inflected? - bug or what is an inflected search?

[Kristin]

I can't tell if this is a bug or if I am missing something. I had thought searching the inflected pulled up everything including the vowels. However, if I go to Gen 6:14 and search the inflected of the word for "pitch" it searches "כֹּֽפֶר" but in this inflected search it finds כְּפֹ֖ר (Ex 16:14) which has different vowels, and כֻּפַּ֣ר (Ex 21:30), which is a third set of vowels, along with כַפֶּר and כַפֵּ֖ר etc.

 

Am I missing something? What is an inflected search, if this is correct?

 

Thank you for any clarification anyone is able to provide,

Kristin

[Gary Raynor]

Hi Kristin

If you insert an = it will search for the exact vowels

"=כֹּֽפֶר"

 

The right click inflected search is defaulting to Do Not Use Exact From but if you chose inflected from (cmd J) in the search bar you have a choice.

This could be a feature request to add to the preference settings (it has a choice of flex or exact in Search Tabs)

 

[Donald Cobb]

Hi Kristen, of course, you're absolutely right. If you get different forms while searching for an inflected search, then by definition there's something wrong. I just tried your search and there is something strange going on here. I would point it out as a bug. 

 

 

Edited May 1, 2023 by Donald Cobb

[Dr. Nathan Parker]

Let me know if I need to file a bug report on this one.

[Donald Cobb]

8 hours ago, Nathan Parker said:

Let me know if I need to file a bug report on this one.

 

Yes Nathan, please do. Thanks.

[Dr. Nathan Parker]

Will Do. Thanks!

[Ken Simpson]

I don’t think this is a bug myself. An inflected form may have vowel changes due to pausal forms or Cantillation or vowel changes due to dageshes or articles or… (add your favourite vowel changing agents).

 

it’s important to have the ability to search for exactly what you want but I think a standard “search for inflected form” should search for the form that you would find “in the inflection chart” (for want of a better term) and disregard the vowel changes that happen for other reasons. 
 

but that’s just my 2c..,,

[Donald Cobb]

1 hour ago, Ken Simpson said:

I don’t think this is a bug myself. An inflected form may have vowel changes due to pausal forms or Cantillation or vowel changes due to dageshes or articles or… (add your favourite vowel changing agents).

 

it’s important to have the ability to search for exactly what you want but I think a standard “search for inflected form” should search for the form that you would find “in the inflection chart” (for want of a better term) and disregard the vowel changes that happen for other reasons. 
 

but that’s just my 2c..,,

 

Agreed, Ken. But that's not the problem here. Take a look at what the analysis comes up with when I do an inflected search on the word Kristen was discussing: 

 

Nombre total de versets = 87

  (nombre total de versets affichés = 87)

 

כֹּֽפֶר" (90 total de mots)

 

Nombre de formes différentes = 7:

(Triple-cliquer sur une forme pour voir ses occurrences)

 

כְּפוֹר־2 (כפר) frost = 2

כפר to cover, make atonement = 71

כְּפַר (כפר) Chephar = 1

כֹּפֶר־1 (כפר) village = 1

כֹּפֶר־2 (כפר) pitch = 1

כֹּפֶר־3 (כפר) henna = 1

כֹּפֶר־4 (כפר) ransom = 13

 

The word כֹּפֶר־2  is a hapax legomenon, so I should only come up with one occurence. It looks like the search is actually doing a root search. I come up with wildly different results using the Andersen-Forbes HT, but still with problems (two occurrences, but one that is a différent word with a different vocalisation.

[Gary Raynor]

Is  כֹּפֶר־2    a hapax legomenon?

pitch @ [HMT-W4 "=כֹּפֶר"] in most English texts returns one hit

If customise a similar search and analysis to @Donald Cobb to include INFLECT, we get כֹּפֶר  nine times which agrees with 9 hits in an exact inflected Hebrew search.

