Searching by Stem (Hebrew and Greek) aka cognate search

Logos is great when you want to search by lemma. However, searching for cognates can be a hit-or-miss affair. Logos can normally approximate a Greek stem with a wildcard search (e.g. lemma:δικ*) but this does not work for Hebrew. The rub is that both Hebrew and Greek have processes by which letters can change/drop off which can foil a * search.

Tracing a cognates through a passage can be a very important part of exegesis. However, with Logos you have to rely on an inefficient wildcard search that may or may not catch everything or compile a list of all the cognate lemmas manually and then run multiple searches. Logos, could you please design a way to make this easy and intuitive?

I've seen this question come up multiple times in the past months and I would love to see Logos mature in this area.

Find more posts tagged with

Comments

Jason Van Vliet

Precisely! I was thinking the very same thing this past week as I was busy with exegetical work. Undoubtedly Logos is busy now working on those "missing features" that need to be implemented yet. But when time & resources allow, this would certainly be a worthwhile project.

Alan Macgregor

Tracing a cognates through a passage can be a very important part of exegesis.

Yes, indeed. At the moment the best you can do to be aware of cognates is actually read the passage several times in the original language, which is fine if you are reasonably competent in that language. However, if your Greek or Hebrew (not to mention Aramaic) is rusty or if you haven't studied Greek or Hebrew but just want to be exegetically aware of what is going on, then reading through is not going to help you. Then, as Kevin said, the best you can do at the moment in Logos is a lemma search which is often going to miss a lot and can produce results which may or may not be accurate as far as cognates are concerned. (And you have no way of knowing!) Then you are thrown back on the tender mercies of the commentators who have a sneaky habit of skating over the very aspects of the biblical text that you most want to understand more clearly. (I call that Macgregor's Law. Well, Murphy has a law named after him! [;)])

The problem is that some cognates are obvious, some are less obvious, and some are positively opaque!

As someone who is not very computer-literate, (I can drive them but I'm no designer/builder/mechanic) I just don't know what the algorithm would look like. But I know I would like Logos to be able to do this – add to Basic, Bible and Morph – Cognate Search.

George Somsel

Logos is great when you want to search by lemma. However, searching for cognates can be a hit-or-miss affair. Logos can normally approximate a Greek stem with a wildcard search (e.g. lemma:δικ*) but this does not work for Hebrew. The rub is that both Hebrew and Greek have processes by which letters can change/drop off which can foil a * search.

I haven't tried this with L4 because I have been thoroughly disgusted with its searching abilities as well as the implementation of notes so that L4 has been nothing more than a nice reading platform. As you noted, you can use the wild-card search in Greek in L3 but not with Hebrew. I rather imagine that this is due to the necessity of including the vowel points in the search (which is how you can use a computer to distinguish מֶלֶךְ from מָלַך).

Kevin Becker

I've added this to Logos' Uservoice forum. Vote for it! (Sorry to the causes I robbed my votes to create it [:)])

http://logos.uservoice.com/forums/42823-logos-bible-software-4/suggestions/682641-stem-cognate-search

Rich DeRuiter

https://community.logos.com/discussion/comment/121794#Comment_121794

I've added this to Logos' Uservoice forum. Vote for it! (Sorry to the causes I robbed my votes to create it )

http://logos.uservoice.com/forums/42823-logos-bible-software-4/suggestions/682641-stem-cognate-search

Thanks Kevin, I just voted the max possible for this one (3 votes).

I've asked for this before and think it should be a part of the standard Bible Word Study report. Absolutely essential for careful, exegetical studies.

MJ. Smith

https://community.logos.com/discussion/comment/121794#Comment_121794

I've added this to Logos' Uservoice forum.

Thank you.

Ron Corbett

https://community.logos.com/discussion/comment/121882#Comment_121882

Does anyone think that the data is already in LOGOS (ready to be mined) with BDB? maybe a search app that goes word for word through a pericope (or larger structural unit) and compiles different lists of "words which share the same roots".

Alan Macgregor

https://community.logos.com/discussion/comment/121896#Comment_121896

I just don't know what the algorithm would look like.

Now I see how it could be done! [I]

Does anyone think that the data is already in LOGOS (ready to be mined) with BDB? maybe a search app that goes word for word through a pericope (or larger structural unit) and compiles different lists of "words which share the same roots".

Thanks, Ron. BDB holds the list of cognates. Presumably these cognate roots are tagged, therefore, it's only a matter of providing this search based on a word, whose root would be looked up in BDB and then the resultant root would be searched in the BHS text, assuming that individual words are root-tagged! Such a cognate searching could be embedded in Enhanced BDB, so it would become available to those had it (that's everyone who has OLL or any of the Scholars packages) or anyone who purchases it as a separate resource. (A bit like BDAG search in L3.)

