I asked Eysteinn about this one. Here's his reply:

> Láta is a verb. Ég skal láta hefna hans = I shall let him

be avenged = I shall have somebody kill his killer. I have
never seen or heard "lát" used of a man's death by violence,
only of a natural death. It would be in the singular anyway.

> BUT - once again you are using this odd text, which seems

to differ a lot from all the regular ones. This "láta"
does not occur in the ÍF text, and no variants are listed.
All the translators seem to ignore it.

> As I've already mentioned, I don't have Faarlund, and don't

know how reliable he is, but personally I see nothing wrong
with this order of words, if we accept láta as a part of the
text.

--- In norse_course@yahoogroups.com, "llama_nom" <600cell@...> wrote:
>
>
> > enda skal eg hefna hans láta
> > indeed I shall avenge his death
>
> I'm not sure about this, but I wonder if 'láta' is the verb: "I shall
> have him avenged." i.e. 'ég skal láta hefna hans'. Otherwise,
> wouldn't the noun 'lát' "death" be singular here? 'hefna' takes
> genitive for the person to be avenged. Compare:
>
> að þú mundir vilja láta hefna hirðmanns þíns
> "that you will want to have your courtier killed"
>
> Vildir þú láta hefna þeirra?
> "Do you want to have them avenged?"
>
> Hún mun vilja hefna láta Bolla bónda síns.
> "She will want to have her husband Bolli avenged."
>
> On the other hand, this interpretation of 'enda skal eg hefna hans
> láta' seems to contradict Faarlund's statement in Old Norse Syntax, p.
> 163 [ http://tech.groups.yahoo.com/group/norse_course/message/6750 ].
>
> "When there are two non-finite verbs in the sentence the normal order
> is auxiliary verb-main verb in accordance with the VO (Verb Object)
> pattern (1); it is not uncommon however for the non-finite auxiliary
> to follow the main verb (2). [...] If the main verb in such a
> construction has a complement, this complement may precede its head,
> yielding a consistent OV word order (3). [...] The complement may
> also follow its auxiliary (4), but it never follows the main verb
> directly (5). [...] The reason this is disallowed may be that the
> main verb plus the auxiliary was reanalysed as one verbal head after
> the OV (Object Verb) order was no longer productive. The
> main-auxiliary collocation consequently behaves as one syntactic word,
> which allows nothing to intervene." (Faarlund: The Syntax of Old
> Norse, 162-163, 8.4.6 "Head-complement order").
>
> 1. mun eigi vilja af þér taka
> 2. því er þú vilt spurt hafa
> 3. þú vilt honum þjónat hafa
> 4. hann mun ráða vilja ferðum sínum
> 5. * hann mun ráða ferðum sínum vilja
>
> So if 'láta' was a verb, our example, 'enda skal ek hefna hans láta',
> looks rather close to (5), which Faarlund says doesn't occur.
>
>
> > Þar var vanur að ganga hafur um túnið
> > There a goat was accustomed to walk about the home-field
>
> Yes, I think that's right: "there a goat was accustomed to walk" = "a
> goat was accustomed to walk there" -- with "there" having its full
> force referring to the location, rather than "there was a goat..."
> There's something in the Old Norse Online Course about this, about
> 'þar' always having that full local sense in Old Norse, never the
> reduced 'expletive' sense that 'there' can sometimes have in English,
> as in "there was once a king."
>