'
...
*teutÄ 'people'
1. In 1996 I have discussed a few words occurring in several IE
languages which were probable loanwords from a substratum language.
This conclusion was based on formal characteristics which rendered IE origin improbable. Now many such loanwords may have shown no formal characteristics that caused difficulties for IE languages; or else, substratum words may have been so well adapted that their foreign origin cannot be seen anymore; I think that *teutÄ is such a word.
2. The forms are well known (Pok. 1084) so that I give only a representative of each group: Goth. þiuda, OIr. túath, Lith. tautà (probably a loanword from Germanic is OCS *(s/Å¡)tuždÑ, Russ. Äužoj 'foreigner'), Osc. touto, and names Illyr. Teutana, Thrac. Tautomedes, Mac.(?) ΤεÏÏÎ±Î¼Î¿Ï (a Macedonian general), Horn. ÏεÏ
ÏÎ±Î¼Î¯Î´Î·Ï (son of *Teutamos) etc. (The father and the son of Bias of Priene were called ÏεÏ
ÏάμοÏ.) Note that these names do not imply that Greek knew the noun. The names are generally regarded as Illyrian. The Homeric *ÏεÏÏÎ±Î¼Î¿Ï is a Pelasgian. (The Pelasgians probably did not spreek Greek or Indo-European, but they may have taken over names from other languages. Note that -αμ- is considered a non-IE suffix.)
> Two further forms are problematic. (Cf. the survey by Polomé in the Encycl. s.v. people.)
3. Hitt. tuzzi 'Heer, Heerlager' is now considered by most scholars unrelated; see Tischler's etymological dictionary.
The other form is NP toda 'heap, stack, hillock' (also found in Sogdian). Watkins objected (1966, 46 n. 39) that the word had no sociological meaning and is therefore irrelevant. Szemerényi 1977, 100 - 108 argued that the Persian word had retained the original, concrete meaning. This would mean that the sociological meaning had not developed in Indo-Iranian. It seems to me improbable that this word continues an old IE word of which there is no trace in old Indo-Iranian or Indic.
4. The Baltic -au- (Latv. tauta/ tà uta, OPr. tauto; we expect iau, OPr. eu) presents a problem, though a problem that is found in other words too. Endzelins' solution (see Stang 1966, 73f) is not attractive (i.e. iau only before front vowel). (Even more improbable is Schmalstieg, in Endzelin 1971, 35: eu > au regularly, iau only from recent eu.) I wonder whether the Baltic word is a loanword (just like Slavic tuždÑ foreigner'), but then from a language that had eu > ou. The nearest such form seems Thrac. Tautomedes. This form is itself also not unproblematic, as Thracian seems to preserve eu. It reminds me of *tauros (no doubt a loanword, hardly from Semitic), beside which we have *(s)teuros in Germanic. This might point to an interchange eu/au. One might also invoke secondary ablaut, for which cf. Kuiper 1995, 7If.
5. It has been proposed that the word is cognate with Lat. tÅtus. This gives a problem with the -Å- (one expects -Å«- from -eu-), so it would have to be a dialectal form. Benveniste (1969 1, 366) considers the possibility that the adjective was derived from the word for 'people'. The other way round seems much more probable, as was proposed by Meid 1965, 293, who assumes a basic meaning 'Ganzheit'. As this etymology of tÅtus is uncertain, as the adjective is found only in Latin, and as there is an alternative etymology (from *teuH- > *toua-, Pok. 1080), I think we should give up the connection, as did most scholars.
6. As far as I see it is generally accepted that *teutÄ is derived from the root *teuH- 'to swell'; e.g. Szemerenyi 1977, 107. Pokorny (1080 - 1085) gives an enormous collection of heterogeneous words and most dubious connections under this root. Most forms are supposed to derive from enlarged forms of the shape *t(e)uC-; these forms are not relevant here. (Note that the forms in -r, -l, -m, -n cannot be roots, unless one starts from shapes like *tuel-, as Pokorny does in a very few, and most dubious instances.) For a sequence *teut- the only evidence is *teutÄ, so that is of no help for us. The 'unenlarged' form in Pokorny's presentation is that seen in Skt. tavÄ«ti etc. I agree with Szemerényi that this is the only form we can use in this discussion. However, this form has a final laryngeal. There is no certain evidence for a form without a laryngeal. (Of course, in earlier days one was very permissive in assuming variant anit.-forms.) The laryngeal is no problem for Germanic, but in Baltic it would have given an acute (the root is circumflex, taut-) and in Italic and Celtic would have given *teua-. This form is contradicted by the Gaulish names with Teut-, Tout- Therefore this etymology must be given up.
