Hebrew search excluding terms giving unexpected results

[Ken Simpson]

HI,

I was playing around with wildcard searches the other day and I did a search that (generally speaking) searches for geminates.

 

so I typed into the search box (2=)??? which looks for any three letter root where the second and third root letter are the same.

 

Looks good, 3062 hits with 235 different forms on analysis.

 

Looking at the results I see a few names that I want to exclude from my list (like אֲגַג), so I check the tagging and I see that the tag for אֲגַג says NOUN proper

 

OK, so I then go the the search menu and put in [ANY -proper] (using the menu - not typing it in)

 

I did the search and got results where there were actually proper nouns. I then realised that using the search menu had not automatically inserted the @ symbol as I expected) - I checked this, and if I selected the [NOUN -proper] from the menu and it did insert the @. Is this expected behaviour?

 

So I manually typed in the @ symbol and did my search.

 

Looking good - 470 hits, 22 different forms, and none of them are proper nouns. (the same result as if I had put in [NOUN -proper] by the way)

 

But wait, a minute ago I had 235 different forms, and now only 22? Does that mean that the other 202 hits are all names? Let's check...

 

[ANY proper]@(2=)???

 

Hmmm, now that only shows me 25 different forms.

 

So when I search for any geminate that does have the tag "proper", and I add the numbers to the search for the geminate that doesn't have the tag "proper" I am still missing 188 forms.

 

What am I missing here? It feels like ANY -proper should be a much bigger group, but the search seems to be constrained to NOUNS even though I am using the ANY tag.

 

 

 

 

 

 

 

[Helen Brown]

It seems that Accordance does look at the parts of speech to which the ANY tag that you defined applies. Since proper only occurs in nouns, it finds the nouns and eliminates the proper nouns. I searched on [ANY -fem] in Gen 1-3 and it found masculine forms, but not any of the particles which have no gender. I posted the results below. I think this logically makes sense, but for you it means that you will have to use the HITS command to eliminate your proper nouns (i.e. search for proper nouns in another tab, and use HITS to remove those results from the search:

[HITS HMT-W4 2]-@(2=)???

 

ADJECTIVE = 46
Masculine = 45
Both = 1
NOUN = 441
(no Gender) = 36
Masculine = 373
Both = 32
PRONOUN = 17
(no Gender) = 3
Masculine = 13
Common = 1
SUFFIX = 90
(no Gender) = 2
Masculine = 80
Common = 8
VERB = 246
(no Gender) = 29
Masculine = 199
Common = 18

[Ken Simpson]

Thanks Helen,

that makes sense, though not "intuitive sense" if you know what I mean.

 

Thanks for the explanation