Introduction (Belletti and Collins 2019)

This paper is the introduction to the volume on Smuggling to appear (OUP) edited by Adriana Belletti and Chris Collins.

Belletti and Collins 2019
Here are the abstracts:

Belletti’s contribution presents and discusses a number of derivations such as passive, causative and passive in the causative voice/si-causative passive, which all involve movement of a chunk of the verb phrase containing the verb and its internal argument, yielding smuggling in Collins’ (2005) sense. The questions of what the engine of a smuggling derivation is and how the relevant chunk to be smuggled is identified guide the discussion. Evidence from acquisition is also considered where derivations involving smuggling appear to be at the same time more complex and more readily available to the developing child. The relevant chunks can be attracted by different types of heads in the clause structure, which all have the property of attracting syntactic movement into their specifier. Such heads may express features of different nature present in the clausal map, such as the passive and causative voice, as well as discourse related features such as the (vP-peripheral) topic and focus features.

Bianchi’s contribution discusses smuggling in relation to the syntax and semantics of certain adverbs in Italian. In past and future perfect sentences, punctual time adverbials like at five o’clock can specify either the Event Time or the Reference Time. In Italian, their interpretation is affected by syntactic position: a clause-peripheral adverbial allows for both interpretations, while a clause-internal adverbial only has the E-interpretation. Moreover, for clause-peripheral adverbials the presence of the adverb già (already) blocks the E-interpretation. It is shown that this pattern can be accounted for under a smuggling analysis, in which (i) the adverbial is merged as a DP in a functional projection intervening between T and the subject in the edge of v/VP P, thus blocking Agree between them; (ii) smuggling of v/VP past the adverbial solves the intervention effect; (iii) an E-adverbial originates in a projection below già (already), while an R-adverbial originates in a projection above it. A compositional semantic analysis is provided for the proposed syntactic structure.

            Bošković’s contribution argues that there is no general freezing ban. As discussed in section 2, smuggling refers to a situation where, in Bošković’s words, movement of α would induce a violation that is voided by movement of a larger constituent β that contains α, which is followed by movement of α. Smuggling thus involves movement out of a moved element, which is traditionally assumed not to be possible (the constraint is referred to as the freezing ban). Rather, Bošković argues that extraction out of moved elements is in fact generally allowed. The cases where such extraction appears not to be allowed involve independent problems concerning labeling. The paper re-examines from this perspective (which allows but restricts the possibilities for smuggling) the smuggling derivations proposed in Collins (2005a,b), focusing on the passive construction, and the smuggling analysis of tough-constructions proposed in Hicks (2009) illustrated in section 3 of this introduction. A modified version of the latter is argued to be superior to the traditional null Op analysis of tough-constructions. Several conclusions regarding the structure of infinitives are also drawn. Furthermore, the discussion in the paper also shows that there is a strong relationship between movement and labeling: unlabeled elements cannot undergo movement, do not function as interveners, and cannot be the target of movement.

Collins’ contribution discusses the dative alternation in English, which relates the double object construction (John gave Mary the car) to the prepositional dative (John gave the car to Mary). On the basis of traditional c-command tests, it is argued that the prepositional dative is derived from the structure underlying the double object construction. If the theme is smuggled over the goal by VP movement there is no violation of locality constraints.

Corver’s contribution examines the phenomenon of M(easure) P(hrase) alternation from a cross-categorial perspective. An illustration of this phenomenon is given by the minimal pair: (i) John is two inches too tall; (ii) John is too tall by two inches. The former features a bare MP, the latter by+MP. Interestingly, clauses permit only one order: *Mary two years outlived her husband; (ii) Mary outlived her husband by two years. It is proposed that the pattern featuring the bare MP is the base order. The pattern featuring by+MP is the derived order. This derived order results from leftward movement of a phrasal constituent past MP. In clauses, this phrasal constituent is a VP which smuggles the subject across MP. The ill-formedness of the clause featuring a bare MP is due to a locality violation: a subject moves across an intervening MP. In non-clausal configurations, this violation does not occur since the (small clause) subject is located higher than MP.

