Functional Grammar

[Thomas Bloor]

Dear Juno

Big question. I know I'm going to regret this. Very briefly and
simplifyiing grossly:

In early Standard Theory TG, sentences are derived in two stages: (i)  by
base rules plus lexis which give the deep structures and (ii) by
transformations operating on these to give the surface structures. All
subordinate clauses (including non-finite clauses) are derived by the same
set of base rules as main clauses and involve embedding by recursion in the
base. Thus the deep structure of a complex sentence prior to the operation
of transformations would involve a number of passes through the base rules
to give a hierarchically organised tree (a  bracketed string) starting with
S(entence) and involving other instances of S, each occurring as, for
example, part of a Noun Phrase or some other unit.  Each S in the deep
structure is likely to have the characteristics of an independent S
(subject first, no ellipsis, etc.) but various transformation rules operate
to modify them (moving relative pronouns to the front of the clause,
raising elements from one clause to another, moving extraposed clauses,
etc.) to give the surface structure.
[see Burt M K 1971 From Deep to Surface Structure (Harper Row);  Huddleston
R 1976 An Intro to English Transformational Syntax (Longman); Akmajian A &
Heny F 1975 An Intro to the Principles of Transformational Syntax MIT
Press]

X-bar theory simplified a lot of this paraphernalia with a much more
general system for all embedding and modification so that sentences,
(clauses) and phrases all conform to the same principles of heads and
complements, and various constraints on movement, etc, largely remove the
need for transformations.
[see Radford 1981 Transformational Syntax CUP;Radford 1988 Transformational
Grammar; Ouhalla J 1994 Introduction to Transformational Grammar (Arnold)]

In SFG, one main difference from TG, and from  most traditional  grammar,
is that a  distinction is made between embedded  and  dependent clauses. An
embedded clause is one which is part of (or replaces) an item at a lower
rank (eg a group) or the same rank (clause). Thus an embedded clause is
rank-shifted because, although a clause, it functions as something lower,
eg Postmodifier in a nominal group; or Subject in a clause.  A dependent
clause is one which is hypotactically bound to a dominant clause in a
clause complex; ie it is subordinate in status to the dominant clause but
not part of it. Dependency and dominance can operate recursively, eg a
clause Beta is dependent with respect to its dominant Alpha, but dominant
with respect to its dependent Gamma. Examples of embedded clauses are:
defining (i.e. restrictive) relatives (e.g. 'The answer THAT HE GAVE was
not helpful), clause as Subject (e.g. 'THAT THIS IS CONFUSING  is obvious';
'it is obvious THAT THIS IS CONFUSING'). Examples of dependent clauses are:
traditional adverb clauses, non-defining (ie non-restrictive) relatives,
reported speech (eg 'Explain this IF YOU CAN'; 'he tried to explain it,
WHICH WAS NICE;  he said THAT IT WAS AS CLEAR AS MUD'. (For a clearer
account, see Halliday 1994 Intro to Functional Grammar (Arnold) Ch 6 & 7;
Bloor T & Bloor M 1995 The Functional Analysis of English (Arnold) Ch 7, 8,
9 & 10]

One way in which most traditional grammar differs from both TG and SFG is
that it does not allow for non-finite clauses, which it usually classes as
'phrases'. SFG treats non-finite clauses as just that, clauses without a
Finite, and old TG derives them by transformations from finite clauses in
the deep structure (though there was a lot of dispute in this area as
everywhere).  Of course, the label 'traditional grammar' covers an enormous
range of different approaches so most generalisations about trad grammar
have to be treated with caution and the same is true to a smaller extent of
TG and to an even smaller degree of SFG.  Also, central to the issue is the
fact that the long-term aims and interests of linguists who use these
models are different.

I knew I would be sorry for starting this. It is a big issue and cannot be
answered simply; you need to do a lot of reading if you are to include it
in your thesis.

Good luck

Tom Bloor



>Dear all,
>
>I am undergraduate student at Stanford University who is writing an honors
>thesis on subordinate clauses in narratives found in EFL textbooks vs. ESL
>textbooks. I have been coding my texts, based upon the methods of
>functional grammar, but just recently realized that functional grammar
>and the traditional generative, transformational grammar are two entirely
>different approaches to syntax. I am thinking of including a brief
>discussion of this in my thesis. Could anybody clarify for me what exactly
>are the differences between the two types of analyses? I appreciate your
>help.
>
>Sincerely,
>Juno Nakamura
>Junior, Linguistics
>Stanford University
>(650)497-6877

Thomas Bloor
Language Studies Unit
Aston University
Birmingham, UK
B4 7ET

Phone:0121 359 3611 xt 4212/4236
Fax:0121 359 2725