b) NP^T+N (T is a determiner)
c) VP->V+NP
d) T^the
e) N->man, ball, etc.
j) Verb->hit, took, etc.
Given this set consisting of six rules, one can generate an English sentences or number of sentences changing only the N and the transitive V, in accord with the situation he speaks about.
There is another representation of generation of a sentence on the basis of the 1C grammar. This is the “derivation tree” diagram, which is as follows:
Sentence
NP VP
T N NP
T N V ‘T
The man hit the ball.
The derivation tree diagram gives less information of how a sentence must be generated than the rewrite rules, because it does not show explicity the order of the generation, whether you must begin with the NP or the VP. But it illustrates the grow-pings of the 1C more clearly.
The 1C model is more powerful than the linear model. The 1C model has certain advantages as a generating of a sentence must proceed.
In spite of certain merits it is open to criticism. If the sentence is expanded, then the rewrite rules become too numerous to hold and generation of the sentence hinders (‘hinda-v, мешать, препятствоватьвыполненногочего-л).
There is another demerit (dii’merit-недостаток) in the 1C model, something the model cannot show that the relations between the elements of the two sentences are different:
sentence Sentence
Np vp np vp
Np be Ap np be Ap
Np be A v-to np be A v-to
John is eager to please John is easy to please.
These sentences have the same derivation tree showing the 1C of the sentences. Only the transformations of the two sentences can show the difference of the relations of-their elements, that is, “John” is subject of “please” in “is eager to please”. But “John” is object in “is easy to please”.
The following transformations will prove it:
1. I (easily) pleases John;
2. We can (easily) please John;
3. Forms to please John is easy; _______
II. John eagerly pleases everybody;
III. John eagerly pleases everybody;
IV. John can please people;
V. For John please people is pleasure;
There for the critical review of the linear theory and the 1C theory proves that their application as sentence-generation models is very limited. But the 1C theory is more fit (fit-годный, подходящий), and kernel sentences must be generated on the model. We must study the 1C model as a means of producing kernel sentences, the simplest sentences of the language.
The kernel sentences are more important as they are. The other kinds of sentences are their transforms and to be studied to generate (build up) sentences by means of a more powerful grammar, the Transformational grammar.
Questions:
I. What was the purpose of language teaching in the previous centuries?
II. What is the purpose of language teaching today?
III. Why has the young child’s speech problem come to light?
IV. What kind of sentences belong to the kernel?
V. Who were the first profounders of the T-grammar?
VI. What linguistic school gave rise to the T-grammar?
VII. What is the model of the linear theory?
VIII. What are the rewriting rules?
IX. What sentences must be generated by the 1C model?
1. Write down a set of 1C rules for generating the following sentences:
A. The boy saw an arrow
B. The sign was an arrow
2. Draw the derivation tree for the following:
I. Mary is curious to know.
II. Kitty was happy to love
Do'stlaringiz bilan baham: |