certain areas (e.g.,?likely to overcontrol or to be highly sensitive to changes in weekly scientific progress in experiments), information that they provide the organization that?is likely to reach such a sensitive individual or group should explicitly and plainly include whatever secondary premises are necessary so as to compensate?for their general competency deficiency. If the individual is so deficient that no explicit caution will compensate, such potentially unstabilizing information should, if possible, be isolated from the incompetent (at least in this area) individual, unless the potential sender, on the basis of his competent general expectations, feels such information will be reasonably, rather than irrationally interpreted. Of course an organization will be more effective the more competent more of its personnel are. There are, however, some "perceptual distortions" inherent in the role system of most organizations. Thus, salesmen are very likely to hold expectations and perceptions differently than basic researchers. Such differences are, in a large part, a result of the demands of the specific roles (and somewhat of the personalities of the men who choose those work roles). To the extent that there are "role biases," these may also be compensated for by information senders. -50- 2.04-14 Thus, an effective organization should be one whose personnel rapidly and completely send information or decision premises to the proper people in the organization. Such complete transformation includes a) "translation," when necessary and possible, to compensate for idiosyncratic and role differences in perception, interpretation, and expectations, and b) explicitness of premises, especially for those premises not held in common by others.12 Measurement. Measuring organizational effectiveness has been a difficult problem for the general organizational theorist (Barton, 1961; Georgopoulos & Tannenbaum, 1957; Price, 1968a). As productivity, it has been relatively readily measured for manufacturing groups (e.g., number of units produced per hour, scrap reduction, safety), but even with such a well structured situation, there are multiple possible indices of effectiveness. For research and development, the effectiveness problem is much more severe. Utilizing a general mathematical model, Horowitz (1963) summarizes and critiques various approaches to evaluating th~economic results of R&D. While these may be useful for retrospective studies, the long time lags involved, often five years or more, restrict the usefulness of such economic approaches for shorter-term research. In addition, Horowitz believes that the economic impact of R&D should be measured at the level of the individual firm, rather than at the broader level of the economy or industry or at the narrower level of the individual research project. Such a level of analysis is virtually incompatible with research aimed at evaluating the effectiveness of groups within the R&D lab or with short-term action research or change programs. But economic results are only the indirect products of a R&D lab. The direct product is information. This is more readily measured, although there are two aspects of R&D that make its control and evaluation difficult (Rubenstein, 1957). The first is the uncertainty about the outcome of individual projects; the second is the difficulty of attributing the "success" or "failure" of a project to R&D rather than to other factors, such as manufacturing, marketing, national economic conditions, etc.?Rubenstein finds information to be the key to the effectiveness of R&D (ibid., 99). "Useful economic results¥ for any R&D program depend on the quality of 12 of course, individuals are also subject to information overload. Competent personnel also will rate the priority of the decision premises they may communicate, the potential receiver's information processing capability, and his rate of incoming information. Such "filtering," however, is to prevent system overload, not to promote individual competitive advantage or defensiveness. -51- 2.04-15 information that the laboratory provides to the organization as a basis for new and improved products and processes." Argyris (1965, 1968c) is also concerned with communication and information exchange, but his focus is more on the concrete individual. Argyris believes that most American organizations suffer from low interpersonal competence, Such interpersonal incompetence hinders communication, His studies (1965: 197, italics omitted) have led him to hypothesize that "researchers (engineers or physical scientists) tend to be predisposed, relatively speaking, toward interpersonal incompetence." Thus, in an R&D laboratory, where many of the decisions are unprogrammed (Simon, 1965) and where information exchange is the sine qua non, accurate communication may be less likely than in other parts of the organization which make programmed decisions or which are less interpersonally incompetent. The interpersonally incompetent are generally uncomfortable with feelings, openness, and risk-taking in interpersonal relationships. This leads to a culture that tends to reward conformity; mistrust and frustration increase; and individuals begin to question the motives of others. These questions, fears, and defensive behaviors eventually act to make technical problem solving less effective (Argyris, 1968b: 354). -52- 2.05-01 2o05 Trust and Related Variables Trust is a word that is used not infrequently in common speech. Both in ordinary speech and in some research, its definition is left implicit. Three explicit definitions of trust1 follow. One is a common dictionary definition; the second is stated in explicit, experimental terms; and the third is phrased in more clinical terms. Trust, assurance, confidence imply a feeling of security. Trust implies instinctive unquestioning belief in and reliance upon something. Confidence implies conscious trust because of good reasons, definite evidence, or past experience. Assurance implies absolute confidence?and certainty. (American College Dictionary, 1960: 1303) An individual may be said to have trust in the occurrence of an event if he expects its occurrence and his expectation leads to behavior which he perceives to have greater negative motivational consequences if the expectation is not confirmed than positive motivational consequences if it is confirmed.... A person gambles or takes a risk when he perceives that his potential gains from taking the risk are greater than his potential losses. ... Hope which leads to an investment of time, energy, self, etc., which is in any way lost or damaged as a consequence of nonfulfilment, is equivalent to trust .... An individual may be said to be suspicious of the occurrence of an event if the disconfirmation of the expectation of the event's occurrence is preferred to its confirmation and if the expectation of its occurrence leads to behavior which is intended to reduce its negative motivational consequences. ... Fear does not necessarily imply behavior to prevent the negative emotional consequences of the feared event. (Deutsch, 1958: 266f) What we here call trust coincides with what Therese Benedek has called confidence. If I prefer the word "trust," it is because there is more naivete and more mutuality in it: an infant can be said to be trusting where it would go too far to say that he has confidence. The general state of trust, furthermore, implies not only that one has learned to rely on the sameness and continuity of the outer providers, but also that one may trust oneself and the capacity of one1 s own organs to cope with urges; and that one is able to consider oneself trustworthy enough so that the providers will not need be on guard lest they be nipped. (Erikson, 1963: 248) The infinitive "to trust" requires an object. This object delineates what specifically is trusted. When the object of the trust is the self, additional connotations enter into the definition of trust. Trust of self may be seen as an important element of ego strength. "The ego is the organizing, governing, co-ordinating center of personality, whose primary tasks are to maintain 1 Throughout this chapter, references are made to additional researchers' conceptions of trust. Additional authors on trust include Brown (1969)È Carrier (1963); Deutsch, Krauss, & Rosenau (1962); Friedlander (1970); Kelley & Ring (1961); Pilisuk & Skolnick (1968); and Rotter (1967, 1970). -53- 2. 05-02 psychic harmony and to 1adapt' to the demands and possibilities of the environment" (Keniston, 1965: 364). However, besides these controlling functions of the ego, ego strength also involves regression in the service of the ego: "the ability of the ego to, as it were, 'shut itself off' and thereby to remain open to the childish, the sexual, the creative, and the dreamlike" (ibid., 365). Trust of the self may be seen as a feeling that one is capable of dealing with the probable contingencies of his environment. This concept of trust does not assume a beneficent environment, only that one can cope fairly successfully with the environment. As with the general infinitive form of trust, trust of the self needs an environment specified. In this paper, it is assumed that trusting behavior is always exercised within a given environment. This environment may be the family, the work group, the industrial firm, the national society, or all humanity. The given environment should be clear from context. In contrast to the environmental specificity of trust of others, when trust of self is referred to in this paper it will not, unless specifically stated, refer to any given environment, but to all probable environments. Open and nondefensive behavior is rationally permitted when the individual trusts his immediate environment and himself. Even assuming trust of self, however, there are still many environments that are not trusted.(cf. Carrier, 1963). For example, while there do exist many international cooperative agencies, cooperative industry trade organizations, and harmonious intra-firm relations, it is also true that current international relations?are being carried out prtmarily in a competitive framework, that severe competition exists within the automobile and oil industries, for example, and that there are fewer top executive positions available than people who would like to occupy them. Although the general thesis of this dissertation is that open behavior?can promote effe~tiveness, it should be recognized that indiscriminate openness?is not being suggested. "The problem for the group /and for any individual/ is?to find a principle of appropriateness that supersedes the simple norm that authentic open communication is always a good thing" (Schein &Bennis, 1965: 298). Some of the variables in Figures 3, 4, and 5 help establish when conditions are appropriate for openness (cf. Goode, 1967). -54- Figures 3 &4 Figure 3. Model of the Formation of Trust Expectations. 3.1 i's Expectations of the behavior of k in j at t 3.2: Predictibility of k's behavior. 3.3: Strength of effect of k's behavior. 3.4: Net discounted intent of k's behavior. 3.5: Certainty of expectation. 3.6: Degree of trust of k in j at t. 3.7: Risk propensity. 3.8: Degree of openness, non defensiveness. 3.9: Behavior. 3.10: Affect Figure 4. Detailed Model of the Determination of Intent. 4.1: Perceieved intent to entity k* 4.2: Predictibility of k's behavior 4.3: Temporal discount factor =f(i,j,k,t). 4.4: Risk propensity 4.5: Altruism 4.6: Net discounted intent of k's behavior i= Focal individual j= Physical and social environment k= Target entity (individual or group) t= Time -55- Figure 5 Figure 5. Parametric Relationships Between Awareness And Effectiveness 5.1: Actual, not perceived, Event (j,k,t) 5.2: Predictibility of Event 5.3: Strength of Effect of Event 5.4 Net Discounted of Intent of Event 5.5 Can Intent Be Increased By More Effort 5.6 Is Increased Effort Worth It 5.7: Increase Effort to Influence Event 5.8 Risk Propensity 5.9 Is Intent : 0 5.10:Awareness of Event Increases Effectiveness 5.11:Awareness of Event Does Not Directly Influence Effectiveness 5.12 Will Search Behavior Begin 5.13 Will Search Be Successful 5.14 Awareness> Frustration & Decreased Effectiveness 5.15 Awareness> Shortrun Frustration & Decreased Effectiveness 5.16 Awareness> Longrun Increased Effectiveness 5.17: Other Parameters Note: Actual Event Depends on Awareness & Influence Attempts; It is a Reactive Entitity. i= Focal individual j= Physical and social environment k= Target entity (individual or group) t= Time -56- 2.05-03 Figure 5 indicates parametric relationships between awareness and effectiveness. Figure 3 shows the relationships between open, nondefensivebehavior, trust, predictibility of others' behavior, intent (beneficient or malevolent) of that behavior, and the environmental setting. Figure 4 indicates in greater detail how risk propensity, temporal discount factor, and altruism influence intent.2 The figures are drawn parametrically; that is, the parametric influence of social entity (individual or group) k in environment (physical and social) j at time t on individual i is actually under consideration. For example, Jeff's (i) trust of his boss (k) in January (t) on matters involving Jeff's task (j) is one specific relationship. Parametric variations could include Jeff's trust in June (a newt), personal issues (a new j), or Jeff's co-workers (a new k). Figures 3 and 4 indicate some of the major variables associated with trust and openness. Table 2 presents an ad hoc typology of additional variables associated with trust, and indicates some research instruments that have been used to measure them. The table has two sets of numbers: one set indicates typology categories, the other set (#) refers to the instrument numbers in the Northwestern SCALES inventory (Kegan, 1970a). Parenthetical notations refer to secondary sources, other uses of the instrument, or related articles. A plus sign (+) after an author's name indicates "et al." References for those instruments with a SCALES number, and for those without such a number, are given in the appendix. The two major factors leading to the development of trust of a specific object k (be it person, group, situation, event, or environment) are: a) sufficient predictibility of that object and b) sufficient non-harmfulness or beneficence (strength of intent) of the object. While these developmental factors are stated in rather cognitive terms (the strictly cognitive aspects of trust may be termed "confidence" in the non-harmfullness of the situation), "trust" is a combination of cognitive and affective conditions. Trust may be seen as a preconscious condition permitting one to enter a situation with minimal defensiveness. Its behavioral manifestation involves opening oneself up to a situation, exposing oneself to the remaining unpredictibility of the situation. 2 The discount factor refers to the temporal value of goals (see section 2.04). Bruner and Tajkel (1961), Kogan and Wallach (1964), Lillibridge and Lundstedt (1967), Lundstedt (1966), and Starr(l969) report studies of risk. Altruism and equity have been discussed in Berkowitz and Friedman (1967), Friedrichs (1960), Krebs (1970), Latane and Darley (1969), Lawler (1968), Morgan and Sawyer (1967), Pritchard (1969), Pruitt (1968), Sawyer (1966), and Schaps (1970). -57- Table 2 Table 2 Variables Associated with Trust, Research Instruments Measuring Those Variables, and Related Articles I. Trust of Self?1.0 General personality FIRO-B, #369?Interpersonal check list, #434 (Altrocchi +, 1960) MMPI-K, #346?Story completion tests, Miller +, 1960. ...._.,.._.?1.1 Ego strength MMPI-Es, #180 (Miles's restandardization, #431) 1.2 Self-acceptance Group semantic differential, Burke &Bennis, 1961. (Sherwood, 1965) Self-esteem scale, #382?Self-rating form, #117?Sentence completion test, Dorris +, 1954. (Bass, 1962; Rubin, 1967) 1.3 Self-actualization?Interpersonal competence coding categories, #111 Short-form measure of self-actualization, #121 II. Trust of Other?2.0 General and political trust 2.1 Alienation, anomia, and authoritarianism items and factor scales, #414 Beliefs, #358?Cynicism toward government and politics items, #418?Faith-in-people scale, #106 Governmental cynicism scale, #410 Interpersonal risk scale, 4~48 (Lillibridge & Lundstedt, 1967) Interpersonal trust scale (Read), #338?Interpersonal trust scale (Rotter), #347?Jungle scale, #413 Norms (COPED form A-4), #349 Political cynicism scale, #409 Political trust index, #450 Risk taking (Kogan & W allach, 1964) (Slovic, 1962) . 2.2 Fear?Fear of rejection, #124 Primary Affect Scale, #429?2.3 Locus of causality, of control Control questions, #112?Feelings of internal vs. external control of fate, #123 I-E scale, 41433?Mach V attitude inventory, #348 2.4 Cognative structure 2.4l_Modes of perception Myers-Briggs type indicator, Sensation-intuition scale, #172 . 2.42 ÊSimplicity-complexity ?Barron-Asch attitude que8tlonnaire, Barron, 1953. (Bruner &Tajfel, 1961) ?Barron-Welsch art sc~le, Barron 1953. . 2.43 ÊCertainity-provisionalism 2.44 Rigidity-flexibility?Dogmatism scale, #380 (Bacchiano +, 1969; Ehrlich +, 1969; Powell, 1962) F scale, Adorno +, 1950. (Deutsch, 1960)?Rigidity scale, #432 III. Openness 3.1 Physiological-psychological 3.11 Strain Index of job-related strain--JRS, #128 Index of psychosomatic symptoms--PS, #130 Scale of premorbid adjustment, #421 3.12 Speech distrubances Speech disturbance categories, #107 . 3.2 ÊOpenness to experience ?Experience inquiry, #344 . 3.3 ÊCommunication behavior ?Behavior categories, Gibb, 1961. 