Creating NON-agreement in searches of definitiveness

[Bielikov]

In the Construct Panel, how do I set up a search for non-agreement in definiteness, such as between an adjective and a noun, or a noun and a pronoun.

[Helen Brown]

This is a tough one! First you will need to use multiple constructs to search separately for the variants, such as definite noun or adjective  followed by indefinite adjective or noun, and the reverse. I don't think any one search will find all of these since definite is not a tag on the words but depends on the preceding article or particle-article. This is my attempt to fond the first case, but it fails since it simply skips the second article rather than eliminating the hit. I am not sure at this point whether this is a bug.

 

[Helen Brown]

Joel helped me but we could not define an accurate search. It really needs to use the Syntax and place the items within a phrase. Also the negative article/preposition needs to be defined in an INTER since putting it in a column as I did just means that the preceding word should not be what is defined there. He tried the search but every case we looked at could be explained because the adjective was part of a construct, or was the complement to an implied verb, or another type of grammatical form. Let me explain with English words:

 

In Hebrew "the big dog" is expressed as "the dog the big."

 

I think you are looking for cases that would be considered "incorrect" such as "the dog big" or "dog the big."

 

But the search kept finding "the dog was big" where was is implied, or "the dog big the very" (not a good example, but I mean to illustrate a construct form of the adjective). The reverse also missed "his dog the big" which is correct  in Hebrew, so possessives and probably other specifiers need to be accounted for.

 

So unless some real Hebrew scholars can help you here, I think an accurate search is basically impossible to define. I am getting rusty in Hebrew grammar.

[Robert Holmstedt]

Using the syntax module, you would set up 2 types of searches that mix syntax and morphology. The 2 types are for the 2 orders: 1) indefinite noun - definite adjective; 2) definite noun - indefinite adjective.

 

Below is the Type 1.

 

Note that this one gets a few irrelevant its due to the complexity of a verse (1 Kgs 6:34) and the inability to specify agreement between the noun and the adjective within the adjunct phrase (i.e., the hits at Job 28:13 [which really must be a morph tagging error -- ארץ is certainly the bound form here] and 1 Chr 19:8 [a syntax error that I've fixed so that הגבורים is an appositive ].

 

Below is Type 2. 

 

Note that this one requires a variation in which an attached (possessive) pronoun version is created. Also note that there is a clear searching error for this one — there are cases of appositives as the hits instead of adjuncts (e.g., 1 Kgs 18:19, "the prophets of Ashera, 400") -- maybe Joel can take a look at this. 

 

And finally, for what it's worth, I discuss the syntax of the lack of definiteness agreement in my book on the relative clause, Holmstedt 2016: 69-77.

[Bielikov]

This is a tough one! First you will need to use multiple constructs to search separately for the variants, such as definite noun or adjective  followed by indefinite adjective or noun, and the reverse. I don't think any one search will find all of these since definite is not a tag on the words but depends on the preceding article or particle-article. This is my attempt to fond the first case, but it fails since it simply skips the second article rather than eliminating the hit. I am not sure at this point whether this is a bug.

 

Definiteness.png

My apologies for not better explaining the purpose of my search in the construct window. I want to look for both demonstratives used as adjectives, "this man," "these women," and so on, in contrast to demonstratives used as pronouns, "this is the man," "these are the women." So a follow-up question would be, what am I doing wrong when I build my construct windows and put in the furthest right column the pronoun, and in the next column to the left of it, the noun, and the search does not respect that word order? I do not have a check on the both directions, by the way. Thanks, Helen. By the way, I have attended many of the webinars that Accordance has put on, and have been loving them.This is a great service for which I am most grateful.

Using the syntax module, you would set up 2 types of searches that mix syntax and morphology. The 2 types are for the 2 orders: 1) indefinite noun - definite adjective; 2) definite noun - indefinite adjective.

 

Below is the Type 1.

Screen Shot 2020-04-24 at 9.12.55 PM.png

 

Note that this one gets a few irrelevant its due to the complexity of a verse (1 Kgs 6:34) and the inability to specify agreement between the noun and the adjective within the adjunct phrase (i.e., the hits at Job 28:13 [which really must be a morph tagging error -- ארץ is certainly the bound form here] and 1 Chr 19:8 [a syntax error that I've fixed so that הגבורים is an appositive ].

 

Below is Type 2. 

Screen Shot 2020-04-24 at 9.22.43 PM.png

 

Note that this one requires a variation in which an attached (possessive) pronoun version is created. Also note that there is a clear searching error for this one — there are cases of appositives as the hits instead of adjuncts (e.g., 1 Kgs 18:19, "the prophets of Ashera, 400") -- maybe Joel can take a look at this. 

 

And finally, for what it's worth, I discuss the syntax of the lack of definiteness agreement in my book on the relative clause, Holmstedt 2016: 69-77.

Robert, wow, this looks very interesting. I will certainly try these out and report back. Many thanks!

[Bielikov]

Using the syntax module, you would set up 2 types of searches that mix syntax and morphology. The 2 types are for the 2 orders: 1) indefinite noun - definite adjective; 2) definite noun - indefinite adjective.

 

Below is the Type 1.

Screen Shot 2020-04-24 at 9.12.55 PM.png

 

Note that this one gets a few irrelevant its due to the complexity of a verse (1 Kgs 6:34) and the inability to specify agreement between the noun and the adjective within the adjunct phrase (i.e., the hits at Job 28:13 [which really must be a morph tagging error -- ארץ is certainly the bound form here] and 1 Chr 19:8 [a syntax error that I've fixed so that הגבורים is an appositive ].

 

Below is Type 2. 

Screen Shot 2020-04-24 at 9.22.43 PM.png

 

Note that this one requires a variation in which an attached (possessive) pronoun version is created. Also note that there is a clear searching error for this one — there are cases of appositives as the hits instead of adjuncts (e.g., 1 Kgs 18:19, "the prophets of Ashera, 400") -- maybe Joel can take a look at this. 

 

And finally, for what it's worth, I discuss the syntax of the lack of definiteness agreement in my book on the relative clause, Holmstedt 2016: 69-77.

Hi Robert, WOW! I just finished replicating your construct window and it is exquisite. It found, for Type I, 686 cases. I was worried at first as it first gave an error message saying there was nothing, but then it when through four or so search times and eventually yielded a mine of information. In one of my simple construct windows I had also done the negation on ALL as it was giving me many of these instead of what I was searching. But that was some while ago on a different search. Many, many thanks for your work, especially now that you are so buy grading papers. Many blessings to you. I will see if I can adapt this, to do the same type of thing with demonstratives used as adjectives.

[Bielikov]

Robert, I did it! I replaced adjective for demonstrative, and I have it working. I am beyond, beyond thrilled. Thanks again for all your work. This will keep me busy for weeks. What reference source can you recommend to me to learn more about the most interesting construct window you developed? I understand about 3/4 of it, but not all. Many thanks from Chile.