punch not < panc

[Stefan Georg]

>So if the standard etymology of _punch_ can be attacked, it's not on phonetic
>grounds.  Anyone for semantics?

Now that I think the phonetic obstacles have been removed (to a degree, at
least; note that in most modern Indian lgs. /pa^nch/ is specifically noted
with a nasalized /a/ - an artefact of the following nasal consonant, of
course - which fact may have contributed to its reflex in English).
Phonetics apart, the remaining problem seems to be that - in Indian
languages other than English - the meaning is "5" and only that. I fail to
find a trace of any possible usage of this word with the sense of "liquor,
drink" othl. Only thing which comes close is /punch-ghar/, lit.
"punch-house" for "inn". Attestation in English is from roughly the same
time as "punch" itself, so it may or may not be Anglo-Indian rather than
autochthonous (ghar = house is OK, though). In the absence of a clear
Indian native word which combines form and content of modern "punch" (at
least approximately) - which we should look for, preferably in Gujarati and
Marathi, for the earliest (English !) attestations seem to be associated
with this region, or Bengali, where the the earliest British stronghold was
- the "five" etymology, which has become unquestioned standard, may be
reclassified as an educated guess.
However, as it happens some time, maybe *one* Indian cook *once* presented
this kind of edifying drink to *one* colonial officer and, on being asked,
he may have said "I call it /punch/, because I put five things into it".
This can easily give rise to the rumour that "The Indians call it like
that" in some travelogue or casino talk. With a cultural/culinary term like
that, this may well be the whole untold story, and every attempt to find a
similar form-meaning association in one of the potential donor languages
may be in vain. Such things do happen, and there are things between heaven
and earth, etymologists prefer not to dream of ...