Hebrew Syntax search with varying phrase depths

[AviShmidman]

Hi,

 

I am very much enjoying the power of the HMT Hebrew Syntax search via the Hebrew Construct dialog. One challenge I have come across, however, is in cases where I want to find a phrase based on two different criteria at two different depths.

 

For example: I would like to search for Predicate Phrases which contain a Verb and NullComplement in the very first level below the P node, and which do not have any relative particle anywhere at any depth within the nodes under the P node. For this, I need to set "Depth=0" in the Phrase node when searching for the Verb and NullComplement, but at the same time I need to set "Depth = infinity" when specifying that I don't want the relative particle. Yet, as far as I can tell, I can only set one value for Depth for the entire phrase item in the Hebrew Construct.

 

Is there any way around this? I've enclosed a screenshot of the query I am running. In the screenshot, the "Predicate Phrase" node is set to Depth = 0, which is correct for the Verb + NullComplement search. However, at the same time, I want to have the "NOT particle relative" applied to any depth within that same Predicate Phrase. Is there any way to do this?

 

Sincerely,

Avi Shmidman

 

[Robert Holmstedt]

Just wanted to jump in and say that I'll answer this tomorrow or Friday. The term begins tomorrow and I'm focused on class prep.

[AviShmidman]

Following up on this...

Roy/Helen - might there be any plans to allow for a search of this nature, specifying different depth levels for the different components of search?

[Robert Holmstedt]

It's been a busy week, but I've finally had time to look at this. 

 

Before I address the relative in your query, it's important to know that the way that depth works, you cannot simply put a complement within the predicate phrase and get accurate results at depth =0. In fact, any complement returned would be a tagging error, since the complement (null or overt) exists within the complement phrase that exists within the predicate phase. The hierarchy must be built correctly and strictly when using depth = 0. That is why the depth function was created, to allow for "squishy" searches, i.e., the kind where you're not sure what the hierarchy is or where you want to cast out more broadly. 

 

So, I'm assuming dept = 0 and will turn to your specific search. It seems to me you're looking for verbs with null comps, but specifically those that aren't within relatives (since null relative resumption exists quite often). In these cases, the relative word never exists within the predicate phrase, but stands at the front of the subordinate clause, so you could simply specify no אשׁר before the predicate phrase and you'll get 29x, but this includes 3 Aramaic examples with די (Dan. 3:32; Ezra 5:14; 6:5), so excluding both gives you 26x: Gen. 4:2; 16:1; Ex. 15:14; 19:4; Judg. 5:25; 1 Sam. 2:5; 7:9; Jer. 12:11; Joel 2:2; Hag. 1:9; Psa. 31:5, 7; 59:7-8, 15; 68:9; 75:10; 85:14; 104:8; 116:1; Job 31:31; 38:21; Prov. 17:5; Lam. 3:5; ; 1 Chr. 8:8; 29:8. (*I only scanned these to make sure they fit the pattern; I did not re-proof them carefully right now*)

 

Here's the search:

 

 

Now, Exod 19:4 is an interested example because the verb has a null complement head of a relative: ‏אַתֶּ֣ם רְאִיתֶ֔ם אֲשֶׁ֥ר עָשִׂ֖יתִי. If you're trying to also exclude these, then it becomes difficult, because a relative clause is an adjunct clause modifying the null complement, but you don't want to negate every possible adjunct. But specifying an adjunct clause (with a negated אשׁר) would exclude any complements that don't have an embedded adjunct clause. So to get the full set of results, you'd need to have two searches: one that includes an adjunct clause without an אשׁר and one that excludes adjunct clauses altogether. I'm still working on these two searches, but at least you get the idea.

 

 

[AviShmidman]

Dear Dr. Holmstedt,

 

Thank you! Yes, precisely as you said, the point of my query was to find verbs with null comps while excluding relative clauses. 

 

I appreciate your detailed and helpful answer, especially regarding the point that the relative word never exists within the predicate phrase, which solves the problem that I asked about.

 

Sincerely,

Avi