"all the faster" (in Latin too)

[Arnold M. Zwicky]

On Jun 27, 2005, at 6:35 PM, Larry Horn wrote:

> ... as to the claim that the non-comparative "all the Xer"
> construction is limited to "bigger" and "faster", I think that's a
> bit too strong a constraint.

that was lise menn's claim.  i certainly accept some other As --
"further"/"farther" for sure, and some others.

> Here are some relevant google hits, all of which sound fine to me.

LONGER

> ... that's all the longer it's ever lasted
>
> ... is that all the longer it's been?
>
> is that all the longer you expect your ride to last?!
>
> Is that all the longer you would last?????

OLDER

> That's all the older she is
>
> ... that's all the older he is going to get.
>
> If that's all the older it is,...
>
> Because that's all the older the earth is...
>
> That's all the older you are, Duck?

BETTER

> Is that all the better you think I can do??
>
> Is that all the better you can climb?
>
> Is that all the better of an argument you can put up?

HIGHER

> That's all the higher it needs to be.
>
> That's all the higher I expect it to play.

TALLER

> ... and that's all the taller he was?

(note preference for zero-relatives over "that"-relatives.  and how
many of these examples are questions.)

> and so on, of course all involving unmarked adjectives denoting the
> positive quantitative scalar element (no, or at most very few, cases
> of "all the lower", "all the younger", "all the shorter" in the
> relevant sense).

i think this is a necessary, but not sufficient, condition for
occurrence in this construction.  that is, the maximum category is
unmarked As denoting the positive quantitative scalar element, and
there might be people who have the construction for all such As.  but
i'm not such a person.  things like "is that all the clearer you can
write?" and "is that all the clearer the weather gets around here?"
and many others are all odd for me.  my guess is that the best As for
this construction are those that are "semantically central" -- if
your language has any As at all, they will denote such properties
(cue reference to dixon)  -- and "everyday" (frequent, not technical
or otherwise registrally/stylistically restricted, etc.).  there's a
nice little research project for someone here, i think.

arnold (zwicky at csli.stanford.edu)