From: dgkilday57
Message: 71573
Date: 2013-11-13
---In cybalist@yahoogroups.com, <dgkilday57@...> wrote:
---In cybalist@yahoogroups.com, <bhrihstlobhrouzghdhroy@...> wrote:2013/11/9, dgkilday57@... <dgkilday57@...>:
(...)
*Bhr.: I fail to understand the treatment of German /d/ < PIE */t/
together with German /b/ /g/ < PIE */bh/ */gh/: aren't they different
isoglosses? Upper German /p/ /k/ for German /b/ /g/ are the regular
outcome of 2nd Sound Shift (in those areas where it included most
developments) and are therefore a more complete version of German /t/
< PIE */dh/ (or Verner Proto-Germanic */d/ < PIE */t/, of course)
[DGK:]
This is indeed a thorny problem. Goblirsch's presentation implies a third consonant-shift, namely PIE *t- > Gmc. *þ- > OHG *d- > UG *t-. However, Weber's data (assuming _Dorf_ and _Ding_ were simply borrowed from standard HG), along with the assertion about _Traube_ and _Tausend_, could equally well be explained by a separate UG development, Gmc. *þ- > *t-, distinct from the shift to *d- in OHG dialects further north (which was incomplete when literacy arrived, since some of the words have th- or dh- for later d-). This would agree with _troppus_ in the Lex Alam. (and REW 8938) being cognate with LG _drubbel_ (/a/-umlaut explains Gmc. *drobba- against *drubbila-).
Unfortunately It. _tasso_ 'badger' and related Rom. words (REW 8606) cannot help us, since _taxo_ 'small four-footed animal' is already in Late Latin (prob. 4th cent.); while the LL form corresponds to Sp. _tejón_ etc., the It. word is homophonous with _tasso_ 'yew' from Lat. _taxus_, and probably came directly from Gothic *þahsus without the LL extension. At any rate all the Rom. 'badger' words appear to result from borrowing before the High German shift occurred.
On the other hand It. _tanfo_ 'stench, stink' is referred by M.-L. (REW 8696) to Langobardic _thampf_ [sic] corresponding to HG _Dampf_ 'vapor, steam', MHG _dampf_, _tampf_ (Upper?), OHG _dampf_, MLG/MD _damp_. This appears to reflect a Gmc. *þamp- with no obvious IE etymology (8th-cent. OHG _dempfen_ looks like a reflex of Gmc. *þampjanaN, not a borrowing from OLG/OD). If the correct Lgb. form is *tampf, we have *t- against LG/HG d- in the same UG word as *pf. Still I cannot prove that there was no third consonant-shift involved.
[DGK bis:]
Meyer-Lübke's citation form is correct. Other Langobardic glossary items show that while the language underwent the High German shift, e.g. _brehhan_ 'to break', _krampf_ 'cramp', _nappja_ 'nose' (cf. OE _nebb_ 'beak'), _trinkan_ 'to drink', and _zann_ 'tooth', it retained /þ/ in _blauths_ 'naked', _slaiths_ 'smooth', and _thrukkjan_ 'to squeeze' as well as _thampf_.
OHG _dempfen_ in this form should not be ascribed to the 8th cent. In fact the earliest OHG has th- in this verb: _(fir)themphit_ 'suffocates' (K. Gl.), _thamfta_ 'suffocated' (Tat.). That the Gmc. root was indeed *þamp- is shown by Old East Frisian _thampene_, _thempene_ 'suffocation' (W.L. van Helten, Altostfr. Gr., Leeuwarden 1890, p. 144).
The upshot is that despirantization of inherited *þ was not part of the High German consonant-shift proper, but occurred later. The Langobardic glosses are early enough that the unvoiced spirant was retained and represented by _th_, likewise the earliest OHG. The material so far discussed suggests that in Franconian, the spirant first became voiced, represented by _dh_, and then became a stop, written _d_. But in Alemannic (and probably Bavarian), it became, at least initially, an unvoiced stop (though having a voiced allophone, as represented in Notker's orthography _ter bruoder_, _únde des prûder_), normally written _t_ and merging with the shifted reflex of Gmc. *d. Thus no third consonant-shift is necessary.