A Grammar of Spoken English Discourse - The Intonation of Increments

88 A Grammar of Spoken English Discourse

that a Martian scientist observing a child learning English, who has just
learned to produce questions corresponding to the associated declaratives
(examples 4 and 5) might hypothesize that:

the child processes the declarative sentence from its fi rst word (i.e. from

‘left to right’), continuing until he (sic) reaches the fi rst occurrence of

the word ‘is’ (or others like it: ‘may’, ‘will’, etc.); he (sic) then preposes

this occurrence of ‘is’ producing the corresponding question.

(4) the man is tall – is the man tall?

(5) the book is on the table – is the book on the table?

However, Chomsky notes that such a hypothesis is demonstrably false. Were
it correct, the question in (7) and not the question in (6) would be
grammatical. The scientist will realize that a more accurate hypothesis
is that the child analyses the declarative sentence into phrases and locates
the fi rst occurrence of is after the initial noun phrase, and then preposes
this is to form the corresponding question.

(6) the man who is tall is in the room – is the man who is tall in the room

(7)* the man who is tall is in the room – is the man who tall is in the room

Yet, the demonstration that the transformation of declaratives into polar
questions cannot be parsed from left to right has little to say about used
language. The derived ‘question’ in (7) does not have the potential to
satisfy any communicative need, and according to our alternate test for
grammaticality is ungrammatical. As a result, a grammar of used language
does not need to discuss it. But if it had to, an explanation for the ungram-
maticalness of the derived ‘question’ in (7) could go as follows: the N
element man, which has itself realized an intermediate state, anticipates
the following P/N element in the room. This anticipation is ‘interrupted’ by
a suspensive subchain. In the subchain the N element who anticipates the
following V element is. Such an anticipation, however, fails to occur as the
E element tall immediately follows the N element. It is an observed fact of
the language that E elements do not occur between N and V elements and
so the derived question in (7) cannot represent a legitimate purpose-driven
increment.
Brazil (1995: 21), himself, states that Chomsky’s demonstration that a
linear grammar cannot generate all the potential sentences of the language
is, ‘one of the least questioned arguments in the literature of linguistics’ but