Tutti Frutti Greek Manuscript Bundle & Computer Comparison

[Enoch]

Ideally Accordance would have all the Greek manuscripts in one module or bundle. And it would be possible to run a comparison on more than 2 mss. Of course "all" is probably not possible, but would it be possible to have a high percentage of the mss all in Accordance?

 

As to advancing from current comparison capacity, how much comparison would be possible? Right now if I bring up 2 mss, I can use the compare function, but I don't know if the results are available in numerical interpretation; i.e., percent of identity. Comparing manuscripts mathematically is an ideal, though obviously requiring a high degree of sophistication to be useful, as each manuscript will have different missing parts, lacunae. And one codex's missing parts will differ from another's. One would like some way of extrapolating what the comparison would be if there were no missing parts.

 

But the abstract idea is to be able to compare all the manuscripts mathematically and come out with results like:

 

X & Y are 90 percent identical and these 2 are the closest in identity of any two manuscripts.

Z & Q are 89 percent identical and these two have an identity rating of 23 (there are 22 mss which are mutually more identical than these out of a total of 6000 mss).

 

The degree of identity for any two manuscripts would be mathematically expressed as percents.

By comparing all manuscripts with each other this way, families could be established objectively.

Or has anyone already done this?

 

Using the Accordance compare function, I don't know if there is any way to expand the comparison beyond 2 modules (what is the maximum number of mss that can be compared simultaneously?);

but when I tried 4 codices in parallel, only the first 2 compared.

 

Obviously in many cases the comparisons would be reduced to individual book comparisons (Romans in X vs Romans in Y).

 

It would be nice to be able to compare more mss simultaneously. And it would be nice to have a huge bundle of texts with improved means of comparison.

Edited October 16, 2014 by Enoch

[Timothy Jenney]

Hi, Enoch!

 

Compare texts can be controlled in Preferences>Compare Texts. Just choose "Multiple Texts" instead of "Single Text." You'll be ready to go!

 

The analysis you'll have to do on your own.

[Enoch]

Hi, Enoch!

 

Compare texts can be controlled in Preferences>Compare Texts. Just choose "Multiple Texts" instead of "Single Text." You'll be ready to go!

 

The analysis you'll have to do on your own.

Much tnx for the info Timothy. I think there must be some limit on how many parallel windows one can open & thus on how many texts one can compare?

 

I changed my preference to multiple texts, but then the first text was no longer compared. I had 4 parallel texts: UBS4-T & 3 uncial mss. With single text, UBS4-T was compared with the first uncial only. When I changed to multiple, only the 3 uncials were then compared, not UBS4-T.

 

As to the analysis, I don't think anyone can do this on his own. It requires a computer program with some kind of mathematical result. Writing such a program is beyond my capabilities.

 

I did google on this, hoping to find that someone had compared manuscripts with computer and come up with some quantitative results; but I did not find it.

 

BTW, is there anyway to change the highlight result upon comparison? When I do it, the highlights look more like line-out than highlight.

Edited October 16, 2014 by Enoch

[Abram K-J]

As to the analysis, I don't think anyone can do this on his own. It requires a computer with some kind of mathematical result. Writing such a program is beyond my capabilities.

 

I did google on this, hoping to find that someone had compared manuscripts with computer and come up with some quantitative results; but I did not find it.

 

Logos's text comparison feature does what you are asking about (i.e., give you a percentage of closeness/divergence), so it's at least possible for a program to do it. I have no idea what goes into that percentage, how it's calculated, etc.

[Timothy Jenney]

Compare texts, when set to "Multiple Texts," uses the leftmost text as the base, then marks all others where they diverge from it. That's the reason the first text has no marks on it.

 

You can change the colors of the highlighting in preferences, but not its style.

 

[Chuckle] Enoch, I wasn't trying to be snide when I suggested you would have to do your own analysis. I just meant Accordance would not do it for you. Frankly, I have no idea how you would count the various differences, let along weigh them for statistical analysis. Interesting problem, though...

 

Have you considered designing your research around CNTTS? It includes many, many texts and has extensive documentation. With its multiple fields, it can be search easily and specifically. Search results can also be displayed as a Hits Graph or a Table, the latter even at a chapter level detail.

[JonathanHuber]

On a barely related note, you come up with some of the funniest thread titles.

[Abram K-J]

Ha! I, too, was admiring the name of this potential bundle.

[Enoch]

Ha! I, too, was admiring the name of this potential bundle.

