Date: Mon, 18 May 1998 10:11:45 -0400
From: Larry Horn
Subject: Re: "Is it raining out?"

At 10:33 AM -0500 5/16/98, Gerald Cohen wrote:
> A 5/14/98 ads-l message asked:
>>Is it due to regional differences that when asking about the weather some
>>say,"is it raining out?" and others simply, "is it raining?"? why include
>>the "out"? it's not raining in.
> ----The "out" here possibly originated in sentences of the type "Is it
>nice out?" (where "out" = "outside" is justifiably used). Possible
>answer (with transference of "out" from the preceding question): "No, it's
>raining out." (where "out" is unnecessary). Then the latter sentence could
>be turned into a question: "Is it raining out?"
> In the arcana of general-linguistic theory is a "Law of Least Effort,"
>according to which speakers use only the minimal effort necessary to get
>their point across. This so-called law is contradicted by "out" in "Is it
>raining out?"
>--Gerald Cohen
As any study of the workings of Zipf's Law (the Principle of Least Effort
in its linguistic applications) is careful to observe--including those of
H. Paul, G. K. Zipf, A. Martinet, J. Haiman, and I like to think my own
work in several papers and a book I've written since 1978--this principle
does not operate in a vacuum and indeed is counterbalanced by other
important functional principles, including iconicity (see Haiman),
expressivity (see Henri Frei [1928] on La Grammaire des Fautes), a
sufficiency principle establishing a lower bound on informative content
(see Horn), and the conversational desiderata relating to information
packaging, which I think are relevant here. As Jerry suggests, the normal
use of "It's raining out" is as a response to a question, although the
neutral query here would have been not "Is is nice out?" but more likely
"What's it like out?" The "out" in the response is destressed; the
variable for which a value is supplied is not "It's raining/sunny in
[location] X" but with "It's X out."