Andreas Johansson writes:
> Quoting Andreas Johansson <[log in to unmask]>:
> > As I mentioned some time ago, I'm working, if fitfully, on pielang (working
> > name: Wa). The sound-changes I've instituted, this far at least, are not very
> > radical, but sometimes they conspire to make closely related forms diverge
> > wildly. The case inspiring this post is this: PIE *kerd "heart" quite
> > regularly
> > yields Wa /Sard/, and the PIE genitive *krdos equally regularly yields Wa
> > /karado/.
> I just saw that Beekes gives the PIE nom. as *ke:rd (with long vowel), which
> would yield Wa /Serd/, making /karado/ look even more anomalous.

Very interesting forms. :-)  I'd probably regularise a lot and have
/Serdo/ (if that is regular), but keep the old genitive in a compound
or idiom, just to make sure the confusion is not completely lost.  (If
course, some words need to be totally irregular or even irregular in
analogy with others although they would yield regular paradigms.)

Given the right sound shifts (i.e., some with a well-defined time
frame and ordering), you could also regularise the stem at some point
in time and still get an irregular form, say /Sarado/, e.g. because
the stem /k/ regularise to /S/ at some point, but then, the shifts
went on.

I quite like irregularity in historical conlangs, too.  If only
Þrjótrunn was a wee bit easier to handle, I'd have more texts in it
and more vocab.  But it's so awefully irregular.  Almost like
Icelandic or something.