DISCOURSE CATEGORIES, CONVERSATIONAL DYNAMICS AND THE ROOT/EMBEDDED DISTINCTION

Abstract

The present work aims to carry out an extensive and in-depth investigation on the use of Focus in Italian, adopting an interface perspective that takes into account the interactions between the different levels of analysis involved in this phenomenon. In particular, this investigations concentrates on the properties of three types of Focus, namely, Information Focus (IF), Corrective Focus (CF) and Mirative Focus(MF), which have been mainly identified and distinguished in the literature for their semantic and pragmatic properties (cf. Ch. 1, §1.3). Hence, the main goal of this study is to verify whether these differences systematically correspond to specific syntactic and prosodic behaviours (in line with F&H’s 2007 proposal on Topics). This analysis aims to identify: (i) the syntactic and prosodic properties characterizing the three Focus types under exam, and (ii) the behaviour of these Focus types in different embedded contexts, since embedded clauses show different behaviours as far as information-structural phenomena are concerned (cf. Ch.1, §1.4.). In order to reach these goals, a written acceptability judgment test has been designed and distributed to 120 native speakers of the Italian variety spoken in Lazio. As it will be shown in the following illustration of data and relevant discussion, while providing interesting and useful data for the characterization of the phenomenon under exam, the results of this test leave some questions open and bring forth new issues, which were not expected during the first phase of experiment design. Therefore, a second set of tests was designed after that, concentrating on production and prosodic aspects, so as to properly address all the relevant issues and carry out an exhaustive analysis on the properties and use of the three abovementioned Foci. In particular, three new tests have been designed and submitted to informants: an oral production test, a forced choice task and an additional acceptability test, focused on constituent extraction from emotive-factive complements, necessary to check their status of syntactic islands suggested by the preliminary results. Hence, based on the analysis of the experimental data obtained from the above mentioned tests, it will be shown that, from a syntactic perspective, the in situ position is the canonical one for all types of Focus, since they are all considered less acceptable when fronted. Thus, it is proposed that: (i) the different semantic and discourse properties that characterize each Focus type are connected to specific p(honological)-features that are encoded (hence, merged as formal features) in the DP functional area and then matched through AGREE by a dedicated probe in the matrix split-CP domain; (ii) Focus fronting is a marked option, which is triggered by an interface requirement signalling the speaker’s intention of endowing the focused element with formal salience (cf. Ch. 3 §3.1.5.2.); (iii) MF seems to be only acceptable with matrix verbs that select a complement clause endowed with assertive force; (iv) all Focus types are considered less acceptable when combined with emotive-factive verbs, and in particular when fronted in the matrix clause, supporting an analysis of emotive-factive complements as syntactic islands.

As for the prosodic properties of Focus, data analysis leads to the conclusion that: (i) the canonical intonation associated with Focus is the H*+L contour; (ii) the raising L+H* tone is due to the formal validation of the ‘discourse salience’ of the focused element; (iii) Focus constituents mark an Intonational Boundary and, as such, they form independent Intonational Phrases, as originally proposed in Frascarelli (2000).

The present work is organized in four Chapters, which are structured as described below. The first Chapter contains a concise but comprehensive review of the literature dealing with the phenomena at issue in this work. It starts with a general description of conversational dynamics, including an oversight on the most recent and influential discourse models. Then, the concept of information structure is explored, with an overview on its two major categories, namely Topic and Focus, and their prosodic, formal and discourse-semantic properties. Finally, the syntactic framework is presented, briefly describing the main concepts of Generative Minimalism, as well as the distinction between root and non-root clauses. The second Chapter is focused on the first acceptability test. Firstly, the main research questions are illustrated, along with the design of the test (i.e., items, fillers, structure, participants and submission methods). Then, the core of this Chapter is presented, namely the analysis of data, with a brief digression on the statistical methods and techniques adopted. Finally, preliminary conclusions are drawn and unsolved issues pointed out, leading to the necessity of additional tests. The third Chapter is thus dedicated to a new set of tests and it is divided into two parts: the first part concerns the production and the forced-choice tasks, grouped together into a single pilot test, while the second part describes the newly designed acceptability test. Similarly to Chapter 2, both parts start with a description of the test design and continue with the analysis of data and relevant results, also including prosodic analysis. Finally, the fourth Chapter presents the conclusions of this study, confronting the results of the different tests used and putting forth a comprehensive proposal taking into account the different research questions.