Jump to content

Stop treating prefixes and suffixes as separate words in Hebrew


TYA
 Share

Recommended Posts

Accordance apparently doesn't allow a quick, simple search of four consecutive Hebrew letters if one of those letters functions as a prefix or suffix.  In the competitor's software (see attached), I can simply type the four letters, put an asterisk (wildcard) on either side, and find all texts where those four letters appear together--*regardless of their grammatical function.*

 

Why should I need to know the grammatical function in Accordance order to perform a simple lexical search of four Hebrew letters of my choosing?  This makes no sense whatsoever.  Here's a great example: I found a word in the competitor's Qumran sectarian manuscripts (which doesn't include an English translation, like Accordance does); and that word is "ma-a-tsat" (spelled "mem-ayin-tsade-tav"), in 1QS Rule of the Community 6:3.

 

I came over to Accordance and figured that I could simply search "mem-ayin-tsade-tav" (and maybe include wildcards on either side too) and find all Hebrew texts where this occurred in Accordance, which would obviously include 1QS Rule of the Community 6:3.  Then I could open the English version of this text, which Accordance wonderfully makes available, and be happily on my way.

 

But not so easy, because Accordance apparently doesn't allow something so simple--searching four consecutive Hebrew letters.  Since the letter "mem" functions as a preposition on the front of "ma-a-tsat," Accordance didn't return any text results when I did "Research" -> "All."  Instead, I had to fumble around, trying to figure out how to get to 1QS Rule of the Community 6:3 (which wasn't as easy as the competitor either, because there is no simple verse dropdown menu in Accordance).  And so on...

 

Bottom line, it would be nice if you could not treat prefixes / suffixes (technically morphemes / glossemes) as separate words when it comes to searching.  They aren't separate "words" in my opinion if they appear together in one whole morphological unit in the text.  And even if you have a workaround for this issue, such as making a "Construct" and then searching it, that is extremely time-expensive compare to the competitor's software, which I also use.

Link to comment
Share on other sites

This would involve a fundamental change and rewrite of all the code, as well as inconveniencing our many thousands of current users.

 

However if you use the Literal search mode it finds simple character strings, but cannot be used with other criteria such as commands or symbols.

  • Like 1
Link to comment
Share on other sites

They are treated as separate because they are. Enclitic pronouns and suffixes are stand alone words.

  • Like 1
Link to comment
Share on other sites

Thank you Helen for the reply.  This gives me the direction to go in.  Blessings.

Link to comment
Share on other sites

As Helen said, you can just do a simple literal search, and then link to the contents of that window for other additional refining.

 

But as Matt said, they are not prefixes as we talk about them in English. They are words. Separate.

I'd be curious for you to give the example of why you want to do a literal search that spans words.

  • Like 1
Link to comment
Share on other sites

  • 3 weeks later...

 

I'd be curious for you to give the example of why you want to do a literal search that spans words.

 

If your definition of "words" here includes prefixes in Hebrew (as it seems to for you and Matt), then I believe I gave a perfect example up above.  I'll repeat it, and add some comments in bold below (not for elevated tone, but just for clarity).

 

"Here's a great example: I found a word in the competitor's Qumran sectarian manuscripts (which doesn't include an English translation, like Accordance does); and that word is "ma-a-tsat" (spelled "mem-ayin-tsade-tav"), in 1QS Rule of the Community 6:3.

 

I came over to Accordance and figured that I could simply search "mem-ayin-tsade-tav" (and maybe include wildcards on either side too) and find all Hebrew texts where this occurred in Accordance, which would obviously include 1QS Rule of the Community 6:3. Then I could open the English version of this text, which Accordance wonderfully makes available, and be happily on my way.

 

But not so easy, because Accordance apparently doesn't allow something so simple--searching four consecutive Hebrew letters.  Since the letter "mem" functions as a preposition on the front of "ma-a-tsat," [[which you are calling a "word"]] Accordance didn't return any text results when I did "Research" -> "All." [[though BibleWorks did find this]]

 

Again Joe and Matt, I didn't come for the grammar lesson.  With all due respect, I don't see why a wildcard character can't be used to substitute any Hebrew prefix, be it a "vav," "bet," "mem," or "lamed," and let me find the lemma I'm searching for in every possible form.

