Jump to content

Repeated Words & Phrases (& Bug)


Randy Steffens Jr

Recommended Posts

1) What is the easiest way to find all verses with any sequentially repeated english word, not including articles and conjuctions?

 

2) What is the easiest way to do the same thing with repeated phrases?

 

I can accomplish #1 with a simple construct. However, in order to include repeated words which are separated by commas, I have to set "within" to 2 words, which floods me with over 1700 hits, most of which are instances of repeated articles & conjunctions. I need to figure out how to get the the result, minus articles and conjunctions.

 

Also, I think the construct search I am doing has revealed a bug. It seems to yield extraneous hits for some spelled numbers. See the attached photo, and notice how "twelve" and "fifteen" are shown to be hits, even though they appear only once. There are other such instances in the hit list. Can anyone duplicate this?

 

And as for #2, I haven't figured out how to expand this search to look for repetitions of 2 or 3 word phrases.

 

Thanks for any help!

 

Randy

 

EDIT: There are other extraneous hits in addition to numbers. Other non-repeated words appear like "Needlework", Ex 39:29 KJV and "brokenhanded" Lev 21:19 KJV.

Window.tiff

Edited by Randy Steffens Jr
Link to comment
Share on other sites

Try dropping the "within" tile and adding a "Inter" Tile with the Char. Tile embedded, like this:


post-407-0-07883100-1381236196_thumb.png

 

And, I do see the same bug with numbers in the KJVS (and NRSVS) (see 2 Kings 25:2 in the screenshot, or 2 Chronicles 13:21 in NRSVS) and with certain other words. e.g., in the NRSVS: needlework in Ex 39:29, crimson in Lev 14:51, scoundrels in 1 Kings 21:13, etc.

 

Edit: I don't see the same bugs in NIV11-G/K/

Edited by Lorinda H. M. Hoover
Link to comment
Share on other sites

Thanks Lorinda, very helpful.

 

How can I do the same thing with short 2-3 word phrases?

 

And how would I modify the search above to find two more widely separated (non-sequentally) repeated words in a verse, still minus articles and conjunctions?

 

Randy

Link to comment
Share on other sites

If you don't want articles, conjunctions, etc. in between repeated words, this works:


post-407-0-67012100-1381238902_thumb.jpg

 

I experimented, but didn't find a way to exclude articles, conjunctions, etc. from the word entries in the columns, and without doing so, a huge percentage of the hits are those kind of words. I'm guessing there's a way, just not one I know.

 

Lorinda

Link to comment
Share on other sites

In terms of 2 or 3 word phrases, I need more details about what you are looking for.

Link to comment
Share on other sites

In a Greek (or Hebrew) tagged text it's easier to do what you want:


post-407-0-45570200-1381239307_thumb.png

 

Although there's a bit of a bug, here, too. See Matt 4:24, where δαιμονιζομένους is listed as a hit, even though it only appears once. This is (I think) because it is "double tagged" because the form can either be middle or passive.

 

Lorinda

Link to comment
Share on other sites

Man ! a grammatically tagged english text would help. But try the attached.

 

post-32023-0-92804600-1381242483_thumb.jpg

 

Thx

D

Link to comment
Share on other sites

I'm sorry Randy, I was in a rush this morning. My suggestion above needs an INTER between the third and fourth words to make then phrases consecutive. I used the INTER char element Lorinda used above. I'm surprised to see there are 23 such occurrences, 17 in the OT.

 

Removing the INTER there are a far greater number.

 

Thx

D

Link to comment
Share on other sites

  • 2 weeks later...

Thanks Daniel and Lorinda! This is very, very helpful! Does anyone know if there is a way to exclude articles, conjunctions, etc. from the word entries in the columns of English construct searches?

 

Randy



And do I need to do anything further to report the bug above?

 

Thanks!

 

Randy

Link to comment
Share on other sites

Hey Randy,

 

My approach to bugs is that if you have a repro. test case then you're good. If anything else is needed I'm sure they'll ask for it.

 

thx

D

Link to comment
Share on other sites

You can exclude words specifically in the columns by setting the word value field to *@-(the, but, and) etc. But without a tagged text I don't know a simple way. In the two 3 word phrases search you just need this restriction on one of each pair of AGREEing words. Another thing I note here is that you often see 4 or 5 word phrases highlighted which makes sense because within 4 words there are two consecutive 3 word phrases. But in Gen 3:8 "of the Lord God" God is not highlighted. I do not know why not. It makes me wonder how the search is done - I wonder if the result is the same on a non-key numbered text. I think this may have been my fault with the exclusions but I'm still uncertain exactly how this search is performed.

