Theory of binding Book

 
TOWARDS AN ALTERNATIVE
THEORY OF BINDING
Abhilasha Jain
and
B.N. Patnaik
Section - V
We now turn to the issue of the binding domain. Following traditonal grammars, we informally treat both pronouns and anaphors as "substitute" terms; in another terminology, as terms that do not have complete independent, intrinsic reference and are dependent on other terms for the same. Although both are dependent elements, there is a difference between the two; an anaphor must be bound within a very limited domain which must be smaller than a tensed clause, whereas, a pronoun can receive its antecedent even from outside the sentence, in the discourse. The antecedent of the pronoun need not even be explicitly mentioned; it may be part of the shared knowledge of the speaker and the hearer. Or it can be associated with a non-verbal behaviour, namely, pointing. For example, voh akalmand hai (he is intelligent) can associated with a deictic or the referrent of the pronoun can be part of the shared knowledge of the interactants or the sentence can be part of a discourse and the referent of voh must be there in context. When none of these conditions are satisfied the pronoun remains uninterpreted and the sentence becomes ungrammatical.
A pronoun behaves differently from an element bearing reference (an R-expression) in the following respect:
Ordinarily an R-expression is free, but, as is well-known, can be repeated under certain circumstances as in the sentence
Mohan likes                              Mohan
    i                                           i
alone." However, a pronoun can never be repeated in this manner as shown by the ungrammaticality of (130).
(130) *voh usko (hi) accha lagta hia
    i     i
he him like-PRES
The ambiguity of a sentence like "Ram believes that he is honest" is, as is well-known, due to the fact that the pronoun can be interpreted as having an antecedent within the sentence and as having an antecedent outside the sentence, in the discourse. As far as sentence grammar is concerned, in principle there can almost always be a discourse in which a pronoun can receive its antecedent. Hence although a dependent element, a pronoun which does not receive reference within a sentence, is not to be treated as an uninterpreted dependent element.
Although a pronoun can get reference from outside the sentence almost always, there are instances one being (131), when the semantics of the sentence rules out such a possibility.
(131) ram ne hari ko uske paise lauta diya
    i     i
ram CM hari CM his money return+PAST
(Ram returned hismoney to Hari.)
Owing to the semantics of the verb lauta (return), the referrent of the pronoun has, of necessity, to be the recipient of the money.
In sum, both, anaphors and pronouns are dependent elements, although in different ways, and as far as sentence grammar is concerned, except for instances like (131) a pronoun may be considered to be a dependent element that can find its antecedent outside its S and as such a pronoun without an antecedent in the sentence is not to be treated as an instance of a dependent element which is uninterpreted.
If anaphors and pronouns are two different types of referentially dependent elements, then this difference can be expected to show up both in (a) the choice of the antecedent and (b) the nature of the domain in which they find their antecedents, i.e. the binding domain. As far as the latter i.e. (b) is concerned, the problem of defining a domain of anaphors boils down to specifying where an anaphor must get its antecedent, a domain beyond which it cannot look for an antecedent and, for pronouns, a domain in which a pronoun must not look for an antecedent. There cannot be a limiting domain for pronouns, since they can have their antecedents from the discourse.
We know that mukhya is the antecedent of anaphors and mukhya is the or one of the (if there are more than one) arguments of the relevant predicate. The smallest domain in which an anaphor must find its antecedent is, thus, what we may call the proposition. Is it, then also, the domain beyond, which an anaphor cannot look for an antecedent? That is, is it necessary to distinguish between a minimal domain in which an anaphor finds its antecedent and the maximal domain beyond which it cannot look for an antecedent? This distinction seems to be necessary because of at least the following:
a) sometimes a propostion can be one of the arguments of a predicate. One such argument is the possessive phrase.
b) the theory is not constrained so as not to allow an anaphor to be the mukhya of a proposition. In such cases, the anaphor's antecedent, of necessity, has to be outside its own proposition.
Let P be the domain in which the anaphor would ordinarily find its antecedent and P we shall turn soon to the precise characterization of P and P - the domain in which it must find its antecedent. Is it posible to merge the distinction by stating that an anaphor must find its antecedent within P ? It really not, because, if P has a mukhya which is distinct from the anaphor and P has a mukhya, too, then, in order to achieve the correct result the notion of proximity must be used and if must be specified that the anaphor must choose the most proximate mukhya as its antecedent. But there is no real gain. Besides, as we shall see in the cases of pronouns and PRO, P and P need to be distinguished and therefore, the attempt to merge the distinction may be abandoned.
 
Theory of Binding Page
 
FeedBack | Contact Us | Home
ciil grammar footer