THE EVOLUTION OF TRUST AND COOPERATION:

IMPLICATIONS FOR TEAMWORK AND TACIT KNOWLEDGE

Gareth R. Jones

and

Jennifer M. George

Department of Management

Texas A&M University

College Station, TX 77843-4221

Abstract

This paper analyses the way that trust evolves in organizations and its effects on organizational cooperation. First, we propose that trust is experienced psychologically through the interactions of people’s values, attitudes, and moods and emotions. Then, based on this analysis, symbolic interactionism is used to describe how trust evolves and changes over time and two distinct forms or states of trust--conditional and unconditional trust--are identified and defined. We then examine the relationship between these forms of trust and two important types of organizational cooperation: interpersonal cooperation and teamwork and the development of organization-specific tacit knowledge

It has been widely acknowledged that trust can lead to cooperative behavior between individuals, groups, and organizations (e.g., Axelrod, 1984; Gambetta, 1988; Good, 1988; Mayer, Davis & Schoorman, 1995; McAllister, 1995). Hence, today, in an era when organizations are searching for new ways to promote cooperation between people and groups to enhance the value they create, it is not surprising that interest in the concept of trust and, in particular, how to promote or actualize trust is increasing (Kramer & Tyler, 1996) . For example, many organizations have sought to increase cooperation between people and groups by reengineering their organizational structures to move to flatter, more team-based forms in which authority is decentralized to "empowered" lower-level employees. In order to promote trust between two or more parties, however, it is necessary to understand how individuals experience trust in another person, group, or organization and how trust evolves between people or groups over time. In other words, it is necessary to understand how trust in others is experienced psychologically before its impact on behavioral expectations and outcomes such as the level of cooperation between people in an organization can be adequately analyzed. The purpose of this paper is to explore these issues.

Using a theoretical framework based on values, attitudes, and moods and emotions, we analyze how people experience trust in a psychological sense and the feelings, beliefs, and meanings that underlie it. Then, based on this framework, we provide an account of the evolution of trust that helps to explain some of the most interesting issues surrounding trust such as how trust spreads among people and is sustained or strengthened or, conversely, how trust comes to be weakened, broken, or dissolved. Based on this analysis we then distinguish between two different forms or states of trust, conditional trust and unconditional trust, each of which has different implications for the level of cooperation found in a group or organization. To illustrate these different relationships between trust and cooperation, we focus on two constructs that are increasingly being used to explain differences in competitive advantage and performance between organizations: teamwork and the development of tacit organization-specific knowledge.

The Experience of Trust

Researchers have devoted considerable attention to clarifying the meaning of trust in different social contexts and the conditions or determinants of trust. For example, trust is commonly treated as an expression of confidence between the parties to an exchange that they will not be harmed or put at risk by the actions of the other party (Axelrod, 1984; Bateson, 1988; Zucker, 1987) or that no party to the exchange will exploit the other’s vulnerability (Sabel, 1993). Mayer, Davis, and Schoorman (1995: 712) recently developed a model of organizational trust in which several characteristics of the trustor and trustee (such as the trustee’s benevolence and integrity) lead to trust in the trustee, trust being defined as "the willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor or control that other party." From this perspective, trust leads to a set of behavioral expectations among people that allows them to manage the uncertainty or risk associated with their interactions so that they can jointly optimize the gains that will result from cooperative behavior.

In discussing how expectations underlying trust affect subsequent behavior, several researchers have alluded to the fact that trust is a multidimensional construct. For example, Driscoll (1978) and Scott (1980) distinguish between the generalized or global aspects of trust and the situationally-specific aspects of trust. Butler (1991) emphasized the multidimensional

nature of trust and identified different conditions under which trust may occur. Barber (1983) argued that trust is a multidimensional concept that encompasses moral, cognitive, and emotional elements. Building on this research, we propose that the multidimensional nature of trust can be described psychologically in terms of people’s values, attitudes, and moods and emotions. That is, psychologically, the moral, cognitive, and emotional elements of people’s evaluations and expectations of others can be thought of as being represented by their values, their attitudes, and their moods and emotions, respectively. Below, we briefly describe the nature of values, attitudes, and moods and emotions and the role each plays in the psychological experience of trust. In doing so, we provide a platform on which to model the evolution of trust between people and groups and to show how trust can be created or destroyed over time.

Values

Values are general and relatively abstract standards that are evaluative in nature and describe desirable ends and means to achieve them (Olson & Zanna, 1993; Rokeach, 1973). Values are organized into a person’s value system which prioritizes values in terms of their relative importance as guiding principles (Rokeach, 1973). A value system guides behavior and the interpretation of experience by furnishing criteria that a person can use to evaluate and make sense of events and actions in the world around him or her. A person’s value system determines which types of behaviors, events, situations, or people are desirable or undesirable.

The experience of trust builds on internalized values such as loyalty, helpfulness, fairness, predictability, reliability, honesty, responsibility, integrity, competence, consistency, and openness (e.g., Butler, 1991; Butler & Cantrell, 1984; Gabarro, 1978; Jennings, 1967, 1971). In essence, what people view as desirable or ideal, that is, their internalized values, condition the experience of trust and are upheld as standards to strive for in the future. An individual whose value system emphasizes loyalty and honesty, for example, will be striving to achieve loyalty and honesty in his or her relationships with others. Values are standards which guide how an individual experiences trust; they embody what an individual is striving for in his or her relationships with others (Lewin, 1937). Values contribute to the generalized experience of trust and can create a propensity to trust (Mayer et. al., 1995) that surpasses specific situations and relationships.

