Heill Llama!

> I was trying to put together some notes a while ago on
pronunciation, aiming to describe the stage of the language that's
reflected by the normalised spelling used in introductory books and
texts that beginners are likely to encounter, that is: appropriate
to the early 13th century.

Yes, Snorri's language.

> A later date, e.g. about 1300 would be another possibility, but
earlier is better for etymology--and it's easier to ignore
distinctions once you've learnt them than to learn new ones.

True. Always good arguments for learning an earlier stage. About the
language of 1300-plus: while studying the changes herefrom is needed
to understand the development of Modern Icelandic, it is unnecessary
for an understanding of ON. It becomes an abstract study of little
or no relevance to students of ON proper. Linguists aside, most folk
learning ON are interested in ON proper, and especially the culture
stage represented by the Viking Age. Odds are likely 100:1 that this
is the stage most ON students are interested in. Furthermore, as the
linguistic developments post-1300 are irrelevant to the ON language
of the period in question, as well as the cultural developments, the
traditional schoarly concensus is to cut the ON-period at Snorri. It
is somewhat artificial, at least linguistically, but it does work in
practice, as it meets the interests of students eye to eye.

> Any earlier than that though and you need specialised knowledge to
apply that pronunciation to the texts as normally printed, e.g. when
á = /a:/, when /O:/, and when either of these should be nasal.

Ok, linguists and ON academics will hold us to the stage described
by the first grammarian, and print their forms using the alphabet
outlined by the first grammarian. Thus, while regular students of ON
have traditionally studied Snorri's language, as it is the language
of his works (containing what most students are interested to learn)
and other contempory works of interest, linguists and academics will
use the same standards applied in other dead languages, demanding a
correct writing-system - thus, as the first grammarian's system is a
'correct' one, accurately representing his own language, this is now
the standard for academics. Modern etymological works, textbooks,
etc. (and other contexts where correct citation is required), are no
longer using the standard spelling printed for editions of Snorri,
for instance. Instead, the first grammarian's spelling is used. So,
the situation is a bit strange. For instance, Finnur Jónsson used ON
normalized early 13th cent. in his popular, and excellent, edition
of Heimskringla, but used the first grammarian's spelling (and older
forms) in academic works (with commentary, etc. - such as his work
on Hávamál, where in the text itself it is called hávamó,l). So, in
reality, this situation is quite old. About nasals: these are shown
by the edition of a super-script dot over the vowels, as the first
grammarian advises - thus, there are 2 series of the 9-vowels long
and short (36 total), one with and one without the nasalizations. In
modern practice, linguists place the nasalization-dots according to
Germanic Linguists - that is, where nasalization would be inherited.
As this works, no problems here. However, as ON short nasal-vowels
always occur where the nasalizing consonant is preserved, there is a
leading school of thought which simply leaves the dots out for the
short vowels, seeing it as unnecessary (a practice I also follow).
Furthermore, it is acceptable to leave out the nasalization dots for
long vowels as well, as it makes citation easier, more practical and
avoids script-problems, but this is avoided in technically 'correct'
contexts. Lastly, the practice of not writing ð is consistent with
the first grammarian's recommendations, who himself wrote only þ, as
did other writers of his time. Of course, he probably had never seen
ð before, as it is believed to have originated in a some Norwegian
monestary around 1200 (Benediktsson, etc.) and spread thence. Modern
academics, in writing only þ, aren't agreeing with the grammarian by
not writing ð, but are reflecting his own writing-style and the fact
that if one writes ð, then one should also write reverse ð (ebh) for
f in medial positions (fara, hafa/habha), which character was really
not used by the Catholic writers themselves, is not found in any ON
standard texts, and has no use by tradition. Sort of a default to þ,
which is, interestly enough, how ð is always written in runes. So,
there are some interesting issues here, indeed.

> Anyway, I got bogged down eventually on a certain details, and now
your posts on Norwegian dialects have opened up whole new cans of
worms :-) E.g. it's tricky thinking of a test for the pronunciation
of 'll' at any given stage in the language, since whatever the
pronunciation, this wouldn't necessarily affect rhymes.