Thanks. I admit to being influenced by a little paper back book that I bought when studying Greek at the U of Minnesota:

 

Tutti i Verbi Greci. As I recall some Italian made a list of classical Greek verbs in their inflected forms, parsed. Thus one could identify rare verbs or odd forms. This might be a valuable work for Accordance to key to for non-Bible Greek works, though methinks Accordance is deficient in that category. Having the Greek corpus is a great lexical asset -- as if one could highlight a Greek word in the NT and have the entire ancient Greek corpus searched with results like a concordance.

 

Thanks Abram for your comment above on Logos. After I have invested so much in Accordance, I hate to have to go to Logos to get what I want. Perhaps Accordance will surpass Logos on this in the near future so I don't need to do duplicate purchase of Greek texts. I reluctantly have recently invested in Logos to get much more of the Greek corpus (especially Perseus) -- their Community Pricing bid system is letting me acquire a lot of Loeb classics incredibly cheap -- I find I can get like 8 volumes for $4! Of course one expects that between these 2 systems, in some aspects each will be better than the other.

Edited October 17, 2014 by Enoch

[Enoch]

Compare texts, when set to "Multiple Texts," uses the leftmost text as the base, then marks all others where they diverge from it. That's the reason the first text has no marks on it.

 

You can change the colors of the highlighting in preferences, but not its style.

 

[Chuckle] Enoch, I wasn't trying to be snide when I suggested you would have to do your own analysis. I just meant Accordance would not do it for you. Frankly, I have no idea how you would count the various differences, let along weigh them for statistical analysis. Interesting problem, though...

 

Have you considered designing your research around CNTTS? It includes many, many texts and has extensive documentation. With its multiple fields, it can be search easily and specifically. Search results can also be displayed as a Hits Graph or a Table, the latter even at a chapter level detail.

But Timothy, isn't CNTTS just a textual apparatus?

And does it cover mere spelling variations? (like epsilon iota for long iota)?

Perhaps there would be some way to convert that data into the sort of comparison I have in mind, if the variants are exhaustive and cover even the trivial.

[Jonathan C. Borland]

The results of such a feature would not be meaningful unless the texts were already tagged, letting the program know that misspelled words (by far the majority of all differences in any raw transcription of any manuscript) were not meaningful. Also, the program would have to know, e.g., that the omission/addition of a definite article or a particle was less important than other additions/omissions, that a sentence accidentally written twice was not as important as the addition of a new sentence, that the omission of DE after a word and addition of KAI before it was really just a single alteration (as opposed to two), that the addition of EAN with the subjunctive instead of a future indicative was really just a single alteration (and not a very important one), that certain disagreements among manuscripts might actually be agreements (e.g. two mss diverge from the base text with the same word but in different tenses or moods), etc. Not to mention, if someone went to all the trouble to tag every single word in a raw ms transcription, he probably also would have evaluated the differences and made his findings available to the academic world. In the end, the importance of the axiom "differences must be weighed and not counted" cannot be overemphasized.

[Enoch]

The results of such a feature would not be meaningful unless the texts were already tagged, letting the program know that misspelled words (by far the majority of all differences in any raw transcription of any manuscript) were not meaningful. Also, the program would have to know, e.g., that the omission/addition of a definite article or a particle was less important than other additions/omissions, that a sentence accidentally written twice was not as important as the addition of a new sentence, that the omission of DE after a word and addition of KAI before it was really just a single alteration (as opposed to two), that the addition of EAN with the subjunctive instead of a future indicative was really just a single alteration (and not a very important one), that certain disagreements among manuscripts might actually be agreements (e.g. two mss diverge from the base text with the same word but in different tenses or moods), etc. Not to mention, if someone went to all the trouble to tag every single word in a raw ms transcription, he probably also would have evaluated the differences and made his findings available to the academic world. In the end, the importance of the axiom "differences must be weighed and not counted" cannot be overemphasized.

Jonathan, I don't quite follow your reasoning. If one is doing textual criticism, then "misspellings" (anachronistic comment?) IMOH, would be the most meaningful variations of all. It seems to me, that a very mechanical mathematical comparison, without the pollution of modern correctors, would be exactly what is needed.

[JonathanHuber]

If one is doing textual criticism, then "misspellings" (anachronistic comment?) IMOH, would be the most meaningful variations of all.

 

Do you think that a spelling difference of word is more meaningful than addition, deletion, or replacement of a word?

