--- In cybalist@..., g-tegle@... wrote:
> To the germanic infinitive *fánhan my etymological dictionary gives
a
> reduplicated imperf. sg. *fe-gánh-. It makes use of Verner's law to
> explain the past participle *fanganá/*fanginá (*[h]>*[h]). But it
> gives no explanation as to *[g] in the *-gánh- stem with which the
> reduplicated imperfect form was made. It says however that gothic
> imperf. sg. faifâh reflects a younger *fe-fanh- form. In other
> germanic languages the *[g] has spread to other forms by analogy
> (Ex.: 19. century german inf. fahen, later fangen).
>
> The dictionary traces the proto-germanic form back to IE *pank´
(with
> nasalized present stem). Where did the *[g] come from in the first
> place?
>
I first have to acknowledge a mistake.
<(*[h]>*[h]).> should have been *[h]>*[g].
The posting in general was unclear, for which I will partially blame
the dictionary, but most of all it was because I am an amateur.
The dictionary is somewhat unclear in its explanations, but it
specifically says that the verb originally had a reduplicated pret.
sg. *fe-gánx-, younger *fefanx- with *-f- after the infinitive. Then
it refers to the explanations given under the norw. verb 'henge'.
I do not find any explanations there as to the origin of *-g- in *-
ganx-. Perhaps the book is indirectly suggesting that the form was
influenced by germ. *xánxan- to wich the dictionary gives a pret.
sg. *xe-gánx, younger *xexanx-. The change *g > *x is supposed to be
due to influence from the infinitive.
The dictionary says that got. pret. sg. faifâh reflects the younger
*fefanx-.
The pret. sg. forms with *g found in other germanic languages are
converted forms of *fe-gánx- due to analogy with pret. pl. and the
perf. participle. So obviously the dictionary finds nothing strange
as to the origin of the preterite plural forms. What I wonder is how
the forms with *-gánx- is attested.
Piotr:
Obviously, *fe-ganx- is supposed to reflect pre-Verner *fe-xánx-,
presumably from regularly reduplicated *fe-fánx- influenced by forms
of *xanx- < PIE *k^onk- 'hang'. The original infinitives and past
participles of both verbs rhymed, so the development of parallel
preterites would not be surprising. I only wonder if there is
sufficient Germanic evidence to posit *fe-xánx- in the first place.
The usual Old English forms are fôn, fêng, fêngon, -fangen (exactly
parallel to the conjugation pattern of hôn 'hang'). The infinitive
derives from contracted *fô(h)an < *fa~x-an- < *fánx-ana-n; the past
participle is the expected fangen < *fang-ana- < *fanx-aná- (Verner's
Law). The only puzzling forms are the preterites.
Class VII verbs in Old English (including residual reduplicated
forms) show no contrast between the preterite stems -- the forms in
question had been levelled out by OE times. The plural fêngon could
represent older *fegangun < *fe-xanx-'-, haplologically contracted.
Håvard:
How do you explain *fe-xanx-'- in the plural forms?
Here the dictionary obviously disagrees with you. *fe-gánx- < *fe-
xánx- (Verner's law) is the origin of only the sg. pret. forms.
However it seem only your view explains the vocalism in the preterite
plural forms. Perhaps I have misunderstood both you and the
dictionary.
Piotr:
The hypothetical 1/3 sg. *fe-ganx-, however, would have lost the
nasal and become pre-OE *fegôx, so it's fêng that must be explained
as analogical.
Could you remind of the reason for *fe-ganx- > pre-OE *fegôx? Is it
same development that has caused ON pret. sg. fekk?
Piotr:
The persistent nasal looks as if it were a root segment rather than a
present-tense infix, which means either that the *pa-n-k^- analysis
is wrong or that *pank(^)- was reanalysed as a simplex morpheme
already in pre-Proto-Germanic times.
Håvard:
According to the dictionary *pak´ comes from *peH2-k´ "fasten, make
fast" found in Vedic pâ´s´ah. m. and in greek present 1st sg. pê´ttô.
*pank´- cannot be a reflection of *pHnk´- because the only possible
zero-grade *pH2n.k´- would have given germ. *fúnh-. The dictionary
goes on: Nasal infixation was normally done to verbal themes not
roots. It is therefore wrong to accept a original form was a root
*pak´- (as Mayrhofer in EWAIA 2 (125) does) where the *-a- is an
original PIE vowel.
I will have to study the further explations more thoroughly.
Håvard