Heill Konráð!

> Incidentally, the form 'þót' above

is most correct here, over against 'þótt' and uncontracted 'þó at'.
The contracted form would occur here, as in verse generally, and the
form þótt is from þót, which form is overwhelmingly dominant in the
homilies, against rare þótt (a sound-analogical form) - it parallels
the correct 'þít/þvít' (þí/því at uncontracted).

So you don't think this could just be the same change that lengthened
/t/ after a long vowel, as in 'sátt', the 2nd past sg. ind. of 'sjá',
or 'fátt', nom.sg.neut. of 'fár'?

> The +i

in 'séi' above is correct, and known to be so (both historically and
confirmed by meter 10th cent.), but is invisible in runes, as runic
writers just show 'i' for 'éi'. The Homilies have only one isolated
example of +i: séim (1st pl. pres.subj.), against séþ, sé (3rd sg. &
pl.), etc., but it wouldn't occur at all if it had been completely
dead. Gutniska still has it throughout, as would 10th century Norse.

The same contraction can be seen in Old English, cf. Beowulf 2649b
'þenden hyt sy', alliteration on 'hyt', and 'sy' = 'síe'. Either the
word was two syllables when the poem was composed, or the poet was
aware of lines like this that had survived in the tradition from the
days when 'síe' did still have two syllables. Or else I'm mistaken
and it's not a C-type line at all...

> Incidentally, the

poem Ynglingatal is not in fornyrðislag, but in a meter whose name I
do not know. It has pairs of alliterating lines throughout, the 1st
with 3 syllables and the 2nd with 4 - cleary, this must have been
one of the ways the master poets of the Viking Age varied/developed
their inherited fornyrðislag.

kviðuháttr? (Like Egill's Sonatorrek.)

> kópir afglapi : es til kynnis kømr
> Here we cannot simply eliminate a 'k', as such change never occured

in the language,

With 'fór ek einn saman' I was thinking of line where the initial
finite verb is unstressed, but in this example there's no escaping it:
the verb has to be stressed because of the alliteration. Unless the
lines were the other way around? But if this is common, I guess
there's no need to assume that it wasn't acceptable to the poet.

> So perhaps 'áudags þûhta mis' is correct after all. PN would have

rendered 'mis galeikáiþ' the same: *mez galîkêþ. It wouls also agree
on non-suffixing of the reflexive pronouns, of course, being even
that much more like Gothic in view of the above.

Well, the point I was trying to make -- probably not very clearly, as
I tend to get distacted and go off on tangents; you might have
noticed... :-) -- is that although the Gothic verb normally requires
dative for the experiencer (just as in ON), the dative pronoun is left
out where it's reflexive + infinitive "I think myself..." (ik þugkja
... +inf.), unlike ON, unless the infinitive clause is replaced by an
ei-clause with the subordinated verb in the subjunctive. The key
examples are:

jabai hvas anþar þugkeiþ trauan in leika, ik mais (Php 3:4)
"if any other man thinks that he (himself is able to) trust in the
body, I (do) more."

sahvazuh izei usqimiþ izwis, þuggkeiþ hunsla saljan guda (J 16:2)
"whoever kills you will think that he is offering a sacrifice to God."

> If you want to see some very 'incorrect' ON that still mostly

makes perfect sense, then Jón Helgason translated some lines into ON-
archaized Modern Icelandic direct from Wulfila. Actually, I find the
examples absolutely wonderful, and have even thought about actually
continuing them further. While 'wrong', they really give us insight
in ON, Go. and the whole issue of their relationship, history and
the language changes involved. Facinating stuff. In fact, I'll post
some examples next.

Sounds intriguing! I look forward to seeing that.

> Ok, so I guess what I am asking is why:
> *áudags ik þûhta (which does sound very good, by the way)
> instead of:

> *áudags ik þûhta mis (wisan)
> *áudags þûhta mis ik (wisan)
> *áudags þûhta mis wisan (where 'ik' is implied)

> What about the *mis (to whom it seems)? Actually, venturing a guess

from the gut:

> *áudags þûhta mis ik (wisan left implied) - option 3 above (of 4)

> That one feels rights. A parallel 5th option is:

> *áudags þûhta mis wisan (where 'ik' is implied)

As far as I can see, all the relevant examples (not that there are
many of them) just leave out the dative pronoun when 'þugkjan' is used
reflexively with the infinitive. I don't know enough about modern
syntactical ideas to know whether there is some theoretical reason why
the dative pronoun *couldn't* be used in such a construction; unless
it's something to do with the fact that the dative and nominative
pronouns would somehow be competing for the role of subject in a way
that struck Gothic ears as unseemly? Actually it would be really
great if we did have an example like one of those you reconstructed
because this would clear up some mysteries about the syntactical
status of these dative experiencers in Gothic. Sadly we don't have
any Gothic speakers to try it out on and see what they think, whether
they'd say "yes, we talk like that all the time", or "sure, we use the
dative pronoun if there's any doubt but usually don't bother if the
experiencer is clear from the context" -- in which case I'd prefer
*áudags ik þûhta mis (both from what I expect of Gothic and for the
sake of the metre) -- or whether they'd just laugh in out faces at the
absurdity of trying to use the dative experiencer pronoun reflexively
like that. For now, all we can say is that 'audags ik þuhta' matches
the construction that is actually attested where the Gothic translator
needed to express this idea, and that it probably isn't due to a
desire to match the Greek that the dative pronoun is left out, since
the pronoun appears elsewhere (in other contexts) where there is no
Greek model for it. So that's why I'd tentatively go with that. And
I think we're agreed the alternative with an ei-clause + subjunctive,
while it might for all we know have been possible in prose in Gothic
(unlike ON), is a bit to clumsy to work as poetry here.

The other possibility, not attested with this verb, but once with
'wenjan' "hope", and arguably once with 'rahnjan' "reckon, consider",
is to add the pronoun after the infinitive:

* audags þuhta mis wisan ik ??

Hmm, now I feel like I'm going round in circles... Anyway, I'm
learning some interesting stuff about Old Norse metre and language
from these discussions, even if we're still unable to settle on even
quite basic issues of Gothic syntax. Right, I'll see if I can figure
out what's up with that link...

LN