--- In norse_course@yahoogroups.com, "Fred and Grace Hatton"
<hatton@...> wrote:

>
> That's what I get for looking in the irregular verb forms pages in

the back

> of Zoega first! Keyri is given as an alternate form for kjósa, but
> obviously the sense didn't fit.
> Grace

Ah, the appendix of last resort... Although actually I've found that
very useful in the past; there are so many word forms in Old Norse
with mutations and assimilations that cause them to look very
different to the dictionary look-up form, it's great to have something
like that to turn to when you get stumped. And yes, 'keyri' is one of
various forms of the 1st/3rd pers. pret. sg. of 'kjósa'. In Zoega's
format for giving the parts of verbs:

kjósa (kýs; kaus, køri and keyri; kusum and kurum; kosinn, kørinn).

Originally (judging by related languages) it was a regular strong verb
Class II, like 'bjóða' (býð: bauð; buðum; boðinn), except that some
pars of the verb had 's', others 'r', according to Verner's Law (see
Gordon, § 71).

kjósa (kýs; kaus; kurum; korinn).

The 's' goes back to forms where the stress was on the preceding
syllable in the dialect of Proto-Indo-European that the Germanic
languages descend from; where the stress wasn't on the syllable
immediately before it, 's' was voiced to 'z' (in Proto-Germanic) which
eventually became 'r' in Old Norse (and English).

In earlier (Viking Age) Old Norse, 'r' from Proto-Germanic 'r' had a
different pronunciation to the 'r' that cames from Proto-Germanic 'z',
and was spelt with a different rune (the one that scholars often
transcribe as a small capital 'R'). This 'R' caused the same sort of
vowel mutation as a following 'i' or 'j', hence Proto-Norse *'kozanaz'

> Viking Age Old Norse *'køRinn' > Old Icelandic 'kørinn' (see Gordon

§ 37).

Once these sound changes were over though, and with the stress fixed
on the first syllable of words, and especially once two types of 'r'
having merged in Old Norse, this left a rather irregular looking verb.
There was a tendency therefore to simplify and use 's' throughout,
even in those forms which historically had 'r'. Also, a more
regular-looking past participle 'kosinn' was invented by analogy with
other Class II verbs.

On the other hand, the old form of the past participle 'kørinn' must
have reminded some speakers of the preterite of reduplicative Class
VII strong verbs like 'sá' "to sow", pret. sg. 'søri' (later also
'seri'). So by analogy with verbs like this, people invented new
preterite singular forms for 'kjósa': 'køri' and 'keyri' < *'kauRi'
(cf. ON 'eyra' : Gothic 'auso' : English 'ear').

Hope that's not too confusing! I find sometimes that knowing the
history of the word forms helps me to remember them (because then they
don't look so random and irregular, if you can see the underlying
pattern), but I don't know if that works for everyone. There might be
easier ways...

LN