3.31 Intimacy (Baxter +, 1963) -58- Table 2 cont. Intimacy-scaled stimuli, #336?Methods to determine the norms of address, #174?Nonlinguistic moves toward intimacy questions, #173?Procedures for determining forms of address, #109?Seating arrangements: questionnaires and a laboratory measure, #lOS Small group ecology, #175 -59- Table 2 cont. 3.32 Input?Authoritativeness and character of a speaker scales, #132 Categories of barriers to communication, #293?Measurement of interpersonal sensitivity, #420 (Berlew, 1961) V eridical perception interview method, #131 (Giffin, 1967) 3.33 Output Disclosure, #359?Indicators of open behavior, #170?Procedures for determining self-disclosure, #110 Self-disclosure questionnaire, #181 3.4 Altruism, helping behavior Trust question, #171 IV. Organizational climate Climate questionnaire (Littwin & Stringer), #364?Executive climate description inventory (Tagiuri &Litwin), #367 Organizational climate description questionnaire (Halpin), #366 Profile of organizational characteristics (Likert), #368 V. Group behavior Group atmosphere scale, #298?Group behavior inventory, #345?Group effectiveness rating scale, #415 Group growth yardstick, #335 Meetings (COPED form A-6), #350 Training group scales, Blake +, 1962. VI. Laboratory games Altruism scale, Sawyer, 1966. Personality/Attitude Schedule, #430 Prisoner's dilemna, Rapoport & Chamah, 1965. (Deutsch & Krauss, 1962) -60- 2.06-01 2.06 Perception and reality Man has undergone many changes since Adam and Eve. The Newtonian Revolution~created~a mechanistic clockwork man. Adam Smith ,, created1/ an economic-rational man, and Freud ~created~ a man subject to irrationality. Today, the General Systems theorists (von Bertalanffy, 1956; Buckley 1968) have given us a man who is an open subsystem in an environment of other open subsystems. Mankind existed before and after Newton, Smith, Freud, and von Bertalanffy, and yet the ideological revolutions of each of these theorists did change the "objective reality" of what man was. Without going into the metaphysics of the nature of truth, it can be stated that if a group of people believe an assertion, for that group at that time the assertion is true.1 This assertion may involve the relative motions of the Sun and Earth, the "true" nature of man as competitive or cooperative, the determinants of human behavior, or the desirability of eating sweatbreads. The model of man presented in section 2.02 identified four elements in the decision-making processes of man's behavior: a) the objective state of the universe and its component parts, b) the individual's preconscious perceptions of the universe, c) the individual's conscious coding of the universe, and d) the individual's selection mechanism. Perception is a key element, a bridge between the "objective" state of the universe and man's awareness of his world; between the other and the self. But perception is not a fixed, static thing; it is an active process. Relating perception to evolution, Gibson (1970: 99) defines perception as "extracting information from stimulation." However, stimulation is not only full of potential information; it is overfull. There are limits to the information-processing capability of animals; they are subject to information overload (Miller, 1965). Perception is an active process, a search for information. A specific search may be terminated when there is sufficient reduction of uncertainty; but an organism also has need for stimulation 1 Discussions with Charles W. N. Thompson have been helpful in the author's thinking about assertions, confidence, and truth. Laing (1960: 36) suggests that "sanity or psychosis is tested by the degree of conjunction or disjunction between two persons where the one is sane by common consent." -61- 2. 06-02 (cf. Bronson, 1968; Butler, 1954; P, Solomon et al., 1961).?Thus, there is a constant dialectic: search and termination, uncertainty and its reduction, drive and satisfaction, figure and ground; for a host of factors influence the organism's selection mechanisms of perception. The influence of environment on visual perception has been shown by Segal, Campbell, and Herskovits (1966). This cross-cultural study found that experience with two-dimensional representations of a three-dimensional reality, experience with the square corners characteristic of "carpentered-worlds,"?and experience with broad, horizontal vistas had an appreciable effect on subjects' perceptions of various optical illusions. Not only does his physical environment condition man's learnings; his social environment can also have an appreciable effect. Commenting on Asch's (1955) classic experiments in conformity, Campbell (1961: 108) states that in many instances, certainly, so called "conformity" behavior is an intelligent part of a rational search for valid knowledge about a fallibly and indirectly known world rather than merely an interest in being like other persons whether or not they are correct. Miller (1969) has reported experiments where the autonomic nervous system can be conditioned, that visceral and glandular responses can be learned.(cf. I Katkin &Murray, 1968). Using a cognitive dissonance framework, Zimbardo (1966, 1967) has been able to show that even such basic drives as thirst and pain are not "absolute," but can be "perceived or felt" in differing ways. He has shown that not only subjective measures of pain, for example, are subject to cognitive control, but that objective measures such as learning performance, galvanic skin resistance, and basal skin resistance are also influenced.2 If even such "objective" stimulations as images of drawings, visceral responses, and pain can be perceived in differing ways, it could be expected that more complex and ambiguous stimuli would also be subject to such "subjective" perception. 2 Zimbardo notes (1967) that cognitive dissonance theory is a face saving theory: that it operates only for those who cannot admit making a mistake when free choice exists. American culture seems most appropriate for cognitive dissonance theory. Hsu (1963: 224) notes that "in contrast to the two Asian societies [China and India, the dynamism of the American society is particularly aided by its ideal of complete freedom of the individual. influenced. -62- 2. 06-03 One of the most complex and ambiguous sources of stimulation in man's world is other men. Argyris (1969) presents a position that even what the professionals, the psychologists, perceive as basic human nature is not an absolute but is conditioned by their own learning, adaptation, and expectations (cf. Collier, 1968; Sarachek, 1968). As an open systems viewpoint would lead one to believe, and as the work of Argyris, Campbell, and Zimbardo affirms, realityáis not a fixed entity, but is subject to the interpretations of it that men make. This notion of "subjective reality" is essential to a consideration of?man and his effectiveness: for himself and for his work organization. Man differs much from the Newtonian mechanistic clockwork model. Man can, and?does, influence his environment, which includes other men. It is not a question of whether he should or should not exert such influence: he does. It is?rather a question of whether he can use his influence to help achieve his goals. There is a substantial literature concerned with the perception of other persons (e.g., Berlew, 1961; Burke & Bennis, 1961; Cantril, 1957; Goffman, 1959; Heider, 1965; Newcomb, 1956; Steiner, 1955). This literature is concerned with how an individual attributes motives, values, and beliefs to another, and how these attributions in turn influence the behavior of both (Maselli & Altrocchi, 1969). Such attributions can have an extremely strong and subtle effect. Research subjects are often influences by their preconceptions and suspicions (Schultz, 1969). Double-blind research designs are a control against the experimenter's influence, often unconscious, on his subjects and their behavior (Barber & Silver, 1968). Rosenthal and Jacobson (1968) have documented how teacher's expectations can influence student learning; Livingston (1969) has commented on the effect of managerial expectations on subordinates' performance and development. Zalkind and Costello (1962: 234) have reviewed the effects of interpersonal perception on administrative behavior, and suggest that effective administration is encouraged if the administrator is aware of the intricacies of the perceptual process and has an accurate self-perception.?In all these examples, the expectations of the researcher, the teacher, and the manager influence their perceptions and the behaviors of the subject, the student, and the subordinate. -63- 2.06-04 Besides influencing his environment, man is adaptive--he learns. The child learns from his genetic coding, his family, his formal educators, and his friends. As an adult, he is further socialized by his associates at work and by his interactions \\lith others. This learning may be considered to be composed of two parts: that which was true of a unique situation and that which is generally true for all similar situations. In a world which changes slowly, a person can generally rely on his learnings of yesterday to be valid today. In a world which changes rapidly (Bennis, 1966) this is no longer true. The walls of the maze run yesterday may have moved today; more effective procedures for achieving the same results may have been discovered; the goals themselves may have been altered; and new individuals may occupy old positions and roles. But to be aware of the changes around him, man must interact with the new situation and he must be open to perceiving it as it is, rather than as he expected it to be. In a changing environment, to be most effective man must test his conception of reality against the reality that exists in the now. Thus again, it is not a question of whether or not man learns, it is one of whether or not he learns correctly--whether his model of the world corresponds, to a good enough degree, to actual reality. Reality-testing can help man become personally more effective. Section 2.02 indicated that an individual develops programs of behavior and cause- effect beliefs. He uses these to help fulfill his basic needs and derivative goals. However, derivative goals can drift away from relating to basic needs and what were once causes can cease to have their intended effects. Testing actions to see if they have their desired results and testing old goals to see if they still have personal value can help an individual to be a more fully functioning person (Rogers, 1957). -64- 2.07-01 2.07 Communication The communication process may be considered to consist of three parts (Weaver, 1949): energy, meaning, and action. The first requiste for communication is the transmission of energy. A broken wire stops the electric current from flowing in a telephone circuit; machinery noise prevents a shout from being heard; poor printing prevents understanding a letter. These examples show man as a rather passive receiver in the communication process. Man also exercises an active role in communicating. He can move his seat to obtain a better view of a drama; he can shut his eyes to miss the disturbing parts of a film. These are conscious and intentional influences on the transmission of energy. In addition, man controls the information he is exposed to by nonconscious means. Because of proximity, habit, and preference, an individual develops systematic biases and selective exposure to information sources (Freedman & Sears, 1965). Some of the problems of implementing the action phase of a communication were discussed as organizational change in section 2.03. Additional questions of action and implementation will be discussed with decision-making in section 2.08, and in section 2.10 with issues of organizational development. For now, it may be stated that a communication is intended to achieve some result, and the communication is not complete if the result or action does not occur. Much research has been directed at the second element of the communication process: the transmission of meaning. The protean nature of man's perceptions,?of both objects and social situations, was discussed in section 2.06. The?remainder of this present section will review additional findings concerned?with the transmission of meaning and focused on the role of trust in communications. The general problem of communication accuracy has been recently reviewed by Mehrabian and Reed (1968). In a world of conflicting and partial information, what and whom one believes is important (cf. Simons; Berkowitz, &Mayer, 1970). Also, feedback processes, if open and honest, may improve information; while biased feedback may lead to false information and poor decisions. Mellinger (1956: 309) found that a communicator who lacks trust in the recipient of his communication tends to be motivated to conceal his own attitudes, resulting in communications which are evasive, compliant, or aggressive. -65- 2.07-02 Defensive behavior may be defined as an act which is "motivated, in substantial part, by an individual's need to preserve a stable perception of the self or the defend the self from perceived attack or potential attack" (Gibb, 196la: 67 ). In his analysis of small group interactions, Gibb (196lb: 142) found that defensive behavior was correlated positively with?losses in communication efficiency: the receiver would read into communications projections of his own anxieties. Conversely, the more supportative the climate, the better the communication: the receiver would be better able to focus on the structure, content, and meaning of the messages. This problem of defensiveness is not limited to laboratory small group interaction. As Gibb further comments (196lb: 143), "speech is so frequently judgmental that there is a reality base for the defensive interpretations which are so common." Non-judgmental interaction has been a central part of Rogers's client-centered therapy. He sees (1952) the evaluative tendency as the major barrier to communication. This tendency, and thus the barrier, is heightened in emotional situations. Real communication occurs when one sees the other's?idea or attitude from his point of view. However, such understanding is often avoided, for it is risky: the risk is the possibility of being influenced, of being changed. Change is not always a threat, but it may be. If an individual trusts the other, he is more likely to remain open to understanding him; if he does not trust him, defensiveness, evaluation, and miscommunication are probable results. In his review of studies of ethos and source credibility, Giffin (1967: 104) found support for "the hypothesis that interpersonal trust is based upon a listener's perceptions of a speaker's expertness, reliability, intentions, activeness, personal attractiveness, and the majority opinion of the listener's associates." In addition to these attributes, trust and nondefensiveness can be based on the nondefensive behavior of the other. As previously discussed, Gibb and Rogers have found that .nondefensiveness or openness on the part of one individual encouraged openness in the other. In their study of superior-subordinate communications in a large public utility, Burke and Wilcox (1969) found support for this reciprocal relation. In addition, they found that greater openness by one or both members of the dyad was associated with increased work satisfaction. -66- 2.07-03 Read (1962) found a negative relationship between upward mobility aspirations of subordinates and the accuracy of their communication of work- related problems to their immediate superior. In addition, this relationship was strengthened when the subordinate perceived his superior as highly influential concerning the subordinate's career, and when the subordinate had low trust in his superior. However, perceived influence was dominant over trust: the highly ambitious would distort information even if he trusted his superior. Improving awareness and communication skills has been a persistent goal?over many years of evolving experimentation with sensitivity training. In addition, sensitivity training has tried to foster climates of trust and openness. Although also influenced by several other factors, in general sensitivity training does achieve these goals for many of the participants,?at least within the T-group (Durham, Gibb, &Knowles, 1967; Gibb, 1970; Harrison, 1967; Stock, 1964). Transfer to other situations is more conditional, but can occur under favorable circumstances (Argyris, 1968a; Buchanan, 1969; Campbell & Dunnette, 1968; Dunnette & Campbell, 1968; House, 1967, 1968; Patten, 1969; Schein & Bennis, 1965). This problem of transfer is further discussed in section 2.10. -67- 2.08-01 2.08 Decision-making The decision-making model used in section 2.02 for the analysis of individual behavior had four parts: a) the objective state of the universe;?b) the individual's preconscious coding of the universe; c) his conscious?coding, and d) his selection mechanism. In applying this model to organizational decision-making, or to any question of effectiveness, a fifth phase is?necessary: e) implementation of the decision. A decision can be defined as "any conceptually held statement of change between two sets of conditions" (Thompson, 1956: 64). Practical human decision- making involves a choice or selection of one behavioral alternative from a?group of such alternatives (Tannenbaum & Massarik, 1950). There has been significant research on all five segments of the decision-making model. A major effect on decisions is the mediation of the alternatives, consequences, and probabilities of various behaviors from the objective state of the world