Dan Sperber  

The myth of 'scalar implicature'  

So called 'scalar implicatures' (e.g. "some implicatures are scalar" said to carry as a scalar implicature [not all implicatures are scalar]) have become a major topic of investigation at the interface between semantics and pragmatics (in the work of, for instance, Carston, Chemla, Bezuidenhout, Breheny, Chiercha, Fox, Geurts, Horn, Katsos, Kratzer, Levinson, Noveck, Papafragou, Sauerland, Schlenker, Spector). I will argue that there is no such thing as a scalar implicature. The phenomena described as such are a motley, and, in any case, are never scalar and rarely implicatures.