--- In
cybalist@yahoogroups.com, "ehlsmith" <ehlsmith@...> wrote:
>
> --- In cybalist@yahoogroups.com, "Joao S. Lopes" <josimo70@> wrote:
> >
> > In Brazil there is some Tupi (Amerindian) redundant hydronyms and
> > toponyms, like River Iguassu, "big water". But I didnt question the
> > redundance, just the meaning, why to call a river "Liquid"?
>
> I don't believe it would require too much semantic drift for a word
> originally meaning "liquid" to become "water" in some daughter
> language (or perhaps the drift might have gone the other way- a word
> originally meaning "water" may have undergone sufficient shifts in
> meaning in various daughter languages that modern scholars might
> erroneously reconstruct the original meaning as "liguid".)
>
from:
Language Contact;
Substratum, Superstratum, Adstratum in Germanic Languages
Dirk Boutkan and Maarten Kossmann:
On the Etymology of Dutch zijpe
a short quote ;-)
"
5. Further etymological connections within IE are even more doubtful.
The standard view is that PIE had a root *sei- with different
extensions (e.g. *sei-b-, yielding the WGmc. evidence under discussion
here). Pokorny mentions the following evidence:
(1) *sei-, *soi- 'tröpfeln, rinnen, feuchf (1959: 889). This would be
the bare root. However, all forms quoted by Pokorny are extensions
from this alleged root. He gives two formations, viz.
"mit l-formans": geographical names: Venet. Silis, Silarus, Ligur.,
Illyr. Silarus, Hisp. Sil; MIr. silid 'tropft, fliesst, lässt fliessen
("teilweise mit sel- "sich bewegen" ... kontaminiert"); OE sioloT
'See'; Lith. séile 'Speichel, Geifer'.
"mit m-formans": Welsh hufen "Rahm" (< *soimeno-); OHG, Mod. HG seim
'Honigseim', OIc. seimr 'Honigscheibe', "ablaut." simi 'Meer', Danish
sima av 'abtraufeln', Westfalian si&mern 'sickern'...
The l-forms are very doubtful from a PIE point of view. The
geographical names are suspect as being of non-IE origin. The MIr.
form shows problems (contamination). The Gmc. form seems isolated as
the Lith. form differs in meaning. Even if the Celtic, Gmc. and Baltic
forms are related, the whole complex is suspect as reflecting North
European substrate material.
The m-forms show a similar picture. The forms (allegedly) containing
*oi show semantic problems and are restricted to Celtic and Gmc. The
alleged zero-grade forms are restricted to Gmc. only.
(2) *seikW- 'ausgiessen, seihen, rinnen, träufeln' (1959: 893). This
seems to be a PIE root, cf. Skt. sincáti, sécate 'giesst aus,
begiesst, be-feuchtet, bespritzt', Gr. hiksai . die:the~nai 'seihen,
filtern' (Hesychius), Serb. CSL sUco, sUcati 'harnen' (Seebold 1970: 390).
Pokorny (loc. cit.) also mentions Gmc. k-forms that simply do not
correspond and must be grouped separately:
(2a) PGmc. *seik-, e.g. OHG, MHG seich 'harn', norw. diall. sikla
'kleiner Bach', OE si:c 'Wasserlauf. These forms have cognates with an
unexplained geminate, cf. Mod. HG sickern, Sw. sikkla 'geifern,
rieseln'. This gemination makes an IE origin unlikely. The Gmc. forms
arc probably cognate with Slavic forms only, e.g. CS1 sUc^o, 'harnen'
(Pokorny p. 894).
(3) *seip-, seib- 'ausgiessen, seihen, rinnen, tröpfeln' (1959: 894).
The alleged variation in root-final consonants is not a PIE process;
hence we must distinguish between two different formations:
( 3a) *seip-. Here we find the 'Sieb'-etymon, OHG sib, Du. zeef, etc.,
which is semantically remote from the rest of the evidence. The
remaining forms quoted directly denote liquids and may belong
together: OHG seivar, MLG, OFris. se:ver, etc. 'Schleim, Geifer',
Serbocroat sípiti 'rieseln, fein regnen'. Note that PGmc. *saibr- (OHG
seivar, etc.) is semantically different from the rest of the terms
quoted. If it belongs to the rest of the evidence we are probably
dealing with North European substrate material.1
footnote
1 As an alternative, Peter Schrijver (p.c.) draws our attention to the
Latin river-name Sabrina that may represent a variant *sabr- beside
*saibr- (seivar, etc.), showing the typical variation *a ~ *ai in
European substratum words (see also Schrijver 1997; 303ff.). Semantic
problems remain, however. At any rate, there is more evi-dence for a
(non-IE) root *sab- in this semantic field, cf. Du. diall. zebbe, zep,
zeb < *sabjo 'drain', the hydronym Zever < *sa(i)bri-, cf. Sabrina,
perhaps the English hydronym Seph (although < PGmc. saf- [< *sap-]).
Istrup has a hydronym Sepp, also serving as an appellative (Teuchert
1944: 174) with yet another root variant *sapp-.
"
and as a sort of conclusion:
"
6. The Germanic form thus turns out to be isolated and probably of
non-IE origin. One can adduce a possible cognate form in a non-IE
language family, viz. the Berber languages. Elsewhere, we have pro-
posed a number of parallels of Gmc. and European substratum words with
reconstructable Berber forms (Boutkan and Kossmann 1999).
In Berber, there exists a widely attested denomination for 'river',
which is asif or, with different vocalism, suf, cf. Tashlhiyt (SW
Morocco) asif 'river with permanent water', Middle Atlas âsif 'river,
river bed', Senhadja de Sraïr (N Morocco) asif 'river', Kabyle (NE
Algeria) asif 'river, river bed', sedentary Touareg (Ghat, Djanet)
asif 'valley', Chaouia (Aurès, E Algeria) suf 'river with water'
(Basset 1961: 170, 341), Mzab (N Sahara) suf 'flowing river', Djebel
Nefousa (W Libya) usef 'river, river bed'. Although the form suf is
only scarsely attested in the modern dialects, toponymy attests to a
greater spreading at a former stage. As a river name (or part of it),
it is found in all regions where so-called 'Zenatic' dialects are
spoken (cf. Kossmann 1999: 30-31 on the Zenatic dialect group), from
the region called 'Souf in Eastern Algeria to 'Oued Souf ech-Cherg' in
the Eastern Middle Atlas (33.13 N, 4.00 W) and 'Souf in the Rif (34.59
N, 3.42 W). Moreover it is attested in Western Algeria (Beni Snous) as
part of the compound plant name syarsuf 'alder-tree', lit.
'river-wood'. In Kabylia, where the form asif is used, this tree is
called asGersif.
The first vowel of asif is constant (i.e. preserved in the annexed
slate), which, depending on the historical analysis one makes, either
implies that the word contained a laryngeal consonant, or that the
original stem was vowel-initial. The abbreviation and subsequent loss
of the initial vowel in Touareg and the Zenatic dialects is
well-attested in other stems where the first consonant is followed by
a full vowel.
The vowel alternation i - u is irregular. It is attested in a few
other words, like alim - lum 'straw', adif - aduf 'marrow' but has not
been explained satifactorily. The original meaning of asif ~ suf is
probably 'river', i.e. either a stream with permanent water or a river
bed which at the moment it is called asif/suf is water-bearing.
"
on the vowel alternation, cf.
http://tech.groups.yahoo.com/group/cybalist/message/37247
Torsten