For Phrase Structure Rules, what about dialects or spoken language that deviate from conventional rules? Carnie states that linguistics uses descriptive methods to understand language, but defining the structure rules in the study of syntax seems very prescriptive to me. Granted, there may be ways to account for the exceptions found in languages, but the aim is still to create generalizable standards for language, without adequate focus on how sentences are used in everyday speech. Perhaps we do need these rules, and our brain can only process language with a rigid structure similar to that used to make trees. The rapid language acquisition children display is inherently dependent on the ability to generalize syntactic rules. Although this often leads to over-generalizations and errors, the entire process is unconscious and does not require awareness of the rules that we automatically absorb.
The big question for me is how to integrate semantics with syntax? Embodied simulation theory (and common sense) postulates that when humans process language, we create a virtual representation of what the language describes. Thus, in the example in Chapter 1, "the cat spots the kissing fishes," a child creates an internal simulation of the situation based on his or her experience in the world. It is not merely a one to one correlation of a sentence with an image, you need to be able to create a representation of the meaning. I recognize that this is a slightly separate issue from pure syntax, as can be seen in the word salad example, where we recognize acceptable grammar even with nonsense words. Yet, we cannot completely separate meaning from structure. For example, in addition to needed to know that gender of a subject in order to apply the correct anaphor, you must also know properties of an object in order to determine to correct prepositions. The sentence “the cat stood in the mouse” does not make any sense if we are assuming that that animals in question are of anatomically correct size. Yet “the cat stood on the mouse” makes automatic sense to a native English speaker, in part because we are able to create a simulation of the situation, and understand a cat can be on top of a mouse, but would never fit inside one.
When we consider syntax alone, without semantics, the over simplified phrase structure rules of noun and verb phrases may provide an adequate model of human formation of sentences. But when we look deeper into how humans actually use language to communicate, the prescriptive nature of syntax fails to truly describe what occurs in reality.
Where do you see evidence that "the aim is still to create generalizable standards for language, without adequate focus on how sentences are used in everyday speech"? The generative thesis is, explicitly, that the rules of syntax already exist (we know them unconsciously) and that the job of the linguist is to model these rules, with the understanding that different groups can follow different rules. In so doing, prescriptive statements about the correctness of rules are never involved, and sentences are only deemed syntactically correct or incorrect relative to some particular standard. As for your concerns about the separation of semantics and syntax, I think Carnie would say "the cat stood in the mouse" is wrong semantically and not syntactically and that syntactic rules are not the only rules that govern sentence formation.
ReplyDelete