Jump to content


Photo

Understanding Hebrew Word Counts


  • Please log in to reply
12 replies to this topic

#1 Abram K-J

Abram K-J

    Platinum

  • Active Members
  • PipPipPipPipPip
  • 1,419 posts
  • Gender:Male
  • Location:Greater Boston, MA
  • Accordance Version:10.x

Posted 09 June 2014 - 09:23 AM

I'm reviewing a commentary right now that says the Hebrew text of Obadiah has "291 words," a figure which is based on Theological Lexicon of the Old Testament's statistics.

 

When I do an asterisk search in Obadiah to get its number for words in Accordance (using a Hebrew text), I see "388 hits," which then in the Analytics shows as "388 total words" (with 167 different forms).

 

I know that Accordance is counting prefixes separately, because the word list in Analytics includes prefixed prepositions and the article as separate "words":

 

Attached File  Screenshot 2014-06-09 10.02.12.png   21.92KB   0 downloads

 

I also know that Accordance is not counting suffixes, because they appear in black (as opposed to red, in which my hits appear) in my Obadiah text.

 

The count in TLOT is "based on the graphic units" in the Hebrew text ("words separated by spaces or maqqēp"). In other words, unlike Accordance, TLOT counts prefixed prepositions and the article together with the word they precede.

 

Two questions:

 

1. Is everything above correct?

 

2. Is there a way to get a word count in Accordance (say, Hebrew Obadiah) such that prefixed prepositions, etc. are not counted as separate words? I.e., to approximate the kind of criteria TLOT uses?

 

Either way, I think I would probably go with Accordance over TLOT on this way of counting "words" in Hebrew, although I would guess that what constitutes a Hebrew "word" has probably served as the content of some dissertation somewhere....


Abram K-J
Pastor, Writer, Freelance Editor, Youth Ministry Consultant, Blogger
Web: Words on the Word


#2 Timothy Jenney

Timothy Jenney

    Platinum

  • Accordance
  • 1,418 posts
  • Gender:Male
  • Location:sunny Winter Haven, FL
  • Interests:a good cup of coffee, sci-fi, playing bass, listing to jazz and the blues, camping, fishing and the great outdoors
  • Accordance Version:10.x

Posted 09 June 2014 - 09:54 AM

Try this. I get 302 words. The count might be a bit different if the lemmas are not specified as "Exact."

 

(=הַ, =בְּ, =וְ, =וְ־0, =לְ, =לְ־0=)- @*=

Attached Files


Edited by Timothy Jenney, 09 June 2014 - 09:55 AM.

Blessings,
"Dr. J"

Timothy P. Jenney, Ph. D.
"Lighting the Lamp" Host and Producer

#3 Graham Buck

Graham Buck

    Silver

  • Accordance
  • 113 posts
  • Gender:Male
  • Location:God's Country (New England)
  • Accordance Version:10.x

Posted 09 June 2014 - 10:34 AM

Was geeking out for a minute, trying to find some articles on this topic. Nothing directly related, but I though you all might enjoy following up on some of these:

 

An HMM approach to vowel restoration in Arabic and Hebrew - http://dl.acm.org/ci....cfm?id=1118641

 

The symmetry of the Hebrew bible - http://www.tandfonli...rnalCode=sthe20

   [this one looks interesting and might have some data that's worthwhile]

 

And as a bonus, has it really come to this?:

 

Wordplay in church marquees - http://www.degruyter...mr.2011.012.xml


Edited by Graham Buck, 09 June 2014 - 10:34 AM.

Graham Buck
Content Development [Ninja]

OakTree Software


#4 Abram K-J

Abram K-J

    Platinum

  • Active Members
  • PipPipPipPipPip
  • 1,419 posts
  • Gender:Male
  • Location:Greater Boston, MA
  • Accordance Version:10.x

Posted 09 June 2014 - 11:07 AM

"Our comparison of the puns found on church billboards versus those found in a general corpus of puns reveals significant differences in structure and content, suggesting that these puns represent a specific type of humor."

 

I want this "general corpus of puns" as an Accordance Tool. :)

 

Tim--thank you, and good idea. I'll try a similar search string. 


Abram K-J
Pastor, Writer, Freelance Editor, Youth Ministry Consultant, Blogger
Web: Words on the Word


#5 James Tucker

James Tucker

    Platinum

  • Active Members
  • PipPipPipPipPip
  • 632 posts
  • Gender:Not Telling
  • Accordance Version:10.x

Posted 09 June 2014 - 05:05 PM

It all depends on what one defines as a word. Using HaKeter application, I see they've listed 291 words consisting of 1119 characters. It's interesting that Accordance is kicking back 387 (excluding brackets in HMT). Something seems askew here.



