Nicole Gotzner will receive 1.4 Mio Euros by the German Research Foundation (DFG) to lead an Emmy Noether Research Group. With the Emmy Noether Programme the DFG wants to give exceptionally qualified early career researchers the chance to qualify for the post of professor at a university by leading an independent junior research group for a period of six years.
In the project "Scales in language processing and acquisition: Semantic and pragmatic contributions to implicature computation (SPA)" Nicole Gotzner and her team investigate the meaning of scalar expressions. Scales are hidden behind many words. If something is "large", it could also be "huge", for example. Nicole Gotzner and her team will develop a new model to interpret such scalar expressions. They will investigate the mental representation of scalar expressions in adults and how children learn the meaning of these expressions in general.
Pragmatic inferences are ubiquitous in communication among humans. For communication to be successful, we need to understand the literal meaning of utterances but also which alternative meanings are not intended. Consider the following exchange between an adult and a child. The adult asks “Would you like to have ice-cream or chocolate?” and the child responds “I would like to have both”. This response contradicts the scalar implicature typically triggered by or, which children do not fully understand until age 7. Scalar implicature is standardly modeled via a so-called Horn scale, which is an association between the weaker or and its stronger alternative and in the mental lexicon. When a speaker uses the weaker expression, the listener may conclude that the stronger alternative does not hold. Horn scales exist across a variety of expressions, including connectives (<or, and>), quantifiers (<some, all>) and various adjective types (<large, gigantic>, <wet, pouring>, <possible, certain>). Despite the pervasiveness of such scales in language, little is known about their cognitive basis and how children learn to associate different terms on a scale during the course of language acquisition. Implicatures have long been studied in formal and experimental pragmatics, yet the mechanisms underlying implicature computation are still a matter of hot debate.
The project SPA will shed new light on long-standing debates about the nature of Horn scales and alternatives, by investigating a large number of such expressions in language processing and acquisition. The overarching goal is to develop a new model of scales and implicature, that accounts for variability among such expressions. We will examine (a) the extent to which a single mechanism underlying implicature computation can be retained for different Horn scales and (b) the kinds of alternatives that constitute the basis for implicature computation. A major focus will lie on the interpretation of adjectival scales, which have been well-studied in semantics but remain underexplored in pragmatics. We will employ a variety of psycholinguistic methods as well as probabilistic modeling tools in order to integrate insights from semantic and pragmatic theory and cognitive science. This project represents the first large-scale attempt at testing how various scales are processed and how semantic and pragmatic representations of scales develop in tandem.