to the perceptions of the individual decision-maker. Some of the literature relevant to the first three phases of the decision-making model has been reviewed in sections 2.06 (Perception) and 2.07 (Communication). The fourth phase, selection, has received attention in the literatures of game theory (Boyle & Bonacich, 1970; Feldman & Kanter, 1965; Gallo & McClintock, 1965; Luce & Raiffa, 1957; McDonald, 1950; Rapoport, Williams, 1954) Rapoport & Chammah, 1965; Taylor, 1965; Vinacke, 1969; and simulation (Cohen & Cyert, 1965; Guetzkow, Alger, Brody, Noel, & Snyder, 1963). The implementation phase has been a concern of many organization theorists (see section 2.03). This present section will elaborate on the research on decision-making, and on the influence of trust and openness in decision-making. Game theory. Game theory deals with "those decision situations in which the decision-maker has an opponent--in which his choice of action is contingent upon what one or more other decision-makers do" (Taylor, 1965: 54). This "opponent" need not be a specific individual; it may be an organization, the stock market, or nature itself.?Traditional game theory and decision-making theories deal with closed systems.1 They do not consider systems, that is games, in which there is a 1 A closed system is one in which no information, energy, or mass can transfer across the boundary into or from the environment. A system for which such transfer can occur is an open system. For further discussions of open systems, see the General Systems literature, for example, von Bertalanffy, 1956, 1962; Buckley, 1968.' -68- 2.08-02 continuing involvement by the players (cf. Shubik, 1963). Usually the game is one play long, sometimes the game lasts a known number of plays or until a specific goal is won. Noel (1963: 101) has noted that game culture--the set of predispositions which develop out of the shared experiences of the game itself--takes time to appear. Guetzkow (1964: 27lf) has said that the participants in the Inter- Nation Simulation did get emotionally involved in their game, but this simulation is much more complex and takes much more time than do the games of traditional game theorists. Also, in contrast to the studies of the traditional game theorists, there is no intrinsic solution to the Inter-Nation Simulation; the players define their own goals. Thus, in traditional games, the results of one game do not influence the?next game; there is no next game. In these non-cyclic, closed-system games, a mathematical morality, or at least frugality,develops: "the sensible object?of the player is to gain as much from the game as he can, safely, in the face?of a skillful opponent who is pursuing an antithetical goal" (Williams, 1954: 23). The best strategy to achieve this safe maximum is the minimax principle:?minimize your maximum possible loss. The minimax strategy minimizes the variance in the game, and is the best strategy that can be followed in non-cyclic, competitive, closed-system games in which the players have no influence on the payoff matrix, that is, no influence on the alternatives or their consequences. Schelling (1960) makes some advance in game theory by considering non-' zero-sum games in which there is mutual dependence among the players. But still, the game theory model considers decisions under certainty, or at best, under risk. Most interesting decisions occur under uncertainity.2a In reality, man is not a passive captive of "the game;" he is an active player, referee, and rulemaker. 2a "Decision under certa1nity is that in which not on1y the a1ternatives in the choice to be made are known, but also each alternative is known invariably to lead to a specific outcome. Decision under risk is that in which the alternatives are known and in which each alternative leads to one of a set of possible outcomes, each outcome occurring with a known probability. Decision under uncertainty is that in which probabilities of specific outcomes are unknown, or perhaps not even meaningful" (Taylor, 1965: SO). -69- 2.08-03 Man differs much from the Newtonian mechanistic clockwork model. Man can, and does, influence his environment, which includes other men. It is not a question of whether he should or should not exert such influence: he does.?It is rather a question of whether he can use his influence to help achieve his goals. Besides influencing his environment, man is adaptive--he learns. Again, it is not a question of whether or not man learns, it is one of whether or not he learns correctly--whether his model of the world corresponds, to a good enough degree, to actual reality. If a man learns poorly, he may be able degenerately to influence his environment so that it more closely resembles his poor model of it. These actions are sometimes termed self-fulfilling prophecies. But man can also have a "good" model of the world, and he can influence his environment to make it better. Thus, man differs from mechanistic systems in that he can elicit responses from his environment, including other men, and he can learn from his experiences. Because of his eliciting and learning capabilities, man can sometimes do better than a minimax strategy--minimizing his maximum possible loss--would allow.?By learning from his experiences, he can realize a long-term gain, even from?a short-term loss. Commenting on the natural process of development in organizations, Blau and Scott (1962: 250f) state that problems appear, and while_ the process of solving them tends to give rise to new problems, learning has occurred which influences how the?new challenges are met¥. Consequently, effectiveness in an organization /can improve/ as a result of accumulated experience. Also, by eliciting certain behaviors from his environment, especially from other people, one can alter the set of alternatives, consequences, and probabilities of various courses of action. Morton Deutsch and his colleagues have done extensive research with one such situation. They have developed games which investigate the effects of trust and suspicion on decision-making. There are social situations which, in a sense, do not allow the possibility of "rational" individual behavior as long as the conditions for mutual trust do not exist. ...Mutual trust can occur even under circumstances where the people involved are overtly unconcerned with each other's welfare, provided that the characteristics of the situation are such as to lead one to expect one's trust to be fulfilled?(Deutsch, 1958: 278f). -70- 2.08-04 Two of Deutsch's (1958: 268f) early hypotheses assert that: (a) As the individual's confidence that his trust will be fulfilled is increased, the probability of his engaging in trusting behavior will be increased and (b) as the ratio of anticipated positive to anticipated negative motivational consequences increases, the probability of his engaging in trusting behavior will be increased. These two hypotheses may be related to an extension of Romans's (1961) general social interaction model. Using constant coefficient, linear differential equations, Simon (1954) derives his "Berlitz" model from a consideration of the problems of motivation and language learning. The correspondence among the Berlitz (b), Romans (h), and Deutsch (d) models can be seen in their principal variables: Tb= amount of time spent in language learning per day?Th= intensity of interaction or communication among a group Td= trusting behavior Ib= pleasantness of learning?Ih= amount of friendliness or group identification Id= confidence that trust will be fulfilled Wb= ease of learning?Wh= total amount of activity of a group member?Wd= ratio of anticipated motivational consequences The similarities among the mathematical Berlitz model, Romans's general social interaction model, and a model based on Deutsch's trust hypotheses do not prove that any of the three is correct. However, the multiple convergence does give added support to each. From his studies, Deutsch (1962: 316) finds that in the initiation of cooperation, the problem of trust and the bargaining problem are much easier to resolve if the potential cooperators are interested in each other's welfare. The development of a mutual interest in one another's welfare is, in turn, fostered by the experience of successful cooperation. He finds that expectation, intention, retaliation, and absolution are key elements in a stable cooperative system (1958: 273). These, in turn, may be related to problems of threat and yielding. The introduction of threat into a bargaining situation affects the meaning of yielding¥¥¥¥The cultural interpretation of yielding (to?a peer or subordinate) under duress, as compared to giving in without duress /is different./ ¥¥¥The locus of causality is perceived to be outside the voluntary control of the individual (Deutsch &Krauss, 1962: 74). -71- 2. 08-05 Loomis (1959: 305) in his work with games of trust, found four conditions as essential for the establishment of a cooperative relationship based on trust. First of all each individual must be committed to reaching some goal where the commitment is such that failure to achieve the goal would cost him more than he would be willing to risk in an uncertain venture. Secondly, each individual must know that he cannot reach the goal without the help of other persons, and that they are able to help him. Thirdly, each individual must know that the other persons are similarly dependent upon him for their rewards. ...Fourthly, each individual must know that the others are each aware that the members are all mutually interdependent; in other words, there must be mutually perceived promotive interdependence. The effects of different individual motivations in games were investigated by Solomon (1960: 230). He defined a conditionally cooperative strategy as one where the first player makes a cooperative choice on the first trial. However, for each succeeding trial, the choice of the first player is reciprocal to?that of the second player: the first playerámakes a cooperative choice only when he receives one from the other player. Solomon found that a conditionally cooperative strategy is most likely to facilitate the transformation of an individualistic to a cooperative orientation, particularly under conditions where there is a high degree of mutual interdependence. Terhune and Firestone (1966, 1967) have studied the relation of personality and group motives to cooperation and conflict using a Prisoner's Dilemma game and the Princeton Internation Game.2b Based on McClelland's (1961) typology, they define three motives in which only one motive is dominant: Naff (abbreviation of need for affiliation): a person above the median on need for affiliation, below the median in the other two. Napa (need for power): a person above the median on need for power, below on the other two. Nach (need for achievement): a person above the median on need for achievement, below the median on the other two needs. From their results with the Prisoner's Dilemna game, Terhune and Firestone (1966:á 40) found that Naches are interpersonally neutral risk-takers, motivated by a self- interest, and desiring to conform to a standard of excellence. Naffs are interpersonally positive, desire friendship with others, but are highly defensive. Nepos are interpersonally negative and strive for control over others.?The difference between individual behavior and group behavior may be seen by comparing these results with those obtained from the Princeton Internation Game (1967: 90): 2b The Prisoner's Dilemma is a situation in which both "players" have dominant incentives to choose strategies that together yield both a less desirable outcome than if both had made opposite choices (Schelling, 1960: 214). The Princeton Internation Game is a human simulation of international relations. -72- 2.08-06 It appears that Nach groups are cooperatively disposed, but that their choice of goals may lead them into conflict. Naff groups desire friendly relations, but their withdraw! tendencies lead them to both minimal cooperation and minimal conflict. Nepo groups are oriented toward controlling others, which gives them a strong potential for conflict, though they may, through subtle exploitation, find outlets for their power needs in cooperation as well. In the group game, the more complex environment allowed the naffs to withdrawl, an alternative not available to them in the Prisoner's Dilemna games, and allowed the nachs and nepos a choice of goals and of means to reach these goals. Since Deutsch's experiments on trust, there have been a series of other investigations using the Prisoner's Dilemna and similar games to study trust (Behavioral Science, passim; Dissertation Abstracts, passim). However, as illustrated by the findings of Terhune and Firestone, there are difficulties?in generalizing even from dyadic to group laboratory games(cf. Nemeth, 1970). Althoug simulations may be closer to the "reality" to which the research findings aim?their generalizations, the external validity of any simulation must be?systematically examined (Guetzkow, 1968; Hermann, 1967). Two critical?differences between laboratory studies and day-to-day life are the time?perspectives and levels or risk involved (Shubik; 1963; Weick, 1965). The research subject does not live ;n the laboratory: his major goals reside outside. !While people do get involved playing an afternoon game for "Monopoly money,"á there are some behaviors that will not be elicited in normal individuals unless the stakes are higher and the effects of the game longer lasting. Risks involving one's career, thousands of real dollars, or another's physical and mental health are difficult to ethically simulate in the laboratory; they occur not infrequently in the "real world." As Simon (1959: 266) has succinctly stated, "while game theory has greatly clarified the issues involved, it has not provided satisfactory solutions." To help resolve this difficulty, studies of decision-making have been?done in ongoing organizations.3 Although a few organization theorists explicitly consider trust as a central variable (e.g., Argyris, 1965; Bennis, 1966; 3 of course, there are trade-offs to such a benefit. The researcher is limited to situations as they exist, and cannot readily replicate or create "alternative futures." Also, there will be many influencing factoq. ~!:tich?cannot be controlled. Difficulties of validity can be reduced from, !or example, the case study design, by using quasi-experimental, natural experimental, and true experimental designs (Campbell & Stanley, 1963). -73- 2.08-07 Crozier, 1964), the issues of trust and openness are found in most studies of individual and organizational behavior only implicitly. Two subliteratures in organizational decision-making quite related to trust are those of participative decision-making and studies of organizational conflict. 4 Participative decision-making. Participative decision-making may be broadly conceived as a mode of organizational operations in which decisions are arrived at by the persons who are affected, or feel they are affected, by them (after Lewin, 1968: 69). The organizational operations may involve the perception and communication of the need for a decision, the generation of alternatives and consequences, and the final selection. As Tannenbaum and Massarik (1950: 410) note, subordinates can participate in the first two of these steps; only the superior can make the final selection. However, with democratic group operations, there is no "superior," at least within the confines of the democratically defined group, and the group may participate in all stages of the decision-making process. One trust issue in participative decision-making is that of suffrage. For an administrative superior to reasonably relinquish some of his control to his subordinates or peers directly implies some trust of them, since he is increasing their possible influence over his work environment and over himself (cf. Lundstedt, 1966). Lewin (1968: 75f) lists the potentials of increased organizational performance, improved technical and administrative systems,?public pressures for subordinates to conform to prior commitments, and?greater goal commonality as motives contributing to manager acceptance of participative decision-making. He finds ego motivation, financial incentives, and closure and the sense of participation as contributing subordinate motives. In sum, for both superiors and subordinates to leave the traditional hierarchical decision-making procedures for those of participation, they must feel that they will better achieve their goals. 