In the first chapter, Carnie introduces syntax as a science that lies between the formation of words and the meaning of utterances. But the foundations of syntax, e.g. answering what it means for something to be a theory of syntax, seem to have to do as much with philosophy as with science. For instance, it struck me that the generative linguist's position - that a syntactic theory aims to model unconscious rules we use to construct and evaluate sentences - relies on concepts that are subject to serious philosophical concerns: one philosophical question we could ask is how it is possible to follow a rule without knowing what that rule is; second, we might challenge the generative thesis on the grounds that the rules of English grammar seem not to be something that each of us knows - each of us can be mistaken and differ about the rules of grammar - rather the rules seem socially determined. Then whose idea of grammar are we trying to model? If we are studying the socially determined grammar, it does not make sense to say we are modeling some rules that are known unconsciously, unless we take the strange step of asserting that social groups can have unconscious knowledge not possessed by any single member of the group.
I suspect similar challenges have been posed and answered previously. I just wanted to point out the philosophical nature of the generative thesis by pointing out that the conceptual analysis of the idea of an unconscious rule is crucial. I believe most of the scientific work traced in chapter one, e.g. the derivation of anaphor rules, can proceed without a settled definition of a theory of syntax, but I would be interested in learning what work (if any) linguists think relies directly on the theoretical underpinnings of syntax.
Moving on, chapter two deals with parts of speech. The author makes the case that the category of speech for a word is determined by its syntactic and morphological distribution. Morphological distribution includes what affixes the word takes. Since we have not yet studied morphology as a separate unit, this made me wonder at which level an affix is defined. For instance, are "tion" and "sion" the same affix when they are pronounced the same? Or are affixes exactly determined by spelling?
Finally, chapter three described how syntax is formalized as a set of rules that determine a hierarchical structure, where the structure is commonly represented as a tree that relates syntactic entities high in the tree to their constituents in the levels below. This chapter in particular reminded me of the connections between natural language syntax and other formal languages studied in Computer Science and Formal Logic. All logical languages I have come across are instances of what are called context free grammars. Such grammars can be written in an efficient and widely used notation called Blauckas-Naur Form, and their complexity properties, e.g. how quickly can a string be parsed, have been extensively studied. I would be interested in knowing what linguists think the corresponding formal properties of natural language are. I think that the operations relating to grammar must be much more complex - there is no way that the rules of English syntax can be written as BNF.
No comments:
Post a Comment