These assertions are consistent with the larger body of literature on trust. For example, Barber (1983) suggests that trust serves to maintain and express the shared values that trust originates from and that shared values help create relationships characterized by trust. As another example, and consistent with the research of Rotter and his colleagues (e.g., Rotter, 1980), Good (1988) suggests that people who themselves are trustworthy (or endorse values such as honesty) tend to view others as being trustworthy (or as endorsing similar values underlying trust). As a final example, theory and research which considers the stable underpinnings of trust inherent in the trustor, such as Mayer et al.’s (1995) notion of the propensity to trust, Rotter’s (1980) notion of a general tendency to trust others, and research looking at trust as a somewhat stable disposition (Stack, 1978), suggest that trust is based in enduring and relatively stable characteristics of individuals. Additionally, it has been suggested that distrust arises out of value incongruence (Sitkin & Roth, 1993). Distrust results from trust being violated and attributions for the violating behavior being made to values that are incongruent with the trustor’s own values (Sitkin, 1995). Hence, just as trust is experienced through the values which support it, distrust is experienced when these values are challenged.

Attitudes

In addition to being experienced in a stable and general manner through values, trust is also experienced in a more specific mode (Butler, 1991). That is, trust is experienced as part of an attitude toward another entity based on knowledge, beliefs, and feelings about the nature of that entity (e.g., McAllister, 1995; Robinson, 1996). Attitudes can be viewed as (a) the knowledge structures containing the specific thoughts and feelings people have about other people, groups, or organizations; and (b) the means through which they define and structure their interactions with others (Anderson & Armstrong, 1989; Kruglanski, 1989; Olson & Zanna, 1993).

As knowledge structures, the attitudes that people form towards each other in an organizational context are likely to contain information concerning the other party’s trustworthiness for at least three reasons. First, most interactions in organizations entail uncertainty; even when another party provides guarantees that a particular behavior will occur, uncertainty may arise about whether they will be able to carry through with the promised act. For example, a supervisor may make promises to a subordinate about an upcoming promotion but the promotion may nonetheless be uncertain due to organizational initiatives beyond the supervisor’s control. Hence, in forming attitudes about others to guide perceptions and behaviors, individuals are likely to include assessments of trustworthiness. Second, interdependence is an essential feature of organizational life (Thompson, 1967). Since individuals have to depend upon one another as do groups and organizations as a whole, attitudes towards interdependent entities are likely to include feelings, beliefs, and knowledge about the entities’ trustworthiness. Lastly, given that social interaction is built on expectations which are partially cognitive and based on past

experience, it is likely people’s attitudes towards others contain beliefs about the trustworthiness of these others based on past experience, knowledge, and interactions (Rempel, Holmes, & Zanna, 1985).

Given that the specific attitudes formed toward other people and towards groups and organizations as a whole are likely to contain knowledge concerning the trustworthiness of the party in question, these attitudes will be an important aspect of the experience of trust. Attitudes will serve to define and structure social interactions and the experience of trust in ongoing relationships. As Lewis and Weigert (1985) suggest, trust develops from cognitive evaluations of the trustworthiness of other people and collectivities. Moreover, and as mentioned previously, while trust is experienced at a general level through values, attitudes are a means through which trust is experienced at a specific level (Butler, 1991; Driscoll, 1978; Scott, 1980) since attitudes are, by definition, object-specific and responsive to ongoing experiences with the attitude object. Hence, attitudes structure the experience of trust in specific ongoing relationships.

Note that attitudes and values interact and can affect one another over time. Attitudes are evaluative in nature, and values are a key determinant of how people come to evaluate other people and organizations. Conversely, changes in attitudes can precipitate a change of values over time, such that, for example, repeated exposure to episodes of unethical behavior can change the degree to which a person believes that "honesty is the best policy." The interactions between values, attitudes, and moods and emotions are central to our discussion of the evolution of trust and are discussed later.

Moods and Emotions

Moods and emotions capture how people feel as they go about their daily activities including interacting with other people; they are on-line feelings which provide people with information about their ongoing experiences and their general state of being (Jacobsen, 1957; Morris, 1989; Nowlis, 1970; Pribam, 1970). The intensity of the affective state is a primary distinguishing feature between moods and emotions. Emotions are intense affective states that interrupt ongoing cognitive processes and behaviors and are tied to particular events or circumstances (Simon, 1982). Moods are less intense, pervasive and generalized affective states that are not explicitly linked to particular events or circumstances. Moods capture more of the day-to-day feelings that people experience which do not interrupt ongoing activities but affect them in more subtle, but also significant, ways (Clark & Isen, 1982; Thayer, 1989). A well-accepted way of describing moods and emotions is in terms of the extent to which they entail positive or negative affect (Costa & McCrae, 1980; Meyer & Shack, 1989; Watson & Pennebaker, 1989; Watson & Tellegen, 1985). Moods and emotions entailing high level of positive affect can be described by terms such as excited, enthusiastic, active, elated, peppy, and strong while moods and emotions entailing high levels of negative affect can be described by terms such as distressed, hostile, jittery, nervous, scornful, and fearful (Tellegen, 1985; Watson, Clark, & Tellegen, 1988; Watson & Tellegen, 1985).

Emotions and moods are fundamental aspects of the experience of trust for at least four reasons. First, the experience of trust embodies affect, whether it be in terms of strong or intense feelings (emotions) or more subtle feelings (moods). People often decide if they trust someone by examining the feelings associated with that person. For example, in the presence of an individual they distrust, a person might be expected to feel high negative affect and be jittery, nervous or even afraid. On the other hand, in the presence of someone they trust, a person might be expected to feel high positive affect and be excited, elated, or even joyful. Moreover, one person’s moods and emotions can quickly effect others, such as when fear or panic develops in crowd situations, so that very quickly "positive affect circulates among those who express trust behaviorally, just as negative affect arises among those who betray or act distrustfully toward each other" (Lewis & Weigert, 1985: 971). At a fundamental level, one of the intrinsic rewards of a trusting relationship (Rempel et al., 1985) can be the positive feelings and emotions generated by that relationship, such as love, loyalty, or a sense of commitment or companionship.

