Flexible Meaning Adjustments in Context: Underspecification vs. Reinterpretation

In view of the combinatorial nature of linguistic meaning, semanticists agree that some version of the Fregean Principle of Compositionality must certainly hold. But there is also consensus that the richness and complexity of meaning components in natural language go far beyond what is determined by purely grammatically driven processes. When interpreting a sentence, we draw on information from context, world knowledge and the acute situation of the utterance. Language interpretation also regularly displays meaning adjustments, smoothly giving rise to creative interpretations and pragmatic enrichments in a variety of phenomena at the semantics-pragmatics interface. One particularly prominent example is complement coercion, that has been on the agenda of both theoretical and experimental work now for more than 30 years. A number of theoretical proposals have been put forward ranging from underspecification accounts to theories assuming dedicated coercion mechanisms. On the empirical side, findings from experimental studies also provide a diverse picture in evidence of various accounts. The workshop aims at forming a coherent picture of the evidence gathered on coercion and related phenomena, synthesizing the different findings, and evaluate theoretical proposals in the light of the available evidence. We want to touch on the questions of how linguistic modeling can encompass all the different types of information used during sentence comprehension, and how we can distinguish between hard limits governed by the language facilities and conventions that can be overwritten. Special focus lies on the limits of online, in situ meaning adjustment. Do (clear) limits exist, what is their nature and can they be creatively extended or overwritten? Or, is language processing accomplished by inherently flexible operations, drawing on sources of many different knowledge domains? How is the influence of different information organized, and how is its impact best observed? The wo...