Osamu Sawada (Mie University)The Japanese negative totemo ‘very’: Toward a new typology of negative polarity itemsAbstract:The Japanese intensifier totemo `very' can intensify the degree associated with a gradable predicate. However, totemo can also intensify a negative modal statement. In terms of polarity sensitivity, we can say that the negative totemo is a negative polarity item (NPI). However, the negative use of totemo has certain puzzling properties that typical NPIs do not have. First, unlike typical NPIs (e.g., minimizers, any), it cannot be within the semantic scope of negation. Second, it has to always co-occur with a negative gradable modal expression (ability/epistemic). Third, the meaning of the negative totemo is not at-issue.In this talk, I will argue that the negative totemo is not a logical NPI, which is licensed by negation or downward-entailing/non-veridical operators (e.g., Klima 1964; Ladusaw 1979; Giannakidou 1998). Rather, it is a conventional implicature (CI)-inducing expression/expressive (e.g., Grice 1975; Potts 2005), which intensifies the unlikelihood/impossibility of a given proposition and refuses to update the common ground (the context set) with the at-issue proposition. Notably, the negative totemo is also different from the wide-scope universal type of NPIs (negative concord items), which are also placed above negation at LF (e.g., n-words in Romance languages), in that the former, but not the latter, is not interpreted/licensed in the at-issue dimension.This talk proposes that there is a new class of NPIs—expressive NPIs or, more specifically, oppositive NPIs—which requires a negative element in order to satisfy its use-condition, and shows a source of variation of NPIs in terms of the semantics-pragmatics interface.
今仁生美 (Ikumi Imani) 名古屋学院大学(Nagoya Gakuin University)日本語のwh表現の意味論的・語用論的分析-2種類の選言をめぐって概要:日本語のwh表現に関して頭を悩ませる問題の一つは、(1)と(2)の違いは何かという問題である(以下では「なにかが」のような助詞がつくwh-表現をwh-p、「なにか」のように助詞がつかずいわばfloatする形のwh-表現をf-whと略する)。
(1) なにかが、落ちてきた。(wh-p)
(2) なにか、落ちてきた。(f-wh)多くの文脈で(1)と(2)はほぼ同じ意味をもち、際立った相違がない。しかしながら、たとえば、(3)と(4)を比較すると、その違いがわずかに出てくる。(3) 喉が渇いた。?なにかが、飲みたい。(wh-p)
(4) 喉が渇いた。なにか、飲みたい。(f-wh)本発表では、wh-pとf-whには2種類の選言のメカニズムが働いていることを示し、(3)と(4)のような場合になぜニュアンスの違いが出るのかを明らかにする。
2種類の選言のうちの一つは、集合の分割から生じる選言である。たとえば、「魚」という属性を表す集合は、サバやタイといったセルによって分割することができる(セルの集合を{A, B, C}としたとき、魚aは、AかBかCの属性をもつ)。本発表の主張は、f-whはこのセル上の量化を行うということである(Alonso-Ovalle & Shimoyama (2014), Sudo (2010)に関連する議論がある)。本論のこの考えに従うと、たとえば、(5)は、魚という集合の中のすべてのセル(の中のすべての個体)が釣れたことになり、語用論的にはほとんど意味をなさない。
(5)*魚は、なにも、釣れた。なお、この種類の選言は集合Aのすべてのセルを量化の領域とするためfree choiceの解釈を生みやすく、このことが(3)と(4)の違いに反映することも論じる。
もう一つの種類の選言は、標準的な選言すなわち[A or B]のタイプの選言である。本発表の二つ目の主張は、wh-pはこのタイプの選言であるということである。本発表では、通常(6a)が(6b)よりも自然であることは[A or B]の性質から派生することを論じる。
(6) a. なにか、おっしゃいましたか?
b. なにかを、おっしゃいましたか?