Date: 03 Sep 2003 10:51:54 PDT From: Gulsat Aygen Subject: Nonfinite Structures in Theory and Change
Miller, Gary D. (2002) Nonfinite Structures in Theory and Change, Oxford University Press.
Gulsat Aygen, Department of Linguistics, Reed College.
Nonfinite constructions have been studied as a means to understand finiteness and clausal architecture in syntactic theory. Conjugated Infinitives (a.k.a. Inflected Infinitives), observed initially in Portuguese, Italian, Hungarian (Rouveret 1980, Raposo 1987) have been one of the major motivations for including agreement features in syntactic theory (Chomsky 1986). Until recently, our definition of "finiteness" has been shaped by evidence indicating certain syntactic domains with constraints on extraction, and the presence of nominative subject. Finiteness and nominative case licensing have been accounted for by the presence tense and/or agreement features within a clause (Rouveret 1980, George and Kornfilt 1984, Raposo 1989, and Chomsky 1986 among others). The book under review is a solid research substantiating this line of research, accounting for the evolution of nonfinite constructions from finite constructions within a feature based analysis. The analysis follows the [+/-Tense] and [+/-Agreement] line of research and is supported by empirical evidence from both synchronic and diachronic analysis of these constructions in Indo-European languages. The novel analysis of prototypical infinitives as belonging to a mood category is one of the major contributions of this book. The theoretical framework adopted is the theory of case as developed by Schutze (1997) with modifications to account for Accusative/Nominative case checking on subjects of nonfinite constructions. The book is an invaluable contribution to both syntactic theoreticians and historical linguists.
The term nonfinite is used with reference to prototypical infinitives, gerundials and participles that have no subject person agreement. Miller argues that these structures belong to a mood category because it is in complementary distribution with other formal moods and co- occurs with modal adverbs. The prototypical infinitive is observed to have five properties: 1. mood category membership 2. complementary with other moods 3. licensing of modal adverbs 4. licensing of control 5. a [+/- N] feature
In terms of diachronic change in nonfinite structures, Miller argues that the most basic type of structurally or functionally motivated change is reanalysis, and assumes three classes of syntactic change (following Willis 1998): 1. performance changes, i.e. increased frequency 2. parametric change 3. non-parametric change
Chapter 1 presents the assumptions underlying the analysis and some working hypotheses about the nature of case and agreement. Schutze (1997)'s Accord and Spec/Head Case checking mechanism is adopted: The phi features and case features are checked locally against the features of a predicate-related head. Both sets of features must be checked at once. Infinitival constructions that can check Nominative in Latin are discussed.
Chapter 2 focuses on nonfinite tense and argues that 'to' and '-ing' in English have equivalent event structures. English infinitives correspond to irrealis and prospective contexts whereas gerundials correspond to factive and completive structures. Reflective verbs (epistemic, declarative, factive) require complements with independent tense. A syntactically interesting note in this chapter regards Ancient Greek infinitive complements: as tense-aspect marked complements of reflective verbs, infinitives have an independent temporal interpretation, whereas as complements of other predicates they are atemporal.
In Chapter 3, Miller follows Landau (1999) in distinguishing partial and exhaustive obligatory control, and discusses theories of control based on the following: 1. binding relation between external argument and infinitive marker 2. movement and trace 3. feature attraction 4. lack of embedded subject position This chapter is an attempt to account for diverse empirical facts from Portuguese, Welsh, Hungarian and West Greenlandic. In the languages studied, PRO is not licensed in subjunctive and other complement clauses. PRO is licensed by the type of mood and/or lack of agreement. The argument is that PRO and lexical subjects are not in complementary distribution, and that it can receive/check any case since it has no specified phi or case features.
Chapter 4 includes a discussion of conjugated and non-conjugated infinitives in Portuguese, Hungarian and Welsh. Miller proposes that languages have two types of infinitives: clauses with inherent phi features allow lexical or pro subjects and yield to conjugated infinitives; clauses with no inherent phi features allow PRO and trace in the subject position that yield to non-conjugated infinitives. A striking example is Welsh where previously nonfinite clauses without tense or agreement are reanalyzed as finite (pp. 90-3). The Greek construction which cooccurs with the particle na and is obligatorily inflected for person an number agreement is analyzed as an infinitival based on the evidence that standard weak crossover tests indicate that the subject can be PRO and the embedded clauses can be selected by a determiner. Miller argues that neither of these properties is characteristic of subjunctives.