 

thx

D

Edited by Daniel Semler
Link to comment
Share on other sites

Alright and here is my masterpiece - well that's probably overstating it.

 


post-32023-0-84525300-1382069113_thumb.jpg

 

Thx

D

Link to comment
Share on other sites

Thanks Daniel, that is excellent. Amazing work! Appreciate your help very much!

 

Does anyone know how to do this in Hebrew?

 

Thanks,

 

Randy

Link to comment
Share on other sites

Try something like this:

 


post-407-0-15197600-1382151877_thumb.png

 

Note that this allows for any number of words between the two identical phrases.

 

It also may miss some repeated phrases that involve particles (conjuctions, articles, etc.) If you want to allow for that, some changes would have to be made.

 

I tried this:

 

post-407-0-94385000-1382152470_thumb.png

 

But it pulls up the same 6 verses, so I'm not sure I constructed it right.

 

Lorinda

Link to comment
Share on other sites

Hebrew ? I have to bow out - I am just learning the alphabet.

Link to comment
Share on other sites

Thanks Lorinda, that's most helpful. How can I learn more about finding repeated Hebrew phrases that involve particles (conjuctions, articles, etc as you showed in screenshot #2 above) seeing that the changes you made still pull up the identical 6 verses?

 

Thanks!

 

Randy

Link to comment
Share on other sites

Part of the problem with my second attempt above is that I forgot to change most of the "within" tiles from 1 to 3. That's why more verses didn't show up. In Accordance's Hebrew searches, prefixed and suffixed articles prepositions and pronouns count as "words" So I needed to leave room for prefixes and suffixes as well as other particles (if that's what you wanted).

 

I tried this today and got more interesting results:

 


post-407-0-58808700-1382457819_thumb.png

 

If you look at the results on the right, you'll see that particles don't get highlighted even if they are identical (for example, the אֶל in the topmost result, Gen 4:8) This is because I excluded all particles in each of the Lex columns. I did this because when I didn't have the NOT Particle tile in the Lex columns, I was getting lots of hits where one of the identical words was an article or the like. You could refine the NOT Particle tile by specifying which type you want to exclude.

 

Similarly, in the INTER tiles, you could specify only certain particles and/or suffix classes that may come in between.

 

Constructs are one of those things that I have to just play with to get them "right."

 

Lorinda

Link to comment
Share on other sites

  • 1 year later...

Hi All,

 

Here’s an outline of a pattern I’d like to investigate. I call it Fivefold Chiastic Epimone. Whilst I’ve found a couple of references in a number of commentaries to Commoratio / Epimone, (eg. Anders Eriksson) and Chiasm (Kenneth E Bailey), I haven’t anywhere found the pattern I illustrate below described as a formal rhetorical structure and formula for interpretation / translation.

 

See what you think:

 

From Matthew, a very formal, minimal version of fivefold chiastic epimone,

 

A - μη κρινετε #1

B - ινα μη κριθητε #2

C - εν ω γαρ κριματι #3

B' - κρινετε #4

A' - κριθησεσθε #5

Matthew 7:1-2

 

James presents a more complex pattern in which the readers are kept waiting

 

μη καταλαλειτε αλληλων αδελφοι

A - ο καταλαλων αδελφου και κρινων #1 τον αδελφον αυτου

B - καταλαλει νομου και κρινει #2 νομον

C - ει δε νομον κρινεις #3

B'- ουκ ει ποιητης νομου αλλα κριτης #4

εις εστιν ο νομοθετης

ο δυναμενος σωσαι και απολεσαι

A' - συ τις ει ος κρινεις #5 τον ετερον

James 4:11-12

 

Paul here has a sevenfold chiasm very close to James and Matthew

 

A - Διὸ ἀναπολόγητος εἶ, ὦ ἄνθρωπε πᾶς ὁ κρίνων· #1

B - ἐν ᾧ γὰρ κρίνεις #2 τὸν ἕτερον,

C - σεαυτὸν κατακρίνεις,#3

D - τὰ γὰρ αὐτὰ πράσσεις ὁ κρίνων. #4

C' - οἴδαμεν δὲ ὅτι τὸ κρίμα #5 τοῦ θεοῦ ἐστιν κατὰ ἀλήθειαν ἐπὶ τοὺς τὰ τοιαῦτα πράσσοντας.

