MT-LXX Analytics

[Λύχνις Δαν]

Hi ya,

 

  I ran into the MT-LXX Analytics by accident and then started trying it out by checking when הֵיכָל was translated as οἶκος. I ran a simple MT-LXX merge and got 15 cases. I checked the MT-LXX Analytics for the HMT-W4 tab and that clearly shows all 15 cases.  Then I checked the analytics for the LXX tab. Here the translation is from οικος to various Hebrew words. But the analytics only show 9 cases. In addition it shows many more words which cannot possibly be translations of οικος. Is there something wrong here ? This is a partial of what I got as there nearly 2000 forms.

 

  First the 7 cases it mapped correctly. Note I have LEX first and then WORD :

 

οἶκος    house = 1967

...

...

...

‏הֵיכָל‎ = 7
   ‏הֵיכַל‎ = 7
‏הֵיכָל   ָא־0‎ = 2

   ‏הֵיכַל   ָא = 2

...

...

...

 

Now some examples that appear wrong :

 

                    וְ = 19
                    לְ = 3

                    עַם = 1

                    עפל = 1

 

Am I reading things incorrectly or is this expected ?

 

Thx

D

[Helen Brown]

I don't follow what you are doing. I searched LXX1 for οικος, and did an analysis and of course most occurrences are for בֵית but there are 7 for הֵיכַל in Hebrew and 2 in Aramaic.

 

 

[Λύχνις Δαν]

What I expected was that I would find the same number of translations for the Hebrew to Greek word and the Greek to Hebrew reported in each of the MT-LXX analyses. For some reason that is not the case which is puzzling to me. So yes I get the same 9 results you show ( I wasn't aware two were Aramaic but I'll look at that). But if you search for the Hebrew word first הֵיכָל I would have expected to find the same number of translations to Greek ֹοικος but I am not finding that. Instead there I find 15 cases of the Hebrew word translated as οικος. I'm trying to work out why that is. In addition there appear to be cases where the analysis shows οικος translates a word that simply isn't anything that could be construed as meaning house at all.

 

So you've done one half of the example I did. Now if you also do the LXX search for הֵיכָל and look at its analysis you should see the discrepancy I am talking about. You will find 15 cases of the translation of הֵיכָל to some form of οικος.

 

Thx

D

[Helen Brown]

OK, I see what you mean, and you have uncovered an inaccuracy in the analytics. That feature can only give you a rough idea, in any case. The final answer will only come by searching for yourself each verse of the LXX which is parallel to an occurrence of הֵיכָל. However, before that, you get much more accurate results using the brilliant MERGE command in the MT-LXX Parallel after performing both individual searches.

 

 

Now we have 19 verses with the results in both versions, which allies closely with the results of the analysis of the Hebrew search. These verses are a good basis for further comparison of the texts.

 

The developers will have to look for the reason that the LXX analysis misses so many hits.

 

See the Help files for details of setting up this search, or this podcast.

You are here: Biblical Research and Analyses > Language Study Tools > Using the MT-LXX Resources

[Λύχνις Δαν]

Yep I'd already done the MERGE which is why I knew exactly which translations ought to exist. I saw from the doc that the MT-LXX analytics  are only be approximate but figured this was a little too approximate.

 

While dev. is looking at it they may also want to check on the capitalization in the Greek also, which is not in the original text.

 

Thanx for your help

 

thx

D

[Joel Brown]

As Helen said, this is a very good approximation, but there will almost always be fluff.  Looking at one of the bad examples,  Gen 36:6.  MT-LXX says:

 

בית/ו τοῦ οἴκου αὐτοῦ

 

Accordance then tries its best to match which elements of the Hebrew word match which Greek words in the phrase, and unfortunately here, τοῦ became בית , and οἴκου became ו.

 

Statistically, getting ~25 / 2000 wrong is a percentage of 98.75%, which is pretty good IMHO.

[Λύχνις Δαν]

Hi Joel,

 

  I was wondering if it was done in this sort of manner.

 

  I am not sure how you determined that about 25 were wrong. But yes if that's all out of the almost 2000 οικος variants that would fine for an approximate algorithm overall. So I wonder why in the case I cited that 6 of 15 cannot be found. Perhaps they are disproportionately represented in the the cases which are harder to identify. I'll have to look over them. Good practice for my Hebrew I guess.

 

Thx

D