Second, the current affective state an individual is in may color the experience of trust. Experiencing positive moods or emotions may cause one to have more positive perceptions of others and see the world through "rose colored glasses," resulting in a heightened experience of trust in another person. Negative moods and emotions may add a negative tone to interactions and may result in people being perceived as less trustworthy than they actually are. Consistent with this reasoning, positive moods have been shown to heighten liking for others and cause people to have more positive beliefs about human nature (Gouaux, 1971; Veitch & Griffit, 1976), to increase helping and generosity (Isen, 1970; Isen & Levin, 1972), and to promote more integrative as opposed to contentious bargaining in negotiations (Carnevale & Isen, 1986).

Third, moods provide people with information to use in forming their opinions and judgments about a variety of stimuli such as the trustworthiness of others (Schwarz, 1990; Schwarz & Clore, 1988). Suppose that a supervisor is in a negative mood when interacting with a subordinate. In trying to assess the trustworthiness of the subordinate in carrying out an important assignment, the supervisor may question the reliability and capabilities of the subordinate based on his or her negative mood. Similarly, suppose a manager is engaged in negotiations with a supplier while in a particularly good mood. The manager may infer, based on the positive feelings surrounding the interaction, that the supplier is to be trusted. Lastly, trust is built on expectations that are, in part, emotional, and when these expectations are broken, strong emotions are often experienced which signal to the individual the violation of trust and the need to attend to the relationship (Barber, 1983; Frijda, 1988).

Moods and emotions, much more so than values or attitudes, contribute to the ongoing experience of trust as they are concerned with current feelings or affective states. Moods and emotions provide people with signals concerning the changing nature of their ongoing experience of trust with other people or in particular situations (Frijda, 1988; Schwartz & Clore, 1983, 1988). More so than values and attitudes, moods and emotions fluctuate over time. Also, in contrast to values which are general in focus and attitudes which are object-specific, moods and emotions are both general and specific. They are general in that the same kind of mood or emotion may be experienced across a variety of situations and relationships, for example, fear or joy. They are specific in that certain people, situations, or relationships may be associated with certain kinds of feelings. For example, a person may feel scornful in the presence of a conniving colleague or enthusiastic in the presence of a highly capable collaborator. The feelings are specific to the person, interaction, or relationship but are general in the sense that the same kind of feeling will be experienced again, across people and situations.

Moods and emotions also interact with values and attitudes to determine the experience of trust. Negative feelings arise from performing activities that are inconsistent with one’s values; attitudes affect how people feel, so that being highly satisfied can lead to the person feeling good or being in a positive mood. Conversely, if a person experiences a negative mood over a prolonged period of time, this can cause attitudes to become more negative, which can induce changes in values over time.

In sum, based on the forgoing arguments, and consistent with the literature on trust, we propose that the psychological experience of trust is multidimensional and involves people’s values, attitudes, and moods and emotions. Values provide standards of trust which people strive to achieve in their relationships with others; attitudes provide knowledge of another person’s trustworthiness; moods and emotions are indicators of the presence of trust in a relationship.

A Symbolic Interactionist Model of the Evolution of Trust

The previous analysis demonstrates that the psychological experience of trust is complex and that the three ways in which trust is experienced (through values, attitudes, and moods) must be considered simultaneously. To understand how trust evolves over time in an organizational setting, it is useful to model trust from a symbolic interactionist perspective (Blumer, 1962; Mead, 1934). Symbolic interactionism is based on two main assumptions that are pertinent to an analysis of trust: first, people act in social encounters based on the meanings that they have learned to associate with them; and second, that these meanings are acquired by interactions with other people such that a definition of the situation is created over time. More specifically, in any particular encounter, two (or more) parties mutually develop and negotiate a definition of the situation. This joint creation of the definition of a social situation involves each party trying to understand the other party’s expectations, needs, and goals. What one party says or does affects the other; communication and behavior patterns are adjusted to fit the unfolding mutually determined definition of the situation.

Importantly, each individual tries to adjust to the other and brings to the situation his or her interpretive schemes as a base from which to build understanding. As Mead (1934) puts it, a person takes the perspective or role of the other in order to call out the same response in the self. This is the basis from which social action is generated. Such role taking is the process by which parties exchange meanings and develop improved understanding of the context. Over time, the richness of the context is a consequence of the parties’ use of more complex or fine-grained interpretive schemes to understand the actions of the other. In this way, they develop new definitions of the situation and the basis of social action changes.

In the context of the evolution of trust, two or more parties are involved in the creation of a joint definition of the situation. Each party brings their own set of interpretive schemes to the situation. To the extent that they use or develop similar interpretive schemes to define the situation, the parties will tend to agree on their perceptions of the level of trust present in the situation, so adjustment to each other takes place. The likely nature of this adjustment depends on the degree of congruence or similarity between the values, attitudes, and moods and emotions of the two parties.

The Creation of Conditional and Unconditional Trust

In the context of the experience of trust, a symbolic interactionist perspective suggests that trust cannot come into being unless the orienting values underlying trust are prominent in one or both parties’ value systems. Suppose there is an obvious lack of congruence between people’s values and beliefs? Then, as Sitkin and Roth (1993) suggest, value incongruence will arise and perceptions that parties to an exchange do not share key beliefs and values will create distrust because parties cannot take the role of the other. Thus, each party cannot take the perspective of the other and trust the other unless they have an orientation toward trust in social relationships based in their value systems. However, there is inherent uncertainty and risk in social relationships that increases the vulnerability of the parties. Taking this uncertainty and risk into account, we propose that trust comes into being in the following manner.

