[gothic-l] Proto Germanic Vowels

[David Salo]

-------------------------- eGroups Sponsor -------------------------~-~>
GET A NEXTCARD VISA, in 30 seconds!  Get rates
of 2.9% Intro or 9.9% Ongoing APR* and no annual fee!
Apply NOW!
http://click.egroups.com/1/7872/8/_/3398/_/968005107/
---------------------------------------------------------------------_->

Ïusteinus melida:

>    Apparently this pgmc. *iu was a umlaut varient of *eu (I was quite
>surprised that there were umlauts happening that far back!), as an example:
>*theudo > Got. thiuda NHG. Diet-, but *thiudiskaz > Got. thiudisks, NHG.
>Deutsch.

   The umlaut eu > iu is only one part of a larger umlaut involving e > i
before a "yod" ([i], [i:], or [j] in the next syllable.  For instance

   PIE *medhja- "middle" (=Lat. mediu-, Skt. madhya-) > PGmc *meðja- >
Common Germanic *miðja- (Gothic midjis, OE midd).
   PIE *sedja- "sit" > PGmc *setja- > CGmc *sitja- (OE sittan)
   PIE *esti "is" (=Lat. est, Gk. esti, Skt. asti) > CGmc *isti (Gothic
ist, OE is)
    And the following is just the same thing:
   PIE *tewtisko- > PGmc *þewþiska- > CGmc *þiwðiska- (Go. þiudisk-)
   PIE *newjo- > PGmc *newja- > CGmc *niwja- (Go. niujis, OE níwe)

    It only _looks_ like a special case because in the various languages
"eu" and "iu" come to be treated as diphthongs with unique properties,
rather than simply a sequence of e+w and i+w.
    *ej would have umlauted to *ij under the same circumstances if it had
not already become *ij or *i:; I guess that it's impossible to tell whether
*ej had already changed to *i: (which would make it insusceptible to
umlaut), before the general e > i umlaut, or if that umlaut took place and
_then_ remaining *ej > *i:, making the variants with and without umlaut
indistinguishable.

    Nor is this the only case of yod-induced umlaut.  Original PIE short *o
became *a in all cases in PGmc (including the sequences *oj > *aj, *ow >
*aw).  But a new Germanic *o arose from lowering of *u before a following
non-high vowel (short or long a, e, o) in the next syllable; and, also
before non-high vowels (though perhaps under somewhat more limited
circumstances) *i > *e.  This new *o then underwent yod-umlaut _back_ to *u
under the same circumstances as those in which *e (whether original or from
older *i) became *i.

   So: PIE *kRno- > PGmc *hurna- > CGmc *horna- (Gothic haurn, OE horn)
       PIE *wLkwo-/*wLpo- > PGmc *wulfa- > CGmc *wolfa- (Go wulfs)
       PIE *pLno- > PGmc *fulna- > CGmc *folla- (OE full, but German voll)

   But: PIE *pLnj- > PGmc *fulnja- > CGmc *follja- > *fullja- (Go. fulljan,
OE fyllan)

   It's interesting to note that, because of continuing assimilations and
dissimilations related to the initial w-, the word for "wolf" in this
history of English has changed from [wolf] to [wulf] and back multiple
times since Pre-Germanic: PGmc *wulfa- > CGmc *wolfa- > OE wulf > ME
wolf(e) > ModE [wulf] (still spelled "wolf"!); but German retains Wolf <
CGmc *wolfa unchanged.

   Unfortunately the history of o and u has been much confused by various
alterations (like the change of o > u between f, w and l in Old English);
and since o and u might appear in the same paradigm corresponding to
different endings, analogical levelling to one form or another was common;
this also occurred with e and i.  There were also other conditions for the
change of e>i and o>u, namely when they appeared before a nasal sound (n,m)
followed by a consonant. (PIE *wento- > PGmc *wenþa- > CGmc *winda-; PIE
*bhNdhono- > PGmc *bundana- > CGmc *bondana- > *bundana- > Go. bundans
"bound".)
   It is therefore even more unfortunate that Gothic, by its merger of i/e
and o/u has made itself completely useless as a source of information for
the early distribution of these sounds!  The distribution has to be
inferred from the muddled evidence of more recent North and West Germanic
languages, and from non-Germanic languages that borrowed Germanic words at
an early date.

/\     WISTR LAG WIGS RAIHTS
\/            WRAIQS NU IST                               <> David Salo
<dsalo at softhome.net> <>



You are a member of the Gothic-L list.  To unsubscribe, send a blank email to <gothic-l-unsubscribe at egroups.com>.
Homepage: http://www.stormloader.com/carver/gothicl/index.html