#6 James Tucker

James Tucker

    Platinum

  • Active Members
  • PipPipPipPipPip
  • 632 posts
  • Gender:Not Telling
  • Accordance Version:10.x

Posted 09 June 2014 - 05:14 PM

OK. I just ran a Perl Script over Obadiah excluding Ketiv/Qere (which is only 1). I get 291 words, in which a case the lemma, with its infixes, affixes, and suffixes constitute the grapheme. Not sure what Accordance is doing here.


Edited by James Tucker, 09 June 2014 - 05:15 PM.


#7 Abram K-J

Abram K-J

    Platinum

  • Active Members
  • PipPipPipPipPip
  • 1,419 posts
  • Gender:Male
  • Location:Greater Boston, MA
  • Accordance Version:10.x

Posted 09 June 2014 - 08:36 PM

Thanks, James--this is helpful.

 

So I think it's clear that Accordance includes (in its understanding of a Hebrew "word") prefixed prepositions, etc.

 

Is one understanding of what constitutes a Hebrew "word" more common than the other? (Perhaps there is already potential imprecision introduced when using the word "word.")


Abram K-J
Pastor, Writer, Freelance Editor, Youth Ministry Consultant, Blogger
Web: Words on the Word


#8 James Tucker

James Tucker

    Platinum

  • Active Members
  • PipPipPipPipPip
  • 632 posts
  • Gender:Not Telling
  • Accordance Version:10.x

Posted 09 June 2014 - 09:12 PM

I think Accordance is counting inseparable prepositions as "words." Yes, it does seem that Accordance is differentiating in its definition of 'word' in that some bound forms are counted. Thus, Tim's search above doesn't account for the Grammar of Hebrew. This comes closer; I don't have time to track down what gives a correct count of 291 (I just included the Aramaic so as to account for improper tagging should there be an incorrect decision). I would take a more concentrated definition of grapheme rather than word. Ideally, it would be best if Accordance gave the researcher the opportunity to define what 'word' is for their own research interests.

 

This comes close:

Attached File  Screenshot 2014-06-09 18.54.06.png   535.5KB   0 downloads


  • Abram K-J likes this

#9 Abram K-J

Abram K-J

    Platinum

  • Active Members
  • PipPipPipPipPip
  • 1,419 posts
  • Gender:Male
  • Location:Greater Boston, MA
  • Accordance Version:10.x

Posted 09 June 2014 - 09:46 PM

Nice search there, James--thanks again.

 

Agreed that it would be nice for Accordance to somehow let the researcher define "word," depending on what they are trying to do.


Abram K-J
Pastor, Writer, Freelance Editor, Youth Ministry Consultant, Blogger
Web: Words on the Word


#10 Graham Buck

Graham Buck

    Silver

  • Accordance
  • 113 posts
  • Gender:Male
  • Location:God's Country (New England)
  • Accordance Version:10.x

Posted 09 June 2014 - 09:47 PM

You know, guys, that sounds like a fantastic "Feature Request"… ;)


  • Abram K-J likes this

Graham Buck
Content Development [Ninja]

OakTree Software


#11 James Tucker

James Tucker

    Platinum

  • Active Members
  • PipPipPipPipPip
  • 632 posts
  • Gender:Not Telling
  • Accordance Version:10.x

Posted 09 June 2014 - 11:36 PM

Consider it requested! While you are at it, please provide an option to turn off canonical references in DSSB-M. It seems to me that DSSB-C is sufficient to alias the verse. That'd be such a time saver! I could do away with my intermediary paste into Nisus Writer Pro run clean DSS Quote Macro and Switch to Mellel and Paste action script.



#12 Graham Buck

Graham Buck

    Silver

  • Accordance
  • 113 posts
  • Gender:Male
  • Location:God's Country (New England)
  • Accordance Version:10.x

Posted 09 June 2014 - 11:45 PM

Two separate posts to the Request forum and you're well on your way to simplified research goodness.


Graham Buck
Content Development [Ninja]

OakTree Software


#13 Abram K-J

Abram K-J

    Platinum

  • Active Members
  • PipPipPipPipPip
  • 1,419 posts
  • Gender:Male
  • Location:Greater Boston, MA
  • Accordance Version:10.x

Posted 10 June 2014 - 06:40 AM

Okay--I'll just link to this post and put in the request for user options to select how Hebrew words are counted.


Abram K-J
Pastor, Writer, Freelance Editor, Youth Ministry Consultant, Blogger
Web: Words on the Word





0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users