Arthur Merin

Neg-Raising, Elementary Social Acts, and the Austinian Theory of Meaning


Arbeitspapiere des SFB 340, Bericht Nr. 78 (1996), 30pp.
DVI (105kb); Postscript (630kb) 1-up; Postscript gzip-komprimiert (140kb) 1-up , 2-up.

Abstract

A quirk of natural languages, noted by St. Anselm, Quine and Hintikka, is `neg-raising' (NR): the apparent self-duality, with respect to negation, of epistemic or deontic complement-taking predicates including believe, seem, expect, should, intend, want.(E.g. `I don't think that A' being near equivalent to `I think that not-A', etc.) Larry Horn discovered that the items in question are `mid-scalar' on scales of `strength'. This notion is shown to be still in want of further explication. Flaws in Horn's standard 1978/1989 account of Neg-raising are demonstrated. A directly act-denoting, kinematic semantics is proposed for the expression paradigms involved. It hinges on a paradigm of Elementary Social Acts---claim, concession, denial, and retraction---defined by decision-theoretic parameters of contexts, their transformations, groups under composition, and quotient groups. `External' and ` internal' occurences of not map to context- and act-type-changing operations defined in terms of dominance (a.k.a. warrant) and preference. Mid-scalars are explicated as instances of a quotient act-type `suggestion', also representable as a mixture, probabilistic or otherwise, of claim and concession. The analysis predicts the effects of external negation on `weak' and `strong' members of scales, as well as NR for mid-scalars. Rather than rely on metaphoric appeal to operator paradigms in modal logic it offers a formal rationale for the obvious analogy. It exemplifies thereby how a semantics for speech acts, directly in terms of operations on context (and thus arguably in the performative spirit of J.L. Austin), will condition much-debated linguistic data, bypassing the traditional route via truth-conditions. The modal actually and its usage in Leibniz are briefly addressed along the way.
Institut für Maschinelle Sprachverarbeitung
Formale Logik und Sprachphilosophie
University of Stuttgart
Azenbergstr. 12
70174 Stuttgart
Germany
arthur@ims.uni-stuttgart.de