"soon before"

[RonButters at AOL.COM]

This is beginning to sound like the psychosociolinguistic situation with 
'positive anymore', about which speakers swear they have subtle and complex 
judgments about acceptability which are not always in accord with their actujal 
practice, either as speakers or hearers of contemporary American English. People 
just don't realize that they simply don't know WHAT the rules of their grammars 
are anymore. How soon before this discussion did we discuss positive anymore?

In a message dated 5/19/05 4:29:39 PM, zwicky at CSLI.STANFORD.EDU writes:


> On May 19, 2005, at 12:24 PM, Larry Horn wrote (in reply to Alison
> Murie):
> 
> >> I have to admit that there is a nice economy to  "how soon before
> >> X  was
> >> Y?" vs. "how much earlier was Y than X",
> >
> > well, "how much before X was Y?" is equally economical, and "before"
> > gives you the ordering...
> 
> "how long before" would also do, up to a point.
> 
> first, only up to a point, since "how long/much before" lacks the
> component of shortness that "soon" conveys.
> 
> i've added two more informants who reject unmodified "soon before".
> both accept "soon after".  one accepts "how soon before" without a
> blink (this is pattern B).  the other is dubious about "how soon
> before" (pattern C), but added that she couldn't think of how else to
> say it: "how long/much before" doesn't convey shortness, and "how
> shortly before" struck her (and me) as ungrammatical (i don't know
> why, since "very shortly before" is ok).
> 
> second comment: "much" as a modifier of temporal subordinators (both
> "before" *and* "after') is hugely less frequent than "long".  in
> fact, "how much before/after" is a good deal less frequent than "how
> soon before/after"!  (actually, i find "much before/after" only
> marginally grammatical:
> 
>     "much before" -how: ca. 66,300  ("soon": 31,900; "long": 1,610,000)
>     "how much before": ca. 4,340 ("soon": 20,300; "long": 548,000)
> 
> [the "much" numbers are actually inflated by a lot of examples in
> which "much" is the last word of some constituent preceding the
> temporal phrase.]
> 
>     "much after" -how: ca. 65,900 ("soon": 2,670,000; "long": 1,030,000)
>     "how much after": ca. 4,170 ("soon": 86,400; "long": 123,000)
> 
> 
> >> but, in spite of that, I *think* I
> >> would still tend to avoid it.
> >> "Soon"  seems to me to have a kind of afterness in it, so I have
> >> no problem
> >> with "soon after."
> 
> well, OED2's main definition of "soon" is "within a short time (after
> a particular point of time specified or implied)", which has both
> shortness and afterness in it (and a reference point).
> 
> so it's not surprising that some people have trouble with "soon
> before".  but it's clear that other people allow (in some contexts)
> an extension of the meaning of "soon", without the afterness
> component, but keeping the shortness component.  the question is: in
> what contexts?
> 
> (further note: if "after" isn't involved in this at all, then there
> are only three patterns to consider: A, B, and C on my earlier list.)
> 
> >> It would be constrained by the fixedness of  the
> >> events.  That is, I wouldn't say, "How soon after your party was
> >> the Fourth
> >> of July?"  but could say "How soon after the Fourth was your
> >> party?" or, of
> >> course, ..."will your party be?"
> 
> this is, presumably, a fact about reference points, not about the
> syntax or semantics of "after" (or "before").  ceteris paribus, the
> Fourth of July is a culturally more salient reference point than
> someone's party.
> 
> arnold
> 
>