01 April 2010

Words for 'sea'

There's no universal word for 'sea' in IE languages. The most widespread (Celtic, Latin, Germanic, Balto-Slavic) one is *mori-, related to Altaic *mju:ri 'water'.

Another widespread root is *seH2l- 'salt, sea'1 > Greek halê 'sea'2, which I think ultimately comes from PNC *q’eɦlV (˜ -ɫ-) 'bitter'3, with an evolution PNC *q’- > pre-PIE *χ- > PIE *s- (by Fournet's Law). This Vasco-Caucasian root is reflected in Basque gatz 'salt', gazi 'salty' + PNC *ts’s’wenhV 'salt'4 > *gas-dane > gazta(i), gazna 'cheese', both Cantabrian loanwords.

Dolgopolsky proposes a Paleo-Eurasian root *dalqV 'wave' (ND 526) reflected in IE *dhelH- > Greek thálassa (Attic thálatta) 'sea'. In a IE substrate language (probably Italoid), this root gives Greek sálos 'turbulent movement of the sea, flushing of the waves', Latin salum 'open sea; sea waves'5.
_____________
1 Without any linguistic evidence, Arnaud Fournet relates this to Kartvelian *zoɣw-. See his document.
2 A feminine word whose masculine counterpart is halós 'salt'.
3 The glottalized stop is retained (with secondary labialization) in Kartvelian *q’wel- 'cheese'.
4 Native outputs from this root are itsaso 'sea', itze 'sea' (an archaic form quoted by Trombetti) + urde 'pig' > izurde, gizaurde 'dolphin' (lit. 'sea-pig').
5 Which again Fournet unjustifiedly relates to IE *seH2l-.

11 comments:

  1. "Germanic has *sajwa- 'sea, lake', related to Kartvelian *zoɣw- 'sea'. "

    Also compared with Urartian /tsue/ > Armenian tsov 'sea'.

    ReplyDelete
  2. I have read that the word "sea" 'sæ' I West-Saxon OE, 'sēo' in OHG, & with many forms of the same etymon in mod. Icelandic. The most common being (mas) 'sjór) nom. sing. But also w/ the older, or by-form, 'sær'.
    sjór declined

    Only in Singular
    Nom sjór
    Acc sjó
    Dat. sjó
    Gen sjóar or the more colloquial 'sjós' based on analogy of the commenest masc/neut sing genitive morpheme -s, cognate to English 's, Latin, -is, Sanskrit, -asya.
    This above-paradigm gives us little insight as to the etymology, but 'sær' does
    Nom sær
    Acc sæ
    Dat sævi fr *sæwe> *sæhwe
    Gen sævar. > *sæwaz >sæhwaz cf Gothic saíhws
    I've seen the theory that "sea" is etymologically related to the verb "see" videre, and cognate, but w/ semantic-shift to Lat. 'sequere' to follow, the Gmc *sehwana, HG, sehen, OE pret/sing/1&2p, séah. It mean "to follow with the eyes" and ocean 'sea' is explained "as far as one can see". It fits well with Grimm's Law.
    Ken

    ReplyDelete
  3. I'm afraid the Germanic word for 'sea' has no cognates inside the IE family and hence no PIE etymology.

    IMHO, the problem is most Indo-Europeanists try to explain everything within IE as coming from PIE, when this is far from being true.

    ReplyDelete
  4. Actually, "thalassa" is a Pre-Greek word.

    It has a common geminate found in other Aegean place names..."ss"

    That is not what makes it Pre-Greek (most likely of Aegean origin), but that word has that very common geminate found mostly associated with the Aegean. That geminate is also found in Tartessos which is in what is now southwestern Andalucia in Spain. That led some to speculate that the Tartessians where Aegean peoples (which is highly doubtful).

    Aegean endings: -nos/ -na, -sarna, -ssos/ -sos/ -sa, -ttos/ -tta, -(i)nth(os)/ -(i)nth(a) (and their eastern derivations -ndos, -nda, nza, and -nzos) (but not -ynth(os), it is derived from Anatolian IE "wuntha") are considered Pre-Greek.

    ReplyDelete
  5. Actually, "thalassa" is a Pre-Greek word.
    Yeah, but this alone doesn't give us a clue about its etymology.

    It looks like your very concept of "Aegean" is a kind of black box, isn't it?

    That geminate is also found in Tartessos which is in what is now southwestern Andalucia in Spain.
    Yes, but this is a name coined by the Greeks upon an indigenous stem *tarte-, which has a continuation at a later time in the ethnonym Turdetani, from *turde-.

