Introduction
Until 1957, the sciences of linguistics and cultural anthropology placed language in the domain society and outside of an individual.
For linguists and anthropologists, human languages were completely determined by the environment in which the child grows up, and there were no limitations on the differences between languages.
In other words, the view that they embraced was that every imaginable language could, in principle, be found somewhere in the world.
On the other hand, in psychology departments, behaviorism of B. F. Skinner, that run in the strand of intellectual tradition of empiricism stretching as far back in history as the work of John Lock, essentially concurred with the views of linguists and anthropologists.
The behaviorist view was that language is completely a matter of the environment and it is acquired through the mechanisms of imitation and negative and positive feedback. In 1957, in his book Syntactic Structures, Noam Chomsky struck at the very core of dominant intellectual traditions by proposing an entirely new view.
In contrast to other linguists, he suggested that language is for the most part innate. Also, in contrast to behaviorist psychologists, he claimed that the mechanisms of imitation or reward and punishment do not play a significant role in language acquisition due to the fact that on the basis of limited, fragmentary and disorganized linguistic input, children infer incredibly complex abstract rules for producing grammatical sentences of a language.
Generative Syntax
The inborn device that enables children to be so effective in acquiring language contains all the rules of the Universal Grammar, which is mirrored in the underlying principles that constrain the variations between human languages.
Chomsky (2002, p. 18) argues that the rules of grammar are independent of meaning and that by studying those rules one can arrive at the underlying computational system that generates the grammatical sentences of all languages, which is the Universal Grammar.
Therefore, on this view, generative syntax sets out on an extremely interesting and extremely difficult journey of making sense of the incredible amount of linguistic data from the languages of the world in order to uncover those universal principles, and the crucial method in this process is discrimination between grammatical and ungrammatical sentences of a language and providing explanations as to why such differences exist.
The results of this quest have been astonishing as the explanatory power of syntactic theories has already crossed the language boundaries and some concepts that are used to explain the phenomena of one language can be used with a great predictive power to account for the data in other languages.
Another interesting component is that this quest has revealed a lot about the functioning of human mind because the kinds of powerful abstract explanations that syntax gives about such a mundane activity as speaking a language suggests that there is a lot more going on in the human mind than what is physically manifested (Carnie p. 5).
There are many of these abstract theoretical notions that linguists use to explain certain facts about languages that have no overt physical manifestation, but their explanatory theoretical power is such that their existence can be taken with great certainty.
Binding
The first concept in this set is the concept of binding. According to generativists, an element binds another element if it c-commands it and if they are co-referential (Chomsky, 1980).
C-command is the structural relationship between two elements in a syntactic tree such that one can be said to c-command the other one if that other element is located in the area of the tree contained by the node that dominates the first element (Reinhart, 1976).
The notion of binding is relevant and effective because it can help in explaining the contrast between the following examples:
- John surprised himself.
- *The picture of John surprised himself.
One can account for this difference by claiming that the reflexive (himself) must be bound by the co-referential expression (John) within its clausal domain (Chomsky, 1980), otherwise the sentence is ungrammatical.
The sentence in (2) is, thus, ungrammatical because “John” does not c-command the reflexive, but in (1) it does. The provision “in their clausal domain” is significant because (3) is ungrammatical even though “John” c-commands the reflexive because the two elements are in separate clauses.
- *John claims that Mary disappointed himself.
Movement
Movement is also one of important concepts that Chomsky (2002, p. 90) observed when discussing sentences of different types, which seem to stem from the same underlying structure and the differences are the result of reordering of elements.
For instance, the example in (4) is an illustration of a construction that is said to be produced by movement. Namely, “what” is said to have moved from the position after the verb, which is marked by brackets, because this is where it gets its interpretation.
- What did John see ?
Now, this kind of claim is not a syntactic proof, but there are plenty of syntactic arguments for this analysis. For example, one can consider the theory of binding briefly described above.
If the wh-words or phrases really are in some way related to the position after the verb, than it would be possible for subjects to bind reflexives in those positions despite the fact that they appear to the left of the subject.
This is because in deep structure, the subject would still c-command the wh-phrase. This prediction turns out to be correct, and the example in (5) illustrates that.
- Which picture of himself did John see ?
This is an example of how syntactic theories and explanations are interdependent and interrelated, which shows how this approach might be capable of deriving a general theory that can explain all the phenomena of language in a uniform manner.
Empty categories
Another very abstract, but incredibly, convincing postulate that linguists make while discussing the issues in the generative approach to syntax is the idea of empty or phonologically null categories.
Namely, if there are specific reasons to assume that there exists an empty category in a certain position in a structure, linguists might make that assumption, but then they are obliged to give a convincing argument for doing so.
One such postulate is the category of PRO, which is a nominal referential element that is usually found in subjects positions (Chomsky, 1981). Linguists postulate that PRO exists on the basis of sentences like (6).
- John wants to become a millionaire.
The problem with these sentences is that both the verb want and the verb become need some entity that “wants” or “becomes”. However, in (6), there is only one entity “John”, which has to be present in both clauses.
In this situation, linguists might assume that there is an unpronounced nominal element PRO in the position marked in the example in (6) that serves as the entity that “becomes” in the embedded clause.
This presupposition can only be taken seriously if it handles additional data well, and this is precisely the case.
For example, there are reasons to postulate that the head of the tense phrase always requires some nominal element to be in the local relationship with it. In (7), the position of the tense head is marked by “(T)”. It requires “John” to be next to it.
- John (T) is running.
That this requirement always applies is supported by the examples in (8) and (9) where semantically, there is no need for a subject, and yet, a pronoun “it” has to be inserted.
- It (T) is raining.
- It (T) seems that John is happy.
The requirement that “it” be inserted here is purely syntactic and has nothing to do with meaning.
The objection that in sentences like (6) subject is simply not necessary can be replied to by the example in (10), where we can see that as soon as the subject of the embedded clause is not co-referential with that of the main clause, it cannot be left out.
- John wants his sister to become a millionaire.
Finally, one additional and quite convincing proof for the existence of PRO comes from binding. Binding theory would require that if PRO is really present in the embedded clauses, it should be able to bind reflexives.
If there is no PRO, then the main clause subject should not be able to bind the reflexives as binding does not cross clause boundaries, and the sentence would be ungrammatical. The conditions imposed by binding turn out to fit perfectly with the theory of PRO as the sentence in (11) is perfectly grammatical.
- John wants to transform himself.
Conclusion
In conclusion, generative syntax enables us to attempt to study the notion as exciting as Universal Grammar with great scientific rigor and precision.
The kind of arguments that linguists make about the abstract concepts that they introduce are incredibly strong and convincing, which is why generative syntax has been so successful.
Finally, the fact that people employ such abstract concepts in their everyday speech tells us a lot about the incredible complexity of human mind.
References
Carnie, A. (2011). Modern Syntax a Coursebook.. Leiden: Cambridge University Press.
Chomsky, N. (1980). On Binding. Linguistic Inquiry, 11, 1-46.
Chomsky, N. (1981). Lectures on government and binding. Dordrecht, Holland: Foris Publications.
Chomsky, N. (2002). Syntactic structures (2nd ed.). Berlin: Mouton de Gruyter.
Reinhart, T. (1976). The syntactic domain of anaphora. Ph.D. thesis, MIT, Cambridge, MA.