A first-order conditional logic with qualitative statistical semantics (Q4381869)

The logic presented here is a qualitative counterpart of \textit{F. Bacchus}'s approach [Representing and reasoning with probabilistic knowledge (MIT Press, Cambridge, MA, 1990)]. Conditionals are interpreted over a standard first-order structure to which an ordering over domain elements is added. A conditional of the form \(\alpha\rightarrow_{(\bar{x})}\beta\) is interpreted as saying that typical (i.e., minimal relative to the ordering) tuples that can be substituted for \(\bar{x}\) and have the property \(\alpha\), have the property \(\beta\) as well. A sound and complete axiomatization of this logic is given. The logic has also a probabilistic semantics akin to the \(\varepsilon\)-semantics (see \textit{E. W. Adams} [The logic of conditionals. An application of probability to deductive logic (1975; Zbl 0324.02002)] and \textit{M. Goldszmidt, P. Morris} and \textit{J. Pearl} [IEEE Trans. Pattern Analysis Machine Intell. 15, 220-232 (1993)]).