[Jonathan C. Borland]

If a simple misspelling or itacism is counted the same as other alterations, then Vaticanus would rank far closer to the TR than it would to Sinaitucus. This in itself demonstrates the meaninglessness of raw counts of discrepancies between witnesses.

[Abram K-J]

 

Do you think that a spelling difference of word is more meaningful than addition, deletion, or replacement of a word?

 

Sometimes!

[Alistair]

I understand when misspellings give different words with different meanings, I think this is what matters to Enoch.

But what about when you have two different words with the same meaning, e.g "dauid" and "daueid" for 'David'? Here the difference is surely trivial?

[Peter Brylov Christensen]

I understand when misspellings give different words with different meanings, I think this is what matters to Enoch.

But what about when you have two different words with the same meaning, e.g "dauid" and "daueid" for 'David'? Here the difference is surely trivial?

 

You know, that reminds me of a very interesting case with the name of Daniel - it is usually spelled "דניאל", but in Ezekiel 14,14 it is for some strange reason spelled "דנאל" (cf. Ez 14,20 and 28,3). While it may just seem like a typo or merely an alternate spelling, it has been suggested that the author of the book of Ezekiel does not refer to the Biblical figure of Daniel at all, but rather Danel (Dnil) from the Ugaritic Story of Aqhat. Of course, when you look at the vowels, it seems that the Masoretes have tried to "fix" this problem by writing "דָנִאֵל" so that it reads "Daniel" and not "Danel". While it's not 100% possible to determine for sure, Ugaritic literary material has found its way into other Biblical texts in other more certain instances - Isaiah comes to mind here - so it's definitely plausible that we're dealing with the Ugaritic hero and not the prophet in this case.

 

At any rate, even a slightly different spelling of a name may be of importance. But I do agree that the different spellings of David yield no real interesting results in this regard though..

Edited October 24, 2014 by Pchris

[Alistair]

 

You know, that reminds me of a very interesting case with the name of Daniel - it is usually spelled "דניאל", but in Ezekiel 14,14 it is for some strange reason spelled "דנאל" (cf. Ez 14,20 and 28,3). While it may just seem like a typo or merely an alternate spelling, it has been suggested that the author of the book of Ezekiel does not refer to the Biblical figure of Daniel at all, but rather Danel (Dnil) from the Ugaritic Story of Aqhat. Of course, when you look at the vowels, it seems that the Masoretes have tried to "fix" this problem by writing "דָנִאֵל" so that it reads "Daniel" and not "Danel". While it's not 100% possible to determine for sure, Ugaritic literary material has found its way into other Biblical texts in other more certain instances - Isaiah comes to mind here - so it's definitely plausible that we're dealing with the Ugaritic hero and not the prophet in this case.

 

Agreed, the Daniel/Danel reference is an important difference.

[Enoch]

If a simple misspelling or itacism is counted the same as other alterations, then Vaticanus would rank far closer to the TR than it would to Sinaitucus. This in itself demonstrates the meaninglessness of raw counts of discrepancies between witnesses.

 

 

Do you think that a spelling difference of word is more meaningful than addition, deletion, or replacement of a word?

 

I am thinking that it is not best to speak of Vaticanus ranking closer to the TR than to Sinaiticus. I think it is more meaningful to speak of Vaticanus ranking closer to one of these: 1, 1^rK, 2^e, 2^ap, 4^ap, 7, 817 than to Sinaiticus. So you are saying that based on spelling, Vaticanus is closer to each one of the mss that Erasmus consulted than it is to Sinaiticus?

 

I am thinking that the best way to compare texts in making any attempt at determining genealogy of texts would be a purely mathematical comparison. It seems to me that it really is hard to know how much tampering or "improving" of the text was done by a given copyist, which tamperings would vary quite a bit depending upon the temperament and beliefs of copyists. And little of this is yes vs no binary behavior. I mean a given copyist might be rated on a scale of 0-10, 0 meaning he copied just what he saw willy nilly, and 9 meaning he had the hybris to know what Paul originally dictated better than Tertius, and 10 meaning he didn't care what Paul said, as the copyist believed he had the Spirit & knew God's truth better than Paul. And then there may be a further confounder by the possibility that a copyist didn't copy what he read at all, but copied what he heard when a reader read from an older ms. And determining matters like this seems to me to involve impossible subjectivity. Thus, back to mathematics.

 

At the very least, I for one would like to know the results of a purely mathematical comparison which doesn't distinguish spelling from any other sort of variation. In fact, as it occurs to me, the first step in comparing might be just to look at spelling differences & see what results from only comparing spelling variations.

