cat (was: Not "catching the wind " ...)

From: dgkilday57
Message: 61813
Date: 2008-11-24

--- In cybalist@yahoogroups.com, "dgkilday57" <dgkilday57@...> wrote:
>
> --- In cybalist@yahoogroups.com, "tgpedersen" <tgpedersen@> wrote:
> >
> > [...]
> >
> > Ernout-Meillet:
> > 'cattus, -i: m., et catta, -ae f. (doublet gattus, gatta):
> > chat, chatte.
> > Attesté avec ce sens depuis Palladius (le terme ancien est
fe:le:s);
> > bien représenté dans les l. romanes;
>
> Palladius has the first attestation of <cattus> 'tomcat', but
<catta>
> as an epicene goes back to Baruch 6:21 in the Vulgate, translating
> <ailouros> of verse 21 of the Epistle of Jeremiah in the LXX (for
> some reason this is appended to Baruch in the Vulgate, which has
only
> 5 chapters in the LXX). The Latin Baruch is one of the books
which
> predates Jerome and was not edited by him, so its composition may
go
> back to 250 CE or so.
>
> > ital. gatto,-a, esp. gato,-a; fr. chat, chatte, M.L.1770.
> > Sur cattus ... quod cattat, i.e. uidet
> > dans Isid.12,2, 38., v. Sofer, p.62.-
> > Dans Martial, 13, 69,1, Pannonicas nobis numquam dedit Vmbria
> cattas,
> > le mot semble désigner un oiseau, peut-être le hoche-queue,
> aílouros,
> > cf. gattula "attagé:n" Orib.
> > La substitution de cattus à fe:le:s doit correspondre à
> l'introduction
> > à Rome du chat domestique, sans doute importé d'ailleurs.
>
> A cat named Krankru is depicted resting on a table or couch in the
> tomb of the Etruscan family Leinie, ca. 300 BCE, at Sette Camini
(TLE
> 235 = CIE 5095).
>
> > Dérivés: cattin(e)us, tardif (= fe:li:nus);
> > catto:,-as, cf. sans doute esp. catar.
> > Le celtique a irl. catt, gall. cath reposent sur *kattos, qui
figure
> > en gaulois comme nom propre Cattos;
> > l'emprunt du mot au latin, admis par M.Pedersen, est donc peu
> > vraisemblable.
> > Le vieux haut allemand a kazza, le v.norr. ko,ttr, le lituanien
> > kâte.~, le slave kotUka. Mais ces mots peuvent provenir, comme
le
> mot
> > latin, d'une langue inconnue. Le "chat" domestique ne s'est
répandu
> > que tardivement dans le monde romain; l'origine en est discutée
> > (Afrique?).'
> >
> > So, if Celtic *kattos is unlikely to have been borrowed from
Latin,
> > but must come from some other language, why must Germanic *kattu-
be
> > borrowed from Latin? What other examples are there of a Latin
> thematic
> > stem being borrowed into Germanic as a u-stem (cf. the -U- in the
> > Slavic word)?
>
> With this particular word <cattus>, borrowing into early Gmc. as a
u-
> stem is not a problem, because many of the Latin sentences
probably
> involved the nom. sg., unlike words for inanimate objects. It is
not
> hard to imagine a Roman merchant showing off his domestic cat and
> saying "hic est meus cattus; meus cattus mures captat et devorat;
> meus cattus arbores noctu scandit ad aves capiendas", etc. It
would
> have been natural for the nom. sg. <cattus> to have been borrowed
as
> a u-stem. Gothic retains <handus> and the like uncontracted, if
> memory serves.
>
> The cat itself might come from Africa, but looking for the word
there
> is a dead end. None of the three Coptic words for 'cat',
including
> the one used in the Coptic version of the Epistle of Jeremiah,
> resembles <catta> at all. My best guess is that Latin <catta> in
> both its senses, 'certain Pannonian bird (the wagtail?)'
> and 'domestic cat', is borrowed from Messapic (or related Illyro-
> Japygian) *katta, a hypocoristic form of a compound whose first
> element *katt- is cognate with Lat. <quassus> 'shaken', and whose
> second (unidentifiable) element means 'tail' or 'rear end', the
> compound meaning 'having a shaken tail' or 'shaking its tail',
much
> like Greek <ailouros> 'waving-tailed'. It is near my bedtime, so
> details of this hypothesis will follow later.

Sometimes I feel like (yawn) Rip van Winkle ...

