Abstracts

Cartography is, among other things, a large descriptive endeavour. Twenty years of intensive cartographic work on the left periphery have led to the discovery of a wealth of fine properties of syntactic structures: ordering in functional sequences, mutual exclusion of distinct positions, freezing effects, etc. These discoveries enrich the empirical basis of syntactic theory, and raise issues of further explanation: how can the discovered properties be related to fundamental ingredients of linguistic computations? In my presentation, I will address some concrete cases illustrating these points.

Abstract can be downloaded here. 

One of the several ambitions of the cartographic approach as developed by Rizzi (1997) is to integrate informational distinctions to the grammatical architecture of languages.  Evidence of the necessary articulation between IS and syntax is straightforwardly provided by overt, morphological informational marking as found in e.g. Yoruba, but also Wolof (e.g. Robert 2010), the data of which are shown to provide support for a cartographic approach. In the following representative example taken from SVO Lamang (Afro-Asiatic, Wolff 2009: 39), the object is sentence-initial due to the negative focus it comes under, as overtly marked by the morphology.

Yén-b-é                                tsóts-i                                                wò

Tongue-NEG1-FOC           cut(PLURACT):IMP-1SG              NEG2

Lit. (It is) not the tongue (that) I keep cutting.

Is such support also found for covert information marking? A case in point is provided by diachronic and typological evidence relating to the Jespersen cycle. The long-standing assumption is that emerging (and declining) negators are used in clauses with a pragmatic value. Recent literature identifies that value as informational, with a discourse-old status produced by the clause being used in so many words in the antecedent context (see the synthesis in Blaxter and Willis 2017). This explicitly activated discourse-old value can be modelled by informational features on the negative marker (see Battlori 2016) checked in the left-periphery. Indirect evidence for this comes from the fact that activated negatives tend to occur in main-clauses. Extension of distribution beyond root clauses is associated to loss of activated value, as empirically demonstrated by the evolution of Medieval French preverbal non in Larrivée (2010). Whether such movement can apply to grammatical phenomena that involve categorical informational partition, such as Double Negation and Metalinguistic Negation, is considered, in view of recent proposals by Martins (2016) and Larrivée (2017). An issue that arises is the lack of island sensitivity of Metalinguistic Negation, and the possible involvement of Freezing (Sobin 2013) is discussed. Both overt and covert informational marking show that informational status can be integrated into the grammatical architecture.

 

In English, shell nouns such as thing, fact, question are regularly used in the left periphery of the sentence with a (variable) definite article and copula be. These semi-fixed N-is constructions have developed pragmatic function: they function to launch utterances and to focalise the following information. Previous research has traces the development of the pragmatic function (e.g. Aijmer 2007), looks at copula reduplication (e.g. Bolinger 1987, Curzan 2012, Massam 1999) or variable article use (Stvan 2014). Keizer (2013, 2016) uses a functional grammar approach and corpus evidence to the construction, including evidence from Dutch, as well while Hundt and Oppliger (submitted) provide a corpus-based study of German and English focalisers from a construction grammar point of view. Surprisingly, there is no prior research on variation across different World Englishes, even though preliminary evidence suggests that article use as well as degree of syntactic integration with the main clause are variable across varieties such as US, Canadian and Australian English. I will use corpus data from the New On the Web (NOW) corpus to investigate variable article use in focalisers across varieties of English as a first (ENL) and institutionalised second language (ESL), looking into the role that contextual factors such as modification, choice of head noun and degree of syntactic integration may play.

 

Abstract can be downloaded here. 

Hungarian is a language which is known for two salient syntactic properties : (i) it is a « free word order » language, in which constituents may appear in apparently unconstrained orders ; (ii) it is a language which « wears its interpretation on its sleeve ». One way of making sense of these two phenomena is to makes liberal use of left-peripheral syntactic positions.

    The talk discusses well-known phenomena of topicalisation, focusing, wh-movement and quantification. It also explores less discussed features of Hungarian, such as the SVO/VSO orders and their relevance to the notion of “subject”.