True, but actual ON speakers said -ll-, not -dl- etc., and this is
shown conclusively by runic writing, where *-tl- for -ll- does not
occur. So, as far as learning ON is concerned, -ll- is the only one
folk need to learn. If modern descendants of ON folk chose to say -
dl- instead, as that is the majority pronounciation in all the west
norse areas where descendants of ON (west) persons currently live,
then that is another question ;) 'Standard' pronounciation of ON (at
least the ON West variety) by ON descendants is, essentially, a
question of familiarity, practicality, often compromise. It really
has nothing to do with ON learning for historical reasons. No one is
saying that we can't pronounce ON as we choose to. If a 'ban' were
implemented against our modern pronounciation, then the ON situation
would degenerate into an aboriginal rights conflict between scholars
or creative anachronists and the actual ON descendants, a bizarre
situation that is, fortunately, never going to happen ;)

> Is it possible to tell whether miðli > milli = midli all along?
How old are spellings like 'valla' for 'varla'?

The change rl>ll is post Shorri, later reversed. miþli/miðli is the
only correct form (originally dative, against accusative á meþal)
for the first grammarian's language, but Snorri would likely have
said 'milli', but been familiar with 'miðli' as a form occuring in
the language of the oldest generation (or perhaps just poems).

> Of course, it would also be good to pin down the details of the
language of the previous century, and see texts where all the
distinctions that were later lost still in place, including nasalised
vowels!

Well, you have the consensus of modern ON academics in electing to
cite forms, standardize pieces from oral tradition, etc. in the 1st
grammarian's spelling (and language) ;) The fact that, essentially,
nothing separates this language (aside from a few contractions, etc)
from the language of the pre-christian period (that is, the last ON
west norse one) is not going to hurt you either, as this actually
_is_ the period ON students are most interested in, generally. So,
one can have it both ways: 1)learn Snorri's language (and standard
ON orthography) 2)use the first grammarian's in academic contexts.
More later ;)

-K


> Anyway, early 13th c. (unless I'm mistaken) would mean distinct ø
and
> hooked o, distinct long ø and æ, but á and long hooked o fallen
> together as /O/ (the pronunciation of long hooked o), and e and
hooked
> e fallen together (but I'm not sure whether the resulting value was
> open or close), and probably ø1 and ø2 having fallen together.
Again,
> I'm not sure what the value of resulting sound would be -- maybe
open,
> since that would make it closer to hooked o, which eventually
merged
> with it -- but maybe it's not possible to be so precise?
>
> According to Gordon, a 13th century date would mean [v] instead of
> earlier [w]. He suggests that the phoneme went through a
transitional
> phase of being a voiced bilabial fricative in the 12th century.
> Evidence for 'v' becoming a fricative comes from the confusion of
> medial 'v' and 'f'. But when exactly do manuscripts first start to
> show this confusion. I'm also wondering about the proninciation of
> 'f' at this stage. Could it still have had a bilabial
pronunciation
> in the 13th century? I think Adolf Noreen mentions spellings such
as
> 'røfr' for 'refr', which might imply rounding of the vowel caused
by a
> bilabial 'f'. But I'm not sure when and where these forms are
from.
>
> A related dilemma: what about the later changes exemplified
by 'kveld'
> > 'kvöld' and 'váru' > 'voru'? Could these mean that 'v' was
still
> either [w] or a voiced bilabial fricative in these positions at the
> time of these changes?
>
> Presumably there was no major difference in quality between long
and
> short vowels at least up to c. 1200 when the lengthenings happened
in
> words such as 'ulfr'. Similar changes in early Middle English
suggest
> that the tense/lax distinction in /I/ : /i:/ and /U/ : /u:/ found
in
> many Germanic languages isn't an inherited feature. According to
> Stefán Karlsson /i/ and /u/ were lengthened and other vowels either
> lengthened or diphthongised before /ng/ about 1300. If that's
right,
> I guess, short /i/ and /u/ would have still been tense in all
stressed
> positions at least up to then. But is this the only possible
> interpretation of the evidence? I was curious to read at that link
> you posted about similar behaviour of vowels in Sognamálit. But
what
> might that imply, a tendency the settlers brought with them, or a
> later innovation at a time when close contacts were maintained with
> the mainland and with these regions in particular?
>
> LLama Nom
>