Messapic was dead or at least moribund by Ennius's time, so it is
unlikely to be the source of Latin <catta> in either sense.
Probably the word in both its senses came from Pannonian. This
language does not appear to belong to the Illyro-Lusitanian group,
in which PIE *H1ek^wo- 'horse' is reflected as *(h)ikko- or *(h)ippo-
. The Pannonian personal names <Ecco> and <Eppo> have been
plausibly ascribed to *H1ek^wo- by Krahe, Pokorny, and others.
These are probably hypocoristics of compound names, showing
hypocoristic gemination, derived from Q-Pannonian *eko- and P-
Pannonian *epo- 'horse'.

<Aquincum> 'Budapest' suggests that Q-Pannonian maintained */kw/.
This place-name is formed like Gaulish *abinko- 'swampy land' etc.
(whence Catalan-South French <avenc>, J. Hubschmid, _Praeromanica_
53-6) but with *akw- rather than Celtic *ab-. However, the Latin -
qu- could simply reflect the substitution of <aqui-> from <aqua> for
the first part of the place-name by Latin-speakers, as we find in
<aquipenser>, one of the variants of <accipenser> 'sturgeon'. The
true Pannonian name of Budapest could well be *Akinkom, with */k/
from */kw/. In that case <catta> can represent a Q-Pannonian *katta
(:), hypocoristic of a compound parallel to Greek <aílouros>, French
<hochequeue>, and English <wagtail>, whose second element 'tail' is
undetermined, but whose first element 'shake, wave, wag' is cognate
with Latin <quatio:> 'I shake', probably *kati- in Q-Pannonian. The
gemination in *katta(:) is then hypocoristic, not due to Pannonian
having -tt- in participles where Latin has -ss-.

Pat Ryan wondered how Arabic <qit.t.-> 'cat' could be explained as a
loanword. The consonants present no problem, since Arabic has /q/
and /t./ from Latin-Romance /k/ and /t/ in several known loanwords,
and Aramaic also has /q/ and /t./ from Greek and Latin /k/ and /t/.
The weakening of /a/ to /i/ is also found in Arabic <difn->, <difl-
> 'laurel', obviously from Koiné Greek <dáphne:> (strictly an Ionic
form, but Attic prose has it due to its prestige in the Epic
literature, like <Odusseús> for pure Attic <Olutteús>). Spanish-
Portuguese <adelfa> 'oleander' (very similar to the laurel, hence
Grk. <rhododáphne:>, Eng. <rosebay> 'id.') comes from Moorish Arabic
<ad-difla:>, showing the antiquity of the weakening. Masoretic
vowel-pointing indicates a similar weakening of /a/ to /i/ in closed
initial syllables of dissyllabic words in Late Hebrew, e.g. Heb.
<Miryam> (Ex. 15:20 etc.), also LXX <Miriám>, but NT Greek <Mariám>
from Palestinian Aramaic (Luke 1:27 etc.); Heb. <gilgal> 'wheel',
but Aram. <galgal> (a reduplicated word). Similar evidence points
to a parallel weakening, not necessarily simultaneous, in
Phoenician. Thus it is plausible that the vowel-weakening observed
in Arab. <difn->, <difl->, and <qit.t.-> occurred not in Arabic
itself, but in a NW Semitic dialect which passed the words from
Greek and Latin on to Arabic. (Moroccan Arabic has forms of 'cat'
with /a/ which might have come directly from Latin, or perhaps
indirectly by way of North African Gothic).

Rick McCallister recalled seeing a Nubian <qadis> vel sim. cited.
T. Bolelli (_L'Italia Diallettale_ 18:179) cites Nubian <kadi:s> and
Berber <kaddîska> in support of the theory that the word 'cat' came
from Africa along with the animal. The earlier objection still
stands, however. Domestic cats would hardly be acquired by
Europeans directly from Nubian or Berber tribesmen. If native
African cats were indeed desired by wealthy Europeans as prestigious
pets, they would pass through the hands of merchants in Alexandria,
where cats had been familiar for centuries, and those merchants
would use one of the Coptic words. I consider it much more
plausible that cats were popular pets among Pannonian children in
the 3rd c. CE, going by their hypocoristic name, and that
legionaries brought the word <catta> back to Italy, perhaps along
with a cat-keeping fad, where it quickly started to displace the
established word <fe:le:s>. Roman merchants then spread <cattus>
into Celtic and Germanic languages, whence it reached Slavic lands
and eventually the Caucasus.

Douglas G. Kilday