Strong Linearity: Three Case Studies Towards a Theory of Morphosyntactic Templatic Constructions
Total Page:16
File Type:pdf, Size:1020Kb
Strong linearity: Three case studies towards a theory of morphosyntactic templatic constructions by Jeffrey Craig Good B.A. (University of Chicago) 1998 M.A. (University of Chicago) 1998 M.A. (University of California, Berkeley) 2000 A dissertation submitted in partial satisfaction of the requirements for the degree of Doctor of Philosophy in Linguistics in the GRADUATE DIVISION of the UNIVERSITY OF CALIFORNIA, BERKELEY Committee in charge: Professor Larry Hyman, Chair Professor Andrew Garrett Professor Johanna Nichols Fall 2003 The dissertation of Jeffrey Craig Good is approved: Chair Date Date Date University of California, Berkeley Fall 2003 Strong linearity: Three case studies towards a theory of morphosyntactic templatic constructions Copyright 2003 by Jeffrey Craig Good 1 Abstract Strong linearity: Three case studies towards a theory of morphosyntactic templatic constructions by Jeffrey Craig Good Doctor of Philosophy in Linguistics University of California, Berkeley Professor Larry Hyman, Chair The focus of this work is on templatic constructions in morphology and syntax, and its central argument is that their form is consistent with the Strong Linearity Domain Hy- pothesis. This hypothesis states that all templatic constructions must be coextensive with some prosodic constituent and that their templatic restrictions must be consistent with the characteristic phonology of that constituent. The introductory chapter lays the groundwork for later discussion, and the primary support for the Strong Linearity Domain Hypothesis comes from three case studies. The first of these is an examination of the ordering restrictions of three verbal extensions found on the Bantu verb stem, the “long” causative, the applicative, and the “short” causative. The second case study is on a minimal-size restriction found in the Chechen core verb phrase. 2 The final case study is on the order of verbs within serial verb phrases, with a focus on Saramaccan. While each of the three case studies on its own offers support for the Strong Linear- ity Domain Hypothesis, taken together they further illustrate how it can model a cline of templatic restrictions, from more specific to less specific, as linguistic constituents go from smaller to larger size. In addition, each of the case studies can be understood as exemplifying some well- known type of templatic construction. The Bantu case study exemplifies morphosyntactic verbal templates, the Chechen case study exemplifies templates conditioned by special cl- itics, and the Saramaccan case study exemplifies a class of templatic syntactic phenomena often given the label “constructions”. The concluding chapter of this work discusses possible explanations for the Strong Linearity Domain Hypothesis. In particular, an evolutionary account is offered wherein it is understood as falling out naturally from a model of the diachronic development of templates in which they arise out of the reanalysis of prosodic restrictions as templatic restrictions. Professor Larry Hyman Dissertation Committee Chair i Contents 1 Introduction: Templates, linearity domains, and strong linearity 1 1.0 Opening remarks . 1 1.1 Terminology for describing linearity . 5 1.1.1 The basics . 5 1.1.2 Some comments on the term template ................ 7 1.1.3 Linearity domains . 26 1.2 Towards a restrictive theory of strong linearity domains . 36 1.2.1 Introduction . 36 1.2.2 Subtheory I: Prosodic constituency . 37 1.2.3 Subtheory II: The notion of characteristic phonology . 44 1.2.4 The Strong Linearity Domain Hypothesis . 54 1.3 Applying the Strong Linearity Domain Hypothesis . 64 1.3.1 Introduction . 64 1.3.2 Example case study I: Serbo-Croatian second-position clitics . 65 1.3.3 Example case study II: Leggbo´ lV- prefix nouns . 71 1.3.4 Another look at Bobaljik (2002) . 77 1.3.5 Local conclusion . 82 1.4 A brief discussion of previous accounts of linearity relations in morphology and syntax . 84 1.4.1 Introduction . 84 1.4.2 Baker (1988) and the Mirror Principle . 85 1.4.3 Embick and Noyer (2001) . 89 1.4.4 Bybee (1985) . 91 1.4.5 Local conclusion . 93 1.5 The three case studies . 94 1.5.1 The role of the case study in the present work . 94 1.5.2 The three case studies . 96 1.5.3 On the bipartite structure of the case studies . 103 1.5.4 The Strong Linearity Domain Hypothesis and “templatic residue” . 104 1.6 Conclusion . 