B' - λογίζῃ δὲ τοῦτο, ὦ ἄνθρωπε ὁ κρίνων #6 τοὺς τὰ τοιαῦτα πράσσοντας καὶ ποιῶν αὐτά,

A' - ὅτι σὺ ἐκφεύξῃ τὸ κρίμα #7 τοῦ θεοῦ;

Romans 2:1-3

 

Paul mixes it up further adding an alternative and gentler verb of chastisement rather than condemnation to encourage his readers to follow his warning

 

A - ει γαρ εαυτους διεκρινομεν #1

ουκ αν

B - εκρινομεθα #2

C - κρινομενοι #3 δε υπο κυριου

B' - παιδευομεθα (#4)

ινα μη

A' - συν τω κοσμω κατακριθωμεν #5

1Cor.11:31-32

 

Once again the pattern of five κρινω verbs is used, though Paul makes a more complex argument by hiding krino within compound verbs and shifting to a semantically milder equivalent. The chiastic pattern is however delivered without a break.

 

John has yet another interesting take in which he leaves the readers to draw their own conclusion,

 

A - ὑμεῖς κατὰ τὴν σάρκα κρίνετε #1,

B - ἐγὼ οὐ κρίνω #2 οὐδένα.

C - καὶ ἐὰν κρίνω #3 δὲ ἐγώ,

B' - ἡ κρίσις #4 ἡ ἐμὴ ἀληθινή ἐστιν,

(A') - ὅτι μόνος οὐκ εἰμί, ἀλλ᾿ ἐγὼ καὶ ὁ πέμψας με πατήρ.

John 8:15-16

 

NRSV readily supplies the final word,

 

A - You judge #1 by human standards;

B - I judge #2 no one.

C - Yet even if I do judge #3,

B' - my judgment #4 is valid;

(A') - for it is not I alone who judge (#5),

but I and the Father who sent me.

 

This time John sets Jesus' own words (#3) as the standard for judgement,

 

καὶ ἐάν τίς μου ἀκούσῃ τῶν ῥημάτων καὶ μὴ φυλάξῃ,

A - ἐγὼ οὐ κρίνω #1 αὐτόν·

B - οὐ γὰρ ἦλθον ἵνα κρίνω #2 τὸν κόσμον, ἀλλ᾿ ἵνα σώσω τὸν κόσμον.

c - ὁ ἀθετῶν ἐμὲ καὶ μὴ λαμβάνων τὰ ῥήματά μου (#3)

B' - ἔχει τὸν κρίνοντα #4 αὐτόν·

A' - ὁ λόγος ὃν ἐλάλησα ἐκεῖνος κρινεῖ #5 αὐτὸν ἐν τῇ ἐσχάτῃ ἡμέρᾳ.

John 12:47-48

 

That's a quick run through on one verb. Since identifying the pattern when writing an essay on 1 Cor 11:17-34, I’ve found the same structure with εχουσια in a couple of Paul's discussions with the Corinthians and τασσω in his writing to the Romans on obeying the authorities ordered by God. The sequence seems to keep popping up, but I'd like to find a way to formally identify it and test writers inside and outside the NT tradition - e.g. do Philo or Josephus use a similar device, or, Aristotle?

 

I've found others besides and could really do with some help to set up a system for recognising these patterns where x = any repeated verb. If Paul, James, Matthew and John are all riffing off the same verb in very similar ways, then there's either an early saying in there or a rhetorical pattern that is used by those writers and potentially others. I'd like to find out which or both is correct.

 

I think this would offer a good opportunity to pursue a literary / rhetorical critical track in my NT Research Methods essay (MA Biblical Studies Research in Sheffield, England), perhaps also looking at the way the pattern of verbal repetition is or isn’t represented by translations.

 

I’ve posted some of the above observations on a facebook forum of Accordance users (LXX Psalms in a year) to find if anyone knows search methodologies within Accordance for identifying strings of words where root = x and the root occurs more than three times within a limited range of verses and Daniel Semmler has given some helpful pointers. I wonder if anyone here has both the fluency with Accordance and the time to teach a total beginner how to write search constructs.

