Focusing on two Romance languages, French and Romanian, we provide a detailed analysis of gapping and present several empirical arguments for preferring a construction-based approach of gapping (with semantic reconstruction of ellipsis) over alternative accounts that rely on movement or deletion. We then study parallelism constraints and show that syntactic parallelism is less strict than what is usually assumed, while discourse parallelism is clearly required. Syntax is not completely ignored though, as each remnant is required to match some subcategorization frame of the verbal predicate its correlate depends on. We show how those core properties can be accounted for within a construction-based framework relying on inheritance hierarchies of typed feature structures, such as HPSG in its more recent versions.

This paper empirically tests the embedding constraints on gapping in Persian. It has been suggested that gapping differs from other kinds of ellipsis in banning embedding. However, the first counter-examples in the literature come from Persian. Following up on previous experiments on embedded gapping in several languages, we report the results of two acceptability judgment tasks. Our results show that, while embedded gapping is overall acceptable in Persian, speakers’ acceptability judgements also vary depending on the semantic type of the embedding predicate, as well as the presence/absence of the complementizer. Data from Persian highlight that, despite the cross-linguistic variation observed with respect to the acceptability of embedded gapping, a general semantic constraint is at work across languages: non-factive verbs embed more easily than factive ones; inside factive verbs, semi-factive (cognitive) predicates embed more easily than true factive (emotive) ones. Moreover, whereas previous theoretical literature indicates no systematic preference for the absence or the presence of the complementizer in Persian, these new experimental data suggest a preference for complementizer drop. To account for the gradience observed in our experimental data, we propose an approach of gapping based on acceptability rather than grammaticality.

This paper discusses the No Embedding Constraint, considered to be a strong syntactic constraint on gapping and a diagnostic for this ellipsis type. As shown by two acceptability judgments tasks in Spanish, the assumptions related to the No Embedding Constraint are not borne out by our experimental results. Embedded gapping is acceptable in Spanish, and seems to be governed by a (more general) semantic constraint. Specifically, some predicates embed more easily than others, confirming, on the one hand, the asymmetry between non-factive and factive predicates, and, on the other hand, the dichotomy between semi-factive and true factive predicates. Embedded gapping and embedded coordinated clauses in general are thus sensitive to the semantic class of the embedding predicate. We argue that our data on embedded gapping in Spanish constitute a challenge for any Small Conjunct Gapping approach (in terms of subclausal/low coordination) and give support to a Large Conjunct Gapping approach (in terms of clausal/high coordination). In this paper, we adopt a constructional fragment-based analysis of gapping, which treats nonembedded and embedded gapping uniformly. More generally, this paper suggests that gapping is more similar to other ellipsis types than has been traditionally assumed.