The Economic Reductionism and Trust (Ir)rationality - Trust Theory: A Socio-Cognitive and Computational Model - page 242

Consider, for example, the already (see Chapter 1) cited definition by Gambetta: Trust is the
subjective probability by which an individual, X, expects that another individual, Y, performs
a given action on which its welfare depends .
As declared, we think that this definition stresses that trust is basically an estimation, an
opinion, an expectation: a belief. It is also quite remarkable that there is no reference to
exchange, cooperation, mutuality, Y 's awareness.
However, it is also too restricted, since it just refers to one dimension of trust (predictability),
while ignoring the 'competence/quality' dimension. 10
Moreover, to express the idea of an uncertain prediction it uses the notion of 'subjective
probability' and collapses trust in this notion and measure. This is quite risky since it might
make the very notion of 'trust' superfluous (see below). Clearly enough, the reliability, the
probability of the desired event, has nothing to do with its degree of quality, and we cannot
renounce this second dimension of trust. When we trust a medical doctor we trust both his
expertise, competence, skills and his taking care of us, his being reliable, trustworthy. Trust
is an, at least, bi-dimensional notion (actually - as we have shown - is a multi-dimensional
construct); Agent1 : 'Why don't you trust him? He is very reliable and well disposed'; Agent2 :
'That's true; he is very willing, but is not expert in this domain, is not well prepared'.
8.3.2 Probability Collapses Trust 'that' and 'in'
Trust is not simply the subjective probability of a favorable event: that is, trust 'that' the desired
event and outcome will be realized. If trust is 'the subjective probability by which an individual,
X, expects that another individual, Y, performs a given action on which its welfare depends ',
this does not only mean that the expected/desired event is an action of a given individual Y .
What it does mean is that 'we trust Y ', 'we trust in Y '. There is much more than a prediction
of Y 's action (and the desire of that action or of its result). There is something 'about Y ';
something we think of Y , or we feel towards Y . Is this just the estimated probability of his act
or of the outcome? This definition does not capture some of the crucial kernel components of
the very notion of trust: why we do not just trust 'that' Y will do a given favorable action, but
we trust 'in' Y , and we see Y as endowed with some sort of qualities or virtues: trustworthiness,
competence, reliability.
8.3.3 Probability Collapses Internal and External (Attributions of) Trust
Trust cannot be reduced to a simple and opaque index of probability because internal or
external attribution of risk/success makes very different predictions about both the trustor's
decisions and possible interventions and cautions.
As we saw in Section 2.7.2 one should distinguish between trust 'in' someone or something
that has to act and produce a given performance thanks to its internal characteristics , and the
10 In Chapter 1 we have added the following criticisms to this definition: it does not account for the meaning of 'I
trust Y' where there is also the decision to rely on Y; and it doesn't explain what such an evaluation is made of and
based on: the subjective probability includes too many important parameters and beliefs, which are very relevant in
social reasoning. It also does not make explicit the 'evaluative' character of trust.