ABSTRACT
The new paradigm psychology of reasoning is characterised by using proba-
bility theory as a rationality framework for human reasoning. In contrast, classical bivalent logic dominated the traditional psychology of reasoning
(Evans, 2012; Evans & Over, 2012). In this paper I advocate mental probabil-
ity logic (MPL), which emerged within the new paradigm psychology of rea-
soning (e.g., Pfeifer, 2006, 2012, in press; Pfeifer & Kleiter, 2005b, 2009,
2010). Among the various approaches to probability (Hajek, 2011), MPL selects coherence based probability logic as a rationality framework for human reasoning. MPL and the coherence approach are “Bayesian” in the
sense that probabilities are subjective and conceived as degrees of belief.
Degrees of belief are psychologically appealing as they are not only formally
well defined in probabilistic terms, but they also elicit psychologically plausi-
ble connotations: What people know is evaluated in terms of degrees of
belief, since everyday life reasoning is usually based on uncertain and incom-
plete knowledge. Although people usually do not explicitly mention degrees
of belief, they are implicitly attached to sentences in everyday life discourse. Contrary to other Bayesian approaches in the new psychology of reasoning
(e.g., Harris & Hahn, 2009; Oaksford & Chater, 2007), MPL does not use
Bayes’ theorem as the key ingredient for constructing the rationality models.
Rather than assuming an uncertainrelation between the premises and the
conclusion (often formalised by variations on Bayes’ theorem), MPL investi-
gates the coherent transmission of the uncertainty of the premises to the con-
clusion. Probabilities are attached to the premises, and the relation between
the premises and the conclusion remains deductive. Consider the probabilistic modus ponens as an example:
