Definitions of Trust: From Conceptual Components to the General Core - Trust Theory: A Socio-Cognitive and Computational Model - page 24

different readings of the second sentence. I trust you can mean I believe you are trustworthy ,
but it can also mean much more than this: in particular, it can mean I have decided to rely on
you, to depend and count on you, since I believe you to be trustworthy. 19 Trust in our model
is not only a doxastic mental attitude, but also a decision and intention , and the subsequent
action based upon these mental dispositions. There is no contradiction between the so-called
'cognitive' and 'behavioral' notion (acting on trust). In contrast, for Hardin, trust 'if it is
cognitive is not behavioral' (Hardin, 2005). 20 In our model the two notions are embedded one
into the other; and the explicit theory of the relations between these layers and of the transition
from one to the other is important.
Based on this view is the idea that in principle one cannot really decide to trust somebody
else; trust 'is not a matter of choice', insofar as we refer to trust as a doxastic attitude. This
is because, as a matter of fact, one cannot decide to believe something 21 ; but this fact is not
true for the second layer of trust, concerning the decision to delegate and the act based upon
this decision. While trusting you I can really 'decide' to take some risk (which I mentally
consider) and to make myself vulnerable to you. So much so that, indeed, I can later come to
regret my decision, and blame myself for that choice.
It follows that trust can be (and usually is) rational on both these two levels: as a belief, and
as a decision and action. Epistemic rationality consists in forming a justified belief on the basis
of good evidence and reliable sources. Decision rationality consists in taking a decision on the
basis of a correct calculation of values (outcomes) and of their (limited) optimization. In our
model, trust can be rational at both levels: grounded as a belief (evaluation and prediction),
and optimal as a choice.
The last important difference between Hardin's analysis and our model is that Hardin
completely puts aside the issue of 'competence', overlooking all those cases in which we put
our trust and reliance on the trustee's ability, expertise, quality of service, and so on. In his
own words: 'I will usually assume through this topic that competence is not at issue in the trust
relationships under discussion' (Hardin, 2005, Introduction). Unfortunately, this assumption
is untenable, since the competence aspect cannot be either marginalized or rigidly separated
from trust in Y 's reliability (see Section 2.2.5 for more details).
1.5.8 Rousseau: What Kind of Intention is 'Trust'?
A very good (indeed, our favorite) definition of trust, based on a large interdisciplinary literature
and on the identification of fundamental and convergent elements, is the following: '[Trust is]
a psychological state of a trustor comprising the intention to accept vulnerability in a situation
involving risk, based on positive expectations of the intentions or behavior of the trustee'
(Rousseau et al ., 1998).
19 This is the meaning of 'trust' in expressions like: “While trusting Y you expose yourself, you risk very much!”;
“My poor friend, how could you trust him?!”; “Trust me! You can trust me!” “OK, I trust you!”.
20 This claim can be correct if 'behavioral' means a behavioristic definition (like in Yamagishi's approach); but it
is wrong if it is aimed at excluding trust as decision and action. Actually 'Action' is a cognitive notion.
21 Nonetheless, there may be trust attitudes based on 'acceptances' and not on 'beliefs'; and 'acceptances' can be
intentionally assumed. In this perspective, I can also 'decide' to trust you, in the sense that I decide to presume that
you are reliable and competent, and to act on the basis of such an assumption, i.e. “as if” I believed you to be reliable
and competent (Cohen, 1992) (Engel, 1998).