Also, the etymology did not give a satisfactory meaning. Meid (1965, 293) frankly stated: "Der eigentliche Sinn von *teutÄ ist unklar;" (and then proposed 'Ganzheit', see section 5). Benveniste (1966 1, 366) stated that távas- etc. means 'strength', and that therefore ("donc") the basic meaning was 'plenitude'; the logic escapes me. Szemerényi is straightforward in assuming a basic meaning 'power', but this is not a very probable starting point for 'people'. De Vries (1962, 613a) objected" "Ein Volk ist doch nicht nur etwas kräftiges." Note also that Szemerényi's interpretation contradicts his own view that NP toda retained the original meaning, so that one should start from 'heap' or the like. (Szemerenyi's comparison with OP taumÄ 'race, family' is not to the point, as the basic meaning here is quite different, cf. Skt. tókman- 'Schossling', Av. taoxman- 'Same, Keim'.) - Note that Pokorny apparently started from a root *teut-.
I conclude that the connection with the root *teuH- is semantically problematic and formally impossible. This means that *teutÄ is isolated, unmotivated.
7. There is no (other) term of PIE date that might have indicated a larger social group than *ueik´-. Thus Avestan has four groups: dam-, dÉmÄna-; vÄ«s-; zantu-; dahyu- (see e.g. Szemerényi 1977, 100), but the last two terms are not used in the same way even in Indic. The only exception may have been a word for the people as a military force. The most probable is *koro- (OP kÄra-, with Brugmann's law, not a vrddhi-derivation), *korio- (Goth. harjis). (Mallory 1989, 124 gives *teutÄ in this function, but I see no evidence for a military association of this word. See now the Encycl. s.v. people.)
> Words for 'people' and the like are often loanwords, e.g. Lat. populus, vulgus, folk, `ÎθνοÏ.
8. For the origin of the word I see the following possibilities.
1) The word is an old PIE word, which was lost in a number of languages. This is possible. It is most probable that Latin lost the word. The development eu > ou seen in Osco-Umbrian is of Proto-Italic date, so that Latin must have had the word. (Earlier, there was the idea that Osco-Umbrian was a group that came only later in the neighbourhood of Latin.) In the same way, if the word was inherited in Baltic, it must have been present in Slavic. But the distribution of the word makes it more probable that, in some way or other, it is an innovation of the western languages. Porzig 1954, 200 speaks of "Neuerungen". Benveniste 1969 1, 366 is also vague when he says that the later Indo-Iranians, Latins(!) and Greeks had left the community "avant que prévalut le terme *teutÄ" in the later western languages. Mallory (1989, 124f) writes: "To what extent... these terms can be extrapolated back into Proto-Indo-European society ... is debatable." In my opinion the fact that the word is found in a continuous group of IE languages in the West points not to a retained archaism, but to some kind of innovation.
2) Perhaps the word was of PIE date, but with a different meaning, and the western languages innovated in developing the meaning 'people'. Again I think that it is improbable that all these languages knew the same development.
3) An innovation in the sense that these languages created the word, i.e. from inherited, IE material, is improbable because the word is isolated, unmotivated. It is possible that it was made from material unknown to us, but this is not probable. Also there are few if any innovations that these languages made together.
4) The possibility that remains is that these languages innovated by adopting a loanword, somewhere in eastern or central Europe.
Above, section 4, I speculated about the possibility that the word had a variant with au beside en, like (perhaps) the word for 'bull'. This would show a non-IE vowel change. But this is very speculative.
Above (§ 2) I pointed to the non-IE suffix -am-.
Thus I think that, though we have no hard, formal evidence for non-IE origin, the conclusion that the word is a loanword from a substratum or adstratum language, is the most probable solution. The considerations were:
1. The word is unmotivated in IE;
2. the word is found in a limited area;
3) the word is found in a continuous area which often shows non-IE loans;
4. there was probably no word for this notion in PIE,
5. words for this concept are often loanwords.
That a loanword appears in Germanic and in Celtic is unproblematic. That such words appears in Baltic and Germanic is seen frequently. Nor is it a problem that Italic participates (as e.g. caput, Beekes 1996.) How we have to imagine the process is another matter. One way is to assume that the word was taken over, when the relevant groups still lived in South Russia, from the Tripolje culture. The word would then not have spread to the eastern part of the IE world. The advantage of this suggestion is that it would explain how the word came to be adopted by almost all western languages.
However, another solution is indicated by Krahe. Ik 1954, 66 he writes: "Eine Verbreitungskarte [der von *teutÄ aus gebildeten Eigennamen] deckt sich nahezu vollständig mit einer solchen der alteuropäische Gewässernamen, so dass an einem sprachgeschichtlich-ethnischen Zusammenhang beider Kategorien kein Zweifel möglich ist." This means that *teutÄ probably belonged to the language of the Old European river names. For Krahe this was Indo-European. We now know that this language was non-IE.'
My own view of what *teuta: is:
*tuN-t-a:
cf
http://tech.groups.yahoo.com/group/cybalist/message/65461
Later extended in the thread starting in:
http://tech.groups.yahoo.com/group/cybalist/message/65502
note esp.
http://tech.groups.yahoo.com/group/cybalist/message/65523
http://tech.groups.yahoo.com/group/cybalist/message/65528
http://tech.groups.yahoo.com/group/cybalist/message/66841
thus foreign to IE that too.
Torsten