4There is also a substantial decision-making literature in political science. Snyder has provided a general statement of the relevance of game theory (1955) and decision-making (1958) to the study of political phenomena; Snyder and Robinson (n.d.) give a comprehensive report of completed and needed studies in national and international decision-making. Decision-makingástudies in other areas of political science have also been made. For example, Schubert (1958) has analyzed judicial behavior; Feldman & Kanter (1965: 643) inventory community decision-making studies; and Angus Campbell, Converse, Miller, &Stokes (1960) report on their research on the American voter. -74- 2.08-08 After participatory decision-making is introduced into an organization, the formal suffrage patterns may be more the dependent variable than the independent. Both superiors and subordinates are likely to adapt their communication and decision-making patterns depending on their motive satisfaction. For example, if subordinates' perception of the need for a decision are often denied, if subordinates do not receive timely feedback on their suggestions, or if they feel they do not receive proper credit for their contributions to decision-making, they are likely to withdrawal to the traditional, hierarchical, passive role. Likewise, if the superior feels that his subordinates are ineffective or uninterested, he is not likely to facilitate their participation. And if the manager tries to increase his status or security by keeping significant information from his subordinates, their potential contributions are rendered less effective. Thus, it is not so much the formal statement by management that participatory decision-making will be employed that will determine the behavior of managers and subordinates; it is the ongoing behavior of each that contributes to the establishment of dyadic equilibrum (cf. Heller & Yukl, 1969). For participatory decision-making to work, superiors must be open to communications from their subordinates. They must admit that others may be more aware of some of the proper questions, and that others may be better prepared to consider alternative answers. And subordinates must trust their superior to not punish, consciously or unconsciously, them for questioning his omnipotence, omniknowledge, and omnipresence; in short, for questioning the traditional, beautrratic, hierarchical values. Based on his review of the literature (include experimental studies in nonorganizational settings, observational studies in organizations, and experimental studies in organizations), Lowin (1968) proposes several hypotheses concerning the effectiveness of participative decision-making. Some of these hypotheses find the effectiveness of participative decision- making to vary directly with the number of administrative levels subsumed by the program and directly with the amount of useful information available exclusively to the subordinate. Effective participative decision-making is seen to vary inversely with the amount of useful information not available to the subordinate, with the urgency of decisions, with the extent to which subordinates can exert no control over productivity (for example, in some machine-paced operations), and with the degree of coupling of financial rewards with participative decisionámaking activities, provided that financial motives are important. -75- 2.08-09 In light of these hypotheses, it con be useful to examine decision-making?in research and development laboratories. Traditionally, decision-making has?been considered a managerial function separate from the function of the operatives of the organization (e.g., Simon, 1944). In this view, the operatives do the work--turn the lathes, shovel coal, fire rifles--and the administration, through its decisions, influences the operatives¥ behavior. But research is itself?a sequential decision-making process, and the function of organizational R&D?is to find scientific knowledge which will be useful to the organization under the environmental conditions with the greatest expected value. Thus, in an R&D lab, or any organization where the final product can be considered as a decision, decision-making is a double purpose model: it can productively be applied to both administrative and operative behavior. Project selection in R&D is an activity that embraces much of the R&D function and to which the decision-making model readily applied (POMORAD, 1966, 1967, 1970). It is through project selection that the R&D technology-- that is, the research process--is applied to an initial idea, thus increasing the probability that the idea will lead to knowledge and products useful to the organization. General phases in the idea flow and project selection process include idea generation, preliminary discussion, revision, formal proposal, initial formal screening, economic evaluation, other evaluations, decision to enter into the portfolio, decision to fund, personnel assignment, in-progress reviews, periodic reviews, and final review. Many of these project selection decisions are not urgent ones. Project portfolios and budgets are often formally prepared with a year1s lead time; a specific research project can take several months or several years. Also, the researcher can exert a great deal of control over his productivity, at least in the intermediate sense. Although he cannot guarantee a solution to his problem, he can work harder and longer at it, including working or thinking when he is not on "company time." Since much of the research process occurs within the researcher's head, in addition to these conscious behaviors, his unconscious motivations exert additional control over his productivity. The administrative subordinate--that is, the researcher--has a great deal of information available exclusively to him. It is he who has direct experience with his research project. While he can, and often is required to, report on his research to management, there can be wide individual differences in the frequency, detail, and completeness of such reports. In addition, such reports are general. For any specific decision, the requisite information may not be in the report but may reside with the researcher. -76- 2.08-10 These factors--nonurgency, control over productivity, and exclusive information of the subordinate--all indicate that participative decision-making could be effective in a R&D lab. Of course, other factors still operate, and also influence potential effectiveness. If the general climate of the laboratory is permeated by traditional hierarchical or pyramidal values (Argyris, 1965, 1966), then participative decision-making in one group or section of the lab will be less effective. Also, important information such as company-wide priorities, long-range goals and marketing plans, and proprietary and confidential information may not be directly available to researchers, but only to their superiors. For researchers to make less suboptimal decisions, they need to know the goals, values, and some data of other organizational entities--other entities competing for the organization1 s scarce resources. Conflict. As with decision-making, there is an extensive literature pertaining to social conflict (e.g., Journal of Conflict Resolution, passim; Mack & Snyder, 1957; Pondy, 1967, 1969; Walton & Dutton, 1969; Wolfgang, 1966). This literature will not be reviewed here, but the relation of trust and openness to conflict will be discussed. Conflict is inevitable in organizational affairs.5 Even if there were agreement as to goals among cooperating individuals and groups, resource constraints will create conflict in priorities and means. In addition, due?to their individual differences in genetic coding, early learning experiences, resultant personalities, and current situations, people will vary in their values and perceptions. March and Simon (1958: 121) find such differences in goals or in perceptions of reality to contribute to intergroup conflict, provided there is a felt need for joint decision-making. The division of labor and specialization found in organizations will create some needs for joint decision-making. They will also create group subgoals and vested interests. 5 All conflicts are not inherently destructive. Deustch (1969: 10) distinguishes between destructive and productive conflict. A conflict "clearly has destructive consequences if the participants in it are dissatisfied with the outcomes and all feel they have lost as a result of the conflict." If they are satisfied and feel they have gained, the conflict is productive. Destructive conflicts tend to expand and escalate. They invoke competitive strategies, misperception, lack of communication, and excessive tension, which "reduces the intellectual resources available for discovering new ways of coping with a problem or new ideas for resolving a conflict" (at 15). Productive conflict may be seen in individual creative thinking and, at the social level, in cooperative group problem-solving. -77- 2.08-11 These factors, together with the organization1 s at tempts to adapt to its changing environment, will produce conflicting demands and ambiguous requirements for behavior (Kahn, Holfe, Quinn, Snoek, &Rosenthal, 1964). There are two major reactions to conflict: open attempts at solution or defensive attempts at avoidance. As Kahn et al. document, defensive avoidance of the core problem often results in derivative problems, and vicious cycles can result. They further suggest (at 389) that to promote organizational effectiveness, "each member of an organization should have the legitimate authority to convene the members of his role set to reach understanding and agreement .....,6 As discussed in several sections above, such open attempts at conflict resolution necessitate trust. In most organizations, especially those with pyrimidal values, there are strong norms against admitting the need for help and against questioning traditional authority--i.e., authority of position rather than authority of knowledge and skill. Further, true attempts at conflict resolution, as differentiated from "winning the disagreement," open the individual or group to the possibility of external influence and change. The parties in the conflict must have faith that they will not suffer undue harm in¥the short-run and that in the long-run their needs and goals will, on the average, be met as often as those of their peers. Homans (1961: 232) has formulated a law of distributive justice: A man in an exchange relation with another will expect that the rewards of each man be proportional to his costs ... and that the net rewards, or profits, of each man be proportional to his investments .... To the extent that a man feels that distributive justice is not being met, he will display anger or guilt, depending on whether the failure is to his disadvantage or his advantage. One problem with distributive justice is that rewards, costs, and investments are all subjectively valued: one man's conception of his investments may differ from another's perception of them. If these differ~ng 6 In his review of marriage research and conflict, Barry (1970: 48) reports studies finding less residual hostility when subjects "are allowed to communicate their hostility toward the instigator of their frustration." Also, "the more hostility is repressed the more dangerous is the conflict when it breaks out." Cf. Thibaut & Coules, 1952. -78- 2.08-12 evaluations can be brought out into the open, it may be possible to reassess investments and to decrease conflict. (But there are some social functions provided by non-explicitness and ignorance, cf. Moore & Tumin, 1949.) Conflict situations may be perceived as win-lose, zero-sum games, or they may be seen as non-zero-sum games providing the opportunity for interacting individuals and groups to help one another contribute to the overall goals?of the organization, while causing minimal interference with each entity's?own idiosyncratic subgoals. If mistrust is prevalent, intra- and intergroup hostility and fighting can result, with defensive, closed positions, limited and distorted communications, and suboptimal decision-making, But if individuals and groups trust one another, they are more likely to engage in the kind of open problem solving that can increase the overall benefits to?all. -79- 2.09-01 2.09 Creativity and control Some lore has it that all one need do to have an effective R&D organization is to find technically competent scientists, equip them with their testtubes and other tinkertoy apparatus, and leave them alone until they discover nylon, atomic energy, the transistor, or nutritionless food. Organized research is not quite so simple. Three significant parameters influencing the relation between the research conducted in the laboratory and the lab's effectiveness are a) the scientist's motivation, b) his creativity, and c) the laboratory's control mechanisms. Motivationo March and Simon (1958: 48) consider a member of an organization to make two important decisions: a) the decision whether to participate in the organization and b) the decision whether to produce at an "appropriate" level. This appropriate level may be that demanded by the organizational hierarchy, by the work group, by some industry-wide peer group, or by the individual himself. They also note that "present satisfactions are often less important in influencing human behavior than perceived relations between present alternatives and future states." These propositions may be related to professionals in intellectual industries, and specifically to those in research and development.1 An intellectual industry is one concerned primarily with the manipulation of abstract concepts or ideas; such an industry may be contrasted with one which manipulates physical objects. The idea may be held by a human, but if the manipulation is primarily on the idea (as in education) rather than on the person (as in medicine), then the industry may be termed intellectual. Two prominent intellectual industries are research and education. Intellectual industries may be seen as a subset of "intensive technologies." In contrast with other technologies such as the "long-linked" and the "mediating," it is more difficult for the intensive technology to seal itself off from 1 Stinchcombe (1965: 162) finds that the entire nature of "labor" is changing; it is increasingly the "production, dissemination, and application of knowledge rather than energy." He further comments that "the system of rewards and punishments appropriate for maximizing the amount of knowledge and intelligence that will be applied in a work role are likely to be quite different from those appropriate for maximizing the amount of energy applied. In particular, the organization of work roles into a career is much more important, and the -80- 2.09-02 uncertainties in its environment.2 Thompson (1967: 40f) proposes that long- linked technologies seek to reduce uncertainty through vertical integration, mediating technologies through increasing the populations served, and intensive technologies through incorporating the object worked on. While incorporation of the object is possible in the construction industry and in medicine, such incorporation is difficult in an intellectual industry: the intellectual industry does not work directly with physical objects, but with ideas. In education, the manipulation of ideas is governed by laws of learning. Unfortunately, all of the relevant parameters for these laws, and some of the laws themselves perhaps, are not known. In research however, even the degree?of certainty present in education is absent. Research involves a deliberate attempt to innovate, and such innovation inherently involves uncertainty.?Because of the inherent uncertainty in research, the measurement of productivity is more difficult and less absolute than in other industries. In an industry where most of the laws and many of the relevant parameters governing production are known, the work of each individual may be measured against the theoretical production for his position (e.g., assembly-line work, typing, dishwashing, etc.). But where the theoretical production function is not known, an individual's production can only rationally be evaluated with reference to some average of his peer1s work. This relativity of performance measure can influence the individual's decision both to participate and to produce. Entry into an occupation in an intensive technology is usually through specialized training (Thompson, 1967: 112). This is especially true of intellectual industries, which characteristically involve "late-ceiling occupations"--that is, occupations in which the individual can find advancement until rather late in his career. Our society will have an increasing demand for professionals (Beckhard, 1969; Bell, 1967; section 2.11). While general fluctuations in economic conditions will influence this demand, many professionals will often be able to exercise some choice regarding their participation in one organization rather than another. 