Parsing error in Lexham Hebrew Bible?

Page 1 of 1 (16 items)
This post has 15 Replies | 4 Followers

Posts 265
Michel Pauw | Forum Activity | Posted: Mon, Sep 24 2018 7:39 PM

Ruth 1:2 has נָשִׁים, which then in the Lexham Hebrew Bible is glossed as noun, common, masculine, plural, absolute.

Now, the form of the plural suffix indeed is masculine, but grammatically it should be considered feminine, as the following (feminine) adjective מֹאֲבִיּוֹת indicates.

Posts 1264
LogosEmployee
Rick Brannan (Faithlife) | Forum Activity | Replied: Mon, Sep 24 2018 9:37 PM

Hi Michel.

I'll report the issue. My basic understanding of the Lexham Hebrew Bible morphology is that it analyzes form and not contextual grammatical function, but IANAHG* so I'll make inquries.

 - Rick

* IANAHG: "I am not a Hebrew guy"

Rick Brannan
Data Wrangler, Faithlife
My books in print

Posts 265
Michel Pauw | Forum Activity | Replied: Mon, Sep 24 2018 9:44 PM

Thanks, Rick. I would love to hear any results, as it affects searching a lot.

I think most people will search for grammatical gender, not for morphological gender. But I can be wrong, so I would love what's the philosophy of Lexham.

Posts 22
LogosEmployee
Jimmy Parks | Forum Activity | Replied: Tue, Sep 25 2018 9:01 AM

Rick is right on the money.

As you noted, Michel, the form is annotated correctly. This word, though it is feminine in the singular, takes a morphologically masculine ending in the plural. For a more detailed discussion see Biblical Hebrew Reference Grammar §24.2. Van Der Merwe mentions this case explicitly in §24.2.(iv).c.

Posts 113
LogosEmployee

FWIW, WIVU annotates as masculine as well, but Westminster annotates as feminine.

I'm not sure either the presupposition to annotate according to form or prototypical function works out perfectly either way. For example, in Job 41:22 (English 41:30) the presuppositions of LHB and WIVU get the form and function of תַּ֭חְתָּיו correct, while Westminster misses both form and function.

Posts 265
Michel Pauw | Forum Activity | Replied: Wed, Sep 26 2018 8:14 PM

Jimmy Parks:
Rick is right on the money.

As you noted, Michel, the form is annotated correctly. This word, though it is feminine in the singular, takes a morphologically masculine ending in the plural. For a more detailed discussion see Biblical Hebrew Reference Grammar §24.2. Van Der Merwe mentions this case explicitly in §24.2.(iv).c.

I know the differences between morphological, syntactical and semantic gender, and I realize נָשִׁים is:

  • morphologically masculine, since it has a masc plural suffix
  • syntactically feminine, since it takes a feminine adjective טוֹבוֹת
  • semantically feminine, since the meaning is 'women'

However, the question is: should Logos grammatically tag words according to their morphological gender, or according to their syntactical gender?

This is where I disagree with the choice being made. Logos has grammatically tagged נָשִׁים as being a masculine word (and there is some reason for doing that, since - as we saw - the form is masculine). However, since one of the basic rules in Hebrew is that adjectives agree in gender with the noun that it qualifies, I think it would be more reasonable to tag it as feminine.

The other aspect that pleads for the syntactical gender is that it then allows you to search for other (syntactically) 'feminine' with masculine endings, which now is not possible.

So my question is: has Lexham a rationale behind their choice for the morphological gender (against the choice for the syntactical gender) or hasn't it been thought through?

Posts 26104
Forum MVP
MJ. Smith | Forum Activity | Replied: Wed, Sep 26 2018 8:44 PM

Michel Pauw:
However, the question is: should Logos grammatically tag words according to their morphological gender, or according to their syntactical gender?

