Heill aftr.

> > The language here is very conservative and anyone reading it

will have to know where to set tails under o/ó and e/é and where
not to.

> How about in words like 'vóru' and 'nótt'; are these without tails,

with the raising of hooked-o after labials and nasals, or was that a
later development?

With tails, I say. vóru is just u-mutation, which I think is what
originally happened to nótt, as well, although is suffered a
different fate (>ó instead of >á).

> > siávar ... hofþafiolunni ... tialds ... fiollum
>
> What is known about the timing of the change from falling to rising

diphthongs? And what about the First Grammarian's recommendation of
the spelling 'ea' in 'earn' / 'iarn'?

11th century, I think. Scaldic verses speak for forms like: þrír -
þríar - þríu; fé fear/féar; etc., but by 1100, the change seems to
be in place, whatever errant spellings say (þriar, þríár, etc.). I
also take the effect of breaking to be ia (e > ia). Keep in mind
that iárn is exceptional in the diphthong-context, as it is from
ísarn by loss of intervening consonant.

> > þeirri
>
> Books printed in normalised spelling that I've seen have variously

'þeirri' and 'þeiri'. I take it from your reconstruction that the
earliest texts have the -rr- form, like the modern language.

Yes. I think what happened was that after R > r, a second r was put
in on analogy with the adjective-declension. I suspect that it was a
rather immeadiate change, but that perhaps it had to fight a long
fight with form like þeiri, þeirar. It´s an unresolved matter, where
I simply go after the oldest written usage.

> > gørva ... reru (sorry, røru is correct, my oversight)
>
> What is known about the timing of these sound changes, the

rounding of /e/ and unrounding. Is there a particular reason why
you have 'ø' in 'gørva' and 'e' in 'reru'.

No, just an oversight. røru is correct.

> Stefán Karlsson in The Icelandic Language(trans. Rory McTurk) just

mentions that "in some contexts ø was unrounded, e.g. øxi > exi"
(1.2.2), but there seems to be a lot of variation here. Is there a
system; is the unrounding limited to a particulatr environment; or
is it fairly arbitrary? It seems to affect both hooked and unhooked
ø; was there a greater tendency to unround one or the other, or were
they equally affected?

In the older language, u-mutation rules. I-mutation comes in second,
as it can be levelled out (staðr, OGut steðr). This does not happen
with u-mutation until u-mutation of á (hooked-ó) disappears, then
the ball starts rolling on changing the vowel system, which keeps
rolling through the centuries. øxi and gørva both show both u- and i-
mutation. The resulting ø is called ø2, instead of ø1 (< only i-
mutation of o or ó, occuring long or short), can only occur short
(see Gordon and others), is written with the a-rune in inscriptions
(against u-rune for ø1), and does not occur as a 10th vowel in the
1st Grammatical Treatise. Perhaps it was levelled into ø1 by this
time, as the sound was statistically rare, as can be shown by the
etymologies of words containing either ø1 or ø2. Hooked ø is, of
course, the classic way of writing ø2, as it parallels the u-mut.
shown in hooked-o/ó.

> Well, that's enough questions for now. Thanks for posting this.

It raises some interesting issues.

One thing to keep in mind about early Icelandic is that is was a mix
of west-norse dialects. It was the same language, yes, but there
were variations in individual words, paradigms, etc., which become
more noticible when one looks into details. The differences, while
not on the scale of the west/east norse division at this time, have
their roots in the dialects of the Icelandic settlers. Thus, engi
maðr, engi mann and øngr maðr, øngvan mann are both correct, despite
the glaring differences in form and declension. Speakers using one
form would likely have their speech from a different area of Norway,
having preserved it until Icelandic dialect-levelling occured, or
something like that ;) On the other hand, sometimes one form is just
younger than another. sonr, for example, is younger than sunr, which
both occur frequently in the earliest texts. When did it arise?
Well, runic inscription always shows sunr, which could be either u
or o, so about all that we can say here is that the change was
occuring, and that until about 1100, final -r still occured in the
nominative of patronymics (Áleifr Tryggvasunr > sonr > son). Did it
happen because of lack of stress in patronymics (-sunr > -sonr > -
son, like heiþríkr > Heiþrikr > Heiþrekr), or was the a-mutation
levelled from gens.sg.and pl.? Or was it something else? For what I
am writing, any form in use around 1100 is fair game, although I do
choose the most conservative ones when I have a choice. When there
is no more conservative form (as with øngr/engi), I let my aesthetic
sense rule, which is this: I like the prose text to show only one
set of forms (for example, only nekkvert or only nekkvat, etc. for
whole paradigm), while I do not want to discriminate against equally
old forms, and therefore might use the variant set in the poetry. I
realize that, as we have no living speakers, one cannot called text
A 'pure-Rogaland speech' over against text B 'pure Sogn-speech', as
the original differences are far too hard to acertain today, and ran
together in Iceland, anyway. Still, the earlier language is nicer, I
think, than the later one, and no more standardized or unreal than,
say, 13th century standardized language. It also has the advantage
of being virtually identical to the real language of the heathen
norse, the changes being so few by 1100 that one can easily throw it
back to 1000 with _almost_ no change, as both inscription and latin
letter suggest. Lastly, I think that folk want to see this language,
to put it bluntly, as folk reading or learning old norse want to see
the real viking language. Folk want to know that the norse once said
not just 'ek em' (I am), but also 'hann es' (he is) and 'þú est',
see the differences between es/sem and en/an (þan in runic swedish),
etc.etc. before they were obscured. Folk might want to learn that
the article was always hinn, and never enn or inn, in inscription,
and that it still occurs in the earliest manuscripts(beside enn/inn)
and even makes a comeback in Modern Icelandic. A host of issues, so
more later.

Regards,
Konrad

> Llama Nom
>