SLE 2016

Napoli, 31 August 2016

Workshop

Negation at the syntax, semantics and pragmatics interfaces: theoretical, empirical and experimental approaches

Convenors: Jacques Moeschler, Joanna Blochowiak and Cristina Grisot, Department of Linguistics, University of Geneva

Description of the workshop

Negation is a very crucial issue as regards the syntax, semantics and pragmatics of natural language. Often given as a criterion for disentangling semantic meaning from pragmatic meaning (presupposition is not sensitive to negation, while entailment is), and generally associated with truth-conditional meaning (whereas metalinguistic uses are not), negation is one of the most important linguistic phenomena at the interfaces of syntax, semantics and pragmatics. Although negation is a very basic phenomenon in natural language, it exhibits multifaceted and complex behaviours in its scope and variety of usages not only across different languages but also within a particular language. In propositional logic, the meaning of negation is straightforward: the fundamental logical property of negation lies in its taking wide scope over the whole proposition and in reversing the truth-value of the proposition. However, the behaviour of negation in natural languages is more complex. Linguistic negation exhibits mainly three properties: (i) scope properties; (ii) the distinction between descriptive and metalinguistic usages; (iii) the derivation of descriptive and metalinguistic negation.

Because of the current growing interest in negation and in its cognitive processing, we would like to gather researchers working on negation in order to develop a comprehensive and a unified picture of this topic. In this panel, we would like to address questions about the meaning of negation, its syntax, semantics and pragmatics, and most importantly, issues that arise at the interfaces between syntax, semantics and pragmatics. We aim at addressing, using different methodologies and theoretical approaches, the following research questions:

1. What do the syntactic properties of negation reveal about its scope properties?

2. What are the main pragmatic uses of negation?

3. How can a general semantics of negation account for its various uses?

4. What are the main domains of variation for negation (in syntax, semantics and pragmatics)?

5. What are the main morphosyntactic strategies used to convey negative meanings?

6. What is the relation between the logical properties of negation and its semantics and pragmatics?

The panel aims at gathering together researchers working on different languages, in different linguistic fields (from syntax to pragmatics, but also psycholinguistics and corpus linguistics among others), taking different theoretical, empirical and experimental perspectives. The presentations included in this workshop will target the following general topics:

  • Logical form and the syntactic structure of negative sentences

  • Expletive negation

  • NEG raising

  • NEG concord

  • Morphological negation

  • Negative Polarity Items

  • The syntax-semantics interface of negation

  • Descriptive and metalinguistic negation

  • Causality in negative sentences

  • The semantics-pragmatics interface and negation

  • Corpus-based approaches to negation

  • Experimental approaches to negation

Introduction: Negation at the syntax, semantics and pragmatics interfaces: theoretical, empirical and experimental approaches

Jacques Moeschler, Joanna Blochowiak, Cristina Grisot

A criterion for descriptive vs. metalinguistic negation

Jacques Moeschler

Abstract

In Horn (1989, 392-413), three criteria are given to differentiate descriptive negation (DN) from metalinguistic negation (MN): (i) MN is not incorporated (*it is impossible for you to leave, it is necessary), (ii) MN is incompatible with NPIs (Chris didn’t manage to solve some/*any of the problems – he manage to solve all of them), and (iii) the mais test (Anscombre & Ducrot 1977): MN is marked by maisSN (Spanish sino, German sondern), whereas DN is signalled by maisPA (Spanish pero, German aber:

In Anscombre & Ducrot (1977), another test is given for distinguishing maisSN from maisPA, which gives rise to another result: maisSN is strengthened by au contraire (‘on the contrary’) and follows DN, whereas maisPA is completed by en revanche (‘in contrast’), and is thus compatible with MN:

(1) OK Abi n’est pas belle, maisSN au contraire quelconque (Abi is not beautiful, but on the contrary ordinary)

(2) # Abi n’est pas belle, maisSN au contraire très belle

(3) OK Abi n’est pas belle, maisPA en revanche très belle (Abi is not beautiful, but in contrast very beautiful)

(4) # Abi n’est pas belle, maisPA en revanche quelconque

In order to account for this distribution, it seems reasonable to look at basic semantic properties of utterances, connecting the negative clause (NEG), its positive counterpart (POS) and the corrective one introduced by mais (COR). In DN, COR entails NEG, whereas in MN, COR entails POS (Author 2013). So it seems that what makes maisSN a test for DN is that all entailments of the negative clause are confirmed, whereas with MN, it is the other way around: only the entailments of POS are confirmed, whereas negation scopes over the implicature of POS.

This means that a very strong constraint is about negation: negation cannot scope over POS implicature, unless it is metalinguistic. So what it means is that when a negation scopes over a scalar predicate, implicatures are not taken into account, the effect being that not-P means less than P. This effect explains the MaxContr effect (Horn 1989), where not-P implicates the contrary of P: so here, not-beautiful implicates ugly, while it is compatible with not-ugly. On the other hand, when negation is metalinguistic, all entailments of POS are conserved (for instance, beautiful entails not-ugly, and negation scopes over POS implicature, because COR is located higher in the scale, implying POS, and defeating its implicature: gorgeous implies beautiful and defeated its implicature no more than beautiful.

These two properties (entailments and implicatures) define the two mais in French, which explains why maisSN is a criterion for DM and maisPA for MN.

References

Anscombre J.-C. & Ducrot O. (1977), Deux mais en français?, Lingua 34, 23-40.

Author (2013), How much ‘logical’ are logical words? Negation and its descriptive vs. metalinguistic uses, in Taboada M. & Trnavac R. (eds.), Nonveridicality, Evaluation and Coherence Relations, Leiden, Brill, 76-110.

Horn L.R. (1989), A Natural History of Negation, Chicago, The Chicago University Press.