 

I hope this is a viable model for investigation. Please let me know what you think.

 

Many thanks,

 

Pete

Edited by peterjsandford
Link to comment
Share on other sites

Hi Pete,

 

  First thanx for posting here with a much fuller explanation. It's so much easier to work with it here than FB.

 

  Second, not trying to fob you off but there are some excellent doc and podcast things you should read on this.

 

  http://accordancefiles2.com/helpfiles/Win11/Default.htm#topics/06_braa/construct_tab_graphical_search.htm

  http://www.accordancefiles2.com/podcasts/p19_constructsearches.m4v

  http://www.accordancefiles2.com/podcasts/p21_greekconstructs.m4v

  The index of all podcasts is at : http://www.accordancebible.com/Podcasts-Index

 

  Finally it is possible to find the examples you've identified with a search not dissimilar to the one posted on FB. But it also shows a number of false positives words with multiple roots. I am not sure this can be fixed actually, but I'll show you what I have and you can see what you think.

 

post-32023-0-56981300-1434163823_thumb.jpg

 

  To explain it a little.

 

  I used Lex items dragged from the left hand pallete because your examples contain both nouns and verbs from the same root. There are five as that is what you are looking for. The Agree elements also dragged and placed from the palette are set to root, meaning each lex must have the same root. (More on this in a minute because there is something funny here leading to false positives.) Finally I have added a WITHIN element and set it to 50 words between the first and last element. I did this primarily because you need to have chapter scope and that will lead to many false positives. I started with 100 but 50 seems enough. It will depend on how many words you think might be in such a construction. I have set the four Lex elements after the first to be * - they have to have something. That will mean it can be any word and the Agree will constrain the words possible by the root. Note that + preceding a word indicates the root. Look at some of the more basic podcasts and help pages on search to bone up on that background.

 

  False positives. There is one very particular false positive case which arises here. The problem comes from compound verbs, which have multiple roots. This example here is a good one :

 

ἀπεκρίθησαν δὲ αἱ φρόνιμοι λέγουσαι· μήποτε οὐ μὴ ἀρκέσῃ ἡμῖν καὶ ὑμῖν· πορεύεσθε μᾶλλον πρὸς τοὺς πωλοῦντας καὶ ἀγοράσατε ἑαυταῖς. ἀπερχομένων δὲ αὐτῶν ἀγοράσαι ἦλθεν ὁ νυμφίος, καὶ αἱ ἕτοιμοι εἰσῆλθον μετ᾿ αὐτοῦ εἰς τοὺς γάμους καὶ ἐκλείσθη ἡ θύρα.”
(Matthew 25:9–10 Novum Testamentum Graece (New Testament in Greek))

 

The problem is the root απο in ἀπεκρίθησαν. It has two roots in the tagging. The same happens later at ἀπερχομένων introducing the new root for "come/go". Ok I believe having typed this now I can fix this.

 

post-32023-0-53510200-1434164756_thumb.jpg

 

This is not perfect but at least it no longer skips along in a chain. It finds all your examples except John 12:47-48 as there are only 4 words with the right root in that case. I only ran it against the books containing your examples. I got 58 hits in case that helps you.

 

Hopefully this helps a bit.

Others will have more thoughts I trust.

And I have a construction to ponder.

 

Thx

D

 

  • Like 1
Link to comment
Share on other sites

Daniel,

 

Thank you for your generous investment in my question. I'm so new to Accordance that even the construct window is a discovery. The podcasts are helpful too.

 

Making a construct that looks like the one in your screenshot only turns up Luke 13:15 and 2 Cor 10:12 both of which feature false positives based on internal agreements. Are we doing the same search? I wonder if you could talk me through the steps. Does Accordance allowed saved constructs to be shared?

 

The 58 iterations you found, were they all on κρινω? Could you send me the list. That frequency of a construct, if expanded across the NT would surely go some way to justifying my hypothesis. My eventual aim would be to identify any instance of any 4+1 or 5 fold repetition of any verb root within any of the texts covered by Accordance.

 

Thank you again for your time and generosity on this question.

 

Pete

Link to comment
Share on other sites

Hi Pete,

 

  I've PM'd (Personal Messenger in this forum) you a message. You can't save constructs themselves but you can save a workspace and share that. I've sent you details in the PM.

 

Tx

D

Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...