 

As I said above, I shouldn't have to know the grammatical function of a prefix (call it whatever you want) in front of a word that I'm looking up.  That's a catch-22, because I didn't happen to know the meaning / translation of this particular word above before looking it up.  If so, how could I find it in Accordance???

 

In BibleWorks, I don't have to know the meaning (i.e. grammatical function of a prefix).  I only have to know the spelling, plus or minus room for a prefix.  Makes perfect sense to me.  No offense, but I would think something as simple as a word processor would have no problem finding a string of characters, regardless of their grammatical function.  Accordance should offer that option, and I hope this simple, real-life example above illustrates it perfectly.

Link to comment
Share on other sites

If your definition of "words" here includes prefixes in Hebrew (as it seems to for you and Matt), then I believe I gave a perfect example up above.  I'll repeat it, and add some comments in bold below (not for elevated tone, but just for clarity).

 

"Here's a great example: I found a word in the competitor's Qumran sectarian manuscripts (which doesn't include an English translation, like Accordance does); and that word is "ma-a-tsat" (spelled "mem-ayin-tsade-tav"), in 1QS Rule of the Community 6:3.

 

I came over to Accordance and figured that I could simply search "mem-ayin-tsade-tav" (and maybe include wildcards on either side too) and find all Hebrew texts where this occurred in Accordance, which would obviously include 1QS Rule of the Community 6:3. Then I could open the English version of this text, which Accordance wonderfully makes available, and be happily on my way.

 

But not so easy, because Accordance apparently doesn't allow something so simple--searching four consecutive Hebrew letters.  Since the letter "mem" functions as a preposition on the front of "ma-a-tsat," [[which you are calling a "word"]] Accordance didn't return any text results when I did "Research" -> "All." [[though BibleWorks did find this]]

 

Again Joe and Matt, I didn't come for the grammar lesson.  With all due respect, I don't see why a wildcard character can't be used to substitute any Hebrew prefix, be it a "vav," "bet," "mem," or "lamed," and let me find the lemma I'm searching for in every possible form.

 

As I said above, I shouldn't have to know the grammatical function of a prefix (call it whatever you want) in front of a word that I'm looking up.  That's a catch-22, because I didn't happen to know the meaning / translation of this particular word above before looking it up.  If so, how could I find it in Accordance???

 

In BibleWorks, I don't have to know the meaning (i.e. grammatical function of a prefix).  I only have to know the spelling, plus or minus room for a prefix.  Makes perfect sense to me.  No offense, but I would think something as simple as a word processor would have no problem finding a string of characters, regardless of their grammatical function.  Accordance should offer that option, and I hope this simple, real-life example above illustrates it perfectly.

Think that makes sense and agree that that would be helpful.

Link to comment
Share on other sites

  

I came over to Accordance and figured that I could simply search "mem-ayin-tsade-tav" (and maybe include wildcards on either side too) and find all Hebrew texts where this occurred in Accordance, which would obviously include 1QS Rule of the Community 6:3. Then I could open the English version of this text, which Accordance wonderfully makes available, and be happily on my way.

 

But not so easy, because Accordance apparently doesn't allow something so simple--searching four consecutive Hebrew letters.

 

Just enter  מעצת in the Qumran Non-biblical Manuscripts module and use a "letters" search. That will find 1QS. 6:3; 8:22; 4Q258. 2:8; 4Q261. f2a_c:3; 4Q263. f1:4; 11Q19. 58:20.

 

If you want to expand your search, you can right-click מעצת  and select Research—>All Texts and find  מעצת in all your texts.

Edited by Mark Allison
  • Like 2
Link to comment
Share on other sites

You can also use the inflected search and try adding spaces to separate what might be prefixes or suffixes, based on the character. So "מ עצת" should also find the word.

  • Like 1
Link to comment
Share on other sites

TYA, I apologize for anything that sounded like a condescending grammar lesson. The point was simply pointing out how the Accordance search functions (not making any linguistic point). But, as you see above, literal searches will precisely find what you want.

Link to comment
Share on other sites

An honest "thank you" to all who are responding here.  (Accordance has amazed me from day 1 with the magnificent support, and I've never experienced anything this good with Bible software).

