ComSyn Lectures

The next ComSyn meeting will take place on Thursday 8 November.

Speaker: Daniel Altschuler (Heinrich Heine Univesität, Düsseldorf)
Title: ‘Sequence of tense’ as absence of cessation
Time: 15:15-16:30
Venue:  Vrieshof 4 / 012

Abstract 
It has become generally accepted in the research on the syntax-semantics interface that some features (e.g. person/number/tense) on bound-variables are not semantically interpreted; they are there for purely syntactic reasons (e.g. agreement). Such a view is adopted in the tense domain due to the so-called sequence of tense phenomenon; a syntactic feature manipulation mechanism (e.g. feature checking or feature transmission) is posited to render a bound past tense semantically inert, i.e. without an anteriority meaning. The arguments for the sequence of tense analysis that we are familiar with rely on an unstated assumption that states have first moments. In this talk I argue that if this assumption is not made, or if is semantically neutralized, we can and should maintain the view that tense features are always interpreted. The outline of the talk is as follows. I first motivate a constraint that correlates the holding of a state at an interval with the truth of a stative predicate at a moment within that interval. The corollary of this constraint is that if a stative predicate holds of a moment, it necessarily holds of a preceding moment, regardless of whether or not the described state has a first moment. I then show how this corollary underwrites the entailment from PRES-ϕ to PAST-ϕ and use this fact to derive a cessation implicature: the utterance of PAST-ϕ implicates that no state of the kind described currently holds. The proposed analysis makes use of an existential quantifier for tense with a domain restriction variable over reference time concepts. The presence of a cessation implicature is predicted in both matrix and embedded clauses when the reference time concept does not—by itself—make PRES-ϕ false. I suggest that the intuitions often used to motivate the sequence of tense analysis are really intuitions about the absence of the cessation implicature.

The next ComSyn meeting will take place on Tuesday 23 October  2012.

Speaker: Allison Kirk (LUCL)
Title: The diachrony of Greek relative clauses
Time: 15:15-17:00
Venue:  Lipsius/2.35c

Abstract
This talk focuses on changes in the structure of relative clauses (RCs) from Ancient to Modern Greek, with discussion of five periods of Greek: Homeric (texts from 8^th century BC), Classical (texts from 5^th century BC), Koine (texts from 1^st century), Medieval (texts from 7^th – 10^th centuries) and currently spoken Modern Greek. From Homeric Greek until middle Koine Greek, headed relatives equivalent to ‘the boy who I saw is tall’ are attested alongside correlatives equivalent to ‘which boy I saw, this one is tall’. In Modern Greek, the former are predominant and the latter are almost obsolete. It has been claimed that the correlative is the ‘predecessor’ of the headed relative in the related language Latin, as a result of a series of syntactic re-analyses (Bianchi 2000). The goal is to provide synchronic analyses of RCs in the various stages of Greek and then determine the pathways for the changes. In particular, I argue that headed relatives from Homeric to (middle) Koine Greek are not embedded under main clause NP ‘heads’, but that the RCs are all adjoined to main clauses in these stages. I argue for a standard head-raising approach to Modern Greek RCs, and that that the changes in structures correspond to changes in the lexico-syntactic properties relative morphemes. In particular, the re-analysis of the old relative pronoun from a maximal projection to a Dº category corresponds to the new embedded head-external RC. This re-analysis also corresponds to the loss of XP ‘splitting’ in RCs. 

Throughout the talk I will address various methodological issues in diachronic syntax and current theories of syntactic change, and when applied to the Greek facts, what they contribute to the longstanding observation that there is a unidirectional ‘cline’ from ‘paratactic’ to ‘hypotactic’ sentence structures (Hopper 1993, among others).

The first ComSyn meeting will take place on 9 October  2012.

Speaker: Jóhannes Gísli Jónsson (University of Iceland)
Title: Case variation: Formal vs. semantic
Time: 15:15-17:00
Venue:  Lipsius/2.35c

Abstract
Variation in case licensing is a robust phenomenon in languages that have a rich case system like Icelandic. The major claim of this talk is that there are two kinds of case variation in such languages. One type is what we can call formal variation, i.e. variation between two semantically equivalent cases. Formal variation is strongly associated with the erosion of idiosyncratic case, as exemplified by the variation between accusative and dative experiencer subjects in Icelandic (so called Dative Substitution). Formal variation can be diagnosed via “mixed” case marking, the possible use of two cases where case agreement is expected to be obligatory. The other type can be labelled semantic variation because it involves a fairly subtle semantic contrast between two case variants, bearing inherent vs. structural case. This type of variation is harder to detect than the first and consequently it is much less studied. Still, I will argue that semantic variation is different from formal variation in some non-trivial ways. Moreover, semantic variation affords a good vantage point to inspect the semantic determinants of argument realization.

Drinks afterwards!