as ADJ of a N as

[Arnold M. Zwicky]

On Feb 18, 2004, at 10:13 AM, Lynne Murphy wrote:

> ...A colleague here read an American student's phrasing "As old of a
> joke as
> this is..." and queried whether the rest of us could say such a thing.
> (Almost) needless to say, I could, but my British colleagues couldn't.

"as" is an "exceptional degree marker", in my terms:

Zwicky, Arnold M.  1995. Exceptional degree markers: A puzzle in
external and internal syntax. OSU WPL 47.111-23.

ordinary degree markers, like "very" and Adj-ly, combine with an Adj
and a bare (determinerless) NP.  exceptional degree markers, like "as"
and "that", combine with Adj and a NP with determiner "a(n)" (and so
can be used, for most speakers, only with singular count Ns; in the
nonstandard (but *very* widespread) american construction, they combine
with Adj and a PP consisting of "of" plus such a NP.

> ...2.  How would you parse the phrase---is it [[As Adj] [of NP}] or
> [As [Adj
> [of NP]]?

if the parsing of "very Adj N" -- e.g. "very big dog" -- is [[very Adj]
N], then the parsing of "as Adj a(n) N" -- e.g. "as big a dog" -- would
presumably be [[as Adj] [a(n) N]], and the parsing of "as Adj of a(n)
N" -- e.g. "as big of a dog" -- [[as Adj]] [of [a(n) N]]].

but there is a move, very clear in abney's dissertation and radford's
article in the Heads volume edited by corbett et al., to treat
modifiers as heads of their phrases (Adj as head in Adj + N, Adv as
head in Adv + Adj, somewhat parallel to D as head in D + N) and to
mirror this claim in the structure of nominals: [a [very [big dog]]],
[as [big [a dog]] (or [a [as [big dog]], with the "a" moved down in the
structure), [as [big [of [a dog]]].  in fact, abney and radford mention
the nonstandard american construction as evidence for the exceptional
degree marker as head in Adv + Adj -- since the Adv and not the Adj
determines how the rest of the phrase is composed.

> 3.  Would you agree with the 'epenthetic' of analysis?  Can you think
> of
> other cases of epenthetic 'of'?

Ns have plain clausal objects -- "the idea that pigs could fly" -- but
NP objects marked with "of" -- "the idea of pigs flying".  this could
be seen as "of" epenthesis, or as deletion of "of" before clausal
objects, or, of course, simply as a difference in marking, with no
insertions or deletions (which is what i'd prefer).

similarly, some determiners require "of" ("a lot of books"/*"a lot
books" -- cf. "a lot bigger"), some don't allow it (*"a dozen of
books"/"a dozen books"), and some show variation ("a couple of
books"/"a couple books").  this could be treated as insertion (or
deletion), but why?  (one answer: some theories require such a
treatment.)

> 4.  Is there anything published on this construction?

see above, and references therein.  there's also a fairly recent
relevant article (with a pretty extensive bibliography):

Seppa"nen, Aimo;  Solveig Granath; & Lars Danielsson. 2002. The
construction ‘AdjP — a(n) — Noun’ in present-day English
syntax. Leuvense Bijdragen 91.97-136.

(yes, about as obscure as my paper!)  and a lengthy discussion on the
LINGUIST list from a few years ago, which might have some further
sources.

there's variation as to which degree words are exceptional and in what
contexts.  for many american speakers, the nonstandard "of" variant of
the exceptional markers seems to be essentially categorical, but i
think there are speakers who have the "of" variant as informal and
spoken, with the standard variant as formal and written.  (undoubtedly
variation within individuals would be good topic for research.  if
someone has pursued this topic, i'd like to hear about it.)

maybe this is more than you wanted to hear...

arnold (zwicky at csli.stanford.edu)