At the beginning of a social encounter, each party does not simply assume that the other person is trustworthy; rather, people suspend belief that the other’s values may be different from their own--that the other person may not be trustworthy. This does not mean that the person is gullible--a gullible person is one who, in the absence of any information of any kind, takes the other’s trustworthiness on faith and assumes the risk of being exploited. The actor in our model just suspends belief that the other is anything but trustworthy and behaves as if the other has similar values and can be trusted.

This paradoxical beginning to the experience of trust is likely because, as Luhmann (1980:72) suggests, trusting another and assuming they share one’s own values is often preferred to initial distrust because distrust "often absorb[s] the strength of the person who distrusts to an extent which leaves him[her] little energy to explore and adapt to his[her] environment ...and hence allow him[her] fewer opportunities for learning." Relatively, trust is the easier option, and for this reason there is a strong incentive to begin a relationship with trust. This is because an enormous amount of time and energy would be taken up in discovering the true nature of the other’s value system leading to a major advantage associated with suspending belief that another person has sinister motives (Deutsch, 1958, 1960).

This is not to say that people do not use their own values to decide whether or not to trust another. Perceptions of value incongruence can quickly lead to distrust (Sitkin & Roth, 1993) . In meeting a stranger, a person’s value system is used to decide if that person is a fit person to transact with. In other words, people use their values to decide if others fall within the "zone of their indifference" (Barnard, 1938) and their values do not appear to be so divergent that they make themselves vulnerable to the other party. This does not mean, however, that each person’s values need to be the same for trust to emerge in a relationship. Rather, we propose that people approach interactions based on their own orienting values of fairness, equity, and so on, and if they have no obvious sense that some form of value incongruence exists, they suspend their beliefs that the other party is not to be trusted.

Assuming, that distrust is not the outcome of the initial encounter, then in the future the experience of trust is determined by the content of the behavioral exchanges between the parties, exchanges which are experienced psychologically through the evolving attitude that is developed toward the other party. If trust is to build up over time, both parties must be able to take the role of the other and exchange and share the feelings and thoughts that structure the exchange relationship--literally, to develop attitudes toward the other reflecting the other as a trustworthy party. As noted earlier, attitudes have this specific focus and provide the engine of the evolution of trust, capturing the meaning of the behavioral exchanges that lead to the development of trust.

What is the role of mood and emotions in the evolution of trust? As just discussed, the ongoing evolution of trust depends on the development of favorable attitudes and expectations through behavioral exchanges. At every exchange point, moods and emotions affect the ongoing experience and meaning of the relationship. First, positive moods and emotions result from successful behavioral exchanges and over time the parties will experience increased feelings of happiness or excitement in the presence of the other person. High positive affect therefore helps to set the scene for the continuing exchange and building of greater trust. Second, the current affective state of each party will affect ongoing interactions and the way in which trust builds up—people in negative moods will be more inclined to evaluate the other party unfavorably, for example. Moods and emotions affect the parties’ immediate perceptions and judgments that they will not be harmed or exploited by the other party and thus, the experience of trust. The extent to which one or both parties experience positive moods, then, enhances the likelihood that the parties will develop shared interpretive schemes and cement the experience of trust.

Especially important are the roles played by moods and emotions as the mechanisms through which both parties are continually evaluating the ongoing quality of the trust experience. More specifically, to the extent that both parties succeed in developing a common frame of reference and can successfully take the role of the other, positive moods are promoted. At significant points in the ongoing relationship such as the successful completion of an important project, both parties will likely experience positive emotions which provides a powerful signal (Frijda, 1988) to them that they have succeeded in building trust, and do, in fact, share the same attitudes and values. At this stage in the experience of trust, the decision to suspend belief transforms into the desire to trust the other as both parties feel secure that they will not be harmed or put at risk by the actions of the other party and that the other is, in fact, trustworthy.

The point at which the parties to an exchange (a) have favorable attitudes toward each other, (b) experience positive affect in the context of the relationship, and (c) feel confident in each other’s values and trustworthiness is the crucial point in the evolution of trust. In order to distinguish between the experience of trust before and after this point, it is useful to make a distinction between the states of conditional and unconditional trust. Conditional trust is a state of trust in which both parties are willing to transact with each other as long as each behaves appropriately, uses a similar interpretive scheme to define the situation, and can take the role of the other. Under conditional trust, attitudes toward the other party are favorable enough to support future interactions, and sufficient positive affect and a relative lack of negative affect reinforces these attitudes. Conditional trust is usually sufficient to facilitate a wide range of social and economic exchanges and is consistent with the idea that one of the bases for trust is knowledge (e.g., Lewicki & Bunker, 1996; Shapiro, Sheppard, & Cheraskin, 1992; Sheppard & Tuchinsky, 1996) or positive expectations towards the other (Sitkin & Roth, 1993).

Unconditional trust, on the other hand, characterizes an experience of trust which starts when an individual abandons the "pretense" of suspending belief because shared values now structure the social situation and become the primary vehicle through which people experience trust. With unconditional trust, the other’s trustworthiness is now taken as a given, based on confidence in the other’s values backed up by empirical evidence derived from repeated behavioral interactions, knowledge of which is contained in one’s attitude toward the other party (Butler, 1983). Also, positive affect increases as positive moods and emotions strengthen the affective bonds between parties and bolster the experience of trust. As such, when unconditional trust is present, relationships become significant or intense, and often involve a sense of identification with the other party (Lewicki & Bunker, 1996; Shapiro et al., 1992).

