This monograph presents a long essay that was serialized in the International Journal of Dravidian Linguistics under the title Some Implications of the Strict OV Constraint for the Typological Analysis of South Asian Languages. This monograph focuses on the specific issue of the distribution of finite and non-finite predicates in complex sentences in the Dravidian languages. The distribution of these two kinds of predicates is neither random nor arbitrary; hence, some grammatical process must govern their patterning. But do these patterns reflect the operation of general typological principles, and if so which ones?
In 1988 1 published The Serial Verb Formation in the Dravidian Languages, which argued for a rule that governs the distribution of finite and non-finite predicates in the Dravidian sentence, no matter how complex it is. The goal was descriptive: to formulate for the first time anywhere a rule that captures the distribution of these two kinds of predicates in the Dravidian sentence. In two reviews of that book and subsequent publications, Hans H. Hock claimed that their distribution followed from a different rule, which he called the SOV Constraint, or SOVC. He believed that his constraint, which essentially correlates finiteness with coordinate and non-finiteness with subordinate constructions, served both as the basis of a description of Dravidian syntactic patterns and as a general linguistic typology to explain those patterns in Dravidian, Indo-Aryan and other SOV or verb-final languages.
As a matter of sound methodology, before we ask questions of a general typological nature, we must first secure accurate descriptions of the linguistic phenomena they purport to explain. After all, typological analyses derive their explanatory adequacy from the many grammars they subsume, and if those grammars are observationally compromised or descriptively flawed, we can hardly expect the typology to succeed.
Since the SOV Constraint rested on scant syntactic argumentation, I set out to determine whether this constraint, its observations and predictions could, in fact, withstand critical examination. The results are presented here. The arguments show that the SOVC consistently misrepresents the syntax of many basic sentence patterns in Dravidian and in other verb-final languages, so it is observationally and descriptively flawed. Moreover, internal contradictions written into the various formulations of the SOVC prevent it from generating coherent descriptions or cogent predications about syntactic patterns in any language. The constraint is riddled by two major problems. First, the typological linkages it purports to make between finiteness and coordination and between finiteness and constituent order are shown to be spurious, often incoherent. Second, its connection with constituent-order principles (e.g., if a language is verb-final, it has postpositions) proves to be illusory. The SOV Constraint thus fails both as a description and a typological explanation of these syntactic phenomena in Dravidian or any other verb-final language. Lacking these shortcomings, the rule I originally devised in 1988 still serves as the basis of a linguistically adequate treatment of Dravidian syntax.
In recent years, serious attention is given to typological studies. Many reputed linguists are pursuing studies in this field. Typological studies on Indian languages including Dravidian have found an important place in the past few decades. The researches done in this field by Jaroslav Vacek, Hans H. Hock, Sanford B. Steever, Umamaheswar Rao, K.V. Subba Rao and the like have given ample stimulation for the younger researchers to take keen interest in typological studies. A systematic typological study of the present-day languages may show that their structural relations and divergences are not confined to the segregated compartments of the so-called language families. Many linguistic features are seen cutting across the language families. A member of a particular language family may share certain features with a language belonging to a different language family. It is possible that some of such features will not be found in other languages of the same family. Whatever it may be, typological studies are not only useful tools for the studies on languages but also they are helpful for making valid inferences on ancient settlements, population movements, major historical events etc.
The present book Dravidian Syntactic Typology by Sanford B. Steever is a revised version of a long article published earlier in the International Journal of Dravidian Linguistics. Specific attention is given in this book on finiteness in Dravidian sentence. In addition, a critical evaluation of the earlier studies in this field from the author's point of view can be seen.
The views and comments expressed in this book are entirely those of the author and not of the International School of Dravidian Linguistics.
The present book authored by Sanford B. Steever is definitely a contribution to the field of Dravidian Linguistics and so we present this work to all those who are engaged in the study and research on Dravidian languages.
A series of publications written over twenty years ago (Steever 1987a, 1988a, 1988b) outlined a program of analysis for complex syntactic structures in the Dravidian languages that focused on the distribution of finite and non-finite predicates within the sentence. This program was intended to anchor a set of synchronic studies on sentence structure in the individual languages and, ultimately, a comparative study of the entire family. In the most developed of these (Steever 1988a: 111-12), a rule governing the distribution of finite and non-finite predicates within the Dravidian sentence was approached in several steps, culminating in Rule (1) below.
(1)a. Each Dravidian sentence has n+1 functionally finite predicates, where n stands for the number of finite predicate embedders (FPEs) and 1 for the Root Domain of Finite Predicates (RDFP).
b. The functionally finite predicate in the RDFP occurs highest and rightmost in the root clause, commands all other predicates and is itself commanded by none. Each FPE induces a dependent domain of finite predicates (DDFP). The functionally finite predicate in each DDFP occurs highest and rightmost in the domain defined by the FPE, commands all other predicates in that domain and is itself commanded by none.
Send as free online greeting card
Email a Friend
Manage Wishlist