Definitions of Trust: From Conceptual Components to the General Core - Trust Theory: A Socio-Cognitive and Computational Model - page 7

This is implicit when a given definition of trust mentions some 'positive' result/outcome, or the
'welfare' or 'interest' of the trustor, and also whenever the notions of 'dependence', 'reliance',
or 'vulnerability' are invoked. In fact, something can be 'positive' for an agent only when
this agent has some concern, need, desire, task, or intention (more generally, a goal), because
'positive' means exactly that the event or state or action is favorable to or realizes such a goal -
whereas 'negative' means the opposite, i.e. a threat or frustration of some goal.
Analogously, whenever it is observed that the trustor makes her/himself vulnerable to the
trustee (see for instance (Mayer et al. , 1995)), the unavoidable question is - vulnerable for
what? Alongside other costs that are intrinsic in any act of reliance, the trustor becomes
especially vulnerable to the trustee in terms of potential failure of the expected action and
result: the trustee may not perform the action

α or the action may not have the desired

result O .
Moreover, it is precisely with reference to the desired action/result that the trustor is 'de-
pendent on' and relies upon the trustee. Also in the famous definitions provided by (Deutsch,
1985) where trust is relative to an entity 'on which my welfare depends', the goal of the trustor
is clearly presupposed, since the notion of 'welfare' refers to the satisfaction of her needs and
desires.
Building on these observations, in the following we will extensively argue for the necessity
of the trustor to be actively concerned , i.e. to have goals at play in the decision to trust someone
or something, as well as full expectations 4 rather than mere beliefs on the future (forecasts).
Another aspect frequently missed is that trust is an evaluation , and more exactly a positive
evaluation about Y . In the cognitive architecture developed by (Miceli and Castelfranchi, 2000;
Castelfranchi, 2000), an evaluation (which is also an attitude) is a belief about some power
(capacity, ability, aptitude, quality; or lack of capacity ) of Y relative to some goal . Thus the

beliefs about Y being able and willing to do his share for achieving O are in fact evaluations
of Y : positive when there is enough trust, negative when there is mistrust and diffidence.
Here it is important to appreciate the intimate relation between 'beliefs', 'expectations'
and 'evaluations'. In the case of trust, the beliefs on the competence and willingness of Y
are both, and at the same time, parts of expectations (since they are about the future) and
evaluations (since they are about Y 's powers and inclinations); moreover, they are positive
both as expectations and as evaluations, insofar as agent X is expecting from and attributing
to Y an attitude and a subsequent performance that is in X 's best interest. In addition, X might
ground the decision to trust also on other positive evaluations of Y , for example, intelligence,
honesty, persistency (on this point, see also Chapter 2).
It is worth noticing that the characterization of trust as a structure of mental dispositions is
not in contrast with the analysis of trust as a decisional process culminating into an action -
quite the contrary. Indeed, it is rather pointless to dispute whether trust is a belief or an act ,
opposing the view of trust as an evaluation or expectation to the idea that trust is a decision or a
behavior (for example, of making oneself vulnerable, of risking, of betting on Y ). The point is
rather that trust has both these meanings , which stand in a specific structural relation with each

...

4 As detailed in Chapter 2 (see also (Castelfranchi, 2005)), by expectation we mean the functional integration of a
belief on the future (forecast) with a goal, i.e. a motivational state. In short, an expectation (either positive or negative)
is defined as the prediction that a state of the world which constitutes one of the agent's goals will either be realized
or not in the future. Thus, both the goal that p and the belief that, at some time in the future, p will (or will not) be the
case are necessary components for expecting that p (or not-p ).