Insert Figure 1 About Here

Our model proposes that there are three principle forms or states of the trust experience: distrust, conditional trust, and unconditional trust (See Figure 1). Note that this model differs from other models of trust in the literature in several significant ways. First, we conceptualize distrust, conditional trust, and unconditional trust, as being three different forms or states of the same construct, the trust experience, rather than being different constructs. Thus, while our model shares with those of Sitkin and Roth (1993) and Fox (1974) the proposition that shared values and expectations help determine the experience of trust, it differs from these models because they conceptualize trust and distrust as separate constructs each of which has a different set of determinants. For example, Sitkin and Roth see the level of trust as determined by met expectations while distrust is engendered by value incongruence. Moreover, our model makes the interaction of values, attitudes and moods the mechanism by which both trust and distrust are engendered. It is the interactions between values, attitudes, and moods that determines whether or not trust exists, and what state trust takes, that is, conditional or unconditional.

Second, our model of trust differs from models such as those developed by Shapiro, et.al. (1992) and Lewicki and Bunker (1996) which propose that there are quite different types of trust, based on different kinds of determinants. For example, Shapiro, et.al. (1992) proposed that there are three types of trust, deterrence-based trust sustained by threat of sanctions; knowledge-based trust sustained by the ability to predict the behavior of others; and identification-based trust sustained by a complete empathy with the other party’s desires and needs, and that each is quite different, or even may be sequential in a Maslovian way (Lewicki & Bunker, 1996). As noted earlier, however, few people have the time or resources to engage in the extensive information processing that would be necessary to monitor the other party’s behavior to apply appropriate sanctions, or to do so in a way that allows one to predict their future behavior or intentions. Furthermore, ultimately, people are unknowable and it is therefore never possible to have complete empathy with their often hidden or inexpressible desires and needs. Hence, the importance of the state of conditional trust which short-circuits information processing and allows people to economize on their cognitive and emotional energy, and forge ongoing relationships that can lead to unconditional trust but which might also end in distrust.

Thus, rather than asserting that different determinants lead to different types of trust, our model conceptualizes trust as a changing or evolving experience in which values, attitudes, and moods operate simultaneously to produce an overall state of trust or distrust. Hence, we view trust as a more dynamic kind of experience, one that can shift or change between trust states. Indeed, one important implication of the analysis is that while conditional trust can change into unconditional trust, it is also possible for unconditional trust to change into conditional trust, or into the absence of trust, that is, distrust. This is reflected in Figure 1 by the two-way arrows between distrust and conditional trust and between conditional trust and unconditional trust. Before we examine in detail how the difference between conditional and unconditional trust affects subsequent cooperative behavior, it is necessary to analyze the ways in which the model proposes that trust can breakdown or dissolve.

The Dissolution of Trust

The ways in which trust can dissolve or disappear over time has received considerable attention in the literature. For example, Butler (1983) found that when one party signals positive expectations or favorable attitudes to another and the other reciprocates those expectations that trust will spiral upwards and when expectations are not reciprocated trust will spiral down. Sitkin and Roth (1993) argued that, when distrust exists because of value incongruence, the use of legalistic mechanisms to try to restore it will fail because they only act at the level of expectations, not values, and may cause an escalating spiral of distrust. The model developed in this paper, because it conceptualizes trust as a construct which has three principal states, and because it focuses on the interactions between values, attitudes, and moods as determining which state currently exists, sheds additional light on the process by which trust spirals down or dissolves.

At the beginning of a relationship, for example, if there appears to be no possibility of experiencing trust between the parties because obvious value differences preclude the ability to take the role of the other, trust will not develop; and distrust will be engendered if obvious value incongruence exists (Sitkin & Roth, 1993). Assuming that trust does come into being, and conditional trust exists, if during the subsequent exchange process, discrepant behavioral exchanges or the violation of mutually agreed upon expectations occurs, trust is reduced (Sitkin & Roth, 1993), putting the trust relationship in jeopardy. In terms of the present model, what were once favorable attitudes toward the other, become increasingly unfavorable, positive affect disappears, and negative affect is experienced. Thus, conditional trust can quickly spiral down and dissolve and distrust will appear at the point where the ability to suspend belief that the other is trustworthy is lost. Thus, because of our interactionist perspective, our model makes distrust, as well as the dissolution of trust, the outcome of violated expectations.

If the experience of trust has reached the unconditional state, however, the dissolution process becomes more complex. When unconditional trust exists, the exchange relationship is infused with meaning and positive affect deriving from the sharing of values. Now short-term behavioral lapses by the other party are likely to be forgiven because shared values orient the parties to the future and condition the prospective exchange relationship. Behavioral lapses, therefore, may not dissolve the trust bond but emotional outbursts will be likely since they are the means through which the experience of broken trust and its accompanying negative affect is signaled to both parties. The wronged party’s negative emotions signal to him or her that the relationship is in need of attention (Frijda, 1988), while the expression of these feelings via emotional outbursts signals to the other party that a lapse in trust has occurred. Thus, emotion is the signal that triggers a reappraisal of the relationship; it brings to consciousness again the question of whether or not one party might be put at risk by the actions of the other. Responses of both parties to this signal, therefore, can quickly change the quality of the experience of trust, and potentially destroy it. In terms of Figure 1, emotional outbursts may cause unconditional trust to shift to conditional trust but, under certain circumstances, emotional outbursts can also cause unconditional trust to shift to distrust.

The strength of the perceived violation that puts trust to the test is a key factor in determining whether unconditional trust shifts to conditional trust or distrust. Sometimes, the magnitude of the discrepant behavior can be so enormous that it precipitates an immediate collapse of trust—the injured party is assured that discrepant values exist, attitudes become unfavorable, and negative emotions signal the end of the trust experience. For example, a small business owner who discovers that a trusted employee has been embezzling funds may experience an immediate breakdown in the experience of trust. In this case, strong negative emotions (Lewicki & Bunker, 1996) are accompanied by a realization of value incongruence and the development of a very negative attitude toward the other party; distrust is immediately engendered (Sitkin & Roth, 1993).

