Taboo replacements

[Jim Rader]

A few comments on the discussion below, though we are drifting rather
far from Indo-European:  as a source for <anaconda>, the Tamil
compound translated as "elephant killer" is cited in Yule and
Burnell's  _Hobson-Jobson_, but retracted more or less in a
parenthetical note in later editions in favor of a Sinhalese word
<henekandaya:>.  As far as I know, no one has ever seriously disputed
the <henekandaya:> etymology.  The Sinhalese word was borrowed into
English, not Portuguese, and the misapplication to a South American
snake was due to confusion among biologists.  The Brazilian
Portuguese word for the anaconda (Eunectes murinus) is <sucuri>, with
many variants, borrowed from Tupi.  If <anaconda> exists in current
Portuguese, it is as a borrowing of an international zoological term.

I don't think Robert poses a really valid objection to Larry's
points.

Jim Rader

> LARRY TRASK:
> >> 	And zebra is supposedly from Old Spanish ecebra, ezevra, etc. said
> >> to be a derivative from equus plus some ending.
>
> >This is one story.  The four dictionaries in my office give five
> >different stories, though all agree that Spanish or Portuguese is the
> >direct source.  The stories are:
>
> >(1) of unknown origin;
> >(2) from <Zephyrus>, the wind god, because of the animal's speed;
> >(3) from some Italian development of a Latin *<equiferus> `wild horse;
> >(4) from an unspecified Congolese language;
> >(5) from Amharic <zebra:> `zebra'.
>
> >You pays your money...
>
> >(1) is undiscussable.
> >(2) looks fanciful to me.
> >(3) seems to have a phonological problem, unless there were Italian
> >	dialects in which /kw/ was reduced to /k/ very early.
> >(4) is hard to evaluate without specifics, but why a Congolese
> >	language?  Zebras are found on the eastern savannahs, not in
> >	the Congolese rain forests.
> >(5) looks good, if the word is real, but how would an Amharic word
> >	get into Spanish and Portuguese?  (Of course, the Portuguese
> >	were all over east Africa early on, but in Ethiopia?)

ROBERT ORR:
>         when you consider what you have to assume to get Tamil yaanai-kolra
> - "elephant killer" 1) borrowed into Portuguese 2) transferred all the way
> to South America 3) become establsihed enough in the language to refer to
> another giant snake, i.e.,  "anaconda" (the exact path taken by stages 2 and
> 3 is open to debate), (5) above doesn't really look like a problem at all.