The question I raise in this paper is why the Japanese counterpart of the sprouting type of sluicing does not appear to show island sensitivity. I argue that this apparent difference is attributed to another well-established one between these two languages: pro- drop vs. non-pro-drop languages. Thus, sluices are not real instances of the sprouting-type but rather of the merger type, which has an overt correlate in the antecedent clause, i.e., indefinite pro in relevant cases. As substantial evidence for this proposal, I demonstrate that indefinite pro actually functions as a correlate to a sluice in Japanese by examining how it behaves with respect to scopal parallelism. Given this, it is predicted that in Japanese sluicing examples, the implicit correlates that are identified as indefinite pro can take wide scope, unlike truly implicit correlates.

We investigate the nature of the idiomatic gapped right dislocation constructions (RDCs), by examining the distinction between the idiomatic and non-idiomatic appendices. The overt idiomatic correlate in the antecedent clause plays a crucial role in identifying the idiomatic appendix of the gapped RDC. That is, the idiomatic appendix in the gapped RDC must hold the idiomatic correlate in the antecedent context. When the idiomatic appendix in the gapped RDC crosses islands such as a relative clause or an adjunct clause, the gapped RDC cannot preserve its idiomatic interpretation even when the verbal idiomatic expression is given in the antecedent clause or question. This is different from the fact that there is a parallelism between the plain gapped RDCs and the fragment answers, which shows a natural consequence under the standard assumption that the appendices and fragments are derived by ellipsis and are island-insensitive due to the repair effect. Hence, we follow the right dislocation approach for the idiomatic appendices in the gapped RDCs, which resolves the issue of the island sensitivity.

This paper is to argue that the apparently fragmentary answer phrase XP right after the polarity answer particle (PAP) such as ung ‘yes’ or ani ‘no’ is not a run-of-the-mill fragment but a right-dislocated (RD-ed) element. Using negative polarity items and indefinites as a RD-ed element, we show that the PAP itself is also a remnant derived from elision of the answering full clause, which in turn provides a right structural context for right dislocation of another XP remnant. We go on further to show that RD-ed elements in the construction at issue display the same pattern of syntactic behaviors as those in the cannonical RD construction, particularly in terms of island effects, the ‘full’ host clause requirement, Case/voice match, and specificational coordination.

This paper has two goals. First, it aims to carefully examine whether a bi-clausal (movement) approach to -yo attachment in Korean Right Dislocation (RD) constructions is tenable. Yim (2013) argues that -yo marking on an RD-ed element lends support to what is called the 'bi-clausal' approach to RD in Korean. We present an issue and data that solicit reconsideration of Yim's view on the syntax of -yo attachment. In particular, the bi-clausal approach endorsed by Yim is shown to have difficulty capturing two types of coordinate structure, namely interwoven dependency and additive coordination. Second, we demonstrate that the availability of these two types of otherwise puzzling coordination in the RD construction involving -yo marking can be captured under a 'mono-clausal' approach, in conjunction with M-K Park's (2009) midway conjunction analysis, where External Remerge, originally suggested by de Vries (2009), is assumed to play a crucial role in building coordinate structure at issue.

This paper provides a unified analysis for the three types of right displacement in English as the clause-final right-displaced element has in common the function of specifying/elaborating on what has been left referentially indeterminate or blank in the preceding clause. We take the surface make-up of the three constructions at issue to indicate that the right-displaced element has coordinate structure relation with the preceding 'propositional' constituent such as vP or TP. Based on this initial conception of the three constructions at issue, we make three points. First, we argue that these constructions involve clausal coordination and their surface forms are derived by eliding the second conjunct except for the base-generated/ leftward-moved element outside it. Second, we show that the economy of structure-building is at work in the formation of the two coordinate conjuncts in the constructions, accounting for the well-known Right Roof constraint or clause-boundedness condition. Third, we also demonstrate that the ban on P-stranding heavy NP shift of a certain argument element follows from the unavailability of an implicit argument to the relevant position of the first conjunct.