Re: The Combos [hj] [hw] and [gw] in Conlangs

Re "gw", Padraic Brown wrote:
>....Of course English has it.

Marginal, all in non-native words AFAICT:  guano, gouache.  Proper names
Guido, Guelph(?), Gwendolyn, Guinivere, Gwathmey, McGuire.  Am I missing
something?

My first reaction to Leo Moser's post was:  Phonemic status, or permitted
phonetic clusters?  Our various responses have shown a mixed bag of both,
with a preponderance of clusters (in terms of the phonologies of the various
langs. cited).

Assuming for the sake of argument that the German ich-laut and Span
{...giV...} are phonetically identical (and approximately "hj", I'll use ç,
c-cedilla here), a Martian might transcribe Germ. "mich" [miç] and Span
"cogiendo" [koçendo] -- but the two sounds have entirely different status in
each language, as further analysis would presumably show.

Engl. /hj/ is very restricted-- only before the tense vowel /u/ or /uw/-- so
that some might even analyze the /j/ as part of the vowel, triphthong /juw/.
That seems to work, especially for British Engl. (/nju:/, /tju:n/ etc.)
though less well for the US; it doesn't apply at all to orthographic "-oo-"
or oddities like "two".  And of course, most of our /hju.../ words are of
French origin-- clearly [ju] was an adaptation of their [y].  ("Hew" is the
only native example I can think of.)   Viewing the [j] as part of the vowel
also explains (historically at least) the /S/ in "sugar, sure".

For that matter, AFAICT, C+j is not a permissible Engl. initial cluster,
except before /uw/.  (again, foreign exceptions like piano, Pierre,
Piatigorsky (sp.?) )

Anyway, obConlang:  None of the clusters in question would be possible in
Kash, not even in fast speech.
Gwr originally had stops /p t k kw, b d g gw/-- but then I discovered that
/w/ could cluster with any consonant, so who needs labio-velars?  /Cwo/ and
/Cwu/ do, however, > /Co:/, /Cu:/.