Low German
2sg geist, 3sg geit, 123pl gaat
2sg steist, 3sg steit, 123pl staat
Joseph Wright An Old High German Primer:
Alemanic
1sg ga:m, ga:n
2sg ga:s(t)
2sg ga:t
1pl ga:me:s, ga:n
2pl ga:t
3pl ga:nt
Franconian
1sg ge:m, ge:n
2sg ge:s(t)
2sg ge:t
1pl ge:me:s, ge:n
2pl ge:t
3pl ge:nt
Otfrid
1sg ga:m, ga:n
2sg gei:st
2sg gei:t
1pl ga:me:s, ga:n
2pl ga:t
3pl ga:nt
and similarly for sta:n, ste:n
beside the longer stems gang-, stand-
(which I think belonged in the plural)
So it looks like the Modern Low German paradigm is faithful to the
original. It is very likely that the -e- is umlauted -a-, so the
original was *ga:i-s(t), *ga:i-t etc, reminicent of Greek phereis,
pherei. Also of the high number of Dutch verbs in -V:i-, zaaien "sow",
waaien "wave, blow", dial. pooien "drink", which in PIE terminology
are long-vowel verbs, which also have that -i extension. Looks like
there was a time when the -i of the primary inflection could be added
directly to the bare verb stem, without endings (my '0sg'?), seems
like Jasanoff has a point there.
But it gets worse.
Low German
2sg deist, 3sg deit, 123pl doot
No umlaut is possible here. The 2sg and 3g forms look like direct
descendants of PIE *dhe:(i)- "put"
And to top off the weirdness:
Old Frisian (Boutkan)
3sg deth
123pl dwath
inf dua, dva, dwa
(from *dhgh-éno-?)
http://tech.groups.yahoo.com/group/cybalist/message/14720 (Miguel)
Modern Frisian
1sg doch, 2sg dochst, 3sg doch, 123pl dogge, inf. dwaan
More confusion
Old Frisian (Boutkan)
3sg stont, 123pl stondath "stand"
3sg gunth, guncht "go"
On PIE *dheh1-/PGerm. *do:
from the Uralic handbook of Péter Hajdú
(quoted from János Makkay
THE EARLIEST PROTO-INDO-EUROPEAN-
PROTO-URALIC CONTACTS:
AN UPPER PALAEOLITHIC MODEL)
the nine glosses on which there is general agreement that they are
loans between PIE and Uralic:
"
xx PIE PFU(Fi) Hung English
1. *wedh- *wetä- vezet leads
2. *wegh- *wiGge- visz carries
3. *doGw- *toke- hoz brings
< *do:- (Fi tuo)
< *toGe-
4. *mozge- *mos´ke-,
xx xxxxxxx *mus´ke- mos washes
5. *dh3k- *teke- te-sz makes
6. *no:mn.- *nime név name
7. *wed- *wite,
xx xxxxx *wete viz water
8. *kot- *kota ház house
9. *sñew-, *so:ne, in sinew
xx *sen-, *si,ne,
xx *son- *se,ne
"
Notice (surprise!) that both PIE *dheh1- and *doh3- are there (except
that the senses are reversed!), and that they both end in a velar, not
a laryngeal. One begins to suspect the long vowel was caused by
something other than a vowel. Perhaps the velar was lost by a rule
that stripped auslaut consonants off in 23sg, with compensatory
lengthening (cf aorist, or rather injunctive).
These two verbs are close to being function, not content words, and
should be borrowed in a situation when the receiving language changes
type as the result of the influence of a dominating language (but
where are the content word loans?).
As examples of how much dei-, do:- is a function word in Low German:
dat he blöden deit 'daß er blutet'
Un as he do todegen tokieken deit
"And as he then skillfully looked there"
denn wunnern deit een sik "for one did wonder"
Söök di ut, wat passen deit "Seek out for yourself what fits"
un wat de allens vun sik wiesen doot
"and what they all show of themselves"
un wat de allens verkehrt maken doot "and what they all do wrong"
wat froher en Rechtsanwalt kosten dee "what an attorney cost earlier"
This is how one might imagine the Germanic weak preterite to have come
about.
As should be obvious, I'm not finished with this material. Maybe other
good people on cybalist may find ways of reducing its complexity?
Torsten