Other times, repeated lapses may be perceived to be minor enough that they can be forgiven in the context of shared values and favorable attitudes, and while trust may now become conditional, it can also spiral up to become unconditional again, as discussed earlier. Indeed, one of the purposes of the emotional outburst, and thus a key role affect plays in the model, is to signal strongly to the other party the need to change specific behaviors in order to rebuild the experience of unconditional trust that has been temporarily lost. Over time, if such changes in behavior do not occur, then negative moods and emotions begin to pervade the relationship and attitudes evolve to be more and more unfavorable in line with observed behavioral lapses. When the change in attitudes, accompanied by negative affect, changes the perceptions of shared values, the interactions between these constructs precipitates a downward spiral and unconditional trust turns into conditional trust. Finally, when trust has shifted from unconditional to conditional and then further deteriorates, a person no longer can take the role of the other and believe in the other’s trustworthiness and the state of conditional trust shifts to distrust. Common expressions such as "that was the last straw" capture how a wronged party, who has sought to recreate or perpetuate a trust relationship through successive behavioral lapses, eventually finds it impossible to keep taking the role of the other and distrust appears as parties feel they have reason to believe they might by exploited by the other in the future. Lacking the ability to take the role of other, trust cannot be reestablished until either both parties willingly renegotiate the relationship or the injured party willingly reconciles him or herself to the violation(s) and is able to restore the former positive attitude toward the other and perceptions of shared values. Our model thus differs from others because all three factors—values, attitudes, and moods—are needed to analyze the transformation of trust between each of the three states in Figure 1, conditional trust , unconditional trust, and distrust.

Forms of Trust and Organizational Cooperation

As noted earlier, it has been widely acknowledged that trust can lead to cooperative behavior in organizations (Axelrod, 1984; Mayer et al., 1995; McAllister, 1995). Using the model of trust developed above, it is now possible to reconsider the relationship between cooperation and trust and examine how conditional and unconditional trust may lead to different kinds of cooperative behavior in organizations. As a case in point, we focus here on two important kinds of cooperative behavior in organizations, interpersonal cooperation and teamwork and the creation and sharing of organization-specific tacit knowledge that, as noted earlier, are receiving increasing attention in the literature. One of the main implications of the model of trust developed here is that the nature of cooperation will be quite different depending on whether conditional or unconditional trust exists in an exchange relationship.

Interpersonal Cooperation and Teamwork

The importance of interpersonal cooperation and teamwork in organizations has been receiving increasing attention due, in part, to the acknowledged importance of helping and other forms of cooperative behavior in the literature on extra-role behavior and organizational citizenship behavior (e.g., Katz, 1964; Organ, 1988; Organ & Ryan, 1995; Van Dyne, Cummings, & Parks, 1995). Additionally, the increased use of self-managed work teams in organizations, combined with the elimination of middle management positions as a result of organizational restructuring, has also highlighted the importance of interpersonal cooperation and teamwork for organizational effectiveness (Cohen, Ledford, & Spreitzer, 1996; Dunphy & Bryant, 1996).

While both conditional and unconditional trust may result in interpersonal cooperation and teamwork, we propose that the nature of cooperation will be fundamentally different under these two conditions of trust. As noted earlier, conditional trust arises when parties can take the role of the other and developing attitudes are favorable enough to support future interactions. As such, the experience of conditional trust is sufficient to facilitate many kinds of social and economic exchanges such as those that exist between coworkers in organizational settings or between business acquaintances (Bateson, 1988). When unconditional trust exists, however, people begin to feel that they are not mere coworkers or business acquaintances but colleagues, friends, or team members. In other words, the existence of unconditional trust fundamentally changes the quality of the exchange relationship. As Mead (1934:276) notes, "The sense of teamwork is found when all are working towards a common end and everyone has a sense of the common end interpenetrating the particular function which he[she] is carrying on." In other words, while the presence of conditional trust allows a group to work towards a common goal, unconditional trust converts a group into a team. With teamwork, what one person does is determined by what all others are doing, and the parties must be constantly alert to the way others are behaving to be able to respond appropriately. When unconditional trust infuses the relationship between people,

shared values result in strong desires to cooperate, even at personal expense. Shared interpretive schemes, derived from the exchange of values, attitudes, and moods and emotions, support self-sacrifice for the collective good.

In terms of the model of trust developed here, when unconditional trust exists, values, in addition to attitudes, are infusing the exchange relationship and people signal their good intentions to one another, promoting positive affect and strengthening the trust bond. Shared values determine behavioral expectations as people are investing in their relationship and looking more to the future than the present when deciding how to behave (Dasgupta, 1988). Through interpersonal cooperation and teamwork, people signal their shared values and strong desires to contribute to the common good (Kruglanski, 1970). In turn, the cooperative acts themselves often make people feel good and stimulate others to act in a similar fashion reinforcing shared values and positive attitudes and affect.

If only conditional trust exists, on the other hand, and attitudes are primarily governing the exchange process, parties to an exchange are likely to cooperate to maintain their own good standing in the eyes of others. However, because they lack the same stake or investment in the relationship that is present when unconditional trust exists, cooperation which entails considerable personal costs (e.g., time and effort) or self-sacrifice (e.g., setting aside personal gain for the greater collective good) is less likely to occur; the shared values that orient the parties to the future are absent so that positive affect is less likely to infuse relationships. Thus, under conditional trust, ongoing cooperation is based on the continued exchange of acceptable expectations determined by ongoing attitudes whereas with unconditional trust, values, as well as attitudes, infuse the relationship so that people feel more invested in the relationship and cooperation is more intensive and mutually reinforcing.