Den Dikken’s contribution defends an analysis of the active/passive alternation sharing with Collins’ smuggling proposal the idea that the participial VP occupies a specifier position above the external argument, but base-generating it in this position rather than moving it there. In both the active and the passive, the VP and the external argument are in a predication structure, with a RELATOR mediating the predication relation. The active voice builds a canonical predication structure, with the VP in the RELATOR’s complement position and the subject of predication as the specifier. In the passive voice, the VP is externally merged in the specifier of the RELATOR and the external argument in its complement. This analysis provides an explanation for obligatory auxiliation, the unavailability of accusative Case for the internal argument, Visser’s Generalization (the ban on personal passivization of subject control verbs), and the restrictions on referential dependencies and depictive secondary predication in passives.

Koopman’s contribution focuses on the syntax of the can’t seem to construction in English, as in I can’t seem to fix this, which present a syntax semantics mismatch, raising the question how and where it should be resolved. The paper establishes that the problem calls for a syntactic solution: there is unambiguous evidence from idioms and absence of aspectual restrictions that the linear order of I can’t seem to fix this must be derived from a merge order where seem is merged higher than not can V, as in it seems I can’t fix this. The paper motivates each step in the bottom up derivation, with crucial insights coming from comparative syntax, i.e. from the verb clustering West Germanic OV languages. The properties of the construction and the restrictions, including intervention, are shown to reduce to structure building Merge (E- and I- merge), in conjunction with general principles (Attract Closest, and the Extension condition). Pied-piping is a central ingredient in the derivation; Remnant movements play a role in ”smuggling” around interveners; a strong intervention effect caused by experiencers can be reduced entirely to a required sequence of Merge, necessary for convergence. Finally, returning to comparative syntax, the paper discusses how the proposed derivation for English can in turn shed light on a syntactic solution of so-called displaced zu in German. It is precisely because this construction is so restricted, that it provides a valuable testing ground for the type of syntax we should pursue. The proposed analysis thus has direct bearings on the architecture of UG.

The goal of Mateu and Hyams’ study is to address two questions: (i) whether the delays in the acquisition of subject-to-subject raising (StSR) seem and subject control (SC) promise are related, as would be predicted by various developmental accounts, and (ii) whether delays are due to limited processing capacity or immature grammatical abilities. Two comprehension tasks reveal two groups of children: (i) below-chance group: they have a non-adult grammar of StSR or SC and processing capacity does not predict performance; and (ii) at-/above-chance group: they have an adult-like grammar of StSR or SC and processing capacity modulates performance. Importantly, no correlation is found between StSR and SC performance – some children have mastered StSR with seem but not SC with promise and some show the opposite pattern, suggesting a dissociation between the grammatical development of StSR and SC, specifically of the mechanisms required to circumvent intervention.

Poletto and Pollock’s contribution analyzes the syntax of interrogative clauses in French and in some Northern Italian dialects (NIDs), including so-called wh-in-situ configurations. They show that their intricate properties can be derived from standard computations (wh-movement and remnant movement of vP/IP to a Top/ground slot) to either the vP Left periphery (Low Left Periphery/LLP) or the CP domain (High Left Periphery/HLP). The question arises of why some languages make use of the LLP or the HLP or indeed both, like French. They argue that in significant cases the morphological properties of the various Wh-words and the surface forms of the sentences provide all the clues required by the language learner and the linguist. Among their various proposals the authors argue that in French movement of interrogative pronouns to the HLP is actually movement to a free relative layer and that the peculiar properties of French que are captured by analysing it as both an interrogative and relative element in conjunction with a ‘smuggling’ analysis of Subject Clitic Inversion (SCLI). They show that many NIDs make use of both the LLP and the HLP and that smuggling is again crucially involved in a number of them. In addition to the fruitfulness of the ‘smuggling’ idea for Romance, the main theoretical result of their chapter is that notions like ‘relative constructions’ or ‘inter­rogative constructions’ are not primitives of the language faculty (Kayne 2015) since in significant cases the derivation of questions activates both the interrogative side of the LLP and the (free) relative side of the HLP.

Roberts’ contribution argues that the lack of SVO ergative languages (“Mahajan’s Generalization”, see Taraldsen 2017) can be explained by the combination of a smuggling analysis of ergative alignments and the Final over Final Condition (FOFC). The smuggling derivation, when the smuggled category is internally head-initial, creates a configuration which violates FOFC. For this reason, SVO and ergativity do not combine in the world’s languages, a notable typological lacuna that has hitherto defied explanation. The implications of the analysis for V-initial ergative languages and for passives are also briefly explored in the paper.