Tom Meadows

I'm a formal linguist who works on syntax and its interfaces at the University of Geneva.

In October 2023 I started a postdoctoral position alongside Anouk Dieuleveut, as part of a SNSF research project on the behaviour of bound indexicals headed by Isabelle Charnavel.

In September 2023 I defended my PhD at QMUL. My  supervisors were Hagit Borer and David Adger.  My PhD thesis examines the connection between clause structure and locality in the relative clauses of Swahili  (Kiswahili), a Bantu language spoken across East Africa.

I have started [[Temporarily-Employed] Grammarian] to write more about syntax and related issues. I'm also starting get some cooking thoughts together on Silvertong. If I'm not doing linguistics, you can probably find me reading cookbooks.

Current research topics

I am particularly interested in three topics: i) how clause structure affects the locality of syntactic dependencies ii) how syntactic dependencies are morphologically realised iii) the internal and external structure of relative clauses

Some of this relates to the contents of my PhD thesis, Size Matters: Clause Structure and Locality Constraints in Swahili Relatives, publicly available in late June 2024. 

Clause structure and locality in Swahili relatives

Swahili is traditionally described as having three types of relative clause, one with the complementiser amba and two types without. I show that three types of relative clause display different locality profiles with respect to long-distance movement. Broadly speaking, relatives with amba display more unbounded movement than relatives without amba. Based on restrictions on their morphosyntax and word order, I propose that relatives without amba are formed by movement to a position lower in the clause such as TP. These relatives are only clause-bound when strategies involving resumptive pronouns are avoided.  To explain why the relatives with movement to TP are more local, I appeal to a position-based locality condition, informally termed the Williams Cycle (e.g. Williams 2003, 2011). 

Deriving a weakened Williams Cycle

In its strongest form, the Williams Cycle forbids all movement which lands lower in the clausal spine than any clause boundaries crossed. This rules out a number of phenomena which seem to be attested: hyperraising to subject, (hyper)raising to object and clause-internal intermediate movement.  I offer a derivation of this condition which is close to Edwin Williams' original proposal, essentially following from the timing clausal embedding relative to other syntactic operations. I am able to rule in certain kinds of intermediate movement by paying closer attention to the role the  Strict Cycle Condition plays in this system. As to why various kinds of raising are exempt, I am exploring the idea that is because they are driven by (featural) requirements of the moving element, rather than something connected to the landing site.

PF-sensitive VP chain resolution in Mandarin Chinese (w/ Qiuhao Charles Yan) 

Mandarin  has a variety of verb-doubling constructions in which two instances of the same verb appear within the same clause. We focus on a specific type of verb doubling, less discussed in the literature, in which the verb and object precede the subject, which itself is followed by another instance of the verb.  We show that the connection between instances of the verb shows all the hallmarks of a movement dependency, which we model as VP-movement to Spec CP. We believe that Mandarin, and potentially other languages, are displaying instances of partial copy deletion, driven in some sense by morphological factors. Verb doubling is closely connected to appearance of certain particles which appear adjacent to post subject copy. We are currently working out how to derive the sensitivity of copy deletion to morphological factors without look-ahead problems.

VP-fronting and the Williams Cycle in Mandarin Chinese (w/ Qiuhao Charles Yan) 

We show that Mandarin has VP fronting to two different positions in the clause, which is most clearly visible when the movement is long distance, i.e. out of complement clauses. Crucially we observe that movement to the lower position is possible out of a very restricted range of complement clause types. By comparison, movement to the higher position is largely unrestricted. We believe that this is  further evidence for the Williams Cycle, a locality condition connects the height of movement in clausal spine to the size of clause that can be moved out of. The higher movement attempts to land, the bigger the complement clause it can leave. The morphosyntactic correlates of these differences in complement clause size are somewhat different in Mandarin, due to the general lack of e.g. tense/agreement marking.

Person, wh-movement and bound indexicals [new and embryonic!] (w/ Isabelle Charnavel and colleagues)

I am investigating restrictions on bound indexicals in relative clauses in English and French, with the classic case being sentences like I'm the only one who did my homework. What's of general interest is that such a sentence admits a bound reading: nobody else did their homework. The literature has largely focused on bound indexicals in short subject relatives, which I believe has obscured some unusual syntactic restrictions. At the moment I'm investigating the hypothesis that indexicals bound short distance must be non-subjects, and their binders must be subjects. I believe this ultimately stems from the underspecification of wh-moved DPs for person, which can only become specified if they land in subject postion / Spec, TP. [More on this soon!]