At 4:25:49 AM on Monday, July 7, 2008, tgpedersen wrote:
> --- In cybalist@yahoogroups.com, "Brian M. Scott"
> <BMScott@...> wrote:
[...]
>> It is an empirical fact that irregular changes occur,
> That's wrong.
Literally, yes, but it's an obvious shorthand for an
obviously correct statement.
> Irregularity is not a property of a set of changes, but of
> a set of changes with respect to a given set of rules.
Indeed. And of course a rule with only finitely many
exceptions can trivially be rewritten as a rule having no
exceptions simply by incorporating the exceptions in the
statement of the rule. But no one in his right mind would
consider the result a satisfactory rule as the word is
usually employed in this context. (I might add that
postulating dialects/sociolects that are not otherwise in
evidence is methodologically no sounder than devising
unsupported chains of sound changes to account for possible
isolated relationships.)
Brian