Not only did the 'b' survive after the introduction of
svarabhakti 'e', it did survive the extinction of the svarabhakti and
the 'r' in some dialects (like Dalecarlian).

kumr > kumbr > kumber > kumb (MI 'kemur')

As you probably know, the 'b' survived because people just didn't
care whether or not the 'b' was a "real" phoneme. People heard a 'b',
so when the '-e-' was introduced, they still heard a 'b' - or they
thought there should be a 'b' - so it suvived. I should mention that
the 'b' didn't survive into Modern Swedish time since the 'mb'
cluster was lost.


Funny stough:

In Northern Sweden (in Kalix not far from the Finnish border) there
is a dialect were they have turned 'mb' into 'pp' (and 'nd' into 'tt'
and 'ng' into 'kk' - lamb/land/langr > lapp/latt/lakk). Thus, 'kumbr'
should have become 'kupp' there, but I am not sure.


Serious stough:

In Mainland Scandinavia (excluding Denmark here) people consider
Icelandic accent to be arcane since it has no tones (i.e. lacking a
grave tone, i.e. rising tone). Did Old Norse have tones? Yes, it did.
In the standard Mainland Scandinavian languages we still
consider 'kommer' (MIce 'kemur') to be a word with one syllable (from
ON 'komr/kumr') since it has an acute tone! Words with acute tone are
old monosyllables, words with grave tone are old double syllable
words. Thus Modern Mainland Norse 'kastar' with grave tone comes from
an old double syllable 'kastar'.
My own dialect has often even preserved the original number of
syllables: ON 'kumr/komr/kemr' has become 'kum/kom/kjem'; ON 'brýtr'
has become 'bryt'; ON 'langr' has become 'laong/lång' etc with just
one syllable.
In mahy cases it is only tone that makes different words distinct. In
(northern regional) Swedish 'huse' with acute accent means 'the
house' (from ON 'hús it') and 'huse' with grave accent is
means 'house' in (old) dative case (ON 'húsi'). Obviously, by looking
at the ON forms one sees why there is an acute tone in the first
case, and a grave tone in the second. Of course, in ON 'hús-it'
and 'húsi' were pronunced by different tones as well.

For me - as a user of tones - the svarabhakti 'e' in 'drömber' is not
especially important. As long as 'drömbr' and 'drömber' are pronunced
with acute tone, there is no significant difference between them.

I would like to hear Hauk read from an old manuscript using
the "arcane" tones! They are by far more important than most
phonetical features of ON. Thus, any serious reconstruction of ON
should be made using the tones.
I have tried to make some recordings using the tones a month ago:
http://groups.yahoo.com/group/norse_course/message/3470
(Not perfect recordings, but I think you'll get the point)

>
> *sigh* Very well, I meant to say Old Gutnish. It is interesting,
> though, that the 'b' should survive even after a svarabhakti sound
> has been introduced between the 'm' and the 'r'.
>
> Kveðja,
> Haukur

Sklär,
SigwardaR