2 A long-linked technology involves serial interdependence. A mediating technology involves the linking of clients or customers who are or wish to be interdependent. In an intensive technology, a variety of techniques is drawn upon in order to achieve a change in some specific object; but the selection, combination, and order of application are determined by feedback from the object itself. (Thompson, ~967: lSf) -81- 2.09-03 But the professional will still have to decide at what level to produce. One may expect to find high levels of need for achievement among professionals (McClelland, 1961). To the extent that an employee has high need for achievement, he is likely to base his decision of level of production on his own abilities and is likely to produce at a high rate, considering his own capabilities. (The strength and direction of social pressure, as well as the employee is level of need for affiliation and need for power will also influence his production.) Thompson (1967: 114) hypothesizes that "where the intensive technology employs late-ceiling occupations (professionals), the inducements/contributions bargain rests on the individual1s visibility among occupational colleagues." One may expect that in the near future the highly qualified professional will have little trouble in obtaining employment, will produce at near his maximum making his high skill level visible, and consequently will receive a favorable inducements/contributions bargain. The less qualified professional has more problems. If he is high in need?for achievement he should produce at near his maximum and should be willing to?make visible his level of skill. (Again, social pressures are important?additional influence.) However, the individual who has lower need for achievement, but high need for power, may not wish to be judged on the basis of his skill.?This individual is much more likely to spend a sizable proportion of his effort?at trying to appear comparatively better than other individuals. Part of this effort may be directed at his own productive work, but part of it is as liable?to be directed at impeding the progress of others. If the work environment is individually competitive, recognizing only relative contributions, there is little to induce the less qualified employee?to devote his efforts primarily toward making those contributions of which he?is able. However, if the work environment is democratically open (Bucklow, 1966) the absolute contribution of each individual is recognized. In this case, it is not solely the individual1s level of production, but his level of production compared with some estimate of his total ability. -82- 2.09-04 The Scanlon Plan (Krulee, 1955a, 1955b; Yoder, 1962), Bavelas's work (Whyte, 1955: chap. 10), and related group dynamics techniques may be seen as attempts to establish "democratic" environments for work groups. These techniques, as well as those of sensitivity training and organizational development (see sections 2.03 and 2ol0) may be seen as methods to free the individual from inappropriate competitive fears and to enable him to produce at an individually appropriate level. Creativity. Even if the researcher is motivated to direct bis efforts at solving the technical problems, rather than directing significant portions of his energy toward defensive or inappropriately competitive behaviors, he must still have the technical competence to produce. Laboratories have two major controls over the technical competence manifest by their members: selection of personnel and management of the organization. Creativity is generally considered an important component of the researcher's technical competence. Dellas and Gaier (1970: 68) have reviewed research on the identification of creativity in the individual. They find that "creative persons are distinguished more by interests, attitudes, and drives, than by intellectual abilities." The cognitive capacities that appear to be most frequently associated with the creative are an above-average intelligence and the effective use of this intelligence, the ability to produce unusual and appropriate ideas, an exceptional retention and more ready availability of life experiences, ideational fluency and the ability to synthesize remote or disparate ideas, discriminative observation, and a general cognitive flexibility. In the realm of personality, a clearly differentiating factor that characterizes the creative is the relative absence of impulse and imagery control by means of repression. This relative lack of self- defensiveness seems to accord to the creative fuller access to his conscious and unconscious experience, and therefore, a greater opportunity to combine dissociate items. An openness to internal and external stimuli is also indicated as a salient characteristic .... Although the creative appears to be subject to considerable psychic turbulence, ... the creative individual is possessed of superior ego strength and a positive constructive way of reacting to problems. ... Iindependence in attitudes and social behavior emerges with striking consistency as relevant to creativity.?Other res~archers in creativity have arrived at similar findings (e.g., MacKinnon, 1967; Torrance, 1965). However, a laboratory will not necessarily be effective even though its personnel are "creative" individuals.3 More than novelty is needed (Jackson & Messick, 1965). What is áneeded are men with creative ideas that can economically be turned into marketable products.4 -83- 2.09-05 While studies on the identification of creative individuals can be of some aid in designing educational systems (Kagan, 1967) and in personnel selection, it is through the management of the organization that significant improvements in the laboratory's creativity may be fostered. 5 3 Other personality characteristics also affect the researcher's effectiveness. Matussek (1962: 24) discusses the influence of personality on research planning. After describing four idealized clinical, but not abnormal, types (compulsive, depressive, hysterical, and schizoid), he concludes "that the multifariousness of personality structures is responsible for the fact that for a given assignment a particular personality type is best suited and for another a different one; that even a perfectly planned project can be impeded by one personality type and advanced by another." Besides the effect of personality on planning, Matussek is also concerned with the effect of planning on the individual. He finds that the more a researcher "can express the main springs of his personality the more he can be human in his work, the better will be his work." Wainer & Rubin (1969: 178) and Schrage (1965) have studied the personality of R&D entrepreneurs. Wainer & Rubin found high need for achievement and moderate need for power associated with high company performance. Need for power and need for affiliation seemed to affect performance through their influence on leadership styles. They see their findings as casting doubt on some of Schrage's results. Schrage found successful R&D entrepreneurs to be "high in achievement motivation, low in power motivation, and high in awareness of self, the market, and his employees" (at 64). However, quite unsuccessful entrepreneurs differed only in lacking awareness. This awareness included both accurate perception of others and self awareness, which was conceived as including awareness of one's impaired performance under stress and the ability to experience anxiety without debilitation. 4 Vonnegut (1963: 17) characterizes a Nobel Prize winning scientist in his acceptance speech. "I stand before you now because I never stopped dawdling like an eight-year-old on a spring morning on his way to school. Anything can make me stop and look and wonder, and sometimes learn." However, this was a pure researcher, and "pure research men work on what fascinates them, not on what fascinates other people" (at 41). A parallel problem to creativity?exists with the diffusion of innovations and the transfer of technology. In?all these areas the critical elements of usefulness and implementation are?often overlooked. 5 The idea flow and info search projects of the Program of Research on the Management of Research and Development (POMRAD) at Northwestern University have been concerned both with the generation of ideas and organizational facilitators and barriers to their communication (POMRAD, 1966, 1967, 1970). At the University of Utah, Calvin Taylor has been associated with a program of research on creativity. See~ for example, Ellison, McDonald, James, Fox, &Taylor (1968); Taylor (1956, 196~); and Taylor, Ghiselin, & Yagi (1967). -84- 2. 09-06 Laboratories generally already try to attract the most creative scientists and engineers they can; this resource pool is fairly inelastic to direct manipulation. But the management of a laboratory can try to create the kind?of organizational climate that will permit more of the creativity that does exist within its members to be used. (In addition, such a climate may attract more creative personnel or may alienate them less frequently.) Humanistic psychology (see section 2.02) would hold that most individuals are capable of creative acts, but are often inhibited by their own defensive mechanisms and by the social situations in which they live. Some research findings concur with this view. Hall and Lawler (1969) gathered data from the directors of 117 R&D organizations and interviewed 291 researchers in 22 of the labs, all in Connecticut. They found several areas in which the human resources of the laboratories were underutilized. In addition, they found that the organizations exerted negative socialization; the labs were training out many of the creative qualities they said they desired. Concerning motivation, Hall and Lawler (at 342) postulate that work can be intrinsically satisfying if three conditions for psychological success are met: a) the individual participates in setting his goals, b) he independently works out the means of attaining these goals, and c) the goals represent a truly challenging level of aspiration. These three elements are also seen as prerequisites for a proper integration of individual and organizational goals. Generally, the Hall and Lawler results support those of Argyris (1965:?230f), who postulates a theory of the deterioration of research and development organizations. Even assuming that researchers are technically competent, Argyris sees their interpersonal competence as lower than that for members of other areas of the organization; this lower interpersonal competence will interfere with the researcher's effectiveness.6 The relatively low interpersonal competence will lead to a lack of awareness and misperceptions in the laboratory, resulting in an organization increasingly tied up in defensive activity that no one wants or seems to be able to alter. Thus, both managers and researchers will come to behave in ways that tend to inhibit, dissatisfy, and frustrate one another. They do this because they are not interpersonally competent to do otherwise, because they do not know other ways of behaving. 6 The higher the interpersonal competence of an individual, the greater is?a) his awareness of relevant problems, b) his ability to solve problems in such?a way that they remain solved, and c) the probability that the problem-solving process involved has not been harmed or negatively influenced (Argyris, 1965: 4). -85- 2.09-07 One possible, but premature, conclusion from such studies is that R&D labs are overcontrolled. If researchers lack involvement and psychological success, let them determine their own research projects.7 If they lack interpersonal competence, let them work on their own. However, this straw man is readily reduced. Although individual research may be a useful mode for exploratory projects where specific goals cannot be outlined in advance, team research can increase the effectiveness of each member (Abelson, 1965). Also, many projects (for example, those in aero-space associated with NASA's space exploration) are too complex for any single individual. In addition, avoiding the problem of interpersonal incompetence also avoids the potential of increasing the laboratory's, and each individual's, resources; it is a suboptimal solution at best. Interpersonal competence can be increased, and psychological success can?be made more likely. Building on discussions in earlier sections of trust and openness, section 2.10 will include comments on increasing interpersonal competence. Section 2.08 discusses some of the issues involved in participative decision-making, but goal setting in an R&D lab involves more than letting?the research set his own challenging goals. It involves the entire issue?of control. Control: theoretical orientations. Much of organizational behavior may be described as the flow of decision premises through a (spacial-temporal) network of uncertainty absorption.8 Using an information-processing model, these decision premises may be seen as input to or as output from some organizational subsystem (individual, group, division, etc.). When considered 7 A related issue is that of job enlargement. In their review of the literature, Hulin & Blood (1968) find the effectiveness of job enlargement to be dependent on the background of the workers. Ford and Borgatta (1970) have identified eight attitudinal dimensions of satisfaction with the work itself. Also, Porter, Alderfer, Lawler, &Hackman (1969) discuss the influence of task factors on job attitudes and behavior. 8 "Uncertainity absorption takes place when inferences are drawn from a body of evidence and the inferences, instead of the evidence itself, are then communicated" (March & Simon, 1958: 165). -86- 2.09-08 as input they may relate to objectives or to cause-effect relationship beliefs; when considered as output they may or may not be used in feedback control mechanisms. 9 These three aspects of control--objectives, cause-effect beliefs, and feedback mechanisms--are central in a discussion of organization in an industrial research and development laboratory. These general aspects may be rephrased as questions of mission, method, and magic.10 Thus, an extreme of non-autonomy 11 would involve another subsystem determining 12 for the focal subsystem all of the focal subsystem's objectives and cause-effect beliefs (prescribed roles may be included in this category) as well as the form and timing of the control portion of the output. The superior subsystem thus determines the entire characteristic equation for this focal subsystem: 9 Note that often lower objectives are derived from higher objectives via cause-effect beliefs. In addition, often the flow from broad objectives to operating objectives passes through a network of uncertainty absorption. What is termed mission or method will depend on the level of analysis used; that in turn on the specific purpose of that specific analysis. 10 Brooker (1967: 6) has defined magic as "those pattern~ of behavior in which conscious attempts are made to manipulate variables in the environment but which include no means of comparingáresults with intents in order to?modify the manipulative behavior." Formal organizations, especially industrial organizations, and their subsystems are considered as definitely making conscious attempts to manipulate environmental variables, usually to result in survival and profit, as well as other states. While this overall objective?is broken down into smaller operating objectives, the connection with magic?or science (briefly, adaptive behavior) still holds. Thus, "magic" questions which subsystems compare the focal subsystem output with objectives (cf. Brooker, 1967: 13). The term "magic" is used for two reasons: it maintains the aliteration and it highlights the distinction between possible and actual feedback control. Alternatively, terms such as empiricism, accountability, or audit, could be used, although these terms emphasize the "science" end of the magic-science continuum. 11 Using a tensor notation, control or non-autonomy can be seen as the relationship V(i,j,k,t), where V has values for objectives, cause-effect beliefs, and comparison feedback. If there is no V(i,k) relationship, then?i can be considered autonomous, or autonomous within the restraints of j and t. However, to the extent that V(i,j,k,t) relationships exist, i is under some control. 12 "Determine" is used when a subsystem makes something happen; "specify" is used when a subsystem tries to make something happen. A determined event will be implemented, a specified event may or may not be implemented. -87- 2.09-09 input, transformation, output, and control loop.13 Of course, the superior subsystem cannot determine all the objectives and beliefs of people, but it can specify these for the rational, behavior aspects of work which the individual will consider legitimate.14 It should be noted that what a subsystem will regard as a legitimate directive is a variable factor affected by a complex of other variables.?At the other extreme, an autonomous subsystem would determine for itself its characteristic equation: it would form its own objectives, develop means to achieve these ends, and compare its output against its objectives to modify its own operations. This is not to say that such an autonomous subsystem is independent15 of other subsystems. Although other subsystems may affect this autonomous focal subsystem (e.g., by altering the focal subsystem's environment), the control aspects of the focal subsystem (program and feedback) are self- determined. Figure 6 shows a general subsystem (entity k=l) and another subsystem (k=2) attempting to control the focal subsystem. In the focal subsystem, U is the universe as presented to the individual's preconscious; I is the individual1 s net input to consciousness; 0 is his behavior output; G is his characteristic equation; G1 is his conscious transformation of total distal input; and H is an operator determining what part of his output is used to compare against proximal input (I) to determine if his objective has been reached. ~ indicates that the subsystem has an active energy source; the "foot" indicates that the subsystem can "move," and thus alter its distal input (U). Control determined by k=2 is shown by dotted lines; control specified by k=2 operates along with other influences in U. G1 may be related to mission, G to method, and H to magic. 13 A characteristic equation of a subsystem describes the output produced given an input. The terminology employed in this section and its diagram are consistent with formal cybernetic standards. (However, the "foot" in Figure 6 is a nonstandard symbol.) See, for example, Benjamin C. Kuo, Automatic Control Systems. Englewood Cliffs, N.J.: Prentice-Hall, 1962. 14 cf. Katz &Kahn's (1966: 204) discussion of .the psychological contract members of organizations accept. A superior subsystem could also specify behavior, or even affect and perception, for which the individual does not grant legitimacy. In this case the specified behavior would not be determined, and likely would not occur. 