Well done, Ron! Now why couldn't I think of that? [:S]

Rich DeRuiter

https://community.logos.com/discussion/comment/121896#Comment_121896

Does anyone think that the data is already in LOGOS (ready to be mined) with BDB? maybe a search app that goes word for word through a pericope (or larger structural unit) and compiles different lists of "words which share the same roots".

They probably are (I haven't checked), but we need this for Greek too, and for both NT and LXX complete vocabulary. A daunting task (IMHO). But I'd bet (if I did that sort of thing), that there are Greek/Hebrew/Aramaic scholars using Logos who have such lists for their personal use, if nothing else. Maybe Logos could 'buy' several of these (by providing free resources, e.g.?), and look at work that's already been done. Then Logos' academics in residence should compile their own database.

When I suggested stem searching earlier, a few work-arounds were suggested, but these proved unreliable and insufficient. (see here: http://community.logos.com/forums/p/5770/45011.aspx) So we obviously need a tool designed for this job here.

As I see it, to do this Logos would need a stem/cognate database that could be accessed by the search module as a filter. To me, and the way I use the original languages, this is a critical function. I would pay to get it, knowing I would get so much out of it.

I do know that sometimes, by themselves, stems are not reliable predictors of meaning, and that sometimes it's difficult to conclude whether some stems are even related merely because they are spelled the same - i.e., there's a bit of debate at times among original language scholars about these things. But if Logos simply included them all anyway, erring on the side of inclusion, we users would have the tools to look at these matters ourselves and come to our own conclusions.

MJ. Smith

https://community.logos.com/discussion/comment/122097#Comment_122097

stem/cognate database

What I would prefer, being a bit more technical, is the roots (rather than stems) in all manuscript languages (Greek, Hebrew, Aramaic, Syriac, Latin, Sahidic - in that level of importance) and (less importantly) cognates in Greek/Latin/English and in Hebrew/Aramaic/Syriac. The roots I see as essential to Logos' primary thrust; the cognates I see as ancillary but very useful.

Rich DeRuiter

https://community.logos.com/discussion/comment/122121#Comment_122121

stem/cognate database

What I would prefer, being a bit more technical, is the roots (rather than stems) in all manuscript languages (Greek, Hebrew, Aramaic, Syriac, Latin, Sahidic - in that level of importance) and (less importantly) cognates in Greek/Latin/English and in Hebrew/Aramaic/Syriac. The roots I see as essential to Logos' primary thrust; the cognates I see as ancillary but very useful.

I see your point, but I want to make sure that when I search a stem (or root), I find all nouns, verbs, adjectives and adverbs with that stem (or root). These related words are called cognates, and that is what I was referring to. I would not be happy with a system that only searched for verb roots (if I initiated a search or word study based on a verb), or noun roots, etc.

I would also want to know that I was getting all the variations of a root with prefixes or suffixes, which a true root search would include.

I'm not at all interested in searching for cognates that cross linguistic barriers, i.e., from Greek to English. While these can be interesting from a linguistics point of view, they aren't very interesting from a Biblical interpretation point of view (which is where I spend most of my time).

MJ. Smith

https://community.logos.com/discussion/comment/122134#Comment_122134

I see your point, but I want to make sure that when I search a stem (or root), I find all nouns, verbs, adjectives and adverbs with that stem (or root). These related words are called cognates, and that is what I was referring to. I would not be happy with a system that only searched for verb roots (if I initiated a search or word study based on a verb), or noun roots, etc.

I would also want to know that I was getting all the variations of a root with prefixes or suffixes, which a true root search would include.

I'm not at all interested in searching for cognates that cross linguistic barriers, i.e., from Greek to English. While these can be interesting from a linguistics point of view, they aren't very interesting from a Biblical interpretation point of view (which is where I spend most of my time).

Okay - we have a difference in terminology. These are the definitions that I am familar with:

root: the form of the word after all affixes are removed
stem: often identical to root but contains thematic vowels if applied
cognate: words that share a common ancestor in an ancestral language
derivative: word derived from another word

It sounds as if in the terminology I learned you want roots and the set of derivatives of that root.

John Brumett

https://community.logos.com/discussion/comment/122152#Comment_122152

There is a Lexicon that already accomplishes this (A New Testament Greek Morpheme Lexicon by J. Harold Greenlee). This shows all the appropriate root forms and Greek words with those forms including prefixes, root words, suffixes, and terminations. It list every Greek Word in the New Testament. I wish this book was in Libronix. The publisher is Zondervan and the publishing date is 1983. Here are 2 sections from this book.