105 2 Causativization and applicativization of the Bantu verb stem: A case study of a morphosyntactic template 107 2.0 Background: Morphosyntactic templates . 107 2.1 Introduction to Bantu case study . 116 2.1.1 Causativization and applicativization in Bantu . 116 2.1.2 The role of this case study in the general analysis . 122 2.1.3 The structure of the analysis . 122 2.2 Applicativization . 126 2.3 Causativization . 127 ii 2.4 Proto-Bantu . 131 2.4.1 The structure of the Proto-Bantu verb stem . 131 2.4.2 The transitive suffix *-i¸- . 135 2.4.3 The causative extension *-ic- and the applicative extension *-id- . 138 2.4.4 Causative and applicative ordering restrictions in Proto-Bantu . 140 2.4.5 Local conclusion . 142 2.5 The relative order of the causative, applicative, and transitive: A survey . 142 2.5.1 Basic results . 142 2.5.2 Lexicalized suffixes . 150 2.5.3 Special forms of the causative, applicative, and transitive when combined . 159 2.5.4 The effect of complex suffix combinations . 164 2.5.5 Ambiguity of semantic scope for some orders . 165 2.5.6 Suffix doubling resulting in different semantics . 168 2.5.7 Languages showing productive AC order . 171 2.5.8 Local conclusion . 180 2.6 Previous analyses of the relative ordering of the verbal suffixes . 181 2.6.1 Introduction . 181 2.6.2 Baker (1988) . 183 2.6.3 Alsina (1999) . 204 2.6.4 Hyman (2003a) . 218 2.6.5 Local conclusion . 225 2.7 The CAT template and the Strong Linearity Domain Hypothesis . 229 2.7.1 Overview of argument . 229 2.7.2 The Bantu verb stem as a prosodic unit . 231 2.7.3 The characteristic phonology of the Bantu verb stem . 234 2.7.4 Applying the Strong Linearity Domain Hypothesis to the Bantu data . 250 2.7.5 Some notes on why CA is attested instead of AC . 262 2.7.6 Local conclusion . 272 2.8 Conclusion . 273 3 The preverbal enclitic ’a and the Chechen verb phrase template: A case study of a templatically-positioned clitic 275 3.0 Background: Clitics and templates . 275 3.1 Introduction to Chechen case study . 285 3.2 Overview of pertinent aspects of Chechen grammar . 291 3.2.1 Nominal inflection and gender classes . 291 3.2.2 Clause chaining . 293 3.2.3 Word order . 294 3.3 The basic structure of the Chechen verb phrase . 295 3.4 The clitics found in the Chechen verb phrase . 297 iii 3.4.1 Introduction . 297 3.4.2 The negative proclitic ca . 298 3.4.3 The proclitic ma . 299 3.4.4 The enclitic ’a ............................301 3.5 The copy infinitive and the verb phrase template . 315 3.5.1 Introduction . 315 3.5.2 The focal use of the copy infinitive . 315 3.5.3 The obligatory copy infinitive and core the verb phrase template . 317 3.5.4 Other uses of the infinitive . 321 3.5.5 Why the copy infinitive does not appear to be an object . 327 3.5.6 Local conclusion . 330 3.6 Previous analyses involving type 5 clitics . 332 3.6.1 Introduction . 332 3.6.2 Marantz (1988) . 333 3.6.3 Halpern (1995) . 342 3.6.4 Preverbal ’a as a second-position enclitic in its prosodic phrase . 344 3.7 The Chechen verb phrase template and the Strong Linearity Domain Hy- pothesis . 346 3.7.1 Overview of argument . 346 3.7.2 The Chechen verb phrase as a prosodic constituent . 347 3.7.3 The characteristic phonology of the Chechen verb phrase . 351 3.7.4 Applying the Strong Linearity Domain Hypothesis to the Chechen verb phrase template . 360 3.8 Conclusion . 377 4 Saramaccan serial verb phrases and uniformity in serial verb ordering: A case study of a constructional template in the syntax 379 4.0 Background: Topological fields and constructions . 379 4.1 Introduction to Saramaccan case study . 395 4.2 Serial verb phrases . 402 4.2.1 Their descriptive properties . 402 4.2.2 A general review of some syntactic analyses of serial verb phrases . 408 4.2.3 Carstens (2002) . 416 4.2.4 Veenstra (1996) . 428 4.2.5 Other explanations for ordering effects: Temporal iconicity and the thematic hierarchy . 434 4.2.6 Local conclusion . 444 4.3 Saramaccan serial verb phrase ordering and the Strong Linearity Domain Hypothesis . 445 4.3.1 Overview of argument . 445 4.3.2 Saramaccan phrasal prosody . 447 4.3.3 The characteristic phonology of the Saramaccan serial verb phrase 473 iv 4.3.4 Applying the Strong Linearity Domain Hypothesis to Saramaccan serial verb phrases . 477 4.3.5 Comments on the attested verb orders in serial verb phrases . 486 4.3.6 Local conclusion . 491 4.4 Conclusion . 492 5 Conclusion: Towards an explanation for the Strong Linearity Domain Hypoth- esis 496 5.1 Introduction . ..