firstname.lastname@example.org (Richard Kettlewell) said:
> Bill Mitchell writes:
> >> Also, SYSLINUX uses an msdos
> >> diskette, so users can change the kernel image on the diskette before
> >> they even install Linux. With LILO, this is not possible.
> >But LILO diskettes can have several alternate (yes, "alternate"; in
> >the sense of "substitute") kernal images. [...]
> Is this a troll?
> My dictionary lists a number of meanings for ``alternate'' as an
> adjective. Only one of them - which is specifically marked as ``US''
> - has anything to do with substitutes:
> 9. U.S. a person who substitutes for another; stand-in
We've gone around this circle at least once, but I don't think
we've reached consensus. Ian J. says that this usage is bad.
Verious people quoted dictionary definitions for "alternate"
in the sense of substitution. Bruce Perins invites people to work
on Ian's Bug#1063 on this. Raul Miller says that he thought there
had been adequate reverences quoted which support this usage, and
asks if the bug shouldn't just be marked "done".
Apparently we have a case here of of British and American cultures
being separated by a (not so) common language.
I'm reading from my desk dictionary, "The New Websters Dictionary and
Thesaurus", published in 1992. The main dictionary section is (c) 1972
Librarie Larousse as "The Larousse Illistrated International Encyclopedia
and Dictionary", revised and updated in 1992. The second definition of
"alternate" is "n. a substitute."
"Websters Ninth New Collegiate Dictionary", (c) 1989, gives three
alternative definitions for "alternate", the third of which is
"al-ter-nate n (1718) 1: ALTERNATIVE 2: one that substitutes
for or alternates with another". The Explanatory Notes section says:
"Secondary variants belong to standard usage and may be used
according to personal inclination." Regarding the "(1718)" in this
definition, the Explanatory Notes section says: "This is the date
of the earliest recorded use in English, as far as it could be
determined, of the sense which the date preceeds. [...]".
In "The American Heritage Dictionary of the English Language", third
edition (c) 1992, the third variant definition of "alternate" is
"serving or used in place of another; substitute; "an alternate plan".
Interestingly, there's a pointer to a usage note for "alternative".
That usage note reads:
Some traditionalists hold that "alternative" should be used only in
situations where the number of choices involved is exactly two,
becuase of the word's historical relation to tLatin "alter" (the
other of two). H.W. Fowler, among others, has considered this
restriction to be a fetish. [...]
Harry Shaw, in his "Dictionary of Problem Words and Expressions",
says, on the other hand, "[...] 'alternate' and 'alternatly' have a basic
meaining of 'by turns', and 'alternative' and 'alternatively' pertain to
some kind of choice. I don't know if Harry Shaw is much of an authority,
his book just happened to come to hand. The bio notes say that he's
been an editor and teacher and has written a couple of lightweight
sounding reference books.
"The Concise Oxford Dictionary", (the aforementioned H.W. Fowler was
an editor of the original Consice Oxford Disctionar) looks at the
language from a British viewpoint, I think. At least the introductory
material in the Eithth Edition, (c) 1964, 1976, 1982, 1990, seems
to imply this. This dictionary gives several variants for 'alternate',
all having to do with a choice between exactly two things, and
several variants for 'alternative', having to do with choices from
a list of two or more things.
British usage seems to be as represented in the Oxford dictionary,
restricting 'alternate' to a choice between two things. American
usage, on the other hand, seems to be as represented by Websters
and American Heritage dictionaries -- allowing usage of 'alternate'
in the sense of choosing items from a list which might contain more
than two things.
Personally, I don't relish being cast in the role of British English
Usage Grammar Poilceman for man pates (and info text? and usr/doc
documents? and program prompting strings?, and ...) for all of my
packages. As far as I'm concerned, the user gets the upstream
package provider's grammar.