Toward a uniform analysis of short answers and gapping

Ellipsis and information structure: Evidence from Romanian gapping

“…This question-answer congruence will serve us as a diagnostic tool to investigate the information structure of coordination with iar. In 3.2 above, we observed the exact same pattern: gapping is appropriate in contexts where the sequence is interpreted as a pair-list answer to an implicit wh-question (Kuno 1976, Steedman 1990, Reich 2007, Winkler 2005, Repp 2009, Johnson 2014 Third, the conjunction iar is only felicitous with symmetric discourse relations. These relations are called resemblance relations (Kehler 2000(Kehler , 2002, parallel relations (Asher and Lascarides 2003) or similar relations (Repp 2016).…”

“…It is not surprising that many scholars propose a formal analysis in terms of QUD (e.g. Reich 2007, Park 2016) rather than a syntactic-based analysis.…”

“…This QUD perspective, based on question-answer congruence, has been proposed as test for the discursive appropriateness of gapping (Kuno 1976, Steedman 1990, Reich 2007, Winkler 2005, Repp 2009, Johnson 2014: gapping would be felicitous if and only if it can answer a multiple wh-question; the discursive pattern in a gapping construction would be a pair-list answer to an implicit wh-question (12). See, in this respect, the quotation from Steedman (1990: 248): "even the most basic gapped sentence, like Fred ate bread, and Harry, bananas, is only really felicitous in contexts which support (or can accommodate) the presupposition that the topic under discussion is Who ate what."…”

“…Two answers can be given: (i) if 7 See also the so-called No Embedding Constraint (Johnson 2014) which does not allow the gapped clause to be embedded within the conjunct to which it belongs: Bîlbîie and de la Fuente (2019) observe that some languages (such as Spanish) accept embedded gapping with non-factive verbs, but less with factive verbs; the explanation is related to a QUD match/mismatch rather than to a syntactic constraint: in embedded gapping with non-factive verbs, as in regular gapping cases, both the source and the target address the same QUD, unlike in embedded gapping with factive verbs where the conjuncts address different QUDs. one follows Reich (2007), the most salient wh-QUD is reconstructed on the basis of the information structure of the initial conjunct (the source clause); (ii) if one follows Schwabe (2000), it is the second conjunct (the target) that determines the information structure of the first one: "It is the second conjunct in particular that gives insights into the syntactic, semantic and information structure of its antecedent. [...] the information structure of the second conjunct determines the information structure and thus the syntax and semantics of the first" (Schwabe 2000: 247).…”

“…with focus followed by contrastive topic. Moreover, the specific constraints imposed by iar in the second conjunct require us to reassess the assumption made by Reich (2007), who considers that the source clause serves as basis for the reconstruction of an appropriate wh-QUD. This reconstruction has to pay specific attention to the information structure of the target clause introduced by iar.…”