[Enoch]

Agreed, the Daniel/Danel reference is an important difference.

Do you think that the original hand that wrote the Ezekiel verse did not have to have דנאל; what I mean is would not the yod have to be a later addition to a text that originally was utterly vowelless? What I mean is, did matres lectionis exist when Ezekiel wrote? Thus the existence or absence of a yod in Daniel would be irrelevant to what Ezekiel actually wrote anyway? Am I correct that Ezekiel always writes David without the yod -- or has someone postulated a Ugaritic DaVD?> Edited October 28, 2014 by Enoch

[Enoch]

I understand when misspellings give different words with different meanings, I think this is what matters to Enoch.

But what about when you have two different words with the same meaning, e.g "dauid" and "daueid" for 'David'? Here the difference is surely trivial?

I understand when misspellings give different words with different meanings, I think this is what matters to Enoch.

But what about when you have two different words with the same meaning, e.g "dauid" and "daueid" for 'David'? Here the difference is surely trivial?

No, I was thinking that distinctive spellings could define text family, as distinctive spellings would be the result of "just copying the text" [after the initial change was made]. I was thinking that our modern notion of correct spelling was probably not nearly as strong in antiquity, just as it probably was not even as recently as Early USA history. Modern English has to stick to a standard spelling because if we all wrote it phonetically, it would be very hard to read around the world (a Southerner would be spelling I as Ah.)

Edited October 28, 2014 by Enoch

[Peter Brylov Christensen]

Do you think that the original hand that wrote the Ezekiel verse did not have to have דנאל; what I mean is would not the yod have to be a later addition to a text that originally was utterly vowelless? What I mean is, did matres lectionis exist when Ezekiel wrote? Thus the existence or absence of a yod in Daniel would be irrelevant to what Ezekiel actually wrote anyway? Am I correct that Ezekiel always writes David without the yod -- or has someone postulated a Ugaritic DaVD?>

 

Interesting point! I went and had a look at it, and the author of the Book of Ezekiel always writes David as דוד except in 34,23 where it is דויד, oddly enough. So it's either a typo or there's no real orthographic consistency for the name of David in contrast to the name of Danel. In the Qumran scrolls, both דוד and דויד are used for the entire Old Testament - the latter twice as much as the former. Sadly, the Qumran scrolls do not cover the parts of Ezekiel where David is mentioned. That also goes for Danel/Daniel.

 

All in all, it is difficult to be sure whether we're dealing with the prophet or the Ugaritic hero in the Book of Ezekiel - however, given that Ugaritic material and mythology has similar parallels in the Old Testament and sometimes seem to have been quoted directly (Again, Isaiah comes to mind), it's not that farfetched.

 

*EDIT*

 

Here's the example I was thinking of:

 

Is 27,1: ביום ההוא יפקד יהוה בחרבו הקשׁה והגדולה והחזקה על לויתן נחשׁ ברח ועל לויתן נחשׁ עקלתון והרג את התנין אשׁר בים

 

On that day YHWH will punish Leviathan, the fleeing serpent, Leviathan the twisting serpent, with his great, heavy and mighty sword and he will slay the dragon which is in the sea.

 

KTU 1.5 col. I, lines 1-8: k tmḫṣ ltn bṯn brḥ tkly bṯn ʿqltn [[š]] šlyṭ d šbʿt rašm tṯkḥ ttrp šmm krs ipdk ank ispi uṭm ḏrqm amtm l yrt b npš bn ilm mt b mh mrt ydd il ġzr

 

As you (Baʿal) have killed ltn (Leviathan) the fleeing serpent, (as) you have finished off the twisting serpent, the seven-headed tyrant, the heavens will loosen when I tear you in pieces. I shall devour you, elbows, blood and forearms: you will certainly descend into the throat of mt (Mot), son of the gods, in the pit of the beloved of il (ʾEl), the hero.

 

Apart from the word serpent (נהשׁ vs. bṯn), all the other roots describing Leviathan are the same for both texts: It's pretty much the same description of the same sea monster. Of course there are some differences: in the Ugaritic mythology, "ltn" was a different name for the sea god "ym" (Yam), who was subdued and killed by Baʿal, the storm god. Then again, God also subdues the sea in the Old Testament

Edited October 29, 2014 by Pchris

[Peter Brylov Christensen]

Oh, and there is to my knowledge no David figure in the Ugaritic corpus (unless you count the root ydd, which means "beloved" just as דוד also can mean "beloved"!)