Re: How to Make Chicken Cacciatore (was: phonetics by guesswork)

On Fri, 23 Jul 2004 15:31:43 -0500, Mark P. Line <mark@...> wrote:
> Clearly, [N] is an allophone of /h/.
>
> They're in complementary distribution, right?
In the English I'm familiar with, yes ([N] is only syllable-final, [h]
only syllable-initial).

Whether they'd (jocularly) be considered allophones of /N/ or of /h/
would be a matter of convention.

I believe it's the standard example that being in complementary
distribution is a necessary but not sufficient condition for two
phones being considered allophones of the same phoneme. I think I
first came across it in the sci.lang FAQ, or possibly in the LCK --
something my Mark Rosenfelder, at any rate.

*looks* it appears not. The phrase "complementary distribution" occurs
only once on zompist.com, in the sci.lang FAQ's introduction to
phonemes, http://www.zompist.com/lang21.html#23 . However,
http://groups.google.com/groups?q=%22complementary+distribution%22+ng+h
is interesting. (So I suppose I came across it in alt.usage.english or
sci.lang.)

Cheers,
--
Philip Newton <philip.newton@...>