minimal pairs (was: PIE e/o Ablaut)

[Robert Whiting]

On Thu, 6 Apr 2000, petegray wrote:

> Hence my point that minimal pairs are not a sufficient criterion - we
> also actually make decisions on the basis of a theoretical structure
> into which potential phonemes fit.

For a while I was beginning to think that I was the only one who
believed this.  Analysis does not stop with a minimal pair.  Minimal
pairs is neither a necessary nor a sufficient condition for establishing
phonemicity.  Minimal pairs is just a useful heuristic for locating
probable phonemes.  But it is useful because it is usually accurate.
A minimal pair that does not indicate phonemicity has to have a good
explanation.  And there are doubtless cases where minimal pairs still
exist in parts of the language even though phonemicity of the sounds is
being lost through phonemic merger.

Bob Whiting
whiting at cc.helsinki.fi