 

And I certainly appreciate the advice / suggestions, both Helen and Mark.  The "letter" (literal) search has been by best method so far, but I'm still having trouble with the example up above.  I don't want to sound argumentative; it just seems that the letter search still isn't giving me any results when I "Research" מעצת in all Hebrew texts.  I'm searching it from the BHS in a window, and selecting "letter" and then "Research - All"  I'm trying it just as it appears here, and also with + signs, and also with * signs.  (No, I really don't know what I'm doing.)  But I still never find this word in any biblical text, including Qumran.

 

If I can say so, "ughhhh."  I don't know why it just can't find it the way BW does.  But if something like this would require too great an effort (coding, etc.), and if no one else is asking for it, I understand.  I will have to figure something out.

 

I'm really appreciative for Accordance large (and growing) library; and sometimes I just like to be able to do a literal spelling search (like above) across all resources.  I don't want to have to take into account the grammatical function of any part of the word (sometimes)--much less a prefix.  Sorry if it seems that I'm griping.  You are all a huge blessing.  Thank you.

Link to comment
Share on other sites

Hi TYA, All,

 

  First, Hebrew - still very weak for me, but I tried a few things and I am not seeing what I would expect.

 

  Secondly, searching for Letters for מעצת in BHS I get a 4 hits, 3 with all letters directly adjacent, not considering vowels, and 1 with a space after the מ.

  Thirdly, Running this again in Research -> All Texts I get hits only in the DHNT, MHNT which are untagged NT texts. I get nothing in any other Hebrew text.

  Fourth, Rebuild the research cache and try again. Still no hits outside of these two texts.

  Fifth, Change to All instead of merely All Texts and I still get no other texts. I get MHNT and DHNT and then non-text hits.

 

  So I do not see how this works in research so that you get hits in tagged texts. Now given what I know about Acc searching this sort of makes sense but it's a bit of an obstacle to doing research like this. Is there some setting that I'm missing to get this working ?

 

  EDIT: OK so if as Mark describes, you drive the Research query from the BHS search via the context menu then an appropriate lex search is constructed for the Research query and that does produce results. I note it also excludes the מ עצת case which seems correct. But of course, it no longer finds hits in the non-tagged texts that I was seeing because the search now requires a tagged text I suspect. It would be a nice enhancement to permit direct entry into the Research to run such queries in Hebrew.

 

  EDIT: There is a somewhat similar set of cases with crasis in Greek. I can search for καγω in Greek in Research and I get two hit texts only despite the fact that καγω appears in many Greek texts I have. If I then try Mark's approach above and search in NA 28 GNT for "χρασισ" and then select an example of καγω and right click -> Research -> All Texts I actually get a Research query for και and not και@εγω@"χρασισ" or something like it which is sort of what I would have expected. Ok the search should be και εγω@"χρασισ" but I don't see that either.

 

Tx

D

Edited by דָנִיאֶל
Link to comment
Share on other sites

Hi TYA, All,

 

  First, Hebrew - still very weak for me, but I tried a few things and I am not seeing what I would expect.

 

  Secondly, searching for Letters for מעצת in BHS I get a 4 hits, 3 with all letters directly adjacent, not considering vowels, and 1 with a space after the מ.

  Thirdly, Running this again in Research -> All Texts I get hits only in the DHNT, MHNT which are untagged NT texts. I get nothing in any other Hebrew text.

  Fourth, Rebuild the research cache and try again. Still no hits outside of these two texts.

  Fifth, Change to All instead of merely All Texts and I still get no other texts. I get MHNT and DHNT and then non-text hits.

 

  So I do not see how this works in research so that you get hits in tagged texts. Now given what I know about Acc searching this sort of makes sense but it's a bit of an obstacle to doing research like this. Is there some setting that I'm missing to get this working ?

 

  EDIT: OK so if as Mark describes, you drive the Research query from the BHS search via the context menu then an appropriate lex search is constructed for the Research query and that does produce results. I note it also excludes the מ עצת case which seems correct. But of course, it no longer finds hits in the non-tagged texts that I was seeing because the search now requires a tagged text I suspect. It would be a nice enhancement to permit direct entry into the Research to run such queries in Hebrew.

 

  EDIT: There is a somewhat similar set of cases with crasis in Greek. I can search for καγω in Greek in Research and I get two hit texts only despite the fact that καγω appears in many Greek texts I have. If I then try Mark's approach above and search in NA 28 GNT for "χρασισ" and then select an example of καγω and right click -> Research -> All Texts I actually get a Research query for και and not και@εγω@"χρασισ" or something like it which is sort of what I would have expected. Ok the search should be και εγω@"χρασισ" but I don't see that either.

