Testing Calculation of Scalar Implicatures in English and American Sign Language

Scalar implicatures have been a subject of much debate concerning their place as part of either the grammatical computation of the semantics of a sentence or as post-compositional pragmatic reasoning. This paper describes three experiments addressing the semantic/pragmatic question of scalar implicatures from a developmental perspective, one in English and two in American Sign Language (ASL). The purpose of this work is three-fold: first, to test previous results in spoken language adult populations with a new experimental methodology to quantify scalar implicature calculation in ASL; second, to provide the first direct experimental comparison between constructions triggering standard scalar implicatures and constructions interpreted exhaustively, using English and ASL as testing languages; and third, to lay the baseline for and present results investigating the developmental issue of whether delayed first language acquisition affects semantic/pragmatic skills by testing deaf non-native signers on the ability to calculate scalar implicatures. Results show that adult native speakers of English and deaf adult native signers of ASL both interpret underinformative sentences in the constructions studied as having a scalar implicature. Underinformative sentences were accepted more often in constructions interpreted exhaustively then constructions which trigger standard scalar implicatures. Preliminary results are also presented for proficient deaf adult non-native signers of ASL, who behaved like deaf adult native signers of ASL, showing that the delayed first language acquisition of this group did not affect ability to interpret sentences with scalar implicatures in a native-like fashion.