Re: Using word generators (was Re: Semitic root word list?)

On Wed, 10 Jan 2007 13:18:03 -0000, Jonathan Knibb
<jonathan_knibb@...> wrote:

>Alex Fink wrote:
>> Then this might lead to the situation where there's some phoneme that
>appears very often among common words and not often among rare ones, or
>vice versa, which is a pretty unnaturalistic situation.
>
>On the contrary, I would guess that this is the norm for such things,
>just as morphological irregularity tends to be lost more easily the less
>frequent the word. The Arabic example fits this pattern (and maybe
>German /tS/ in 'Deutsch' is sufficiently unusual in the rest of the
>lexicon to count too...?). I suppose a single very-high-frequency word
>could support the persistence of the phoneme in more obscure words.
Perhaps I've made too broad a generalization.  Now that you mention it,
isn't there a lect of English in which "gone" has a unique vowel not found
anywhere else?  Still, these examples seem to be confined more or less to
isolated words, and I'd be surprised to find a natlang in which some phoneme
occurred in, say, 20% of common words but only 1% of rare ones, ruling out
borrowing as an explanation.  (Hmm, that's starting to sound like a pretty
weak claim now.)

Although it's not quite the same idea, I do remember reading some source
which suggested that although more common words tend to be shorter they
don't display any bias towards less marked phonemes.  This was, I think,
noted in a discussion of the idea that inflectional morphemes _do_ prefer
unmarked sounds; sorry I don't remember the source more usefully.

>> Well, maybe it's not that unnaturalistic; I guess you could justify
>this kind of variation in phone frequency by saying the less common
>group of words are borrowings from a language with a different
>distribution.
>
>Again, I would argue precisely the converse - I would rather justify the
>appearance of an unusual phoneme in a low-frequency word by (more or
>less synchronic) borrowing. The first example to come to mind would be
>initial /sf/ in English (a cluster, not a phoneme, but the analogy
>holds), which is restricted AFAIK to a small group of low-frequency
>Greek loans. I'm sure there are examples of languages with phonemes
>restricted to loans, of which some probably only occur in unusual and
>even perhaps ad-hoc loans, but I don't know any off the top of my head.
>(Urdu from Arabic? Welsh from English?)
I suppose I wasn't clear -- that's the situation I was trying to allow for,
as an exception.  Introduction of phonemes or new phonotactic possibilities
by borrowing, which thus only occur in loans, is entirely commonplace.
Treatments of English phonology which make /x/ a phoneme have a clear
example there.

Alex