Return of the minimal pairs

[Larry Trask]

Somebody -- I think it was Paul Cohen, but I'm not sure, and my apologies
if not -- wrote this:

>>> And I would judge the Traskian point above about /oi/ never occurring
>>> in native English words, even if true, to be irrelevant to the
>>> discussion:

Let me clarify this.  The point I was trying to make is this: not everything
which is *true* of English is linguistically significant.  It is true that
/oi/ never occurs in native English words, but I cannot see that this is a
linguistically significant statement about English -- in great contrast to the
absence of word-initial [eng] in English, which I firmly believe really *is*
linguistically significant.

Take another case or two.  The rules governing the possible word-initial
consonant clusters in English clearly permit the initial clusters /skl-/
and /gj-/ (second = US /gy-/).  Yet we do not have such words in the
language, apart from the Greek-derived technical term 'sclerosis', the
obscure and French-derived heraldic term 'gules', and the obsolescent and
possibly expressive word 'gewgaw' -- 'obsolescent', because my students
don't know it.  However, I cannot see that the general absence of these
initial clusters is a significant fact about English phonology: it is
merely a historical accident, no more.

And I was arguing that the seeming absence of initial [D] in English
lexical items is likewise a historical accident, and not a significant fact
about English phonology.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk

Tel: (01273)-678693 (from UK); +44-1273-678693 (from abroad)
Fax: (01273)-671320 (from UK); +44-1273-671320 (from abroad)