Volume 42, Issue 3 - Special Issue: Case, agreement and ellipsis; Guest Editors: Arantzazu Elordieta and Laura Vela-Plo

The interest in such diverse topics such as agreement and ellipsis, and their interaction with several components of the language system (syntax, pragmatics, phonology…) has increased in the last decades, as new approaches emerge to enable more precise analyses of the mechanisms underlying agreement relations and elliptical constructions. This special issue gathers a selection of six papers from the 32nd Colloquium on Generative Grammar and from the Workshop on Case, Agreement and Ellipsis immediately following the 32CGG (University of the Basque Country UPV/EHU, April 2023) which focus on two main topics: (i) agreement and related notions, such as feature specification (or underspecification) on heads, the nature of φ-feature valuation and the agreement mechanism itself, or a case of predicate ellipsis where a clitic appears to participate in agreement licensing; and (ii) prima facie elliptical phenomena, such as Righ-Node Raising (analysed through the lenses of discourse), an LF-copying approach to ellipsis, where ellipsis in the main clause is licensed by an adverbial clause, and a syntactic approach to symmetric predicates under VP-ellipsis. This introductory paper serves as a state of the art as well as guide for the readers of this special collection and offers an overview of the topics dealt with in each contribution.

In configurations where inflected and uninflected infinitives may alternate in Portuguese, the option without overt agreement may be incompatible with adjectival predicates, passivized verbs, or reflexive clitics and is subject to a puzzling dialectal variation among speakers. This paper argues that the infinitival form without agreement may be analyzed by speakers as involving a T head with no ϕ-features, a T head with just a number feature, or a T head with just a person feature. Each of these specifications has different consequences for minimality computations when adjectival predicates, passives, or reflexive clitics intervene between T and the subject, yielding the dialectal variation found.

In this paper we present an approach to agreement where different agreement processes occur at different levels. We argue that φ-feature valuation is never the trigger of a checking relation between a head H and an NP, but the effect of a syntactic relation that must be triggered by some independent operation. We analyze in detail characteristic cases where agreement is not syntactic but induced in the morphological component or in processing. We also show that all these cases are indirectly determined by syntactic derivations where the φ-feature matrix in the head H is not fully specified, leaving room for additional, extra-syntactic, operations to tamper with agreement.

This paper documents and analyzes a predicate ellipsis alternation found in contemporary Spanish, addressing some challenges it poses for current theories of ellipsis licensing with the Minimalist Program. Spanish predicate ellipsis occurs in two variants: (i) lo-insertion , where an overt proform (the neuter clitic lo ) attaches to the auxiliary preceding the silent gap, and (ii) lo-omission , where predicate ellipsis is licensed under polarity focus (typically marked on the stranded auxiliary). Both variants arise from ellipsis of the main predicate – and its arguments – and are licensed by the same restricted class of auxiliaries. To account for their shared distribution, this paper argues that predicate ellipsis stems from a single [E]-feature, in both the lo-insertion and the lo-omission context. Concretely, the presence of a clitic (in lo-insertion ) and its absence under polarity focus (in lo-omission ) are analyzed as two alternative strategies to satisfy the semantic requirements of a single [E]-feature, without encoding these contexts for ellipsis directly in [E]’s syntax. This analysis is further extended to predicate ellipsis constructions in Germanic languages, which employ strategies for predicate anaphora akin to both lo- insertion and lo- omission, albeit with a different distribution.

Despite decades of research on Right-node Raising (RNR), the phenomenon has remained elusive, seemingly defying the laws of syntax as we know it. This paper argues that this is so because RNR is not a bona fide syntactic phenomenon at all, but rather a paratactic configuration composed in discourse. RNR configurations consist of an unbounded sequence of potentially incomplete expressions whose prosodic realization provokes the expectation of a continuation, which is furnished by a sequence-final expression which, depending on the type of RNR, is either sentential or fragmentary. It is shown that this analysis can reconcile seemingly contradictory properties of RNR constructions, crucially by abandoning the long-standing assumption that RNR configurations are constructed in sentential syntax.

While implicitly assumed, the restriction that the antecedent of ellipsis cannot be base-merged within the ellipsis site has not to our knowledge been explicitly stated or analysed. In this article we discuss a pattern of ellipsis in Spanish, resembling gapping but we argue distinct from it, where ellipsis in the main clause can find its antecedent in some but not all adverbial clauses. The adverbial clauses that cannot provide the antecedent for ellipsis are, we argue, those that would have to be interpreted within the ellipsis site at LF. We argue that this pattern and its restrictions supports LF-identity theories of clausal ellipsis, and also provides evidence that remnants of clausal ellipsis move at LF.

This paper focuses on the behavior of symmetric predicates under ellipsis. Building on Stockwell (Stockwell, Richard. 2020. Contrast and verb phrase ellipsis: Triviality, symmetry, and competition . UCLA. Doctoral Dissertation), we examine the phenomenon of participant and transitivity switching under VP-ellipsis, which symmetric predicates permit but non-symmetric ones do not. While Stockwell takes this as evidence for a semantic identity condition on ellipsis, we show that semantic equivalence is too weak. Specifically, cases of participant or transitivity switching involving focus marking are wrongly predicted to be acceptable. We argue instead for a syntactic identity condition on ellipsis and sketch a new analysis of symmetric predicates as unaccusatives.