15 Charles F. Douds introduced me to Jacques (1956) notion of dependence, roughly measured by how long it will take for a change in another subsystem to affect the focal subsystem,á and how strong the effect will be. -88- Figure 6. Components of E: Generalized Subsystem (k=l), Showing It Under Some Control of Another Subsystem (k=2). A general subsystem (entity k = 1) and another subsystem (k = 2) attempting to control the focal subsystem are shown. In the focal subsystem, U is the universe as presented to the individual's preconscious; I is the individual's net input to consciousness; O is his behavior output; G is his characteristic equation; G' is his conscious transformation of total distal input; and H?is an operator determining what part of his output is used to compare?against proximal input (I) to determine if his objective has been reached. ~ indicates that the subsystem has an active energy source; the "foot" indicates that the subsystem can "move," and thus alter its distal input (U). Control determined by k =2 operates by dotted lines; control specified by k =2 operates along with other influences in U. -89- 2.09-10 The importance of these aspects of control (mission, method, and magic;?or program and feedback) in organizational behavior need not be discussed at great length. Obviously, how the organization's subsystems are controlled 'vill largely determine the adaptiveness and effectiveness of the organization. Control and freedom in R&D. There is a basic tension in organizations between freedom and control (Blau & Scott, 1962: 242f). Thus, while free communication flow facilitates problem-solving and decision-making, it can hinder coordination. While a researcher is likely to be highly motivated for those projects he chooses himself, these projects may not contribute to the organization's effectiveness. The individual researcher, project group, and laboratory need freedom to involve themselves in their work, but they also need control to ensure that their work will contribute to the goals of the organization (cf. Hughes, 1968; Smith, 1970). Based on the characteristics of organizations in general, of R&D labs in particular, and of the creative, innovative process, several patterns of effective control mechanisms can be postulated. These relations are not hypotheses to be tested in this study, but are indicated as potential points of leverage and change. For any subsystem, there are patterns of autonomy that are more effective than other patterns.16 A pattern of autonomy is the degree of autonomy the focal subsystem has in its network of interrelationships. It may be specified as a V(i,j,k,t) matrix, where V refers to the three factors of autonomy--mission, method, and magic.?Thus, the relative specification and determination of program (means and methods) by R&D or by manufacturing, marketing, and top management would have to be considered. Also, under consideration would be the reports that R&D must make that ~used to determine later R&D operations (e.g., project effectiveness summaries to receive financial support next year). Reports that are not used are a part of program, not of feedback: they are magic. Note that some operational objectives 16 Litterer (1970) discusses the necessity for R&D to have a structure different from that of traditional bureaucratic production organizations. His discussion uses results from studies of Lawrence & Lorsch (1967). They find á that subsystems develop differentiation related to their specific requirements and subenvironments. Further, in high producing organizations, integrating devices were more effective. -90- 2.09-11 may be, in an objective, causal sense, unrelated to main objectives. Thus, feedback at one level becomes magic at the next level. Assuming technically and interpersonally competent personnel, a high degree of self-organization is both feasible and effective. Note that this does not mean that the self-organized subsystem reports to no other subsystem or that its objectives and programs are independent of other subsystems, but that the focal subsystem itself determines these lines of control. While at first it may appear that it is inconsistent to speak of?an autonomous subsystem under control aspects of another subsystem, it should?be remembered that the autonomous group has the power to alter or sever these control aspects. For example, while an R&D group may ask marketing for research projects (objectives), as long as R&D can determine on its own criteria when to accept which projects, then the R&D group will be autonomous in this respect. The term "control aspect" is used to indicate this voluntary means of organization. "high degree" may be taken as an absolute measure, or?relative to most current practice. Three more relations involving autonomy may be postulated. Assuming technically and interpersonally competent personnel, if?only one of the three factors can be relatively autonomous, it is most effective (and most feasible considering current organizational power distributions) to have the methods factor (cause-effect beliefs) autonomous in an R&D group. Assuming technically and interpersonally competent personnel, the feedback factor of control (magic) will be the least feasible to make more autonomous. Assuming technically and interpersonally competent personnel, a non-extreme amount of autonomy with respect to feedback will yield a more organizationally effective subsystem than either extreme of autonomy. Subsystem operational objectives are liable to drift away from organizational objectives if some feedback control is not exercised. It is likely that feedback control will be coupled with superior objective setting. (an instance partially without such coupling would be employee evaluation where the employee sets his own career goals. Likewise, free-time research may be evaluated, but the researcher may choose any research topic.) An excessive number of reporting requirements, or even a few, may degenerate into magic, rather than being used to compare output and modify subsystern behavior. -91- 2.09-12 The case of mixed magic-science (feedback) deserves special attention. In this instance, a superior subsystem may have a certain objective A (e.g., keeping a strong power position) and may institute a reporting' program for a subordinate subsystem. However, the superior subsystem would explicilty announce an objective B (e.go, conducting good research, profitable to the firm). In this case, as viewed from the subordinate subsystem the reporting system is magic, as viewed from the superior subsystem and the actual (rather than announced) objectives it may be scientific (feedback). A non-extreme amount of feedback autonomy balances the tendency for superior subsystems to develop operating objectives (or comparison criteria) at deviance from the overall organizational objectives with the tendency for subordinate subsystems to deviate. These three control aspects of organizational behavior--mission, method, and magic--are an important influence on individual and organizational effectiveness. In any study of organizational behavior, it may often prove useful to obtain at least some data on the control matrix. In industrial R&D, control variables may be expected to play a central role. The main operating objectives of an industrial organization usually revolve around short-run profit. Long-run aspects of organizational objectives are often served by the R&D laboratory. (Of course, top management also has long-range functions.)?By a Gresham's law of planning, short-run, programmatic behavior often becomes dominant over longer run, unprogrammed behavior. Thus, an industrial R&D lab may be dominated by other departments of the organization. This domination of R&D may lead to laboratory and organizational control procedures which have adverse effects on the behavior and performance of the lab and, in turn, dysfunctional consequences for the rest of the organization. When control procedures are employed which consider the nature of the task at hand and of the people working at that task, laboratory and organizational effectiveness should be improved. The author has proposed, above, that a decrease in industrial magic and an increase in adaptive control and behavior are important aspects of such increased effectiveness. Further, it has been suggested that an atmosphere of trust, rather than of defensiveness, is conductive to proper organizational control. Keeler's (1966: 19Sf) study of eight different laboratory environments supports some of the above postulates. He concludes that "an organization must have elements of both control and freedom to get meaningful creative activity." -92- 2.09-13 More specific conclusions include: a) centralization of decision-making reduces the opportunity and motivation for creativity; b) high levels of internal communication, especially through informal channels are conducive to increasing motivation to be creative; c) high levels of coworker empathy foster creative activity; d) high levels of empathy and warmth of interpersonal relationships give psychological protection to an individual in an unpredictable environment. This protection fosters creativity. e) responsible leadership is a desired organizational control mechanism that is function to creativity. Summary of ~ effective R&D climate. Based on the general organizational considerations discussed so far, and on the specific situation of research and development laboratories, several conclusions can be made concerning what an effective climate for an R&D organization would be. To begin with, technically competent personnel are needed. In addition, creativity is desired. Some of the significant characteristics of the creative person are a general lack of defensiveness, an openness to internal and external experience, high ego strength, tolerance of ambiguity and complexity, and independence. While technical competence is a basic necessity, it is not a sufficient condition for an effective lab. Even having "creative" individuals is not sufficient, for the management and climate of the organization can significantly modify, either favorably or unfavorably, the behavior of its members. So that the problems the researchers work on will contribute to the organization's effectiveness, there must be effective communication of the organization's goals and environmental constraints. This will generally imply some degree of feedback control of researchers by laboratory management, and of the laboratory by the parent organization's management. In any organization's operations, conflicts will arise. The greater the interpersonal competence of the lab1s personnel, the more effectively they will be able to resolve such conflicts. Interpersonal competence will allow consideration of both the technical problem and the feelings of the individuals involved. -93- 2.09-14 To facilitate researchers' motivation, participative decision-making?can be employed. To permit greater amounts of scientists' energy going?into productive work, defensive attitudes and behaviors need to be controlled and minimized. An atmosphere which permits trust and openness to develop?and which fosters their growth will thus tend to promote effectiveness.17 Throughout the organization, trust and openness are the processes that allow inquiry, reality-testing, and adaptation. 17 such a laboratory climate is similar to that of some university research groups. Grinnell (1967), Ritti & Goldner (1969), and Shepard (1954) discuss some problems and successes with such groups. Trust and openness can be even more critical for newly formed project teams or research groups. As Stinchcombe (1965: 149) comments,á"new organizations must rely heavily on social relations among strangers. This means that relations of trust are much more precarious in new than old organizations .... Grinnell (1967) discusses how Creative Interprofessional Collaboration--which is characterized by a high degree of trust; mutual respect, commitment, and confrontation; and situational adaptability-- can facilitate interprofessional research. -94- 2.10-01 2.10 Organizational development This dissertation reports the results of a multi-goaled research project. Two of the goals were to understand better the role of trust in organizational behavior and to test explicit hypotheses involving trust. Another set of goals involved organizational development. These included testing the applicability of the research procedures for an ongoing system of "humnan resources accounting" and the relevance to the participants of the chosen measures. Thus far, this chapter has focused on a discussion of trust. This section will discuss organizational development: what it is, its relation to trust and openness, and some strategic issues in organizational development (OD). What OD is: !-group beginnings. Organizational development grew out of?the attempt to apply T-group technology to change programs in organizations.?T-groups were acc~dentally discovered during a 1946 workshop on interracial?relations (Benne, 1964). The workshop participants, with research observers, spent much of their formal program time in small groups in which the training focused on the analysis of back-home problems. In the evenings the training and research staff had meetings during which the process observations of the groups were discussed. Conference participants began to attend these staff meetings, and became quite involved along with the staff in analyzing and interpreting group behavior. Thus, an evolving discovery began: Group members, if they were confronted more or less objectively with?data concerning their own behavior and its effects, and if they carne?to participate nondefensively in thinking about these data, might achieve highly meaningful learnings about themselves, about the responses of others to them, and about group behavior and group development in?general (Benne, 1964:á 83). Out of this original 1946 staff group grew the National Training Laboratories (NTL), an organization associated with the National Educational Association. NTL held annual training conferences and experimented with various technologies. The "T-group" (T for training) evolved as a small group training procedure in which the participants would utilize their own behavior in that setting ("here and now"'data) to help them learn about themselves and about the properties of groups. "T-group," "sensitivity training," and "encounter group" are differing terms applied to a similar learning situation--the intensive group experience (Rogers, 1967b). Ever since its beginnings in 1946, T-group "technology" has been experimented with and has grown to include structured experiences similar to theater improvisation (Spolin, 1963), psychodrama (Moreno, 1946, 1969), Gestalt techniques (Perls, 1969), nonverbal exercises (cf. Duncan, 1969; Mehrabian, 1969; Mill & Ritvo, 1969), and other -95- 2.10-02 procedures to help the participant learn. One of the goals of these potential ingredients in sensitivity training programs is to break through the barrier of intellectualization and verbalization to facilitate the participant experiencing his mm behavior and affect (cf. Gunther, 1968)¥1 Thus, áone of the major goals of the intensive group experience is to allow the participant a supportive environment in which he may experiment with new ways of behaving, and to facilitate his development of new, more effective, ways of perceiving his reality (cf. section 2.03). T-groups have other goals--explicitly value-laden, normative goals.?Salient among these normative goals are a spirit of inquiry, authenticity in interpersonal relations, expanded consciousness and recognition of choice, and the conception of collaborative authority relationships (Bennis, 1962a). The theory of T-groups does not state that everything should always be questioned, that constant awareness is universally good, that authenticity is uniformly beneficial, or that collaboration is the best working relationship. The theory does state, however, that there are times when these values and behaviors are appropriate. Since many people in our culture do not hold these values, the T-group is a time When the participant can experience behavior based on these open norms. Thereafter, he should be aware of his additional options and better able to choose when they are appropriate for him and his situation. Trust and openness are critical and central variables in T-group theory. Inquiry and expanded consciousness require openness to various forms of input, both from the self and from the environment. Inquiry and choice combine in individual responsibility; trust of self is required to accept that responsibility--to question, to choose, and to act. Trust of "the other" facilitates collaboration and the openness of authentic interpersonal relations. What OD is: Definition and goals. Most writers on and practitioners of organizational development have their own definitions of what OD is. Although the specific formulations may differ, there is a substantial agreement 1 "I knew a young man once who wanted to write a novel, and the first time he kissed a girl, he said to her anxiously, 'Now tell me in your own words exactly how that felt.' Later he gave up literature altogether in favor of experience. It was (he said) more interesting in itself, and a great deal easier to do." (John Bowen, New York Times, section VII, page 2, 4 September 1966.) -96- 2.10-03 concerning the core of OD. Consolidating many of these specific definitions (Bennis, 1969; Burke & Hornstein, 1970; Gardner, 1964), this author would propose the following definition of OD: Organizational development is an educational strategy employing experienced-based behavior in order to achieve a self-renewing organization. OD is based on behavioral science findings and hypotheses (What is?OD?, 1968). These generally stress the value--for both the individual and?for his work effectiveness in the organization--of confrontation, choice, and collaboration. From these initial theories, OD seeks to facilitate organizational members removing the barriers which prevent the release of human potential?within the organization. Specific process goals of many OD programs might include: creating an open, problem-solving climate; supplementing the authority of status with that of competence; building trust; reducing inappropriate competition and fostering collaboration; developing reward systems which recognize both organizational and individual goals; locating decision-making and problem-solving responsibilities close to the information sources; increasing the sense of "ownership" of the organization and its objectives; and increasing self-control and self-direction for organizational members. OD is a form of the classic "scientific method" (Bronowski, 1965; Kaplan,l964: 28) applied to organizational behavior by the organizational members themselves. And like the physical sciences, OD is subject to what Kaplan terms "the law?of the instrument: ...give a small boy a hammer, and he will find that everything he encounters needs pounding." The objectives of OD are not to institute any particular style of behavior, but to help the individual have more options concerning his behavior, more flexibility in choosing these options, and increased knowledge of probable consequences of his options. Although OD grew out of the T-group technology, OD is B£! a T-group in an organizational setting. OD is an approach to change. It seeks to foster a spirit of inquiry and collaboration; it works towards involving the members of the organization in designing and implementing their self-renewing organization. What OD is: Organizational Practice. In the early 1950's some business organizations saw the T-group as a technique to áimprove, especially_ among top and middle management, interpersonal relations and thus organizational performance. Initial "change strategies" involved sending an individual to?an off-site "stranger lab," a T-group composed of people previously unknown to one another. Such residential stranger laboratories usually lasted one or two weeks. While significant 'individual learnings sometimes resulted from -97- 2.10-04 participation in such a stranger lab, the organization itself was rarely changed. 