>Very interesting suggestion, thank you very much. Such a reading would make sense, the question is if we can make the grammar work. A plain "they didn't see anything" would be "ní-accatar ní" as co- is dropped whenever another conjunct particle precedes this verb. Provided we have a relative clause with antecedent of time, we would end up with nad-n-accatar ní unless I am mistaken.
>So far I see no way to squeeze such a reading from the mss. Unfortunately
The funny thing is that if we were sitting in a modern Irish class doing this translation, I'd know exactly what's happened here. The problem really resides in an oddity of English - the word 'nothing', which apparently was in the language prior to 900AD. But having a negative pronoun like 'nothing' is rare, most languages can only express the idea as 'they didn't see anything/something' rather than 'they saw nothing'. But when people go to translate such a sentence from say modern Irish into English, they will inevitably twist it into the 'nothing' form of sentence. Which is logical - until you notice that probably most of the class suddenly believes the word 'ni' means 'nothing' instead of 'something'. I've even heard people argue vehemently that it means 'nothing'. I guess people don't tend to do grammar analysis on the fly. And I suspect the fact that 'ní ' looks like the verbal negative has already set the brain up for expectation of a negative pronoun too.
I cant say that this is the case here, although this manuscript is apparently a relatively late one, so could this confusion be at play? would be rather fun if it was ...