| "33.6 Though the relative and interrogative pronouns and adverbs beginning with wh are identical in form, it is possible in most cases to tell whether a clause is relative or interrogative. What
is relative in 'I insist on paying what it has cost,' but interrogative
in 'I insist on knowing what it has cost.'" |
|
Either Jespersen
wrote this in a moment of madness, or he possesses some clairvoyant
power that leaves me in the dust. More on this later, but first:
I like Huddleston's term. He calls these
what structures
the fused relative constuction.
Construction, not clause. And he's quite clear that he doesn't like the word
clause for these.
Example:
I paid what it cost.
The paraphrase is
I paid that which it cost.
And only the underlined part is a (relative) clause.
what it cost
is the fused relative construction: it is itself a Noun
Phrase (containing a sentence S) (in the sense given in Transformational Grammar) (I think this is what Clive is calling a noun
clause, by the way), and it features the
antecedent
that fused to the relative pronoun
which to form
what (=
that which).
________
As for the claim that it is usually possible to tell which of these are
relative and which interrogative, even when the words that compose them
are identical ( !!! ), I cannot agree. Of course, unlike
Huddleston, Jespersen calls the whole structure -- including the
what
-- a relative clause (or an interrogative clause, depending on some
ill-defined rules of context). So for Jespersen, one must appeal
to semantics (and I'm
supposing this means the verb) to disambiguate between 'relative'
clauses of this type (fused relative constructions) and 'interrogative'
clauses of this type. If it is the verb that makes the difference
between the relative type and the interrogative type, then Jespersen
thinks that the verb
pay for takes a relative and
know takes an interrogative.
Now note that all the examples given by the original poster have the verb
show. What would Jespersen do here? Would he say that
show
takes a relative and say all examples are of relative clauses (i.e.,
Huddleston's fused relative construction)? Or would he say that
show takes an interrogative and say all examples are of interrogative clauses?
Or -- and this gets very complicated -- would he say that he does not
base his decision on the verb itself at all -- and so some of these are
relative clauses and others are interrogatives? If so, what is
the basis of such decisions? And how would he have decided?
I'm inclined to think that it is impossible to base the decision on
anything syntactic, because the constituent words are identical.
One can only make an arbitrary ruling on which verbs take relatives and
which take interrogatives, and I wonder if there can be any really good
reason for making such arbitrary rulings.
I'm inclined to say that all the original examples are fused relative
constructions AND embedded interrogative clauses. (And that if
there is some complicated way to disambiguate the two, it is probably
not worth the trouble to try!)
CJ