A Dual-Mode Model of Cooperation in Risk Management
According to the Japanese social psychologist, Yamagishi, failure to clearly distinguish trust from confidence is a "major source of confusion surrounding discussions of trust." During the past year, we have tried to clarify the distinction between trust and confidence, first, by reviewing a large number of studies drawn from many disciplines and, second, by constructing a general, dual-mode model of cooperation based on trust and confidence. In this brief report, I outline the elements of our model. An expanded treatment, with references, is available.
We assume, as many dual-mode theorists do, that the two processes operate simultaneously and that interaction between the two consists primarily of associative thinking affecting rule-based thinking, though effects in the other direction are also possible; i.e., the associative system can be overruled. One immediate implication of this is that trust should affect confidence, but only when confidence is rule-based ("rational," "analytical," etc.). But when confidence is assumed (i.e., associative), there is no social uncertainty, and trust is irrelevant. Thus, whenever trust is in play, it should affect judgments of confidence.
Our model depicts two pathways to cooperation, the upper via trust, the lower via confidence. At the far left of the model, the information perceived by a person is divided into two types, that which is judged to be relevant to "morality" and that which is judged relevant to "performance." This division of information, although central in studies of impression formation, has been overlooked in most studies of trust and confidence. The importance of this distinction is demonstrated, first, by studies that show that persons tend to organize impressions of others along two dimensions, social desirability (morality) and intellectual desirability (performance), and, second, that morality information tends to dominate performance information. By "dominate" we mean that, to an observer, morality information is more important and that it conditions the interpretation of performance information. For example, given positive morality information, negative performance is judged much less harshly than it would be if the morality information were negative. The elements of our model are aligned in parallel pairs for trust and confidence:
a. Perceived "Amplitude" of (Morality/Performance) Information. The judged degree to which the given information has (morality/performance) implications.
b. Perceived "Valence" of (Morality/Performance) Information. The judged degree of positivity/negativity of the given information. (a. and b. combine to form c.)
c. Attributed (Values/Performance). The (values/performance) attributed by the observer to the other.
d. Salient Values/Salient Performance History. In the case of values, these are the values that are currently salient to the observer--which may be the product of existing social trust relations. In the case of performance, this is whatever history of relevant performance that is currently available to the observer.
(c. and d. combine to form e.) e. Value Similarity/Perceived Performance. Value Similarity is the judged similarity between the observer's currently salient values and the values attributed to the other. Perceived Performance is the observer's interpretation of the other's performance; note that this is a product not only of c. and d., but also of Social Trust, element g., below.
f. General Trust/General Confidence. General Trust is generalized interpersonal trust, the belief that most people can be trusted. General Confidence is the performance-based counterpart of the values-based General Trust: the belief that things in general are under control, uncertainty is low, and events will occur as expected.
(e. and f. combine to form g.) g. Social Trust/Confidence.
h. Cooperation. Any form of cooperative behavior between a person and another person or group of persons, or between a person and an organization/institution.
Among the key features of our dual-mode model of cooperation are these: 1) It shows how social trust is based on morality-relevant information, while confidence is based on performance-relevant information; 2) It shows how, in times of low social uncertainty, when morality information isn't relevant, social trust doesn't play a role in cooperation; 3) It shows how social trust becomes important in times of uncertainty, when morality information is relevant; and 4) It shows how social trust affects judgments of confidence via effects on perceived performance.
A number of important, testable hypotheses can be derived from this model; for example, a reinterpretation of "trust asymmetry," a widely-accepted "fact" of trust--that trust is hard to win, but easy to lose. In terms of our dual-mode model of cooperation, "trust asymmetry" collapses the two modes to one: Instead of morality information and performance information, there is simply information; instead of trust and confidence, with trust affecting confidence, there is simply trust. Most seriously, trust asymmetry ignores the role of existing attitudes and beliefs, which enter our dual mode model as Salient Values and Salient Performance History.
Contact Timothy Earle