PDA

View Full Version : Searching for the same word repeated



Keith McAliley
04-24-2009, 10:41 AM
Many apologies if this is too dumb a question for this esteemed forum. The searching thing doesn't come easily to me yet.

I want to search (preferably in a Hebrew text) for all the instances of "X of Xs." In Hebrew, it's a singular common noun followed by the plural of the same noun, preceded by the definite article. I know what the results will be ("lord of lords," "holy of holies," "song of songs"), but it bugs me that I can't construct the search. Help?

Peace,
Keith McAliley
Bridgeport Christian Church
Frankfort, KY

SkipB
04-24-2009, 12:21 PM
Keith
I am not as proficient as others around here in the GSE. But I put together this query that found 51 verses of the construct singular followed by a plural of the same lemma. It will give you a place to start and then you can tweak it if I've missed something.

Ben Spackman
04-24-2009, 12:37 PM
You can do this from the command line as well
'#1@nc?p* #1@nc?sc*

#1 tells it to use the same root. Interestingly, if you cut everything else off, just '#1 #1 it finds verbs AND nouns. (Not sure if it catches mem-preformatives and other things yet, I assume it doesn't.)

Edit: Look carefully at your results though. This also catches ketib/qere changes which aren't the construction you're looking for.

Keith McAliley
04-24-2009, 09:45 PM
Ben's solution works but gives me a couple of "Error: Too Many Parantheses!" messages. Why? And when I played around with '#1 * #1 in an English version I got so many of these messages that I had to shut down the program.

But I'm glad to have found out about #1. And thanks, SkipB, for the query. I've had BW for five years and I've NEVER figure out that ASE or GSE or whatever it is now. :)

Peace,
Keith

Keith McAliley
04-24-2009, 09:48 PM
Oh and '#1 #1 #1 in an English text REALLY drives it crazy. It gives me a message I'd never seen before ("BibleWorks has encountered a severe error . . .")

So I'm not sure my computer is ready for the #1 thing.

Peace,
K

Ben Spackman
04-24-2009, 09:59 PM
I don't get those messages on either my Mac (under emulation) or my older PC- P4 2.4, 1.2 Gig Ram, WinXp.

The number thing works with more too. You can do things like '#1 #2 #1 #2

Glenn Weaver
04-27-2009, 09:10 AM
Hi Keith,

To fix the error problem that you experienced, I suggest that you close BibleWorks, then hold down your shift key while restarting BibleWorks. This opens the BibleWorks Trouble-Shooting Options window. I suggest that you check the following boxes under Option 1:
--Delete the BW800.INI file
--Rebuild Version Data
--Delete Context Files

Then click "Perform the selected options and run BibleWorks." This should fix the problem. Keep in mind that deleting the BW800.INI file will remove any custom search limits that you have created, and any other custom configurations you may have set.

If this does not help, please contact Customer Service for further help.

DavidR
04-27-2009, 03:42 PM
I'm not getting any errors using the #1 ... type search in BW 8.0.008w.2, searching in NRS. For instance, searching for '#1 #1 #1 finds exactly 4 verses in the entire Bible (Isa 6:3; Jer 22:29; Rev 4:8; Rev 8:13). I also find that it doesn't matter what search operator I use: '#1 #1 #1, .#1 #1 #1; and /#1 #1 #1 all produce the same results.

But note that this is only able to find multiple consecutive instances of any (unspecified) word in the same verse. The original request was to find "X of Xs". This type of search would find them in a morphologically marked version (like WTM)--and much else besides. But suppose you wanted to look for the repetition of a specific word ("king of kings," for example); or suppose you wanted to find multiple non-consecutive instances of a word (specified or unspecified); or suppose you wanted to do some of this in an English version. Can any of that be done from the command line? Or would it definitely require the GSE?

Glenn Weaver
04-27-2009, 03:56 PM
Hi David,

The "#1 #1" type of search is called a "lemma agreement" search. It is discussed in Chapter 25 of the Help file under "Lemma Agreement". It runs a little different than usual Command Line searches, which is why it doesn't matter if you use the period, slash, or single quote mark. (In fact, the Help file suggests using the semicolon for these types of searches. You will need to use the semicolon for the search example given below.)

In order to find the same words in a verse even if they are not consecutive, include an asterisk plus number of intervening words between the two lemma agreement operators. For example, to find two identical words (or lemmas in a morphologically-tagged version) that can have up to 3 intervening words, type ";#1 *3 #1".

This search will search for lemmas in morphologically-tagged versions, or it will search for exact words in other texts, including English versions.

DavidR
04-28-2009, 09:41 AM
Thanks, Glenn. Usually I'm good about reading manuals and help files, but I seem to have lost my way.... :o

I actually had tried out the ";#1 *3 #1" type of search, and it works well, though it produces lots of results, since it finds verses with 3, 2, 1, and 0 words between the identical words, and you get a lot of verses with repeated instances of "the" and "and," etc.

This does seem likely to be more useful in morphological databases like WTM, BNM, etc. And if one wanted to find instances of specific words repeated in a verse, the GSE would seem to be the way to go.

Glenn Weaver
04-28-2009, 09:45 AM
Hi David,

You are correct that the GSE would give some more flexibiliyt than using the Command Line. But there are ways to do these things on the Command Line if you wish. For example, you can also use the new word-lists search feature on the Command Line to search on lists of words. In this way you can search for specific words that repeat, though I don't believe it would be possible to say that they are the exact same words, only that the two words are in the list of words.

MBushell
04-28-2009, 01:06 PM
Hi All,

I believe (hope) the # search crash (ot "too many parentheses") problem has been fixed. Try the latest exe (8z). The problem was with the "compound word form" feature. It was removed from BW8 but a definition file was leftover (default.cws) and was causing problems. You can delete that file if you wish, but it should no longer cause a problem.

Mike

MBushell
04-28-2009, 01:12 PM
Thanks, Glenn. Usually I'm good about reading manuals and help files, but I seem to have lost my way.... :o

I actually had tried out the ";#1 *3 #1" type of search, and it works well, though it produces lots of results, since it finds verses with 3, 2, 1, and 0 words between the identical words, and you get a lot of verses with repeated instances of "the" and "and," etc.

This does seem likely to be more useful in morphological databases like WTM, BNM, etc. And if one wanted to find instances of specific words repeated in a verse, the GSE would seem to be the way to go.
David,
Keep in mind that you can use morph codes if using the # parameter with a morphology version. For example, #1@v* #1@v* will fild only repeated verbs.

Mike

DavidR
04-28-2009, 06:27 PM
[QUOTE=MBushell;17962]
Keep in mind that you can use morph codes if using the # parameter with a morphology version. For example, #1@v* #1@v* will fild only repeated verbs.
[QUOTE]

Yes, that's why using this feature on morphologically-tagged versions seems like the most useful way to employ it. It's still fairly hypothetical for me, but I can imagine it being helpful sometime.

Keith McAliley
05-01-2009, 05:25 PM
Thanks. Updating to version 8z cleared it up.

What a fun feature! I'm glad I asked.

Peace,
Keith