Inconsistency in tagging LXX to GNT

[Ken Simpson]

Hi, I think I have seen this before, and it may have been reported on the forums before, but I was just performing a comparison between LXX and GNT language when I noticed a very significant issue (IMHO).

 

The aorist of λέγω, εἰπον is tagged as such in the GNT (i.e. the lemma of εἰπον is reported as λέγω) but in the LXX the form εἰπον is also reported as the lemma.

 

Given the frequency of this word it would seem to be a very significant problem in doing INFER searches for example, or any lemma search for λέγω in fact.

I note that the lemma ὀράω is correctly reported as having the aorist form εἰδον.

 

It’s clearly not a bug, and I imagine it would be a big job to change, but it seems to me to be a significant issue.

 

Interestingly λέγω is recognised as the underlying lemma in Prov 25:7 where ῥηθῆναι is the inflected form in the LXX

Edited October 25, 2014 by Ken Simpson

[Helen Brown]

We are not the people responsible for tagging the LXX, we simply receive the fully tagged database. It's my guess that our scholars want to stick to their scheme and the LXX scholars prefer theirs. Unless they agree on how to tag these words, and one of them changes their entire system, we are left with the inconsistency.

[Λύχνις Δαν]

Interestingly it does seem to report λεγω as the root (though alas not in all cases either) and thus a root search may help. Of course you would have to know where and when you needed this. I guess that INFER's ability to drop or add a word may help with phrase matching.

 

tx

D

Edited October 25, 2014 by Daniel Semler

[Robert Holmstedt]

Those darned scholars and their differing systems!

[Ken Simpson]

Those darned scholars and their differing systems!

 

Amen!

 

Helen, thanks for the info. I am certainly for the system Rex and Dr Mounce has applied in the GNT.

 

I had it in my mind that Rex was also responsible for the LXX, but as you point out, that is not the case. Drs Taylor and Wheeler may well be unwilling to change, and may have good reason not to. But to me it seems inconsistent at least.

 

Daniel, yes infer can deal with a one word difference, but INFER compares lemmas, not roots, and if you have to drop λέγω as your term, then you are dropping a very significant term. It really does seem to question the thoroughness of an INFER search.

 

Oh well, that’s v sad.

Edited October 25, 2014 by Ken Simpson

[Helen Brown]

Maybe we can program this one case to be regarded as equivalents in INFER. Maybe not right now, but after the current round of releases.

[Ken Simpson]

Hi Helen, thanks. It took me this long to realise there was an issue, so I am sure that I can wait.

 

Thanks again.

[Anthony Pyles]

This has definitely come up before, and I think makes for a significant limitation not just in using an awesome tool like INFER but even in simple matters like using searches to generate vocabulary lists (e.g., a list of words that occur in LXX but not in GNT).

 

The difference in tagging has also been mentioned before. But surely there is a way to rectify this? E.g., add λέγω into the tagging info for εἶπον, etc., so that it's picked up in a search?

 

In the mean time, between the various GNT and LXX modules (GNT-28, Swete, etc.) is it possible to get a GNT/LXX combo that are tagged with the same lemmas?