Dutch sentence focus constructions on the semantics-pragmatics interface : a case study

This paper reports on an ongoing investigation of the semantics and pragmatics of the Dutch Syntactic Inversion with Filler Insertion Construction (henceforth: SIFIC). Formally, the SIFIC is characterized by a non-canonical syntactic sentence structure, whereby the subject follows the verb, which in turn is preceded by the adverbial pronoun er, as illustrated in (1) and (2). (1) Er valt sneeuw there falls snow ‘It is snowing.’ (2) Er loopt iemand op het dak there walks someone on the roof ‘There is someone walking on the roof.’ The Dutch SIFIC, along with its counterparts in other languages, e.g. English (Birner & Ward 1996, 1998), has received attention for its particular topic-comment structure and focus-background articulation. In Lambrecht’s theory of information structure, which fits particularly well within the Construction Grammar framework (cf. Leino 2013), it can be analyzed as an instance of a sentence-focus construction (cf. Lambrecht 1987, 1994, 2000, 2001). Sentence-focus constructions are constructions on the sentence level, formally marked by morphosyntax and/or prosody, that convey a particular (procedural) meaning that can be conceptualized as ‘theticity’ (Sasse 1987, 1995, 2006) or as ‘sentence-focus’, i.e. a lack of pragmatic presuppositions attached to the subject or predicate (Lambrecht 1987, 1994, 2000). However, the question whether ‘sentence-focus’ has to regarded as the purely linguistic, non-defeasible and encoded meaning (semantics) of SIFIC or as a sense generated by encyclopedic knowledge, default inferences and implicatures (pragmatics) has hitherto not been addressed. Following the need to differentiate between these different kinds of linguistic content, as argued for by (neo-)Gricean pragmatics (Atlas 2005, Grice 1989, Levinson 2000), Relevance Theory (Sperber & Wilson 1986) structural functionalism (Coseriu 1985) and some proponents of Cognitive Linguistics (Zlatev 2007, 2011), this paper raises the question whether ‘sentence-focus’ belongs to the semantics or the pragmatics of the Dutch SIFIC. In order to shed light on this issue, this study reports on a corpus research on the SIFIC in both spoken and written Dutch, whereby the various possible uses of the SIFIC were analyzed. SIFIC tokens were extracted from the SoNaR-corpus Hedendaags Nederlands and annotated for various factors, including topic-comment structure, focus-background articulation and referential givenness. SIFIC was attested with and without a ‘sentence-focus’ reading, which indicates that ‘sentence-focus’ cannot be considered to be the encoded, non-defeasible, meaning of the Dutch SIFIC. This study thus confirms the cross-linguistic attested finding that sentence-focus constructions are often not dedicated to the expression of only ‘sentence-focus’ (cf. Matic 2003, Karssenberg 2016, Karssenberg et al. 2018, Sasse 1995, 2006). Matic & Wedgwood (2013) argued on the basis of similar findings that focus constructions simply do not exist, because ‘focus’ is never semantically encoded in linguistic constructions. The analysis proposed in this paper departs from the analysis proposed in Matic & Wedgwood, by arguing that sentence-focus constructions do exist, be it on the level of generalized conversational implicatures (pragmatics) rather than on the level of encoded semantics. This ties in with previous observations that the meaning of many constructions on the sentence level should be analyzed as a meaning on the pragmatic level rather than as encoded semantics (cf. Coene & Willems 2006, Willems & Coene 2006).