In support of the advantages of unconditional trust for interpersonal cooperation and teamwork, we propose that the sharing of values characteristic of unconditional trust will also lead to broader role definitions, communal relationships, and more help-seeking behavior in organizations. Broader role definitions, communal relationships, and help-seeking, in turn, also are promotive of interpersonal cooperation and teamwork (e.g., Anderson & Williams, 1996; Clark, Ouellette, Powell, & Milberg, 1987; Morrison, 1994). Hence, unconditional trust is proposed to have both direct and indirect effects on cooperation and teamwork (Figure 2).

Insert Figure 2 About Here

How broadly or narrowly individuals define their work roles has been shown to influence their levels of cooperative behaviors such as organizational citizenship behavior (Morrison, 1994). When unconditional trust exists, the sharing of the values which underlie trust is likely to lead to individuals broadly defining their roles. Shared values will result in individuals wanting to cooperate and do whatever they can for the common good; hence, roles will be defined broadly to include whatever acts they are capable of performing which contribute to common goals.

Many exchange relationships in organizations are driven by a quid pro quo orientation in which individuals cooperate and help each other to compensate for past help received or in anticipation of future help needed. Cooperation driven by conditional trust is likely to involve such a calculative mentality and as long as behavioral expectations are realized, attitudes sustain the cooperative behavior. These kinds of exchange relationships have been contrasted with communal relationships (Clark & Mills, 1979; Mills & Clark, 1982). In a communal relationship, people not only feel obligated to help each other and cooperate but also want to and feel responsible for doing so, leading to increased helping and cooperation (Clark et al., 1987). The shared values underlying unconditional trust guide people to strive for communal relationships characterized by helpfulness and responsibility and contribute to the development of such relationships. Communal relationships, in turn, are likely to promote interpersonal cooperation and teamwork (Clark et al., 1987).

Lastly, whether or not cooperation and teamwork occur depends upon there being opportunities for their occurrence. While help-seeking can lead to cooperative behavior (Anderson & Williams, 1996), there are sometimes powerful constraints against seeking help. For example, people may refrain from seeking help because they do not want others to think they are inadequate, they do not want to be indebted to another person, or they do not want to feel threatened by being dependent upon another person (e.g., Brehm, 1966; Nadler, 1991; Greenberg, 1980; Walster, Walster, & Berscheid, 1978). Unconditional trust is likely to remove these constraints against seeking help. When trust is unconditional, feelings of indebtedness are irrelevant as shared values underlie cooperation and cause people to want to cooperate. Seeking help is not threatening under unconditional trust because interdependence is seen as a positive force, not something to be avoided, and shared values and positive attitudes ensure against attributions of inadequacy. Hence, we propose that individuals will be much more likely to seek help under conditions of unconditional as opposed to conditional trust. And, seeking help contributes to cooperative interpersonal behavior and teamwork (Anderson & Williams, 1996).

The Development of Organization-Specific Tacit Knowledge

Recently, many researchers have argued that one of the sources of an organization’s competitive advantage stems from the extent and richness of its tacit knowledge; knowledge that is unique or specific to an organization which contributes to superior performance (Prahalad & Hammel, 1990; Wernerfelt, 1984). More specifically, tacit knowledge is the unspoken, implicit knowledge, present in the interactions among organizational members and embedded in their exchanges, that contributes to organizational effectiveness (Polyani, 1962, Kogut & Zander, 1992). In trying to understand how tacit knowledge is created, researchers have focused on an organization’s capabilities, and in particular, on the extent to which an organization is able to effectively mobilize its human resources. For example, Amit and Schoemaker (1993:35) argue that an organization’s capabilities derive from "developing, carrying, and exchanging information through the firm’s human capital." Capabilities are embedded in the skills and knowledge of organizational members and the interactions between them, particularly in teams of the kind discussed above. Similarly, researchers have suggested that accumulated interactions and exchanges between members of different functions (Grant, 1991; Prahalad & Hamel, 1990) enhance an organization’s capabilities and gives rise to tacit knowledge or knowledge-based "invisible assets" (Itami, 1987; Nelson & Winter, 1982; Winter, 1987).

One of the central characteristics of organizational capabilities that contributes to their value and the generation of tacit knowledge is the extent to which they are unique and cannot be imitated. To the degree that other organizations can copy or mimic a focal organization’s capabilities, these capabilities will not be a long-term source of competitive advantage. Hence, attention has been focused on the identification of the "isolating mechanisms" which make it difficult or impossible to imitate organizational capabilities (Rumelt, 1984). One of these isolating mechanisms is trust (Barney & Hansen, 1994). Based on the analysis of trust developed in this paper, we propose that the development of the unique capabilities that lead to tacit knowledge depends on the existence of unconditional, rather than conditional trust, between people.

Conditional trust develops as people exchange information and develop positive attitudes

toward each other that allow them to take the role of the other. Through conditional trust, people develop stable expectations of each other that routinize their interactions and make them predictable and reliable. However, the emergence of these stable, routine interactions is a necessary but not a sufficient condition for the emergence of tacit knowledge. This is because such routine interactions can be systematized into sets of rules and procedures which, by definition, can be copied and imitated by other organizations (Granovetter, 1985; Mahoney, 1996). While the organizational capabilities and learning which stem from such interactions are important, they are not the unique kinds of capabilities that lead to a sustainable competitive advantage because they can be bought or transferred to another organization.