 

כֹּפֶר־1              (כפר)                village = 1

                                                כֹּפֶר = 1

כֹּפֶר־2              (כפר)                pitch = 1

                                                כֹּפֶר = 1

כֹּפֶר־3              (כפר)                henna = 2

                                                כֹּפֶר = 1

                                                כְּפָרִים = 1

כֹּפֶר־4              (כפר)                ransom = 13

                                                כָפְר = 1

                                                כָּפְר = 1

                                                כֹּפֶר = 6

[Donald Cobb]

Thanks Gary. I would distinguish between occurrences of the same word with a particular inflection and homonyms. Kristen can chime in here, but if I were looking for the different inflections of a specific word, which I think is what she was doing, I wouldn't include homonyms. And again, the search as Kristen did it, turns up results that shouldn't be there.

[Kristin]

I feel pretty confident at this point that we are dealing with a bug, since the results are not the same.

1) My original post was concerning that I was searching for the inflected of "pitch" in Gen 6:14, and I was getting a bunch of uninflected results, which @Donald Cobb also experienced.

2) @Gary Raynor, whether or not this is a hapax legomenon or not is how this all started when I first started searching for it. The ESV thinks so, as it gives it a -2, and I was trying to do research if I agreed.

3) The reason I think this is a bug though is because I am getting different hits, and we all seem to be. When I first tried it I got the uninfleced form, then I tried it again and only got Gen 6:14, and now I am trying it yet again and my only hit is 1 Sam 6:18. That has a -1 and does not include Gen 6:14. So that for sure is a bug if Gen 6:14 is not included in an inflected search of a word in the verse.

Ps - @Donald Cobb, I am just seeing your response as I was getting ready to post this, yes, I agree that the search should not include homonyms. The homonyms should be noted with a -1, -2 etc, but distinguished as such in an inflected search. I would like to mention though, that my search in the original post was pulling more than homonyms, as it had been including all sorts of verses with totally different vowels, which is not an inflected search as I had done.

Kristin

[Gary Raynor]

@Kristin To solve this, do a lexeme search.

Accordance is powerful - it does however (as with all powerful tools) have a steep learning curve to unlock its full potential.

I enjoy challenges such as this!

 

you can also choose a construct search to achieve the same result but an inflected search on its own (correctly) does not distinguish between homonyms.

[Kristin]

25 minutes ago, Gary Raynor said:

To solve this, do a lexeme search.

 

Hi @Gary Raynor,

Thanks for the idea, but the issue is that I had needed the inflected, not the lex.

 

26 minutes ago, Gary Raynor said:

you can also choose a construct search to achieve the same result

 

That is a good idea. I will try giving a construct search a try. I had always thought searching the lex or inflected was more accurate than a construct search, but it would be worth trying giving that the inflected obviously has a bug.
 

28 minutes ago, Gary Raynor said:

but an inflected search on its own (correctly) does not distinguish between homonyms.

 

Pulling homonyms would be totally fine as I had been looking for homonyms. The issue is that the so-called "inflected" search of כֹּֽפֶר had also pulled כַפֶּר and כַפֵּ֖ר and a bunch of other words, all with different vowels, which are not homonyms, yet had been coming up in the inflected search.

Kristin

[Gary Raynor]

Hi @Kristin

A combined lexeme and non-exact inflected search would be as follows:

"כפר"@=כֹּפֶר־2

The full equivalent search bar equivalent of the right click word construct search is:

"כפר"@=כֹּפֶר־2@ ‎[NOUN common masculine singular absolute]‏ @ ‎[COMPLEMENT]‏

I suspect that accordance is simply taking the top row which is INFLECT part out of the Construct Search!

Thus, not meeting a significant number of users' expectations of a Search For: Inflect Search.

[Gary Raynor]

To meet expectations, I think Search For: Inflect should either produce:

"=כֹּפֶר"

or provide a choice of exact or non-exact inflect searches.

This will not solve searching for homonyms, which would require more Search For: options.

Sometimes a lot of research needs to take place to achieve some goals (I have been frustrated many times) - thankfully we have this forum which is very helpful!