Degree Attenuators: Attenuating polarity items in the degree domain

Natural languages have a wide variety of lexical items that express degree or quantity, or that modify expressions of degree and quantity. An important aspect of many of them is that they are polarity sensitive.  This project, which builds on the results of the DFG-funded project ‘Scales and their Structures’, investigates degree attenuators (DAs), a class of scalar polarity items that are characterized by their attenuating effect (Israel 1996).

Paradigm examples of this class are the negative polarity item much (e.g. Fred *has / doesn’t have much money) and the positive polarity item fairly (e.g. Fred is / *isn’t fairly tall).  The defining characteristic of DAs is that they are comparatively uninformative, in that in the contexts where they are licensed, they produce weaker assertions than salient alternatives.  For example, ‘doesn’t have much money’ is weaker than ‘doesn’t have money/any money/a dime’, and ‘fairly tall’ is weaker than ‘very/extremely tall’.  This attenuating or weakening character is a surprising property: better known polarity items – e.g. any, ever and minimizers such as (lift) a finger – have the opposite effect of strengthening the utterances in which they occur, and some prominent theories have taken such strengthening effect to be a necessary component of polarity sensitivity.

DAs constitute a large natural class, including not only vague quantifiers/quantity adverbials (much, far, Japanese anmari) but also precision regulators (about, ungefähr) and a variety of adjectival degree modifiers (e.g. all that, sonderlich, ziemlich). Yet while DAs lie at the intersection of two extensively researched topics in semantics, namely degree modification and polarity sensitivity, they have received relatively little attention in either tradition.

The goal of the Degree Attenuators project is to develop a comprehensive semantic theory of the class of DAs.  The project approaches their polarity sensitivity via an in-depth investigation of their scalar semantics, drawing on insights from the ‘Scales’ project. As such, this is not only the first large-scale study of attenuating polarity items, but also the first to investigate polarity sensitivity from the perspective of a degree-based semantic framework.



Funding period


Principal Investigator

Dr. Stephanie Solt