 

Tx

D

As TYA has stated, this type of search is very difficult for Accordance to do. I think TYA has a good point from the software point of view.

Link to comment
Share on other sites

Yep, agreed. At times what one really wants is a simple grep. It's been raised multiple times before now. Letter (literal) searches go some way in addressing the issue but it remains a limitation for some cases. The Research feature also adds a new twist to the problem.

 

Thx

D

Link to comment
Share on other sites

I searched for the inflected form in Research: "ם עצת"and got hits in the Hebrew Bible as well as Qumran. "מעצת" with or without the quotes got hits in BHQ and Talmud (both are untagged). Hebrew NT gave no results, the form does not occur.

Link to comment
Share on other sites

So TYA, were you able to get this to work? It appears as though Helen is able to do what you want.

 

A.D.

Link to comment
Share on other sites

Dear A.D.

 

That's nice of you to reach out.  No.  I tried looking up how to do an Inflected search in the Accordance Help menu.  The instructions there seem straightforward there, yet I don't have the options that the instructions are telling me to select.  This is cumbersome.

 

I don't want to kvetch too much, but it seems like Accordance has a dozen different kinds of window panes ("Notes," "Research," "Bible" etc., which is great in and of itself,) BUT they all behave differently.  They all have different options, so that just when I'm trying to figure out how to perform a search in one window, I open another window which behaves differently.

 

And while reading the Help page to try and figure out how to do Inflected searches (per Helen's recommendation), am I now correct that Accordance doesn't even search tagged and untagged texts the same way?

 

So what I was frustrated about before was that I can't do a simple, straightforward, uncomplicated search of "abcd" together (in Hebrew, and allowing for any prefix, OR any string of characters to either side).  NOW, it seems that Accordance produces inconsistent results based on *tagging*?  What?

 

Honestly, honestly.  It seems that I'm further away from a simple, successful ability to search universally throughout all (say, Hebrew) texts.  Guys.  I fully understand that each software functions differently.  One software has you press "B" to accomplish something, and another software uses "C" instead.  I get that.  No problem.

 

And BW has complicated searches that would take me long to learn, and it also has quirks which irk me to no end.  But I will tell you what: I can step up to the same Command line, choose "Hebrew language," and type any string of characters (with wildcards on either side; and yes, I do this almost every day, very often), and it will find that string in *all* texts, tagged or untagged.

 

(Oh, and yes, with an extremely efficient way to review those search results in multi-version popup windows and results which allow popup-able display, which I know is another topic I posted, but oh so important to me.  It is the way I sometimes research 4+ hours per day in Hebrew, Greek, and Aramaic texts.  (And yes, I can do more complicated searches when necessary).

 

But I'm getting more tied into a pretzel here than I ever could have dreamed with Accordance.  Again, each window seems to function differently.  Do I have to remember how each and every kind of window functions (i.e. what search options / filters / language, etc.) it will give me?  Which window will allow me to launch a search?  Which window will allow me to do one thing, or another?

 

Okay, if so, maybe I can accomplish that with time.  But what then about the inconsistent search results.  One person comes to this thread and reports that they find "a,b,c" in their search results.  Another person finds "b,d,f."  Another person finds "x,y,z."  And I'm scraping by to just find "c."

 

I was attracted to Accordance primarily because of its larger library of texts.  (That potentially makes me a good customer, obviously).  And indeed I'm super excited about the possibilities.  And yes, I understand that Accordance has bent over backwards for others, such as BW users like myself.  I don't forget this, and I'm truly grateful.  I can't imagine that the program ever existed *without* "LiveClick," but it is great.

 

But if you are telling me that I can't search 1) a single Hebrew string of characters and 2) find it *consistently* across all Hebrew texts, regardless of whether they are tagged or not, I... am left speechless.  That's where the real power is--at least for those whose study primarily centers on lexical analysis and thorough textual criticism.  The search is key.

 

Sorry if anyone is offended.  I just appreciate... help, I guess.

