Trust, Control and Autonomy: A Dialectic Relationship - Trust Theory: A Socio-Cognitive and Computational Model - page 206

Table 7.1 Mental Ingredients in Weak-Delegation (pseudo-formal description)

a) The achievement of τ is a goal of Ag 1 .
b) Ag 1 believes that there exists another agent Ag 2 that has the power of achieving τ .
c) Ag 1 believes that Ag 2 will achieve τ in time and by itself (without Ag 1 's intervention).
(if Ag 2 is a cognitive agent, Ag 1 believes that Ag 2 intends to achieve τ .).
d) Ag 1 prefers 22 to achieve τ through Ag 2 .
e) The achievement of τ through Ag 2 is the choice (goal) of Ag 1 .
f) Ag 1 has the goal ( relativized (Cohen and Levesque, 1987) to (e)) of not achieving τ by itself.

Table 7.2 Mental Ingredients in Mild-Delegation (pseudo-formal description)

a ≡ a; b ≡ b; d ≡ d; e ≡ e; f ≡ f; (referring to a , b , d , e ,and f as described in Table 7.1)
c ) Ag 1 does not believe that Ag 2 will achieve τ by itself (without Ag 1 's intervention).

g ) Ag 1 believes that if Ag 1 realizes an action α then it is be more probable that Ag 2 intends τ .

But Ag 2 does not adopt Ag 1 s goal that Ag 2 intends τ .

h ) Ag 1 intends to do α relativized to (e ).

We consider 'Potential for relying on' and 'Decision to rely on' as two constructs temporally
and logically related to each other . 23
M-Delegates is the operator for representing mild delegation .
M-Delegates(Ag 1 Ag 2 ) represents the necessary mental ingredients of trust shown in

Figure 2.12 and resumed in less formal way in Table 7.2.
We consider in Table 7.2 ( a , b , c , d and e ) what agent Ag 1 viewsasa Potential for relying
on agent Ag 2 ; and ( f , g and h ) what Ag 1 viewsasthe Decision to rely on Ag 2 . 24
S-Delegates is the operator for representing strong delegation . So the expression S-
Delegates(Ag 1 Ag 2 τ

τ

) represents the necessary mental ingredients of trust as shown in Figure
2.13 and resumed in less formal way in Table 7.3:
We consider in Table 7.3 (a ,b ,c ,d and e ) what agent Ag 1 viewsasa Potential for
relying on agent Ag 2 ; and ( f , g and h ) what Ag 1 views as the Decision to rely on Ag 2 .
For a corresponding analysis of adoption, and for how the kind of interaction between client
and contractor influences the adoption itself see (Castelfranchi and Falcone, 1998).

22 This means that Ag 1 believes that either the achievement of τ or a broader goal g' that includes the achievement
of τ , implies Ag 1 to be dependent on Ag 2 . Moreover (d) implies Ag 1 's goal that Ag 2 achieves τ .
23 As for weak delegation it is interesting to analyze the possibilities of Ag 2 's mind. We should distinguish
between two main cases: Ag 2 knows W − Delegates ( Ag 1 Ag 2 τ )and Ag 2 does not know W − Delegates ( Ag 1 Ag 2 τ ;).
In other words, a weak delegation is possible even if the delegee knows it. Either this knowledge has no effect (the
achievement of Ag 1 's goal is just a side-effect known by Ag 2 ) or this knowledge changes Ag 2 's goal: Ag 2 can either
arrive at spontaneous and unilateral help or to a reactive, hostile attitude.
24 In analogy with what we have said in weak delegation, also in mild delegation we should distinguish between
two main cases about the possible mental states of Ag 2 : Ag 2 knows M-Delegates(Ag 1 Ag 2 τ ) and Ag 2 does not know
M-Delegates(Ag 1 Ag 2 τ ) . So, it is possible to have a mild delegation even if the delegee knows it and if consequently
it changes its own behavior to favor or to hamper the success of it.