John Brumett

https://community.logos.com/discussion/comment/122152#Comment_122152

Rosie Perera

https://community.logos.com/discussion/comment/121794#Comment_121794

I've added this to Logos' Uservoice forum. Vote for it! (Sorry to the causes I robbed my votes to create it )

http://logos.uservoice.com/forums/42823-logos-bible-software-4/suggestions/682641-stem-cognate-search

I've now reassigned one of my votes from a less important feature to this. Let's get it up onto the first page of feature suggestions! Surely all the votes for Full Feature Print Functions, Smart Tags, and Sentence Diagramming can be reassigned, now that those features are nearly complete and will be shipping in 4.1 in the near future.

Daniel Luke Gonzalez

Went ahead and added my votes. If I remember correctly, Bibleworks has this feature. I should be able to use Logos for all of my Exegetical needs and not have to ask a friend with Bibleworks to help me do a search.

Rich DeRuiter

Bump.

Ran into this limitation again today, twice in one verse.

Any way to get this done?

George Somsel

https://community.logos.com/discussion/comment/122121#Comment_122121

stem/cognate database

What I would prefer, being a bit more technical, is the roots (rather than stems) in all manuscript languages (Greek, Hebrew, Aramaic, Syriac, Latin, Sahidic - in that level of importance) and (less importantly) cognates in Greek/Latin/English and in Hebrew/Aramaic/Syriac. The roots I see as essential to Logos' primary thrust; the cognates I see as ancillary but very useful.

I have what may be good news. It appears that an old-fashioned wildcard search will work. I need to test it a bit more, but first looks are favorable.

David Knoll

https://community.logos.com/discussion/comment/270875#Comment_270875

It seems I never said [Y] on this. I did vote for it on uservoice of course.

Forrest James

https://community.logos.com/discussion/comment/170171#Comment_170171

John, from a year ago. Has there been any updates at logos or elsewhere, so that full cognate searches are available online, or is Greenlee still the best option. Thanks. fob James

Kevin Becker

https://community.logos.com/discussion/comment/308939#Comment_308939

Forrest, welcome to the forums.

The Lexham Analytic Lexicon (which is in most base packages) has been updated with cognates so one can use that list to construct a search to catch all Greek cognates to a specific word but Logos doesn't have a true cognate search yet. Please add your votes to the Uservoice item on this if you haven't already.

Donovan R. Palmer

https://community.logos.com/discussion/comment/308984#Comment_308984

I would sure like to see this built in and for it to appear as part of the guide tool!

Alan Macgregor

https://community.logos.com/discussion/comment/312361#Comment_312361

You and me – both! [Y] [Y]

Forrest James

https://community.logos.com/discussion/comment/308984#Comment_308984

Kevin, I got the 1983 Morpheme Lexicon by Greenlee. (For $3 plus 4.99 shipping!) It is most helpful for looking at the cognates manually. For root words with many cognates --- would a Lexham search pick up most of the cognates, even though often the cognates do not begin with, or even contain, the same letters as the root. I am working with "didomi," for instance, which is the root and has 33 cognates from the BAGD, none of which begin with "did..", and only a few of which contain "did..". So, with the Lexham software, if I want combinations of "didomi" cognates and logos (root "lego") cognates (of which here are 73), I would have to search manually one combination at at time? Or does Lexham do more than this? Also, does the Logos software allow you, in a search, to specify how close the words appear next to each other in the text? With didomi and logos, I am mainly interested in their cognates, if any, that appear immediately next to each other. But if Logos provides for an expansion of the search to one, or two, or any specified number of words apart, that would be helpful as well? Rgds. Fob James

Kevin Becker

https://community.logos.com/discussion/comment/314321#Comment_314321

Forrest,

This post by Mark Barnes describes in detail how to achieve a cognate search in Logos using the list in the Lexham Analytical Lexicon. It is possible to use proximity operators to specify distance between one cognate set and another with a format like this (lemma:x, lemma:y, lemma:z) WITHIN 5 WORDS (lemma:a, lemma:b, lemma:c). Of course, this type of search becomes rather unwieldy when you are dealing with words that have a lot of cognates (not to mention might take a very long time to run). For more details about the capabiliites of the Logos search engine see: http://wiki.logos.com/Detailed_Search_Help

Denise Pass

This resource in Logos has been very helpful with listing Greek cognates as well as language equivalents in Hebrew, Aramaic, etc.: The Lexham Analytical Lexicon to the Greek New Testament. Logos Bible Software, 2011.