    ReplyDelete
  6. (large gasp) It is not a blackbox. The Aegean Languages (Minoan, Etruscan, Lemnian, EteoCretan, EteoCypriot, and Raetic/ Camunic) of which I am more than certain you are aware of...are the prime candidates for those unetymologizable words in Greek and Latin...It is hard to tell in the Anatolian IE languages...because they had contact with several other non-IE languages (Hatto-Kaskian, Hurro-Urartian, Kartvelian), but it is thought there is definitely some Aegean influence there (borrowings, loans, toponyms)...the amount is debated and the direction of the influence also (Anatolian IE to Aegean or Aegean to Anatolian IE). As we know now...the Aegean peoples came from Western Anatolia. Herodotus said Lydia (geopraphically speaking, that IS Western Anatolia) and, as has been verified, he wasn't wrong at all. So there had to be some contact between Anatolian IE and Aegean.

    There are words in Etruscan with similar endings as those mentioned in my previous post...that is where my extrapolation comes from...I dare not mention Mr. "you know who".

    So I will show you this blog...
    http://minoablog.blogspot.com/2010/08/pre-greek-place-names-of-aegean.html

    http://iedo.brillonline.nl/dictionaries/content/greek/loanwords.html

    I bet I know your reaction already.

    Hahaha, and I just saw your comment on the paper Mr. Fournet posted about the Pre-Greek substrate on Scribd...you really have a professional dislike of him...hahaha

    Also the other commentor, Mr. Doig, has a blog that I found accidentally while looking up "thalassa"...

    http://www.proto-germanic.com/2011/06/greek.html

    ReplyDelete
  7. And thank you for the Tartessos etymology...I figured those words had some connection...but wasn't sure.

    ReplyDelete
  8. The Aegean Languages (Minoan, Etruscan, Lemnian, EteoCretan, EteoCypriot, and Raetic/ Camunic) of which I am more than certain you are aware of...are the prime candidates for those unetymologizable words in Greek and Latin...
    But by no means the only one, especially in the case of Latin. As Etruscans weren't autochtonous to Italy, their language also absorbed loanwords from pre-Italic languages.

    It is hard to tell in the Anatolian IE languages...because they had contact with several other non-IE languages (Hatto-Kaskian, Hurro-Urartian, Kartvelian), but it is thought there is definitely some Aegean influence there (borrowings, loans, toponyms)...the amount is debated and the direction of the influence also (Anatolian IE to Aegean or Aegean to Anatolian IE).
    IMHO, they happened in both directions. I also consider Etruscan and the other Aegean languages to belong to the Vasco-Caucasian macro-family, as well as Hurro-Urartian and possibly also Sumerian and Elamite (but not Kartvelian, for example).

    Anyway, you should bear in mind that also Aegean words should have an etymology, so simply saying that the Latin word A is a loanword from Etruscan or Greek B is from Pre-Greek is only a part of the solution, but NOT the solution.

    Hahaha, and I just saw your comment on the paper Mr. Fournet posted about the Pre-Greek substrate on Scribd...you really have a professional dislike of him...hahaha
    He's really obsessed with Hurrian, around of which has made several crackpot theories, including a "decipherment" of the Phaistos Disk.

    As we know now...the Aegean peoples came from Western Anatolia. Herodotus said Lydia (geopraphically speaking, that IS Western Anatolia) and, as has been verified, he wasn't wrong at all. So there had to be some contact between Anatolian IE and Aegean.
    I think you're confusing "Etruscan" with "Aegean". Although Etruscans were Aegeans, certainly not all the Aegeans lived in NW Anatolia nor sailed to Italy.

    ReplyDelete
  9. I've removed the references of the Germanic word for 'sea', as I've written a new article about it.

    ReplyDelete
  10. Why is my reference still up here..what i didn't explain, was semantic shift from "sequor" (deponent "I follow") to Gmc *sehwana (sem. shift, "to see", as in "to follow with the eyes", that part is solid etymologically..yes, it's a stretch to "sea", and I never claimed it as fact. But there is a chance that the same lexeme that makes "sea" is found in other IE languages but the reflexes are so dramatically sem. shifted as unrecognizable. Cf. The theory of verb-person semantic-shift. The confusion between PGmc *wiz, Gothic, weis, Ger. wir, we, with the 2nd p.pl vous, vos, uos, Slav. vi, where the 1st & 2nd persons seemed to have been fluid. Cf, sanskrit "vayam" to "we" but the oblique forms of "we" us, our, from OE(Ingvaeonic, w/ loss of nasals before spirants, in NGmc, but before S/Z only) wé, ús, úre,úser, from PGmc *uns, *unsar-, evidenced by Gothic uns, unsar, Mod HG uns/unser, Hollandic/Frankish ,ons, onse, all from PIE *ns (the oblique form of the 1ppl) hence Latin nos, noster, from earlier *nosser, cf. Port. nossa....food for fought

    ReplyDelete