2 In the late 1950's a few organizations had external consultants design and run T-group programs specifically for their companies. These were not ''stranger labs''; the participants had definite connections with one another. Some of the designs employed were "family labs," where the T-group was composed of an intact working group;á"cousin labs," where the participants were of the same hierarchical rank but from different work groups; and "diagonally sliced labs," where individuals from different ranks and different work groups met. 2 There is now a substantial research literature on the individual and organizational effects of various kinds of T-groups and other experienced-based learning designs. Comprehensive reviews include Argyris, 1968a; Buchanan, 1965, 1969; Campbell & Dunnette, 1968; Cooper & Mangham, 1970; Dunnette & Campbell, 1968; Durham, Gibb, &Knowles, 1967; Gibb, 1970; Gottschalk &Pattison, 1969; House, 1967, 1968; Lake, 1968; Lake, Ritvo, & O'Brien, 1969; Lieberman, 1970; Mangham & Cooper, 1969; Mouton & Blake, 1969; Patten, 1969; Schein & Bennis, 1965; Stock, 1964; and Yalom, 1970. French, 1969, offers a concise review of OD; Steele, 1970, presents a beautiful comment on OD research. Reports of specific programs appear frequently in the Journal of Applied Behavioral Science, occasional publications of the NTL Institute, and the Training and Development Journal. There has also developed a large, popular literature on T-groups. Examples include Business Week, T-school for behaviorism, 10 January 1970; Family Circle (Lee Parr McGrath & Joan Scobey), A marathon assult on your hang-ups, November 1969; Glamour (Mid Turk), How to get to know someone deeply in 90 minutes, February 1969; Glamour (Marilyn Mercer), Sensitivity training: What happens when it goes wrong--and when it goes right, February 1970; Life (Jane Howard), Inhibitions thrown to the gentle winds, 12 July 1968; Look (George B. Leonard),?The man and woman thing, 24 December 1968; Mademoiselle feature, Finding yourself-- the great American hang-up, May 1970; Medical World News, Group search for healthier emotions, 27 March 1970; New York (Linda Francke), See me, feel me, touch me, heal me: The encounter group explosion, 25 May 1970; Ne~week, The group: Joy on Thursday, 12 May 1969; Playboy (Ernest Havemann), Alternatives to analysis, November 1969; Psychology Today, passim; Saturday Review (Max Birnbaum), Sense about sensitivity training, 15 November 1969; Science ~ Technology (Herbert E. Klein), T-groups: Talk, training, or therapy?, June 1968; Time, Human potential: The revolution in feeling, 9 November 1970; Tod~ Health (Ted J. Rakstis), Sensitivity training: Fad, fraud, or new frontier?, January 1970; Jane Howard's best-selling Please Touch (New York: McGraw-Hill, 1970); and the movie Bob and Carol and Ted and Alice. Some of the popular and trade journal articles seem to emphasize the controversial nature of T-groups and to áshow a surprising ignorance of the published research, other popular articles give a more "profession.al" report. Responsible critiques of sensitivity training techniques include William H. Blanchard, Ecstasy without agony is baloney, Psychology Today, 3, 8, January?1970; Marvin D. Dunnette, Should your people take sensitivity training, Innovation, number 14, 42-55, September 1970; Martin Lankin, Some ethical issues in?sensitivity training, American Psychologist, 24, 10, 923-928, October 1969; NTL Institute, Standards for the use of laboratory method, n.d.; and Everett L. Shostrom, Group therapy: Let the buyer beware, Psychology Today, 2, 12, May 1969. -98- 2.10-05 During the next decade, both internal and external change agents and consultants researched and experimented with various T-group related programs. By 1970, an extensive organizational development technology had been developed. Among the techniques potentially useful in an OD program are the T-group, team building labs, intergroup problem-solving sessions, techno-structural interventions, data feedback, and skill training (Burke & Hornstein, 1970). Some OD consultants have developed approaches to OD which are now closely associated with them: Beckhard and the confrontation meeting (1967, 1969); Blake and Mouton (1969) with their Organizational Grid; the focus on organizational interfaces, differentiation, and integration of Lawrence and Lersch (1967, 1969); Schein's (1969) process consultation; and Walton's approach to third party consultation (1969). These are only potentially useful techniques and approaches. The appropriateness of any given technique or approach will be influenced by the specific characteristics of the OD agent (cf. Nylen, Mitchell, &Stout on the "helping relationship," 1967: 112-117), his client group, and situational factors (e.g., Harrison, 1970). One danger in OD is the attempt to achieve?an organic, self-renewing organization through mechanistic means (Siegel, 1969). Bob Euritt has commented (NTL OD Network meeting, Minneapolis, October 1970) that even the well known "organic problem-solving" approach Shel Davis (1967) used at TRW Systems might well have failed at another division of the company. There are pressures to make routine and standardized the initially nonprogrammed, innovative approach (Simon, 1965). A good OD program will be aware of "standardized" tools and approaches, but will remain open to the evolving characteristics and needs of the situation at hand. Planning and spontaneity. Many conceptions of organizational development see it as an essentially planned effort (cf. Beckhard, 1969; Burke &Hornstein, 1970). The author believes that elements of both planning and spontaneity contribute to successful change efforts.3 Specifically, while the overall goals of an OD effort (for example, inquiry, authenticity, collaboration, and choice) and some of the strategy (for example, involving top management in laboratory learning) may be planned before the organization becomes intensively involved with the OD program, it is likely that the program will be more effective if many of the tactics and micro-strategies evolve during the development of the change program (cf. Duhl, 1967). 3 The author's thoughts on "planned" and "spontaneous" approaches to change were stimulated by discussions with William c. Schutz and with Daniel Shurman. -99- 2.10-06 Spontaneity permits maximum utilization of the resources available at the moment to reach goals appropriate to the situation at hand. Planning permits the accumulation of resources (money, skill, knowledge, and commitment), but may also be seen as a defense (sometimes appropriate, sometimes inappropriate) against?the future (cf. Perls, 1969). Many "management science" or "operations research" change programs tend to be planned in great detail, yet often they encounter resistance and unanticipated dysfunctions (cf. Greiner, 1967; Riland, 1968; section 2.03 on "rational" and "real" models of man). On the other hand, some rather open-ended change programs encounter difficulties which potentially could have been anticipated (Kegan, 1970b; Shaevitz, 1970; Rogers, 1967a1 anticipates some of the difficulties reported by Shaevitz). The blend of planning and spontaneity appropriate for any given change effort will vary with the change agent, his clients, the organization, and the environmental conditions. However, even with all of these factors, a guiding proposition may be formulated: Planning should be relied upon when the change agent feels he and his system will not otherwise have the resources (money, commitment, skill, etc.) necessary to effect the desired change. Spontaneity may be relied upon when it is felt such resources will be available without planning at the needed time. -100- 2.11-01 2.11 Environmental~ cultural influences?This chapter has focused on a theory of trust and openness in organizational behavior. Always implicit, and sometimes explicit, has been the environmental and cultural background of this theory: high technology organizations in contemporary American society. This section will discuss trust as it relates to probable changes in the American environment and to differing cultures. Probable futures for America. At least for the next several decades, American society will probably become increasingly complex and increasingly interrelated. Because of the manifold effects of any single action, individual laissez-faire philosophies are likely to be replaced by policies involving larger social aggregrates (cf. Crowe, 1969; Hardin, 1968; Mesthene, 1968; Wagar, 1970). The Commission on the Year 2000 forsees that "domestically the United States is becoming a communal society rather than a contractual one" (Bell, 1967: 977). As society becomes more interdependent and populations live more densely, the correlation between a man's well-being and that of his neighbor increases (but, Cf. Milgram, 1970). School decay, pollution, misuse of natural resources, and the threat of nuclear and conventional warfare are all contemporary?problems whose solutions must benefit the entire community if any individual?is to benefit (cf. Benedict, 1970). The Commission has further identified several long-term trends, interrelated trends which seem likely to continue at least to the turn of the century (Kahn &Wiener, 1967: 706). These include increasingly sensate, secular, humanistic cultures; the institutionalization of change, especially research, development, innovation, and diffusion; increasing affluence and leisure; population growth; urbanization; increased capability for mass destruction; increasing tempo of change; and the increasing universality of these trends. Perhaps one of the main causal variables for these trends is our society's increasing affluence and consequent decreasing need to be concerned with individual survival. 1 A society of individuals concerned about each one's own 1 Keniston (1965: 440) comments on poverty and humanity: " ... only those who have been able to conquer poverty and tyranny have energy to cultivate their full humanity." While our society as a whole may be affluent, some individuals in it are poor. However, as the above trends continue, one man's poverty will increasingly influence the well-being of his fellow citizens. Thus it will be to the benefit of both poor and affluent to ensure survival for both. -101- 2.11-02 individual survival "produces a world of mutual threat, which reaffirms" the?need for competition for survival (Shepard, 1965: 1119). However, when individual survival is not a salient problem, other, "higher" needs may be attended to and the individual may more readily approach Rogers' ideal of the "fully functioning person'' (1961). Also, when individual survival is not a problem, the individual may focus his concern on issues involving a longer time perspective or larger aggregrates of individuals. These societal trends are also manifest in the industrial organization Increasingly, business is finding it must broaden its concern from immediate profit to include issues affecting the society as a whole (Ansoff &Brandenburg, 1969; Bennis, 1966; Drucker, 1969; Golembiewski, 1969; Haire, 1967; Henderson, 1968; Keezer, 1968; Tannenbaum & Davis, 1969). And internally, modern organizations are again confronting change. In 1912 when Frederick Taylor and Scientific Management were investigated by Congress, many Americans were centrally concerned with their physical needs: food, shelter, and clothing. Even in 1939 during the time of George Mayo and his Hawthorne studies, the concern with the physical needs was large. But with the growing prosperity of the United States following World War Two, lower needs became relatively satisfied and higher needs gained greater prominence. From initial concerns about salary, physical working conditions, and retirement security, many organizational members are now concerned about the quality of their working life (cf. Burger's discussion of automated affection, 1969;?also Gardner, 1963; Townsend, 1970), In general, these probable futures for American society and industry will support or intensify the relationships of trust discussed in this chapter. Change, interdependence, and the need for increasing utilization of an individual's potential are three elements of the probable future. All three can be significantly influenced by trust and openness. In a changing environment, continual reality-testing is necessary to determine what old knowledge is still appropriate and which beliefs should be altered to meet áthe new conditions. Sufficient self trust is required to remain open to "what is" and to the possibility of the necessity for personal change. When people are interdependent, trust of one another is likely to make their work on tasks of mutual concern easier and more effective. This is not to say that one should always trust those whose work or goals may interact with his own. It is to say, ho~ever, that if trusting relationships can reasonably be established and maintained, more optimal solutions are likely to be found. -102- 2. 11-03 And the process of finding those solutions is likely to be more enjoyable and fulfilling for the participants. With increasingly interdependent, changing situations, solutions to?problems will become more difficult. To find solutions will increasingly require greater resources, especially human resources. Thus, in the future, it will be increasingly necessary to tap previously unused human potentials. In many cases, this will entail trusting another that he has contributions to make and that he?will make them. Again, total lack of control is not recommended. What is suggested is trusting others--especially formal subordinates, liaison groups, and?technical personnel with training different from one's own--enough to support?their problem solving efforts and enough to test their effectiveness with such support. Cultural considerations. While this chapter is oriented toward?contemporary American society, some readily available cross-cultural research?may be used to further probe questions of external validity (cf. Roberts, 1970). In their study of political attitudes and democracy in five nations, Almond and Verba (1965) measured "trust" using Rosenberg's original (1956) Faith-in-People scale. They found (at 214) "the Americans and British tend to be consistently most positive about the safety and responsiveness of the human environment, the Germans and Italians ¥:-.ore negative, and the Mexicans inconsistent." While in the U. K. and the U. S. faith in people, or trust in people-in- general, is correlated with one's propensity to form political action groups, this relationship does not hold in the other three nations. Almond and Verba interpret this by seeing general social trust translated into politically relevant trust in the U. K. and the U. s., while such translation does not occur in Mexico, Germany, and Italy. That is, in the latter three countries, politics is seen as "a separate, independent sphere of activity for which one has a set of political attitudes not particularly grounded in general social attitudes" (at 230). This finding may be generalized to imply that increased trust will not necessarily lead to increased effectiveness unless the focal individual or group perceives the trusted other as a potential coparticipant in his efforts. Edward Banfield (1958: 83) has studied an Italian village to answer why?there was an "absence of organized action in the face of pressing local problems." He found the major explanation to be an ethos--"amoral familism"--which states -103- 2. 