Arguments, Conflicts, and Decisions - Conflict and Multimodal Communication: Social Research and Machine Intelligence - page 125

relative impact of each factor is also linked to verifiable predictions: if the decision
of truncating the argument is inspired by excessive costs (factor A above), one would
expect that manipulating perceived costs would have an effect on such decision;
if, on the contrary, the choice is due to increased dangers of negative side effects
(factor B), reducing or enhancing the likelihood of such dangers should affect
argument termination; finally, if what matters is duration-induced skepticism on
the chances of reaching an agreement (factor C), the arguers should be sensitive
to manipulations that reduce such skepticism. Moreover, different factors may
also prompt different strategies on how to terminate the argument: for instance, if
argument termination is motivated by a desire to avoid antagonizing the counterpart,
generously conceding the point and negotiating an amicable truce are the most likely
options; on the contrary, if the argument is ended out of dissatisfaction for one's
chances of overcoming opposition, an abrupt interruption is more probable, with no
concessions or niceties toward the counterpart.
Before turning to consider the role of conflict in decisions on argument termi-
nation, it is important to defuse a potential objection to the hypotheses ventured
so far. At this point, in fact, some readers might feel that these predictions on
argument termination, and the proposed analysis of argumentative decisions in
general, rely too much on expected utility theory, which is, after all, a debated
model of human choice (for an early, balanced, and still authoritative review of
its merits and limits, see Schoemaker 1982 ; for a recent discussion of its theoretical
implications, see Secchi 2011 ). However, expected utility is used here only for the
sake of simplicity, since it allows for a very straightforward illustration of what
kind of predictions can be generated on argument termination, once we assume that
arguers have to make a decision about it. But the tenability of a decision-theoretic
perspective on argumentation is largely independent from the particular theory of
rational choice that arguers are supposed to apply. Whereas specific predictions
will change according to different theories of choice, this does not undermine the
case for the relevance of decision making for argumentation. On the contrary, as
extensively discussed elsewhere (Paglieri 2013b ), one of the key advantages of
marrying argumentation and decision making is in the variety of competing theories
of rational choice that become available for modeling the strategic rationality of
arguers.
Still, with respect to argument termination, adopting a different theory of choice
might well result in different predictions. For instance, one could speculate that
arguers decide whether to terminate an argument by comparing the two options
(continue vs. quit) according to various criteria (e.g., quality of the relationship with
the other party, social image, self-esteem, etc.), which they learn by experience
to rank in terms of their importance in guiding their choice, and thus consult
lexicographically: that is, they start the comparison from the most important
criterion, then proceed in order of descending importance, and stop the procedure
as soon as one criterion allows to discriminate between the options (Gigerenzer and
Goldstein 1996 ). So, with an arguer who gives greater stock to preserving good
relationships than to make a strong impression as a competent speaker or boost
his/her self-esteem, any discussion that endangers the quality of the relationship