Christopher Vance wrote:
>
> "When the (parent) process has _vfork_'ed a child process, the parent
> is typically suspended until the shared address space is released by
> the child when it _exit_'s or (more usually) _exec_'s a new program."
>
> Some of these uses can be avoided by circumlocution, but the language
> gets very stilted if that is always required. Font differences can
> sometimes be adequate to avoid the separator, but not always. You may
> find some people using a hyphen instead of the apostrophe, but some
> sort of separator is almost required for legibility. In this context,
> plural "s" is only one of the affixes of interest.
Exactly. This is a case where usefulness in practice is gradually
overwhelming the "Thou Shalt Not" prescriptivists.
Linguistically what is going on is that the users of this convention
are adding to the writing system a convention for indicating a
zero-stem-formation process that takes an alien lexical item
(or a non-standard lexical representation of a native lexical
item, as in the 70's, 80's, 90's cases) and
marks it as accepting regular morphological affixes for tense,
plurality, nominalization (_vfork_'ing), and so on. The communicative
usefulness is in marking off the derived stem, as if it were
bracketed, so that "A's" is recognized as [{A}+plural suffix] and
not as "As".
I guarantee you, the Oakland Athletics would be very upset if someone
tried to take away their apostrophe:
http://oakland.athetics.mlb.com/NASApp/mlb/oak/homepage/oak_homepage.jsp
For that matter, the users of Myers-Briggs Type Indicators might also
get upset if they had to give up their S's, N's, T's, and J's. ;-)
--Ken
