Richard Hudson, On 02/04/2009 15:17:
> Re the first construction "if ... then ...", I have to admit that I've
> never really thought much about it, but I guess there must be some kind
> of dependency between the words, just as there is between "both .... and
> ..." and "either ... or ...". It's an interesting construction because
> "if" depends on the higher verb; e.g. in (2) "if" depends on "ought"
> (2) If you think it's going to rain, then you ought to stay inside.
> But "then" clearly depends on "if" (because you can omit "then" but not
> "if"), so this is a slightly complicated example of mutual dependency.
> (Not unprecedented, incidentally - also found in relative clauses, as
> you'll find in
> http://www.phon.ucl.ac.uk/home/dick/enc/syntax.htm#relative-clause.)
> Thanks for drawing it to my attention!
It's not really a WG-specific issue (tho I guess mutual dependency is pretty specific to WG), but I wd analyse the root of "if X, then Y" as "then", and "then" as having two complements, "if" and "Y". That analysis gives the observed cooccurrences and accounts for the word-order. (If WG had specifiers, one might say that "if" is the specifier, "then" is the head, and "Y" the complement.)
--And.
|