Re: Why there is t- in German tausend "thousand"?

2013/11/14, Brian M. Scott <bm.brian@...>:

> At 7:33:31 PM on Wednesday, November 13, 2013,
> Bhrihskwobhloukstroy wrote:
>
>> 2013/11/14, gprosti <gprosti@...>:
>
>>> --- In cybalist@yahoogroups.com, Bhrihskwobhloukstroy
>>> <bhrihstlobhrouzghdhroy@...> wrote:
>
>>>> 2013/11/13, gprosti <gprosti@...>:
>
>>>>> The question is *how* statistically probable it is that
>>>>> two words with the same meaning, which share five
>>>>> phonemes in the same sequence, are in fact two
>>>>> completely (historically) separate forms.
>
>>>> *Bhr.: 100% probable.
>
>>> It's 100% probable -- i.e., physically or logically
>>> necessary -- that two semantically matching words with a
>>> long, matching sequence of phonemes, but one non-matching
>>> phoneme, must have no historical affinity whatsoever?
>
>> *Bhr.: if they irreducibly differ even in just one
>> phoneme in the root, especially word-initially, they ARE
>> different by definition
>
> English </i:/conomics>~</ε/conomics>, </i:/ther>~</aɪ/ther>,
> </sk/edule>~</∫/edule>, </s/ism>~</sk/ism>,
> <hoofs>~<hooves>, <r/u:/te>~<r/aʊ/te>, <r/u:t/>~<r/ʊ/t>,
> <v/ɑ:z/e>~<v/eɪs/e>.

*Bhr.: <hoofs>~<hooves>, <r/u:/te>~<r/aʊ/te>, <r/u:t/>~<r/ʊ/t>,
<v/ɑ:z/e>~<v/eɪs/e> have no word-initial variations;
</i:/conomics>~</ε/conomics>, </sk/edule>~</∫/edule>,
</s/ism>~</sk/ism> are loan-words with diamesic interference, so
variants are completely regular in this case; </i:/ther>~</aɪ/ther>
are originally diatopic variants, so what's wrong? If You are able to
find diatopic variation for th- / t- in OHG, You're welcome

>
>>> Even if historical affinity doesn't include these cases,
>>> the probability is still not 100%, because the laws (=
>>> tendencies) of sound change are not laws of mathematics
>>> or physics.
>
>> *Bhr.: they are logical laws, otherwise one could not
>> demonstrate whether a given etymological hypothesis is
>> correct or not
>
> One can’t, with complete certainty. And they are certainly
> not logical laws, unless you mean only that they are
> reasonable, in which case your ‘otherwise ...’ is a non
> sequitur.
>
> Brian
>
>

*Bhr.: they're logical insofar the system-internal transformations,
just like meaning of words, are neither mathematical nor physical
laws, but nevertheless subject to non-contradiction principle: if e.g.
English "brain", in a given context, translates "Gehirn", it does so
in every identical context (without being a mathematical or physical
constraint); similarly, if Proto-OHG <th> (or its Germanic antecedent)
is transformed into <d> in a given context (e.g. in the word for
'1000'), it isn't transformed into <t> in the same contexts, as it is
inevitable to conclude as long as there are no traces of a source
dialect for a loan.
The question is: if one can find a possible source for *Old* Upper
German /d/ > /t/, that would solve the problem; until that isn't (yet)
realized, a lautgesetzlich PIE alternative imposes itself