Chapter 5 includes the properties of two non-matrix mood formatives in West Greenlandic where Case is argued to be compatible with PRO subjects. Lexical and pro subjects require the presence of agreement morphology. Miller shows that assignments of non-structural case to the clause can block the checking of structural case inherently.
Chapter 6 argues that one core type of SC is verb/-tense in French, Portuguese and English. The presence of subject-predicate geometry is the justification of its clausal nature. The contrast between English and French in terms of reflective verb infinitival complements is accounted for the a ]+Accord] in French that allows Nominative case checking within the lower clause, and the [- Accord] in English, that requires the subject to check its Accusative case in the higher clause.
Chapter 7 argues that ECM constructions is an innovation in Old English (like French and Dutch) in disallowing ECM by requiring reflective verb infinitive complements to be [+Accord]. Miller argues that the [- Accord] setting was reinforced by new causatives that were reanalyzed from object control verbs.
Chapter 8 is a historical overview of English infinitive structures. Infinitive 'to' is argued to belong to a category M (infinitival marker) and is predicted to be different from "for". for NPs are argued to be reanalyzed as to INFs.
Chapter 9 discusses the properties of "-ing" and t-less infinitives. Miller proposes that the only difference between a participle relative and a perception verb complement in English is that participle relative adjoins to D/NP, whereas, a perception verb complement is selected.
Chapter 10 develops a theory of English gerundials and argues that "- ing' is underspecified for feature [N]; when it is positive, DP becomes a poss-ing construction. Chapter 11 presents a historical account of the English gerund. Chapter 12is significant in that it accounts for the spread of English from nominal to clausal categories. The idea of the morpho-supercategory is presented.
Miller (2002) argues that clauses with inherent phi features yield conjugated infinitives and that clauses that lack phi features yield prototypical infinitives. He proposes that a mood category is crucial to the analysis of infinitives. His empirical evidence is based on both synchronic and diachronic data from Indo-European languages.
This book presents a thorough, interesting, and informative work from a novel perspective. It presents an encompassing study of nonfinite structures. From a syntactic point of view, it argues for a very interesting account of inflected infinitives which I believe, deserves to be followed up with further research on non-Indo-European languages. The theoretical implications of this book provide a novel analysis of Agreement as Mood. Miller (2002) provides a line of research on finite/non-finite structures in terms of relevant features on the relevant functional heads. Mood being involved in nominative case licensing is a brand-new idea that could help us account for the clausal architecture of non-tense/non-agreement structures in English subjunctives as well as structures/languages with no overt tense and/or agreement. This approach could also clear the view in syntactic studies where mood and tense, modality, and tense are equated depending on the properties of languages. The infamous Tense/Aspect/Modal (TAM) label utilized for such ambiguities in studies on functional categories could be re-studied for a better understanding of the clausal architecture cross-linguistically. Although the analysis presented in the book is consistent in terms of its framework, and valuable to readers working on any framework, it could benefit from recent work in syntactic theory, particularly Chomsky (2000) and Pesetsky and Torrego 2001 for Nominative case analysis, Huang (1989) for the discussion of control structures. I strongly recommend this book to linguists working on subject-case licensing and finiteness as well as to those interested in historical syntactic change in finite/non-finite constructions.
Aygen, Gulsat. 2002. Finiteness, Case and Clausal Architecture. Ph.D Dissertation. Harvard University. Cambridge.
Chomsky, Noam. 2000. Phases. Cambridge: MIT Press.
George, L. and Jaklin Kornfilt. 1981. Finiteness and Boundedness in Turkish. in Binding and Filtering, ed. Frank Heny, p 105-128. London: Croomhelm Ltd.
Huang, C.-T. James. 1989. "Pro-drop in Chinese: a generalized control theory" in Jaeggli O. and K. Safir (eds.) The Null Subject Parameter, Dordrecht: Kluwer Academic Publishers, 185-214.
Pesetsky, David and Esther Torrego. 2001. Tense to C: Causes and Consequences. In Ken Hale: A Life in Language. ed. Michael Kenstowicz, MIT Press, 355-426.
Raposo, E. 1987. Case Theory and Infl-to-Comp: The Inflected Infinitive in European Portuguese. LI 18, 85-109.
ABOUT THE REVIEWER:
ABOUT THE REVIEWER Gulsat Aygen did her undergraduate work at Bogazici University in Istanbul, Turkey, and received a PhD in linguistics from Harvard University. Her interests include syntax, syntax/semantics interface, clausal architecture of finite and nonfinite constructions, particularly Tense, Aspect, Mood and Modality in Turkic languages. She is currently a Visiting Assistant Professor at Reed College.