Re: COMMENT PLEASE (WAS:Conlang Journal and being a fish)

Roberto Suarez Soto wrote:


>On Sep/20/2002, JS Bangs wrote:
>
>> bota > bot > boT
>> gota > goda > gora
>> Which is wrong, for obvious reasons.
>
>        Not so obvious for me :-) Could you elaborate on this, please?
>
Theoretically, in historical linguistics, sound change is exceptionless--
that is, if X (in this case /t/) undergoes change >Y in environment Z (in
this case /o__a/, or more generally, between vowels), then all instances of
X will undergo the change.

Thus in the examples given, there is no reason at all why -ota > -oT in one
case, but ultimately > -ora in the other. All things being equal, the
sequence -ota has to develop in one way or the other; it can't be both.

One could imagine scenarios in which the examples would be possible, but
that would require more specific conditions, or would require that one know
more about the history of the language than is given here.  There might be,
for ex., undetected/unknown differences in the accentual pattern, perhaps in
some earlier stage of the language; or it could be that /bota/ is a unitary
lexical item while /gota/ is actually polymorphemic, maybe an inflected form
?/go-ta/; that would make a difference in the statement of the environment,
and could make a difference in how the sounds change.  One form or the other
might be borrowed from another dialect or (un)related language; one or the
other might be part of the learned/religious/upper-register vocabulary.

The concept of exceptionless sound change, in the Real World of historical
linguistics, is somewhat too strong, though it remains axiomatic.   Things
like Jesse's examples do in fact occur, and are a real bother....... At last
resort, there's always "unexplained".

Reply