The answer is fairly clear: do you want to do a morphological search or a syntactic search? I want Logos tagging to be as "by the book" as possible rather than their hiding all the interesting quirks from me. Case in point are individual people where the evidence in the text is ambiguous regarding gender but Logos blythfully hides that fact from me ... making some commentaries/dictionaries more opaque.

Orthodox Bishop Hilarion Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."

Posts 113
LogosEmployee

Michel Pauw:

The other aspect that pleads for the syntactical gender is that it then allows you to search for other (syntactically) 'feminine' with masculine endings, which now is not possible.

Michel: I'm not sure what you are suggesting makes this search possible. If the words are tagged as feminine, then you are no longer able to search that they have masculine endings. The only way to do both things at the same time would be to have the words tagged both according to form and function.

Michel Pauw:

So my question is: has Lexham a rationale behind their choice for the morphological gender (against the choice for the syntactical gender) or hasn't it been thought through?

As it stands, though, the Logos morphology does allow you to get at more of what you are looking for than a tagging based on syntax would allow. For example, try the search "<LogosMorphHeb ~ NCMPA?> BEFORE 2 WORDS <LogosMorphHeb ~ JFPA?>" . Now, imagine that the first word in the sequence had been tagged according to syntactic gender. You would just be searching for feminines occurring before feminines. That search would return all kinds of ordinary feminine before feminine results that you are not interested in. So, tagging by form allows you to search for the mismatch better than tagging by syntax or semantics. This could be considered part of the rationale, unless I am misunderstanding you entirely.

Posts 113
LogosEmployee

For the sake of clarity, here is what the same search looks like in Westminster: <WestMorph ~ ncfp?+S???E?> BEFORE 2 WORDS <WestMorph ~ afp?+S???E?>. Note all of the occurrences of nouns with ות endings mixed in the with the nouns with ים ending due to the tagging of the nouns according to syntactic/semantic gender.

Posts 1003
EastTN | Forum Activity | Replied: Thu, Sep 27 2018 8:50 AM

MJ. Smith:

The answer is fairly clear: do you want to do a morphological search or a syntactic search? I want Logos tagging to be as "by the book" as possible rather than their hiding all the interesting quirks from me. Case in point are individual people where the evidence in the text is ambiguous regarding gender but Logos blythfully hides that fact from me ... making some commentaries/dictionaries more opaque.

I'm not sure it's all that clear.  You learn that some feminine words are irregular, and take a masculine ending in the plural, in first year Hebrew.  נָשִׁים was the first example of that in the class I took.  Unless someone really is looking specifically for words with a hireq yod mem ending, tagging "women" as masculine seems like an unhelpful result.

The parallel I would offer would be third declension nouns in Greek. You can't reliably determine the gender based on the endings alone - you have to memorize it separately for each word.  This is a case where you can't tell the accurately determine gender of a Hebrew word based on the ending alone - you have to memorize it.

Posts 265
Michel Pauw | Forum Activity | Replied: Thu, Sep 27 2018 8:48 PM

Jeremy Thompson (Faithlife):
Michel Pauw:

The other aspect that pleads for the syntactical gender is that it then allows you to search for other (syntactically) 'feminine' with masculine endings, which now is not possible.

Michel: I'm not sure what you are suggesting makes this search possible. If the words are tagged as feminine, then you are no longer able to search that they have masculine endings. The only way to do both things at the same time would be to have the words tagged both according to form and function.

(...)

As it stands, though, the Logos morphology does allow you to get at more of what you are looking for than a tagging based on syntax would allow. For example, try the search "<LogosMorphHeb ~ NCMPA?> BEFORE 2 WORDS <LogosMorphHeb ~ JFPA?>" . Now, imagine that the first word in the sequence had been tagged according to syntactic gender. You would just be searching for feminines occurring before feminines. That search would return all kinds of ordinary feminine before feminine results that you are not interested in. So, tagging by form allows you to search for the mismatch better than tagging by syntax or semantics. This could be considered part of the rationale, unless I am misunderstanding you entirely.

