Re: X-Bar Theory

>hehe, thanks, please go into detail, im 15, so i really and just getting my
>feet wet (trying to soak up all knowledge i can find while im at it too!)
Oops, I'll try to be as explicit as I can (just keep in mind that I'm not an english
native speaker, there will be mistakes as I was taught it in my language, so
please be leniant). Here, I will use "language" to refer to "the specificly
human capacity to communicate", and "tongue" to a "particular instanciation of
the language in a given social group" .

Generative Grammar (GG) is a theory initiated by Noam Chomsky (around 1950, depending on
which event or book you consider) : the main ideas are :
1 ) language has a an abstract structure which has the rigor of mathematics.
 2) Human beings can produce an infinite number of sentences with a finite number of elements.
3) Behind the "performance" (sentences they produce in every day life, even when
stressed, drunk or tired) there is a "competence", what people are *able* to
do. GG will just work on competence.
4) Competence can *generate* grammatical structures. Grammatical structures are
sentences that sound correct according to a native speaker of a given tongue.
So an algorithm should be able to do so.


Then, GG is a syntagmatic grammar, so structures are recursively made of phrases. A
sentence (S) has basically a structure such as S = NP + VP (Noun Phrase + Verb
Phrase).
For instance,
"the man plays in the garden"
can be analysed as
S = NP+VP

NP= Det + N (Noun)
Det = 'the'
N = 'man'

VP = VP + PP (here VP gives VP and not V because 'in the garden' is a modifier
and not a complement, so it can't be a sister of V. PP is Preposition Phrase)
VP = V
V= 'plays'
PP = P + NP
P = 'in'
NP = Det + N
Det = 'the'
N = 'garden'



The interesting thing, according to me , is that contrarily to a lot of theories,
GG has no predicate structure (but they will be introduced later with the theta
role module dealing with semantics). A lot of theories would analyse :
play(the man, in the garden), more or less.
Here is a strictly syntagmatic GG, but Chomsky realized that it was not good enough.
Consider :
(i) John is easy to please.
(ii) John is eager to please.

You may feel a strong difference, even if the structures lool identical. According
to Chomsky, here the difference remains at a more abstract level (the Deep
Structure), and there must be some operations (transformations) that give what
we sees (the Surface Structure). So in Deep structures we have :
(i)' [e] is easy to please John.
(ii)' John is eager to please.

[e] is an abstract empty position, that can be occupied by something (and that
*must* in english to have a grammatical sentence), either a phrase or an
expletive pronoun. So we can have
(ia) It is easy to please John
(ib) John is easy to please.

but
(iia)* It is eager to please John (agrammatical structure)
(iib)John is eager to please.


The X-bar theory is a module that gives a straightforward (in GG only ;))
representation of Deep and Surface Structures. In (i), the second level N''
(the equivalent of NP) complement of the V rises (not sure of the word) to a
first level N'' position (abstract, isn't it !?)
I would really need to draw syntactic trees to make you understand, but I can't.:((

I'm sorry I cannot carry on, but here are some ideas in GG. I should have spoken
of Universal Grammar, principles and parameters, but I can sup up in post three
years of generative grammar learning ;). Just keep in mind that this theory has
been given up, mainly because it is too powerful (even if the theory was
strongly constrainted, by restricting transformations to movement operations),
and it doesn't deal with semantics (according to Chomsky, syntax is the core of
the grammatical theory), or at least in satisfying way. A good book to begin
with is certainly 'Understanding morpho-syntax' (a black and white book) where
you will learn more about syntax in general : then you might be able to read
some more specific stuff. But do not try to read Chomsky : it is borring, hard
and linguist-oriented. Even Dik will be too hard at this time, but when you
have read a few general books it will certainly be easier for you. According to
me, you should try to read functional theories r!
ather tha



n generative ones : they try to catch the semantics (the meaning) and can offer
you staightforward tools to analyse your language (The classification of States
of Affairs in Dik is just the
best thing I read in syntax at this time).

Julien.

Replies