Help needed understanding search results in BHS text

Michael Burer

When I search for the Hebrew word אֶחָד I get lots of hits:

When I limit this to feminine forms I get no hits:

Two issues:

1. Other Bible search software which uses this same database on this same text shows 977 occurrences of this word, not 701. Why is there a discrepancy? The Logos search appears to be missing some hits.
2. Why does the search for feminine forms return nothing? There are definitely feminine forms of this word in the text.

Thanks in advance for help in working through this issue.

Vincent Setterholm

It's hard for me to test this one, because that is a very ancient version of the Westminster morphology database, and I don't have that on my PC anymore. The resource you're using was the 2.0 version of the Westminster database and the data must be at least 17 years old, since it was included in our 1996 base packages. (Don't be fooled by the 'resource' date in the support information area - that just refers to the last time the resource passed through the compiler, not the date of the database it was compiled from.) If I'm reading the vowel points right (they're not coming through very well on this screen), you're looking for the numeral 'one'. The latest version (4.18) of the Westminster database tags 976 instances in 739 verses, and you'd use the string: lemma:אֶחָד@u?f to find the feminine forms (273 in the 4.18). ('u' being the code for a numeral, rather than 'a' for adjective, the ? to ignore whether the word is tagged as cardinal or ordinal - all instances of this lemma are tagged cardinal, and the 'f' for feminine.)

The 4.18 is ready for release, so information should appear on our website shortly. But the last version of Westminster we released was the 4.2 (rather old now, but still much better than the 2.0), and the string you constructed seems to work on that version (if I'm reading your screenshot correctly), and that gets 276 hits. So there have been 3 gender reassignments between the 4.2 and the 4.18.

The old 2.0 version has a lot of issues besides just being very old tagging. That resource was made before we had support for homograph numbers or tables to do really good linking into lexicons. So I think you'll find the newer versions function much better.