Bug with clauses search? Or my mistake?

[Robert Holmstedt]

You are trying to search for unspecified clauses within a single higher clause. This is simply too complex for the search programming at the current time.

Edited January 13, 2023 by Robert Holmstedt

[Robert Holmstedt]

The searching for nulls is still not perfect (cue plea to programmers!).

In the cases in Gen 1:7 and 1:16 there is null subject followed by a null predicate immediately preceding the preposition. This has apparently caught the search, even though it was set up for *only* null PREDICATES.

The problem is greater for the two cases in Gen 1:11 and 1:12, since there is only a single null item in proximity to the prepositions. 

(Repeated plea to programmers to address the thorny null issue.)

 

[Robert Holmstedt]

There's a bug here. I get many hits from the search, but it misses most of the ones shown in the old screenshot and has numerous mis-hits. 

[Robert Holmstedt]

Yeah, that's a definite bug. Should be hundreds of each. 

 

[Robert Holmstedt]

That's not the fault of the syntax database; we had to build on the Westminster morphology and how they determined the tagging of names (not always consistent, if I remember correctly) affects the syntax hits. 

[Accordance Enthusiast]

Thank you so much for the very clear explanation!

 

On 1/14/2023 at 8:01 AM, Robert Holmstedt said:

two equal level clauses tagged under one clause node

 

So good to know and understand this.

 

Thank you

Shalom

[Accordance Enthusiast]

On 1/14/2023 at 8:02 AM, Robert Holmstedt said:

This is simply too complex for the search programming at the current time

Thank you very much for the clarification.

 

I am hoping that this would be improved / made possible in future?

 

Shalom

[Accordance Enthusiast]

On 1/14/2023 at 8:01 AM, Robert Holmstedt said:

The bullets will be searched as distinct clauses like any other distinct clause.

 

Excellent. So, they're distinct but not subordinate.

 

Thus, would Gen. 45:8 be three or two distinct clauses all together?

 

I reckon it's three, but not absolutely sure.

Edited January 15, 2023 by Anonymous

[Accordance Enthusiast]

Also, it seems that there is some confusion in the nomenclature used in Accordance. Either that, or a bug?

 

If I set search depth to 0, it still finds hits across distinct phrases, as shown above also:

 

 

 

So, though it has the depth set to 0, the allow hits in "sub-clauses" allows the search to find the hits across two phrases on the same level. 

Though not subordinate as @Robert Holmstedt said above, this is still called a "sub-clause" in Accordance?

 

You see, if I set this to "don't allow hits in sub-clauses", ONLY THEN is Gen 45:8 excluded from the search:

 

 

 

 Now it gives the correct results. Because sub-clauses on the ZERO level are also excluded. 

 

Thus, in Accordance, these bulleted clauses are called sub-clauses, but on the 0 level.

[Accordance Enthusiast]

The other thing is that the automatic construct collapses these "distinct clauses" into one clause:

 

 

The result is:

 

 

With only one clause at the top, though it should have been either 2 or 3?

 

By the way, this search finds no results at all. 

 

 

Edited January 15, 2023 by Anonymous

[Accordance Enthusiast]

So, my conclusion is that these bulleted clauses are so called "sub-clauses" on the same level of syntax, and are thus often treated as one continuous clause, at least in searching. If I am wrong, please could someone correct me.

[Accordance Enthusiast]

I tried duplicated the first clause in this verse:

 

Here is my search:

 

Results don't include Psalm 2:8

 

[Accordance Enthusiast]

It seems the results can't and don't find any hits where bullets appear in the clause level.

 

Is this a bug or how should I do this?

Any advice would be appreciated!

[Robert Holmstedt]

Gen 45:8 has no sub-clauses. There are 2 clauses in the screenshot that is shown. There are an additional 3 clauses in the rest of the verb. Those three have ellipsis of the verb, which is precisely why we contain distinct clauses within one N label (this happens a lot of poetry, of course, but also in narrative, as we see in this verb and I have also published about).

The search programming is not supposed to see beyond the "bullet" points at all and so the results in the first part of Gen 45:8 are a bug, plain and simple. 

[Robert Holmstedt]

As I have replied in another post, you cannot do this. The database was never designed for this. This is an attempt to search beyond the clause boundaries of clauses at the same level. Such a search would only work when the "antecedent" function is finally programmed in (it is currently not). Even then, it would only find examples of across-clause-boundaries ellipsis, such as verb gapping, or multiple relative clause stacking on a single relative head. 

 

[Accordance Enthusiast]

2 hours ago, Robert Holmstedt said:

the results in the first part of Gen 45:8 are a bug, plain and simple

 

Thank you very much for the clarification.

 

So, I was trying to figure out something that's really only a bug.

[Accordance Enthusiast]

Here is the search. Accordance complains that it is invalid.

 

Is this a bug?

[Accordance Enthusiast]

If so, does anyone know how I can find the first part of Ps. 2:8 in a construct search?

 

Thank you

[Dr. Nathan Parker]

Possibly. By the way, if you’re the same person posting multiple threads about syntax searches, it would be easier to combine more of your requests into a single thread. That would make it easier for Robert to answer any syntax-related questions you have, as well as I’d have a consolidated thread I can show to developers on actual bugs. Seeing a bunch of separate posts on the forums from an “anonymous” user tends to look like we’re being spammed. I can merge all of your threads into a single thread if need be. In the future, try to consolidate as much as you can into one or a couple of threads.

[Dr. Nathan Parker]

I went ahead and merged everything into a single post. Any other potential syntax bugs you find, please append them to this post. That’ll give us a cleaner way to report issues to developers.

[Accordance Enthusiast]

On 1/16/2023 at 8:42 AM, Anonymous said:

If so, does anyone know how I can find the first part of Ps. 2:8 in a construct search?

 

Thank you

 

Please, I am trying to confirm whether this is another bug, and whether this can be done.

 

I want to find the first part of Psalm 2:8 in a construct search. Is it possible to find the first part of Ps. 2:8 in a construct search?

[Dr. Nathan Parker]

I’ll leave that one for Robert to look into.

[Accordance Enthusiast]

Good day Mr. @Robert Holmstedt.

 

I am trying to confirm whether this is a bug, or whether it is simply not possible.

 

I want to find the words "שאל ממי" in Psalm 2:8 in a syntax search in the applicable clause and phrases.

 

Is this possible? I've tried manual searches and automatic construct searches and can't find Ps. 2:8 in my results. 

 

If you could confirm whether this is a bug on Windows or whether I am doing something wrong, or whether this is just not possible I would really appreciate it.

 

Shalom.

 

[Accordance Enthusiast]

On 1/16/2023 at 6:32 AM, Robert Holmstedt said:

Gen 45:8 has no sub-clauses. There are 2 clauses in the screenshot that is shown

 

Thus, the following screenshot shows only one clause? I'm wondering why the first two words are not under the bullet if they are part of the clause. 

 

[Robert Holmstedt]

The first two words are not "under" the clause because they are at the conjunction level. 

 

On Ps 2:8, I'm running into some obvious searching bugs that have appeared since I last put the syntax searching through the paces (I increasingly simply use the underlying database). I was able to get Ps 2:8 as a hit on the two searches below. But -- critically -- NOT when I put both the complement AND adjunct into the search (and this SHOULD work).

Additionally, when I looked just for the imperative verb followed by a null subject, I got similar hits (e.g., Deut 4:32, Judg 18:5, 1 Kgs 2:20; Jer 18:13) but NOT Ps 2:8. Clearly another bug. I've done some digging and the bugs are not in my database. 

 

 

Edited January 17, 2023 by Robert Holmstedt