Socio-Cognitive Model of Trust: Basic Ingredients - Trust Theory: A Socio-Cognitive and Computational Model - page 57

additional information and evidence, but just by making our decision: 'I have decided to trust
you, (although )'. This does not only mean that I have decided to rely on you (trust as act ),

but it can also mean that I am assuming something about you (expectations, evaluations), and
that I am testing you out: precisely those assumptions will be confirmed or invalidated by your
behavior (see Section 2.6). Sometimes we even delegate a task to Y , we take a risk, in order to
get information about Y 's trustworthiness; we put him on test.
Trust beliefs obviously can be just implicit ; not explicitly represented and considered in
X 's mind. They can be just presupposed as logical conditions or logical entailment of some
(explicit) belief. Suppose, for example, that X has such an evaluation about Y :' Y is a very
good medical doctor' and suppose that this evaluation comes from Z 's recommendation (and
X trusts very much Z ), or from her direct practical experience. In this evaluation X implicitly
assumes that Y is well prepared (competent on the subject), and also technically able to apply
this doctrine, and also reliable in interaction, he takes care of you and of your problems. All
these evaluations, or these possible pieces of X 's trust in Y (the fact that X trusts Y for being
prepared, for taking care of, etc.; the fact that X trusts in Y 's expertise, attention, etc.), are just
implicit beliefs; not necessarily explicitly derived and 'written' in X 's mind (just 'potential'
(Levesque, 1984; Castelfranchi, 1996; 1997)), and/or not explicitly focused and taken into
account.
It is also important to remark that there are many forms of trust which are not based on such
an explicit and reason-based (argumentative) process we have presented in previous sections.
They are rather automatic; not real 'decisions' or 'deliberations'. These are forms of Trust
'choices' and 'acts' based on some 'default rules' (positive evaluations are just implicit). The
rule is: 'Except you have specific signals and specific reasons for do not trusting Y and rely
on him, trust him'.
So the lack of distrust is the condition for trusting more than the explicit presence of trust
evaluations (see Table 2.2; where 'Not (Believe q)' denotes the absence of such Belief in X 's
mind).
These forms of automatic or by-default trust are very important, not only for characterizing
generalized dispositions of people or affective trust attitudes, but also in other domains. For
example, trust in our own natural information sources (our memory, our eyes, our reasoning)
and, frequently even in social information sources, is suspect-less, is automatic, by-default,
and doesn't need additional justification and meta-beliefs. (Castelfranchi, 1997).
Moreover, to 'believe' is in general not a real 'decision', but is certainly the result of some
computation process based on sources, evidences, and data. But - in a sense - to 'come to
believe' is an act of trust:

...

trust in the belief (you rely on it and implicitly assume that it is valid, true), and
Implicitly trust in the sources and bases of the belief; and
(just procedurally and fully implicitly) even trust in the machinery or 'procedure' which
outputs the belief.

Table 2.2 By-default Trust

IF Not (Believe X Not (Trustworthy Y τ C))

THEN Trust (X Y C τ g X ) will be 'naturally' over the threshold for delegating