The cancellability test for conversational implicatures

Abstract

Many people follow Grice in thinking that all conversational implicatures are cancellable. And often enough, they use this insight as a test for conversational implicatures. If you want to find out whether something is a conversational implicature, the test has it, you should ask yourself whether the thing in question is cancellable; if you find that it is not cancellable, you can infer that it is not a conversational implicature. If you find that it is cancellable, you can infer that it might well be a conversational implicature and that you should now do further testing. Various philosophers and linguists have questioned the test however. Some have held that Grice's cancellability claim is subject to counterexamples and that the test is not reliable. Others have argued that even though Grice's claim can be defended against the examples in question, the test is not as informative as people have hoped. This article provides an overview of the recent discussion on the cancellability test and suggests a way forward.

Processing Non-at-Issue Meanings of Conditional Connectives: The wenn/falls Contrast in German

“…These data taken together indicate that the weak bias created by falls is a secondary, i.e., non-at-issue, content that is lexically triggered but contextually cancellable, and more precisely, an implicature ( Grice, 1975 ; Potts, 2005 ; see Zakkou, 2018 for discussion of the reliability of the cancellability test). Here, a natural question arises whether it is a scalar implicature.…”

Separating the evaluative from the descriptive

“…Arguments in favour of either the semantic or the pragmatic separabilist position heavily rely on linguistic intuitions about how thick terms expressing thick concepts can be used. Fortunately, experimental linguistics provides the means to test the intuition that the evaluative component is merely communicated via conversational implicatures, namely the cancellability test (Grice, 1989, but see Blome‐Tillmann, 2008; Sullivan, 2017; Zakkou, 2018 for discussions of the test's limitations). If the pragmatists are correct and the evaluative aspect is only conversationally implicated, cancelling the evaluation should not lead to a contradiction.…”

Odwoływalność i mówienie nie wprost

“…Jest to chyba jednak mniej istotne niż zwrócenie uwagi na istnienie tego rodzaju pragmatycznego związku. To nie 16 Warto też skomentować ciekawą propozycję Zakkou (2018), zgodnie z którą sąd p komunikowany (conveyed) przez użycie zdania Z przez mówcę w kontekście K jest odwoływalny zawsze i tylko wtedy, gdy istnieje kontekst K* taki, że (i) jest podobny do K przez to, że ten sam nadawca używa Z, oraz (ii) różni się od K tym, że nadawca używa odwołującej frazy  ale nie p  lub  ale nie chcę przez to implikować, że p  w taki sposób, że (a) cała wypowiedź nadawcy jest prawidłowa (felicitous). Taka definicja pozwala jednak uznać za odwoływalną treść semantyczną.…”