Verb-second and the verb-stranding verb phrase ellipsis debate

Verb-stranding verb phrase ellipsis (VPE), when a verb is stranded outside of the VP-sized ellipsis site in which it originated, has been identified in a number of languages (Irish, McCloskey 1991; Hebrew, Doron 1991, Goldberg 2005; Greek, Merchant 2018; Uzbek, Gribanova 2019, i.a.), and has been invoked productively in analyses investigating the position to which verbs move and the timing of verb movement in the grammar. Recently, Landau (2018, 2019, to appear) proposes a phase-based negative licensing condition on head-stranding ellipsis that precludes verb-stranding VPE altogether. He claims that apparent verb-stranding VPE must be reanalyzed either as Argument Ellipsis (Oku 1998; Kim 1999; Takahashi 2008), or a clause-sized ellipsis that strands main verbs (Gribanova 2017). This article approaches this debate through an analysis of head movement and head-stranding ellipsis in the Indic verb-second (V2) language Kashmiri, arguing that Landau’s phase-based approach encounters empirical challenges in accounting for variation in the presentation of ellipsis in V2 languages and requires an unconventional approach to V2, at odds with recent accounts of Kashmiri V2 (Bhatt 1999; Munshi and Bhatt 2009; Manetta 2011) and mainstream views of V2 generally (e.g. Holmberg 1986; Travis 1991; Vikner 1995; Zwart 1997). While the present article argues in favor of the standard account of ellipsis (Merchant 2001, 2008), we affirm the important contribution of Landau’s work in identifying challenges facing any account of head-stranding ellipsis licensing. At issue is the larger question of whether and how verb-stranding ellipses can be used to better understand head movement.

The Role of UID for the Usage of Verb Phrase Ellipsis: Psycholinguistic Evidence From Length and Context Effects

We investigate the underexplored question of when speakers make use of the omission phenomenon verb phrase ellipsis (VPE) in English given that the full form is also available to them. We base the interpretation of our results on the well-established information-theoretic Uniform Information Density (UID) hypothesis: Speakers tend to distribute processing effort uniformly across utterances and avoid regions of low information by omitting redundant material through, e.g., VPE. We investigate the length of the omittable VP and its predictability in context as sources of redundancy which lead to larger or deeper regions of low information and an increased pressure to use ellipsis. We use both naturalness rating and self-paced reading studies in order to link naturalness patterns to potential processing difficulties. For the length effects our rating and reading results support a UID account. Surprisingly, we do not find an effect of the context on the naturalness and the processing of VPE. We suggest that our manipulation might have been too weak or not effective to evidence such an effect.

Expanding the typology of Verb Second VPE

Mysteriously, Verb Second (V2) languages are known to exhibit auxiliary-stranding verb phrase ellipsis (VPE) but to lack verb-stranding VPE, even though the inflected verb must leave the VP (Mikkelsen 2006; LaCara 2014). Sailor (2018) claims that VPE bleeds V2; the feature that drives ellipsis (on T) is introduced derivationally prior to the feature driving V2 (on C). Only languages with verb movement triggered by T, as in Hebrew (Goldberg 2005), exhibit V-stranding VPE. This chapter offers evidence that Sailor’s approach is on the right track; the Indic language Kashmiri is a V2 language in which auxiliary-stranding and V-stranding VPE co-occur, because T is independently a trigger for V movement (Munshi and Bhatt 2009). The findings support a particular approach to the timing and interaction of the major operations in the grammar and suggest that any approach to V2 must account for the variation in the presentation of VPE in V2 languages.

Voice Mismatch Tolerance in Verb Stranding Verb Phrase Ellipsis

Verb Phrase Ellipsis, in particular, is taken to be very rare in languages other than English. However, recent literature has pointed out that a Verb Phrase Ellipsis-like construction does in fact exist in other languages, but may be masked due to the fact that the main verb raises to INFL in such languages, a process known as Verb Stranding Verb Phrase Ellipsis (VSVPE). This paper addresses two main issues: 1) whether such a construction in Hebrew patterns with VPE in English or with Pseudogapping; via an examination of voice mismatch tolerance following Merchant (2013) 2) After establishing that Hebrew is a VSVPE language and controlling for external variables such as independent object drop, this paper tests which constituent in particular is targeted in the ellipsis process. It is concluded that VSVPE languages target vP for ellipsis, not VP, nor PP, as opposed to recent accounts. In other words, they pattern with Pseudogappig in including higher constituents (vP, not VP) in the ellipsis) more than they do with VPE. Other Verb-Raising to INFL languages are predicted to behave similarly with respect to which constituent is targeted for ellipsis.

A Neural Network Approach to Verb Phrase Ellipsis Resolution

Verb Phrase Ellipsis (VPE) is a linguistic phenomenon, where some verb phrases as syntactic constituents are omitted and typically referred by an auxiliary verb. It is ubiquitous in both formal and informal text, such as news articles and dialogues. Previous work on VPE resolution mainly focused on manually constructing features extracted from auxiliary verbs, syntactic trees, etc. However, the optimization of feature representation, the effectiveness of continuous features and the automatic composition of features are not well addressed. In this paper, we explore the advantages of neural models on VPE resolution in both pipeline and end-to-end processes, comparing the differences between statistical and neural models. Two neural models, namely multi-layer perception and the Transformer, are employed for the subtasks of VPE detection and resolution. Experimental results show that the neural models outperform the state-of-the-art baselines in both subtasks and the end-to-end results.

Antecedent-Contained Deletion

This chapter introduces the linguistic phenomenon of Antecedent-Contained Deletion (ACD): a type of construction in which a site of Verb Phrase Ellipsis (VPE) is contained in the antecedent from which it derives its interpretation. The chapter reviews theoretical approaches to resolving interpretation in ACD structures (drawing primarily on the covert movement operation of Quantifier Raising (QR)), and the accessibility of one or more sentential interpretations when the site of ellipsis is embedded in non-finite and finite clauses. Behavioural responses from offline judgement studies with children and adults, as well as online studies with adults, provide data bearing directly on these theoretical accounts.