Morphological remodelling

[Jurgis]

> Can anyone on the list think of examples of morphological remodelling
> within a verbal paradigm based on a single ending? The only secure example

There're cases in Lithuanian dialects when fragments of former athematic
endings have been reinterpreted as parts of the present stem.

In the case of Lithuanian verb miegoti "to sleep", there two remodellings of
this kind attested.

1) -m- was introduced from 1st sg. miegmu ("I'm sleeping") after it had been
remodelled from 1st sg. miegmi (cf. dirbu "I'm working"):
1 sg. miegmu
2 sg. miegmi
1 pl.  miegmam
2 pl. miegmat
3 sg./pl. miegma

2) -t- was introduced from 3rd sg./pl. athematic form miegti ("sleeps"). The
paradigm itself got i-stem endings probably because 3rd sg./pl. athematic
miegti reminds of an i-stem form (cf. 3rd sg./pl. tiki "believes"):
1 sg. miegciu (c = ch; miegciu < *miegtju, cf. tikiu "i believe")
2 sg. miegti (stress on the 2nd syllable, cf. tiki "you believe")
1 pl. miegtim (cf. tikim(e) "we believe")
2 pl. miegtit (cf. tikit(e) "you believe")
3 sg./pl. miegti

Source of the forms: Z. Zinkevicius, Lietuviu dialektologija, Vilnius:
Mintis, 1966, p. 349.

Larry Trask mentions:

>Watkins draws attention to such remodelings in Celtic, Polish and Persian,
>at least, all based upon the third-singular forms.  The observation that
>such remodelings are typically based upon third-singular forms has been
>dubbed 'Watkins's Law'.

Could anyone provide data / reference on how typical / frequent remodelling
based on 1st sg. is?

Regards,
Jurgis Pakerys
Vilnius University