Satoshi Tomioka (Delaware)
Beyond the well-known patterns, the IEs have several layers of complexity, making the phenomenon a moving target. Many of those challenges have been documented. Among the languages that exhibit the IEs, each language seems to have a (slightly) different set of interveners although there are definitely some (e.g., NPIs, negative quantifiers) that are very common. Within a language, not all interveners are equal. In Korean and Japanese, for instance, universal quantifiers induce weaker effects than NPIs. The difference in strength also seems to matter in the amenability of IEs in some structural configurations. Another twist is that IEs re-emerge unexpectedly with marked prosody. For some interveners in Japanese and Korean, IEs are elicited if they get focal accents (no matter where they are placed structurally). We should also consider the languages that do not show IEs. They come in two varieties. Italian, for instance, a Wh-phrase and a fronted focus phrase are incompatible, regardless of the word order. On the other hand, Egyptian Arabic and Amharic seem to lack IEs altogether. In the blurred landscape of IEs, one thing is clear: we have more questions than answers. What generalization(s) should we aim to capture? Is a universal solution possible at all, or should we start with a language-specific solution? Even within one language, a similar question arises: should one shoot for a uniform solution, or is a hybrid account something one should consider? How should we deal with the gradability of IEs? Is it a central or peripheral issue? The talk is much more about asking those questions than giving answers. I will present some new empirical data (including experimental results) to emphasize the importance and urgency of those research questions.