On Some Symmetric Constituents in Japanese
Ryoichiro Kobayashi
July 2017

A well-known {XP, YP} problem brought up by Chomsky (2008) is now a hot topic of debate. He claims that every Syntactic Object (SO) must be labeled for interpretation at interfaces. Labels are determined by Labeling Algorithm applying to an SO at the timing of Transfer phase by phase. There are two ways to label such symmetric constituents: Labeling via movement (i) and via feature-sharing (ii). In English, the label of XP = {DP, vP} is determined as vP since the DP internally merges with TP, rendering the lower copy invisible from LA. Another {XP, YP} problem arises with the merger of DP and TP. The label of YP = {DP, TP} is determined by so-called labeling via feature-sharing: the relevant DP and TP share φ-features, which undergo Agree. It labels YP as <φ, φ>. For labeling via feature-sharing, Chomsky (2013) assumes that Agree is necessary between XP and YP. It follows that this strategy is available only in languages with agreement. Then, what would the label of {XP, YP} structure in φ-featureless languages like Japanese (Fukui 1986, 1988) be like? This is the central question the current study will deal with.
Format: [ pdf ]
Reference: lingbuzz/003582
(please use that when you cite this article)
Published in: to appear in the Proceedings of SICOGG 19
keywords: labeling, categorial features, feature-sharing, anti-labeling features, syntax
Downloaded:194 times


[ edit this article | back to article list ]