More on tattoos

[Ross Clark]

[Yoram Meroz writes:]

> As it last stood, the etymology required the Tahitian /tatau/ to become the
> english /tatu/. Two unsatisfactory possibilities stand:
>
> 1. The French spellings of the word follow the same shift, from
> Bougainville's tataou to the later tatoue, first manifested in the 1778
> translation of Cook's journeys (according to the Tresor de la Langue
> Francaise). One can surmise that the English Tattow got transliterated to
> the w-less tatou in French; that the pronunciation followed the spelling
> (like the German taetowieren, /tetovi:rn/, never mind the umlaut), and that
> the French pronunciation was later re-exported to England (and other
> European countries) and phonetically spelled tattoo. This is a complex, but
> plausible scenario, but the spelling 'tattoo' in English occurs as early as
> 1774 (according to the OED), well before the first French occurrence of
> 'tatoue'. So there are problems here, unless someone can suggest a
> different chronology.

I've been thinking about this a little since the previous discussion
too. The only idea I have to add is the possibility that the /au/ >
/u:/ change was the result of influence from that other famous
Polynesian word that entered English at the same time, namely "taboo"
(Polynesian *tapu). Since there's essentially no semantic connection,
one would have to appeal to some kind of association via a common
context of interest in the South Seas, and via actual documents such
as Cook/Hawkesworth.

>
> 2. Several dictionaries (the New Shorter OED, Webster's, Random House)
> claim tattoo comes from Marquesan /tatu/. As it happens, though, tatu does
> not seem to be a Marquesan word with this meaning. I checked Dordillon's
> 1904 dictionary, as well as a few other sources, and they all agree on the
> word _tiki_ meaning both the action of tattooing and its result. _Tatu_ is
> only mentioned as some kind of an illness (with a cognate in Mangarevan).
> POLLEX does not mention _tatu_ meaning tattooing in any Polynesian language
> (Jeff Marck, if you are reading this, does the irregular shift /au/>/u/
> occur anywhere in East Polynesia?) I have no idea, then, where the
> dictionaries I mentioned got the idea for a Marquesan source.

Yikes! Three major dictionaries all citing the same nonexistent word?
What can I believe in?

FWIW, the recently published 1799 Marquesan vocabulary of William
Pascoe Crook has

            tigge       punctures, figures, images

I liked the juxtaposition of "punctures" with the other two, an
unexpected connection of meanings which points straight to tattooing.
But then I began to wonder whether it could be a misreading for
"pictures"?? One would want to check the MS. *tiki in PN usually
means just "picture" or "image".

In any case I can't find anything like tatu or tatau in either Crook
or the other early vocabulary (Robarts). I haven't checked either
writer's ethnographic accounts, though.

Can someone with the latest OED handy check to see whether they have
an actual citation connecting this with the Marquesas? If not,
someone should check what their authority is. I've had some success
in the past in convincing them of errors in their etymologies. And
they are emailable these days....

Ross Clark