It does seem you understand the issue. However, such an proximity search is not what I am looking for, since it will return lots of irrelevant hits. What I would like to be able to achieve is:

  • @ncmpa INTERSECTS *ים, which returns all syntactically masculine nouns with the masculine suffix *ים
  • @ncfpa INTERSECTS *ים, which then should return all and only syntactically feminine nouns with the masculine plural suffix *ים
  • @ncmpa INTERSECTS *וֹת, which then should return all and only syntactically masculine nouns with the feminine plural suffix *וֹת

Check both out in Ruth (otherwise it might take a while), for the Lexham Hebrew Bible. The first works, the last two don't. 

This would be way more precise (actually, precisely what I wanted) than the search you suggested. As far as I can see it would also be the only way to exactly find these feminine nouns that have masculine plural suffixes.

Posts 265
Michel Pauw | Forum Activity | Replied: Thu, Sep 27 2018 8:59 PM

MJ. Smith:

Michel Pauw:
However, the question is: should Logos grammatically tag words according to their morphological gender, or according to their syntactical gender?

The answer is fairly clear: do you want to do a morphological search or a syntactic search? I want Logos tagging to be as "by the book" as possible rather than their hiding all the interesting quirks from me. Case in point are individual people where the evidence in the text is ambiguous regarding gender but Logos blythfully hides that fact from me ... making some commentaries/dictionaries more opaque.

That looks clear, but it isn't. It would be, if Logos would tag nouns with their syntactical gender, but I don't think they do. So we have to stick with the morph searches.

Again, if the morph search would adopt the syntactical gender for nouns, then these searches would be possible AND helpful.

  • @ncfpa INTERSECTS *ים, which then should return all and only syntactically feminine nouns with the masculine plural suffix *ים
  • @ncmpa INTERSECTS *וֹת, which then should return all and only syntactically masculine nouns with the feminine plural suffix *וֹת

The question is: can someone imagine a usage of the masculine plural suffix tag for feminine nouns? If so, then we need to weigh the (dis)advantages. If not, then I don't see why we should not change the tagging.

Posts 113
LogosEmployee

And, it turns out that you are a better searcher than I am Smile. I will bring up this issue with our core texts team to hash out and see what conclusions we come to.

For the time being, though, if this is a search that you need and Westminster is available to you, it is currently possible: <WestMorph ~ ncfp?>  INTERSECTS *ים

Posts 265
Michel Pauw | Forum Activity | Replied: Fri, Sep 28 2018 6:16 AM

Jeremy Thompson (Faithlife):

For the time being, though, if this is a search that you need and Westminster is available to you, it is currently possible: <WestMorph ~ ncfp?>  INTERSECTS *ים

Unfortunately I don't have the Westminster. But I would love to hear the outcome of your team discussion. 

Posts 265
Michel Pauw | Forum Activity | Replied: Fri, Sep 28 2018 6:16 AM

Jeremy Thompson (Faithlife):

For the time being, though, if this is a search that you need and Westminster is available to you, it is currently possible: <WestMorph ~ ncfp?>  INTERSECTS *ים

Unfortunately I don't have the Westminster. But I would love to hear the outcome of your team discussion. 

Posts 113
LogosEmployee

Hi Michel:

I just wanted to give a brief update. Our consensus was that we don't want to violate our morph tagging principle of annotating based on form for only one issue. If we handled issues like these, we would want to do so more comprehensively. For example, annotating based on function would almost certainly mean changes to our preposition tagging, participle tagging (i.e., functioning as nouns/adjectives), etc. I am sympathetic to the idea of wanting to have better morph tagging since my background is in Hebrew language. So, I will continue to look at what a more comprehensive approach might look like. But, there are no guarantees since this would be a matter of coming up with a proposal, gauging interest, etc.

Page 1 of 1 (16 items) | RSS