Edited by TYA
Link to comment
Share on other sites

In regards to my statement above about inconsistent search results among various people reporting on this thread, that may be putting it lightly, because I don't even get consistent, or predictable search results by myself when using the "Research" -> "All" option for a simple "Letter" (literal) search.

 

I neither get thorough, universal search results when searching for Hebrew, or for English content.  Please see these screenshots illustrating that both Hebrew and English searches are providing limited, inconsistent / paradoxical search results.

 

This simply should not be, and it seems to be robbing my prospects of being able to grow my Accordance library joyfully and accomplish straightforward, thorough searches throughout the library of texts and modules.

 

That is one of the most powerful, effective things about BibleWorks (at least as it relates to Bible modules, since "LiveClick" does excel BW in the fact that it returns for results for multiple kinds of modules).  But BW is thorough and consistent when it comes to original language texts (including English), and that is what I expect from Accordance.  It can't be asking too much.

post-35231-0-76098900-1536301082_thumb.jpg

post-35231-0-17762600-1536301088_thumb.jpg

post-35231-0-15370900-1536301093_thumb.jpg

post-35231-0-83710900-1536301099_thumb.jpg

Edited by TYA
Link to comment
Share on other sites

TYA, using your search example for "common" in BDB Complete I see you found 123 flex hits. My search returns the same and includes the one you stated was missed.

 

I don't mean to sound rude so please excuse me if you know this. But are you aware of the hits arrows at the bottom of the pane that allows you to jump to the next occurance of the searched word? Accordance does not show only the text that was searched for. It shows the entire body of the text and then allows you to jump to the next occurance of the result using these arrows.

Link to comment
Share on other sites

I confirm that Accordance finds the word "common" in both flex and exact searches in the English content field. (See screenshot.)

post-29215-0-46513600-1536326100_thumb.png

 

 

Link to comment
Share on other sites

Hi TYA, You cannot actually do a search in two steps like you tried. What happened is that the second search was run as an entirely new search. It hit an earlier hit for "common" than the one you saw on the page. The issue of a search in page (what is currently on-screen) has come up before but no such feature exists. If you want to search for ܐܶܢܳܐ in Syriac and common in English then after doing the first search go back up to the search box and click + (at the far right end of the box and add another search field and switch that to English Content and enter common and hit enter. Then you will get :

 

post-32023-0-07152800-1536326351_thumb.jpg

 

Thx

D

 

 

 

  • Like 2
Link to comment
Share on other sites

TYA, you'll notice that D's screenshot does include the word "common" in the next entry whereas mine does not.

 

The reason for this is because I limited my search to just paragraphs. That is, I told Accordance, "Only find these words when they occur together in a single paragraph."

 

You can see that from the little box at the top right corner of my search (marked in red) here.

 

post-211-0-77810600-1536326589_thumb.png

Link to comment
Share on other sites

But are you aware of the hits arrows at the bottom of the pane that allows you to jump to the next occurance of the searched word?

 

Yes sir, but I appreciate you pointing it out, since I am new at this.  You and Timothy are both correct: Accordance did find the word "common."  I was mistaken, because I assumed it would find the closest (proximity) occurrence, which was right there on the screen.

 

So I was wrong to say that Accordance wasn't finding the words.  But on the other hand, I think this still proves the need / value for a "Find on Page" apparatus, which would allow me to specifically search *from the current location in the document,* which is what I meant by "my typical two-step search process."

 

I did try using the option "Search within each Paragraph," thinking that this would limit my search for "common" to that region of the document, which would have essentially accomplished what I want.  I tried both "Search within each Paragraph" and "Search within each Article," but I was taken up to the top of the document in both case, so it would seem that this doesn't accomplish my need.

 

As to the "Find on Page" request, Accordance has already stated that this is on their list of updates for the next major release.  (Big thank you for that).

Link to comment
Share on other sites

If you want to search for ܐܶܢܳܐ in Syriac and common in English then after doing the first search go back up to the search box and click + (at the far right end of the box and add another search field and switch that to English Content and enter common and hit enter. Then you will get :

 

attachicon.gifsc.jpg

 

Thx

D

 

Dear Daniel, what am I doing wrong?  Your process seems like it would accomplish what I'm looking for, but when I do what you say, it still throws me up to the top of the document (i.e. first occurrence of "common").

post-35231-0-79682700-1536337277_thumb.jpg

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
 Share

×
×
  • Create New...