We propose that in order to have the capability for real synergy among organizational members that leads to the development of tacit knowledge which cannot be translated into rules or routines, unconditional trust is required. Synergy is achieved when knowledge and information are freely exchanged, one’s own ego and needs are subjugated to the pursuit of a common goal, a high level of confidence is placed in others, and interactions are intense (Ghemawat, 1991; Mead, 1934; Simmel, 1950). Unconditional trust and the sharing of values is necessary for synergy for at least four reasons. First, knowledge and information are not likely to be exchanged freely when a party cannot be sure about the moral basis of other parties’ actions or the values that are prospectively guiding their behavior. Positive attitudes characteristic of conditional trust will promote the sharing of knowledge to accomplish ongoing tasks but little more (Williamson, 1985). When people are unsure about others’ values, they will refrain from free exchange of knowledge and information because there is uncertainty concerning how others will use this information (e.g., for their own benefit) and because possession of knowledge (particularly if it is hard to come by) is a source of power (Fama & Jensen, 1983a,b). Under conditions of unconditional trust, however, shared values underlying trust provide the assurance that knowledge and information will be used for the greater good, and that one need not exercise power to protect one’s own interests (Figure 3).

Insert Figure 3 About Here

Second, individuals are unlikely to subjugate their own needs and egos to pursue a common goal when only conditional trust exists because they cannot be sure if others will be doing the same. There is always the potential that some individuals will passively shirk or free ride on the efforts of others (Holmstrom, 1979) or that individuals will seek their own gain at the expense of the greater good (e.g., Alchian & Demsetz, 1972; Hill, 1990). The sucker effect (e.g., Jackson & Harkins, 1985; Kerr, 1983) suggests that, in such circumstances, people will withhold of themselves so as not to be taken advantage by others. However, the shared values underlying unconditional trust provide individuals with greater assurance that others will act in good faith and will be guided by the same prospective standards as oneself. Hence, unconditional trust is more likely to lead people to subjugate their own needs and egos for the greater good because of the greater assurance that others are doing the same.

Third, shared values underlying unconditional trust also provide individuals with the high degree of confidence in each other necessary for synergy. Positive attitudes characteristic of conditional trust may provide a certain degree of confidence in others, but this confidence is guarded in that one can never be sure about another’s real intentions or ultimate goals (Dasgupta, 1988). The sharing of values promotes high confidence in others because one can be assured of others’ ultimate intentions and goals or objectives (Bateson, 1988). Hence, people are more likely to make the organizational-specific investments that result in synergies.

Lastly, intense interactions are both effortful and can leave one vulnerable. Vulnerability can come from, for example, investing in a course of action, intensely arguing a position, and then needing to acknowledge it was mistaken, admitting ignorance or a lack of understanding, or having one’s own arguments used against one. The shared values underlying unconditional trust provide people with the assurance that their efforts and contributions will be recognized, put to good use, and reciprocated. Consequently, vulnerability is a much less relevant issue because of

the good intentions and prospective standards the parties are abiding by. Under conditional trust, however, simply having positive attitudes towards others may not be enough to warrant the high involvement and vulnerability characteristic of intense interactions.

The tacit knowledge that results from real synergy is often taken for granted by organizational members (Itami, 1987; Polyani, 1962). But it is difficult and extremely time consuming to translate into standards or procedures or to explain to people who are not involved in the interactions themselves. Thus, the intense interactions in teams, facilitated by unconditional trust, are both the generators and actualizers of tacit knowledge. The positive affect that envelopes relationships based on unconditional trust also is likely to contribute to synergy and the development of shared tacit knowledge. For example, research has found that positive affective states promote social interaction and creativity (e.g., Clark & Watson, 1988; Isen & Baron, 1991; Isen, Daubman, & Nowicki, 1987), two important contributors to tacit knowledge development.

Thus, we propose that when unconditional trust is present in relationships, organizational members are more likely to develop and accumulate a store of tacit knowledge which can give an organization a competitive advantage. A real source of competitive advantage deriving from organizational capabilities is an organization’s ability to create the conditions that allow its members to experience unconditional trust and thus build up a stock of shared tacit knowledge.

Conclusions

We believe that the theoretical distinction between conditional and unconditional trust is meaningful and, rooted in the interactions between values, attitudes, and moods and emotions, has the potential to explain why organizational cooperation does and does not occur, as well as different kinds and degrees of cooperation. Conditional trust is likely to be sufficient to promote those cooperative behaviors that do not entail significant personal costs or self-sacrifice. However, why would members of an organization perform cooperative behaviors that are arduous and time consuming and entail considerable self-sacrifice and no tangible reward? When unconditional trust exists, we propose that people want to perform these acts, and that the psychological experience of unconditional trust, accompanied by the positive feelings which flow from its actualization, is its own reward. At the organizational level, the reward from unconditional trust is the competitive advantage that accrues from an organization’s ability to reap the value added produced by teamwork, synergy, and the development of tacit knowledge.

However, the development of trust (conditional or unconditional) is a function of an organization’s ability to create the setting within which trust can develop over time. Does the work environment and context promote positive attitudes and moods and emotions? Does the organizational culture endorse and encourage the expression of the values underlying trust (Fiol, 1991)? Are individuals given the opportunity to explore shared values? Does the organization’s structure provide the appropriate set of task and reporting relationships that facilitates the development of positive attitudes and moods?

An important distinction needs to be made between organizations which strive to create unconditional trust versus organizations which are content simply to operate with a predominance of relationships built on conditional trust. While significant advantages such as heightened cooperative behavior may accrue from unconditional trust, unconditional trust does not come without costs. Unconditional trust takes time and effort to build and maintain. For example, a relationship of unconditional trust between a supervisor and a subordinate necessitates that the supervisor is supportive of the subordinate and looks out for the subordinate’s well-being (Eisenberger, Fasolo, & Davis-La Mastro, 1990). Some organization may not be willing to make investments in their members or provide the kinds of organizational support which would promote the development of unconditional trust. Or some organizations may want or need the flexibility to lay off employees as economic conditions warrant. Under these kinds of conditions, unconditional trust is not a viable option and conditional trust will remain the most common state

of trust found in organizations.

References