Project description
Project goal
Indeterminacy is not an exception in language, but can be found at all linguistic levels. It not only occurs frequently, but also represents a central mechanism in grammatical change as well as a frequent phenomenon in empirically based grammatical classification. Theoretically sound and empirically robust analyses of grammatical indeterminacy phenomena are desiderata of empirical grammar research: although indeterminacy phenomena are a fundamental characteristic of grammatical change and the classification of grammatical units, existing studies often lack uniform theoretical frameworks for phenomena such as ambiguity, polysemy, and vagueness. Moreover, there is no widely accepted empirical operationalization of indeterminacy phenomena. Despite this, these indeterminacy phenomena can be classified using approaches from theoretical semantics. In this project, theoretically based classifications will be transferred to the conception and methodology of empirically-based grammatical research.
The aim of the project is to develop and establish a sound theoretical framework for the empirical analysis of indeterminacy in grammar. The foundation builds on the canonical typology approach (Brown et al. 2013), which has already been established for the empirical operationalization of theoretical grammatical concepts.
Quantitative analyses of grammatical indeterminacy phenomena will be used to test and refine the classification criteria established in semantics, most notably those proposed by Pinkal (1985).
The empirical basis of the project including exemplary cases of the concept will be created using various studies on frequency, form, and resolution of grammatical indeterminacy. On this basis, a classification concept will be developed within the framework of canonical typology, which claims to be applicable to both grammatical change and classification phenomena.
Research question
The project's approach can be illustrated by sample sentences in (1) and (2), which contain both ambiguous and vague structures and can be located in the field of grammar:
(1) Er hat erzählt, dass er viel Staub gewischt hat. ('He told that he did a lot of dusting.')
(2) Sie ist am Wochenende viel gefahren. ('She drove a lot over the weekend.')
According to the classification in lexical semantics (cf. e.g., Pinkal 1985), ambiguity arises when an expressions allows for two distinct readings. These readings must be disambiguated in some way (e.g., through context), as both readings cannot be processed simultaneously. In case there is a conventionalized reading, which is usually in focus, even if both readings are theoretically possible, we speak of polysemy, a weakened form of ambiguity. With regard to the part of speech, viel 'a lot' in (1) can be classified both as a determiner of the noun Staub 'dust' [viel Staub]NP 'lots of dust' and as an adverb that is not part of the NP [viel]ADV[Staub]NP 'a lot of dusting'. At the morphosyntactic level, the structure can therefore be classified as ambiguous; different readings allow for different analyses of the syntactic elements, which go hand in hand with different word type classifications (as either a determiner or adverb). A reading in which both classifications co-exist is not possible. Context commonly determines whether viel needs to be classified as an adverb or a determiner. This structural ambiguity results in ambiguity between nominal and verbal quantification on a functional level, as viel can both quantify the amount of dust nominally and the activity of dusting verbally.
In addition, (1) and (2) illustrate the vagueness of viel at the lexeme level. According to Pinkal (1985), vagueness can manifest in multiple readings. As an adverb, viel allows for iterative ('frequently'), durative ('long'), as well as locative ('far') readings depending on the context in (2). These readings are related to each other in vague structures. Vague readings cannot always be distinguished from one another, and can be combined in a generalized reading without the need to specify or disambiguate them (cf. e.g., Lakoff 1970; Pinkal 1985). The wording viel gefahren 'drove a lot' initially only conveys that the activity of driving was carried out to a significant extent. There are several possibilities as to how this high degree could be achieved: she might have driven fifteen times, covered 2,000 km, or she might have driven for 36 hours. While context can possibly clarify and resolve this ambiguity, this is oftentimes not the case.
The project investigates how phenomena of indeterminacy in grammar, as previously outlined, can be operationalized for empirical analyses and theoretically specified based on empirical data. To achieve this, it draws on existing resources and approaches from semantics and empirical gramma research whenever possible, developing a concept within the theoretical framework of canonical typology. This framework will allow for the empirical testing of theoretical assumptions on grammatical indeterminacy.
As an exemplary area of investigation, the project examines empirical data on indefinite quantifiers such as viel ('a lot'), wenig ('little'), and mehr ('more'). It herewith addresses the question which types of indeterminacy characterize its development and classification and explore the mechanisms through which these manifest.