Servicenavigation

Detailseite

Zurück

Projekt Druckansicht

Degree Attenuators: Die Semantik abschwächender Polaritätsausdrücke aus dem skalaren Bereich

Antragstellerin Stephanie Solt, Ph.D.
Fachliche Zuordnung Allgemeine und Vergleichende Sprachwissenschaft, Experimentelle Linguistik, Typologie, Außereuropäische Sprachen
Förderung Förderung von 2012 bis 2022
Projektkennung Deutsche Forschungsgemeinschaft (DFG) - Projektnummer 214369516
 
Erstellungsjahr 2021

Zusammenfassung der Projektergebnisse

The project ‘Scales’/‘Degree Attenuators’ had as its focus the semantics and pragmatics of expressions such as fairly tall, (about) 50 people, not much coffee and more beautiful than Zoe, whose meanings involve reference to or comparison between points on a scale. The project comprised two phases. Phase 1 investigated the semantics of such expressions by studying the properties of the measurement scales and measure functions that underlie their interpretations, yielding new insights into the nature of scalar meaning. Phase 2 focused on a particular class of scalar expressions, namely degree attenuators – polarity sensitive scalar modifiers that have a weakening or attenuating effect – and used findings from the first project phase to develop a novel account of their behavior. As such, this was the first large-scale project to approach the topic of polarity sensitivity from the perspective of a degree-based semantic theory. In both areas, the project made substantial contributions to semantic and pragmatic theory. The most important of these are highlighted here. New insights on the ontology of degree. Most centrally, Phase 1 yielded a deeper understanding of the range of scale structures relevant to natural language. Two previously unrecognized scale types were identified. First, drawing on insights from measurement theory and from the psychology of number cognition, it was shown that scales with a semi-ordered structure play an important role in the grammar. The distinction between totally and semi-ordered scales was shown to underlie contrasts between the superficially similar quantifiers most and more than half, and is proving to have explanatory value in other domains (e.g. modality). Secondly, it was shown that assuming scales that directly track proportions of a total yields an account of otherwise unexplained phenomena relating to expressions of proportion. Conversely, it was demonstrated – contra prior claims in literature – that the interpretation of gradable adjectives cannot be understood in terms of an ordinal scale derived from an ordering on a comparison class. Finally, it was demonstrated that scale granularity is best modeled in terms of sets of alternatives to a measure expression. Degree-based analysis of Q-adjectives. The adjectives of quantity many /few /much/little have been the subject of considerable research in semantics, but the focus has primarily been their quantificational use. In work finalized in Phase 1, it was demonstrated that analyzing Q-adjectives as degree expressions allows a unified account of their semantics across the range of positions in which they occur. This has become a standard theory of Q-adjectives, which is extensively cited by authors working on this class cross-linguistically. Two sources of scalar subjectivity. Adjectival subjectivity is the focus of a large body of research in semantics, with attention recently turning to the subjective interpretation of comparative o constructions such as The Picasso is more beautiful than the Mir´. The project established that there are two distinct sources of such subjectivity, namely multidimensionality and judge/experiencer dependence. These findings, and the experimental methodology they are based on, have contributed significantly to ongoing research in semantics and philosophy on the topic of subjectivity. New sources of polarity sensitivity. Polarity items with an attenuating or weakening effect are not well handled by theories that link polarity sensitivity inherently to semantic strengthening. The central achievement of Phase 2 was a novel semantic/pragmatic theory of attenuating polarity items in the degree domain, on which their polarity sensitivity derives from competition with simpler alternatives. Degree attenuators are modifiers that, due to their weak and vague semantics, produce assertions that are not meaningfully different in informativity from the simpler unmodified form; polarity-based distributional restrictions then arise as a consequence of contradictory manner implicatures. This framework was succesfully applied to a range of degree attenuators, including approximators (about), much words and moderate-degree modifiers (fairly ). This work thus demonstrates the role of vagueness and simplicity-based reasoning as sources of polarity sensitivity. A novel empirical finding was that bare cardinal numerals exhibit a contextual variety of polarity sensitivity, being infelicitous in the scope of negation in neutral contexts, but acceptable when the numerical value is contextually salient. It was shown that this pattern can be derived from a discourse constraint that numerical expressions describe a convex region in the space of answers to the current Question Under Discussion, pointing to (non)convexity as an additional source of polarity sensitivity.

Projektbezogene Publikationen (Auswahl)

 
 

Zusatzinformationen

© 2024 DFG Kontakt / Impressum / Datenschutz

Textvergrößerung und Kontrastanpassung