Morphosyntactic parasites as underspecified heads: On the double supine construction

Sammanfattning: The purpose of the present study is threefold. Firstly, I wish to investigate the phenomenon of double supines in colloquial Swedish. In doing so, I hope to achieve my second purpose, which is to take a first step towards a better understanding of a particular kind of complementation of which the double supine construction forms a subset. I will refer to this complementation as parasitic complementation. A parasitic complement has at least two characteristic properties: (1) The expected infinitival verb form that is prescribed in the corresponding standard construction is replaced by a surface form, where features seem to be copied from an adjacent superordinate verb. (2) The overt morphology of the parasitic verb never has any influence on the LF interpretation and appears to be purely cosmetic. The present study is restricted to Swedish complements of modal verbs (bare complements) where the complement verb is parasitic on a superordinate verb in the past participle form used in perfect tense formation (the so-called supine form). I refer to this as the parasitic supine construction. This type of parasitic complementation is frequent in variants of the Mainland Scandinavian languages and Faroese. I show that the parasitic supine construction is highly problematic for grammar models which combine a view of syntax as restricted to the path between the Lexicon and LF with a strong lexicalist approach to word formation. The third aim of my study is to show that the parasitic supine construction, as well as other types of parasitic complementation, offers ample support for the following three theoretical assumptions: (i) Syntax, understood as determining word order and inflectional form, involves the path to PF as well. (ii) Word formation, understood as the creation of complex syntactic heads, is not limited to the Lexicon. (iii) Phonological features are added post-syntactically by the insertion of Vocabulary items into the terminal nodes (Late Insertion). These assumptions are based mainly on ideas from Halle & Marantz (1993) coupled with insights from Acquaviva (1998) and others. The study is organised as follows. The basic properties and the distribution of multiple participle constructions are introduced in chapter 1. In chapter 2., I discuss the participle form called supine, used in formation of the perfect tense/aspect in Swedish and distinct from the past participle. In chapter 3., I describe how parasitic supines are identified aiming at a definition of a parasitic complement. Chapter 4. is devoted to my analysis of the construction. It involves a discussion of the theoretical assumptions needed to account for the facts, an outline of a grammar model, and an analysis where I consider the parasitic supine step by step from the lexical resources to each of the interface levels LF and PF respectively. Chapter 5. is an attempt to place the construction in a broader perspective. More extended parasitic complements in Mainland Scandinavian are discussed and similar or related constructions in other Germanic languages are summarized. The chapter ends with a brief discussion of the role of Autonomous Morphology in parasitic complementation.