![Sample our Language & Literature journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5941/TOC-Subject-Sample-2021-Language-Literature.jpg"})
ABSTRACT
This article examines wh-questions with conjunction, i.e. ATB-questions and wh&wh-questions, both of which have been analyzed as elliptical structures in a few different languages. Based on the morphosyntactic and semantic properties of these structures in Greek the article argues against an analysis employing ellipsis. Moreover, it proposes that these complex wh-structures involve multidominance. Specifically, one or more syntactic objects are remerged with Parallel Merge, but they are spelled-out only once, giving rise to discrepancies between meaning and form. According to this approach, ellipsis in wh-questions with coordination can be an epiphenomenon of multidominance.
Disclosure statement
No potential conflict of interest was reported by the author.
Notes
1 Glossings of Greek sentences follow Leipzig conventions, additional category COMP ‘comparative’. Note that in Greek, GEN apart from other functions (possession) also marks indirect objects.
2 The complement of the verbs proiγúme ‘precede’ and épome ‘follow’ in (8a) is a genitive DP. By contrast, the verb akoluθó ‘follow’ in (8b) requires an accusative complement.
3 I.e., the ungrammaticality which arises when Condition C of Binding Theory is violated (Chomsky Citation1982, Citation1993).
4 (13) is acceptable if the possessive clitic tu/tis ‘his/her’ refers to an extrasentential antecedent.
5 An anonymous reviewer finds (19) less acceptable than the equivalent structure with single-word wh-phrases. Such a judgement could be attributed to the fact that in wh&wh-questions there can be no intonational break before the conjunction, which might explain the preference for a single-word wh-phrase preceding the conjunction.
6 Gračanin-Yüksek (Citation2007, 140) uses the term ‘it-reading’ for the interpretation in (20b).
7 It should be mentioned that there are other studies which resort neither to ellipsis nor to MD (one of the most recent being Blümel Citation2017). Due to space limitations, I concentrate in what follows on the analyses employing ellipsis.
8 In what follows, elided constituents are enclosed by angle brackets. Deletion of lower copies is indicated by strikethrough.
9 On the contrary, CitationHa’s (2008a) analysis is asymmetric, whereas Salzmann’s (Citation2013) allows for consistent reconstruction in the first conjunct but only partial reconstruction in the second. For further details, the interested reader is referred to the works cited.
10 These include an ES feature for sluicing (Merchant Citation2004, 670), an ERNR feature for Right Node Raising (Ha Citation2008b), an EVPE feature for VP-ellipsis (Aelbrecht Citation2010, 174), an EPSG feature for pseudogapping (Gengel Citation2013, 131), an EMCE feature for modal complement ellipsis (Aelbrecht Citation2010, 124) and an Ef feature for fragment answers (Merchant Citation2004, 675).
11 I.e., the appearance of a pronoun in an ellipsis site, in the position corresponding to an R-expression in the antecedent (Fiengo and May Citation1994).
12 There are speakers who find (32a) better than (14b). I attribute this to the adjacency of the wh-DO to the verb in (32a).
13 Also called ‘external remerge’ (de Vries Citation2009) and ‘horizontal sharing’ (Gračanin-Yüksek Citation2007).
14 Regarding the structure of &P, I follow the view that the two conjuncts occupy the specifier and the complement position within the phrase headed by the coordinator (Zoerner Citation1995, among others). At the same time, though, I take this phrase to be exocentric in the sense of Chomsky (Citation2013). For example, the wh&wh-question in (35) is a CP rather than a &P.
15 There is an intermediate landing site for the wh-phrase, i.e., the left periphery of the two vPs. For expository reasons, this intermediate position, which would involve both Merge and Remerge, is not depicted in the diagram.
16 Assuming vP is the locus of voice. By contrast, the derivation proposed by Citko (Citation2011, 55) for English involves shared T and v heads, because Tense and voice mismatches are not attested in this language.
17 The alternative approach taken by Agüero-Bautista (Citation2001) is also compatible with the current analysis.
18 Asarina (Citation2011) offers an alternative account in the framework of Distributed Morphology (DM) (cf. Bjorkman Citation2013). Her analysis of RNR constructions in Russian combines MD with DM.
19 Multiple Agree has been defended by Grosz (Citation2015) for RNR, where one Probe agrees with two Goals (cf. Hiraiwa Citation2001). The derivation in (35), however, instantiates multiple Agree between two Probes and one Goal. I leave this case open for future research.
20 In TP1 the verb merges with the wh-DO ti ‘what’, thus forming VP1. In VP2 the verb functions as a VP, to which the adverb is adjoined. The possibility for an element to share XP and X⁰ properties is allowed along the lines of Chomsky’s (Citation1995) bare phrase structure.
21 The intermediate landing sites of the wh-phrases are ignored here as well, as they do not affect the analysis.
22 For the sake of clarity, multiply dominated constituents are repeated in both CPs.
23 Note that we may not assume that the wh-subject pjos in (42) or who in (43) is shared by the two CPs, as this would give rise to a multiple question in CP2. However, the interpretation of wh&wh-questions is clearly incompatible with a multiple-question reading of CP2. An anonymous reviewer observes that (42) differs from (37)/(38) in that its two CPs do not share all their nodes below CP-level, since pro appears only in CP2. This difficulty can be overcome if one assumes that in a well-formed wh&wh-question all nodes other than the wh-phrases have to be shared, provided that they are phonologically realized (for a discussion of this topic, see Sinopoulou Citation2019b).
Additional information
Notes on contributors
Ourania Sinopoulou
Ourania Sinopoulou holds a BA in Classics, an MA and a PhD in Theoretical Linguistics, all from the University of Athens. In her doctoral dissertation she investigated wh-structures with coordination in Greek. Her main interests include wh-structures, movement, and multidominance. At present, she is an academic fellow at the University of Athens.