The Scalar Presupposition of even: Support from the Improvement Effect on Extreme-Adjective Comparatives
Abstract:
In this talk we examine observations, so far unaccounted for, regarding the fact that the even-like particles in several languages improve the degraded status of comparatives based on extreme adjectives (EA-comparatives). (i) and (ii) illustrate that in English:
(i) ? Godzilla is more gigantic than Mothra (cf. Godzilla is larger than Mothra).
(ii) Godzilla is even more gigantic than Mothra.
We argue that this examination can contribute to debates regarding the scalar presupposition of even, and in particular that it supports a degree-based characterization of this presupposition over the traditional unlikelihood-based one. To show that we follow existing claims which attributed the original degradedness of EA-comparatives, as in (i), to violating pragmatic constraints when attempting to distinguish between degrees on non-salient portions of scales, with which extreme adjectives associate. We propose that such violations are avoided in the presence of even, as in (ii), due to the degree-based scalar presupposition it triggers, which requires that its prejacent, p, to lead to a higher degree than its alternatives, q, on a mapped scale associated with a gradable property. With EA-comparatives the accommodation of this presupposed mapped scale supports contexts where the distinctions between extreme degrees become significant, thus eliminating the difficulty of comparing degrees on a non-salient portion of the scale.
We support our proposal with novel observations showing that two other constructions which were independently argued to involve mapped scales as well, are similar to even in also improving the status EA-comparatives. In contrast, we argue that trying to explain the improving effect of even on EA-comparatives using an unlikelihood-based scalar presupposition faces issues and is unable to explain this effect with the two other constructions.