11-04 that one should "maximize the material, short-run advantage of the nuclear family [and} assume that all others will do likewise." The extreme poverty, high mortality rate, and low social status of the region contributed to the origin of the ethos. Banfield (at 158f) concludes that amoral familism must be modified in at least three respects in order for organization, and thus economic and political progress, to be possible: 1) The individual must define self or family interest less narrowly than material, short-run advantage . . . . 2) A few persons, at least, must have the moral capacity to act as leaders. . .. 3) Voters and others must not destroy organization gratuitously or out of spite or envy; that is, they must be willing to tolerate it when it does not interfere with them. However, even if trust and time perspectives were increased, the village's "isolation and relative lack of resources give it a comparative disadvantage which no amount of cooperation can overcome" (at 157). Although the culture is different, this lack of corporate action in the Italian village may be directly related to impediments to effectiveness in American organizations. Rather than an "amoral familism," many organizations seem to be governed by an "amoral work-groupism" or "amoral individualism." Such an ethos would see other work groups or individuals within the organization as competitors. If trust and perspectives were increased, the organization's members may shift to view one another as generally engaged in collaborative efforts. If competition exists, often it will be against "nature" or other external organizations; when internal competition exists, it should be "appropriate" (Beckhard, 1969). Pearlin and Kohn (1968: 478) studied social class, occupation, and parental values in a cross-cultural project involving the United States and Italy. Although they found Italian parental values to be more adult-centered and American values more child-centered, the relationship between social class and parental values was similar in both countries. They saw occupations--"how the structure of work imposes constraints and imperatives" on the worker-- as the critical dimension of social class causing this relationship. Three aspects of occupation--closeness of supervision, type of work (with things, people, or ideas), and degree of self-reliance required for the job--we~e measured. Each of the three was independently related to fathers' values--self-control versus obedience--for their children; the combined effect of the three work aspects accounted for a "very large part of the difference between middle- and working-class fathers' values." -104- 2.11-05 From this study it may be inferred that the structure of his work has?a great deal of leverage on an employee's values. If an organization wishes to promote a climate of trust and openness, it cannot rely on "importing" individuals who hold trusting values and placing them in an unchanged, mistrusting work situation, for they are liable to shift towards the prevailing values of the organization. To promote trust and openness, the work structure and the climate of the organization, or of the group under study, must support the desired values. -105- 2.12-01 2.12 Hypotheses The underlying general research area of this dissertation is that concerning trust, humanism in organizations, and the relationship between?the individual and his organized society. From this general area, two specific research question@_ are addressed: a) what are some of the individual and organizational correlates of trust, and b) what are some measures which may indicate the effectiveness of organizational development and sensitivity training programs. Two core hypotheses underly this research. They are: CH 1. An open and trusting environment within certain types of organizations will lead to increased organizational effectiveness and to increased individual effectiveness. CH 2. For certain types of organizations, proper sensitivity training will lead to the fostering and support of an open and trusting environment within the organization. These core hypotheses have significance in three areas: effectiveness, sensitivity training, and methodology. Increased organizational effectiveness is generally a desired goal. In addition, there is growing recognition that greater concern for the individual qua human, a fostering of a conception of the dignity of man, is a critical need for our culture. Sensitivity training and organizational development programs are increasingly being used in organizations; their effects have not yet been adequately researched (cf. note 2, section 2.10). These programs are potentially powerful agents of social change. With concern both for efficiency and efficaciousness, additional research can be useful. Finally, the procedures and methodology developed during this study can contribute to longitudinal studies, administrative experimentation, and organizational development programs. The core hypotheses may be stated as more operational hypotheses, reflecting the measures used in this study. H 1. The more an individual trusts his work group and the more he generally trusts others with whom he interacts during his work, the greater will be his self-actualization. It may be expected that both trust of work group and trust of others will additively contribute to self-actualization, although trust of work group should be the stronger effect. The suggested (this is not part of the hypothesis) causal sequence is: actual working conditions moving toward the elements of -106- 2.12-02 self-actualization, perception of trust, perception of self-actualization. The elements of self-actualization (Bonjean &Vance, 1968: 300) are self- determination, independence, variety, challenge, a long time perspective,. awareness and control over self, and equal or superordinate position relative to peers. H 2a. The more the members of a group trust the group, and the more they generally trust others with whom they interact during their work, the more effective the group will be in its goal accomplishment. This hypothesis assumes that the members of the group have not lost touch with reality, are not in a state of "blissful ignorance," and are perceiving the demands of their task. A subsystem out of touch with its task demands cannot long survive in a suprasystem which is in touch with its reality. The group's rate of transformation (an unaware group may be transformed to awareness or the group may not "survive") may be expected to be more rapid the closer to that group aware interacting "others" are. Typical "others" are supervisors, department heads, presidents, stockholders, and customers. H 2b. The higher the group trust of itself, the greater the group's effectiveness toward achieving its own goals. H 2c. The higher the group trust of others with whom they interact, the greater the group's effectiveness toward achieving the organization's goals. H 3. A "proper" organizational development program will increase individuals' feelings of trust toward their own work group and toward others while keeping the individuals aware of the demands of their tasks. A long time period may be necessary for such global results: hostile groups may exist within the organization. Short-run results of proper sensitivity training and organizational development programs may be expected to be the increasing of trust of one's own group, and with a longer time, of trust toward some other interacting groups. The adjective "proper" in H3 needs additional comment. As indicated in section 2.10, sensitivity training is not a panacea; it is a tool which can be well used or quite misused. Schein and Bennis's (1965: 216) five-step model for diagnosing the state of a target system indicates significant preconditions for the success of an organizational development program. These preconditions include a) the OD change goals being appropriate to the target system; b) the -107- 2.12-03 target system being culturally prepared for the OD program; c) key people being involved and committed; d) members of the target system being adequately prepared and oriented to the program; and e) voluntary participation being insured. In addition to a receptive target system, the change agent must be competent. In sum, a "proper" organizational development plan may be seen?as one which competent applied behavioral scientists believe to have met the significant preconditons for success. This test of a "proper" OD program may be made before the effectiveness of the program is measured, thus avoiding problems of tautology and contaminated judgment. Due to the time span of this dissertation study and to the procedures for selecting participating groups (see section 3.03), a rigorous, formal test of H 3 is not possible. However, for two groups (POl and P02) with relatively well defined OD programs and with at least a year's data, some explorations of H 3 can be made. Section 3.07 describes the development of the scales used in this study; section 3.10 presents an evaluation of the scales. Following are the major variables of the three operational hypotheses, together with their indicators. Individual trust of his own and other groups Q07 Individual self-actualization QOS Group trust of itself and other groups Q07 Group effectiveness: own & organizational QlS OD program Ql7 & interviews Significant parametersters mentioned throughout this chapter include the type of organization, its technology and market structure, and top management knowledge of and support for OD and sensitivity training. Many other variables were measured in this study. These constitute an implicit hypothesis: that there will be interesting relations among these additional variables. These additional variables include demographic items (measured by Qll): age, sex, years of formal education completed, academic degrees, years in present organizational position, years with the company,?number of people reporting to the participant, and the participant's hierarchical index; personality variables (measured by Q02): openness to experience, ego strength, and defensiveness; confidence in the information network (Q06); openness in communication (Q08); evaluation of past decisions (Q09); and intergroup relationships (Ql2). During the course of this study, additional, secondary hypotheses were -108- 2.12-04 formally stated. Such hypothesis conceptualization was stimulated by the editing of returned questionnaires for completeness. Some questionnaires evoked the thought: "Variable x seems unusual, could it relate to variable y?" However, variables x and y were not compared until after the hypotheses were formally stated. Thus, although developed after this study began, these secondary hypotheses may still be considered as a priori. HS 1. a) There is a positive correlation between the time taken to complete the questionnaire and the part~iipant's defensiveness (K score on Q04); and b) a negative correlation between this time and his trust of people (Q07). HS 2. Individuals dropping from the study will have extreme scores on some dimension. Probable dimensions and directions include long completion time, high defensiveness, low ego strength, low trust of people or of own work group, and low self-actualization (cf. Hamilton, 1968). HS 3. Over a period of time, with proper organizational development, the difference between self and other disclosure scores (Q08) should decrease. Figures 1, 3, 4, and 5 contain many implicit propositions. Those of Figure 1 include: HS 4. Defensiveness (Q04-K) and openness to experience (Q03) are inversely correlated. HS 5. Openness to experience and self-actualization (Q05) positively correlate. HS 6. Self-actualization and group effectiveness (Ql5) positively correlate. HS 7. Self-actualization and organizational effectiveness (Ql5) positively correlate. HS 8. Ego strength (Q04-Es) and disclosure (Q08-self) positively correlate. HS 9. Decision evaluation (Q09) and effectiveness (Ql5) positively correlate. In the conceptual development of this research, many propositions were conceived or encountered in the literature. Some of these have been incorporated into the discussions of this chapter. Others have not been directly studied?in this research. However, inasmuch as they may be helpful in other organizational studies of trust, they are listed here as tertiary hypotheses. -109- 2.12-05 HT 1. Inappropriate defensiveness restricts an individual's functioning. Many schools of psychotherapy agree that the individ~al needs some defense mechanisms (e.g., Ford & Urban, 1963). However, when defensive measures are employed unnecessarily (i.e., when "reality" indicates no actual threat, cf. Kaplan (1964: 128) on intersubjectivity--a procedure to test reality), the individual becomes less fully functioning than he might be (Rogers, 1961); he becomes more dead and less alive (Perls, 1969). HT 2. For strong organizations whose products compete in?the marketplace on the basis of inherent, technical differences from other products and whose members have fairly high levels of need for achievement, an internal atmosphere of openness will be beneficial to the organization. HT 3. For organizations that are in less strong positions, whose products compete on the basis of gimmicks or nontechnical differentiation, or whose members have relatively low levels of need for achievement, some degree of openness will be beneficial, but the nature of the threats to the organization-- from both internal and external sources--must not be forgotten. Individuals with a high need for achievement may be expected to be motivated toward and receive satisfaction from task accomplishment at an individually challenging and reasonable level. In contrast, individuals with a high need for power may be expected to be motivated toward and receive satisfaction from successful competition (cf. McClelland, 1961). In addition to such relations between personality and competition, competition is also manifest in organizational behavior. Goldner (1965) reports a study on demotion; Steinmetz (1968) discusses the problem of firing a manager; Fair (1966) considers the possibilities for increased use of corporate spies. McCarthy (1965: 69) presents a study of R &D's response to market crises in a highly competitive oligopolistic industry. His results indicate that in the cases studied, product innovations were made with a primary objective of threatening the market position of rival firms. HT 4. "Organizations formed at one time typically have a different social structure from those formed at another time" (Stinchcombe, 1965: 154). HT 5. An organization with an "intellectual, rational" social structure will more easily foster openness than will an organization with a "power" social structure (cf. Stinchcombe, 1965). HT 6. Stinchcombe's "modern" industries will approximate -110- 2.12-06 the "rational" social structure; the "railroad age" industries will approximate the "power" social structure. "Power" orientations tend to employ ,.,in-lose, zero-sum philosophies; "rational" norms, on the other hand, place greater emphasis on the possibilities of creation and synergy. Hypotheses HT 4 through HT 6 extend the concepts of HT 2 and HT 3--trust and competitive environments--to relatively enduring factors in the organization which may have been influenced by its historical founding. HT 7. Work concerned with the creation and distribution of information will be more sensitive to these trust propositions than will work concerned with the production and distribution of physical objects. HT 8. Work involving people will be more sensitive to these trust propositions than will work involving physical objects. Traditional values of possession make the sharing of intangibles less likely?to be considered a win-lose, zero-sum transaction, than the sharing of physical objects. HT 9. The differential effects of trust and openness will be more pronounced for managers than for bench workers. To the extent that the work of managers involves more unprogrammed decision- making and more interpersonal communication and coordination than that of bench workers, the effects of trust and openness on communication and information exchange should be more pronounced for the managers (see section 2.04). HT 10. Managerial styles in one culture may have different results when applied in a different culture. Cultures differ, as do some of the psychological underpinnings of them (Hsu, 1963; Kroeber, 1948). And these psychological differences may mean that people in one culture will interpret situations and stimuli in a manner different than individuals from another culture. For example, Lewin (1968: 79) refers to several studies indicating some effects of cultural differences on the effectiveness of participative decision-making. Pfaff and Pfaff (1967) discuss problems of a manager adapting to a foreign culture. Finally, Prasad (1969) discusses forces changing the traditional paternalistic authority system in Japanese industry. HT 11. The easiest method of influencing individuals, organizations, and the society in the United States to develop and maintain open and trusting behavior is through occupational organizations. Section 2.11 discussed the,cross-cultural research of Pearlin and Kohn (1966). This study indicated that a man's work structure had an extremely strong effect on his values for his children.