I can't really see the problem; according to the written evidence and
the geographical distribution of its reflexes, High German 2nd Sound
Shift took place as follows:
2nd third of 7th century:
1) t > #ts-, -s- (tid/Zeit-line)
2) k > -x- (maken/machen-line)
3) p > #pf-, -f- (pund/fund/Pfund-line)
2nd half of the 8th century:
k > #kx-
8th-9th century:
1) d > t
2) g > k
3) b > p
9th century:
θ > d
Therefore porta > Pforze (8th century porta, borta; 9th century
phorta), like Pfoertner (12th century portināri, Middle High German
phortenaere), after postconsonantal /t/ > /ts/ but before word-initial
/p/ > /pf-/ (a Romance †porda is unconceivable, lenition of voiceless
plosives doesn't work in postconsontal environments), while Middle
High German phorze (in contrast with phorte) before postconsonantal
/t/ > /ts/ but North of pund/fund/Pfund-line.
Place-names of Latin origin were obviously subject to Old High German
sound-changes from their very beginning, so they show pure
geographical distribution, viz. (from portus):
Porz am Rhein (near Cologne), 1160 Porze
Pforzen on Wertach, 897 Forzheim
In Pforzheim on Enz we even see 7th century Porza, 1067 Pforzheim,
1257 Phorzein, yet 1338 Portzheim, 1357 Phortzhein, 1689 Pforzen.
Note in until 12th century Romance-speaking Moselle valley 776 Porto
Pigontio, 1098 Pizport, now Piesport an der Mosel
In Langobardic:
t > ts by 573
p > f by 581
d > t by 584
b > p 6th-7th century
k > x by 643-670
g > k by 670
In Crimean Gothic:
d > t (tag, alt, plut, but handa, fyder)
b > p (plut, but bryder, broe, brunna)
k > -x- (ich) (no #k- > †#kx-: kommen)
g > k (rinck, but goltz, geen, singhen)
θ > d (bruder, but goltz, statz, tzo)
2013/11/11,
johnvertical@... <
johnvertical@...>:
> ---In cybalist@yahoogroups.com, <grzegorj2000@...> wrote:
>>
>> There are interesting forms, especially among Latin borrowings, with
>> results of the Shift different than expected (and thus they can be
>> compared with the "take" and "up" problems with the 1st Shift mentioned
>> above).
>
>> Romance (< Celtic) *pauta "paw" has yielded German Pfaute (with p > pf
>> while -t- unchanged).
>> Also Latin porta --> OHG pforta (not *pforza), modern Pforte "gate".
>> Latin imputare --> OHG impfitōn "inoculate", now impfen, with -t-
>> preserved but then irregularly omitted.
>
> These are rather interesting - but rather than evidence for highly different
> datings of p > pf viz. t > z, they look like they would have been acquired
> from a Romance variety where medial stops had become voiced in certain
> positions (*pauda, *porda, *impudare)?
>
> Rather than an actual sound change, any change of this sort also might have
> been a loanword substitution of some kind. The Romance voiced/voiceless and
> the Germanic fortis/lenis contrasts in stops are different, and cannot be
> linearly equated. If medial fortes in pre-HG were already aspirated (as
> suggested by their eventual affrication), perhaps the lenes were found here
> a more acceptable substitute for the voiceless unaspirated medials of VL.
>
> _j.