Is there an easy way to search for hapax legomena in the NT books?

[A McPherson]

Dear fellow forum users,

I have been directed by "Jay" in Accordance to ask this question:

Message: Can you tell me to whom I would direct an important and time-sensitive question regarding the best way to search on hapax legomena in NT /NT Pauline epistles? I found an article online (from a LOGOS user) who had presented every hapax legomena in the NT into a Word document—which I now have. For my PhD dissertation, I would like to know how easy it is to search in Accordance for hapax legomena in the NT, especially Colossians and the gospel of Luke / Acts? Sincerely, Alistair McPherson, PhDc (candidate)

[Dr. Nathan Parker]

Are you needing this in Greek? If so, I may move this to Original Languages for you. You're likely to get an answer there quicker.

[mgvh]

Easy peasy!

In NA28, set search for Words

Use the COUNT command and set to 1

I got 1940 hapax

BTW, that is a lexeme search (not an inflected form search)

Edited November 25, 2023 by mgvh

[Dr. Nathan Parker]

Thanks! I may have to do a QuickTip video on this one too. :-) 

[Lawrence]

@A McPherson

 

If you're interested in just a subset of the NT (e.g. just Colossians), be sure to do mgvh's search on the whole NT first, then choose what you want from (say) Colossians. Otherwise, you'll get the words that appear once in Colossians but there's no guarantee that they don't also appear in (say) Philippians.

[mgvh]

To get hapax in Colossians that are hapax only in Colossians, add a Range:

 

To quickly get hapax in Colossians that are hapax in the NT as well, use RANGE in the command line.

[COUNT 1] [RANGE Col]

[Steven S]

5 hours ago, mgvh said:

To quickly get hapax in Colossians that are hapax in the NT as well, use RANGE in the command line.

[COUNT 1] [RANGE Col]

 

I've always been curious if Accordance guarantees a particular order of operations in the search bar. I assumed the above search works as intended because it applies COUNT to the entire NT first, and then extracts the RANGE. However, reversing the order of the commands (i.e., [RANGE Col] [COUNT 1]) produces the exact same number of hits for me.

 

So I'm guessing Accordance has a preferred order in which individual subcommands are executed. Does anyone know if this is documented somewhere?

[Lawrence]

@Steven S

 

I don't know whether a canonical order is documented somewhere, but "[x] [y]" is equivalent to "[x] <AND> [y]" in Accordance.

That's effectively set intersection, which would explain why order doesn't matter in that case.

 

To impose an ordering, do mgvh's trick (which is cool!).

Another idea is to do a search in one tab, then a second search in another tab that references the first tab. You may be able to daisy-chain the tabs to do deeper sequential searches. With these more funky searches, you'll need to know what you're doing - or just have fun playing with it.

Edited November 25, 2023 by Lawrence

[Steven S]

@Lawrence, thanks for that summary; it was very helpful!

 

43 minutes ago, Lawrence said:

"[x] [y]" is equivalent to "[x] <AND> [y]" in Accordance.

That's effectively set intersection, which would explain why order doesn't matter in that case.

 

Yeah, I think the way I was originally thinking of this search was as a pipeline of operations that incrementally whittle down the available search space. That is, beginning with [1] all words in the NT, to [2] all hapax in (1), to [3] all words in (2) that appear in Colossians. Obviously, in such a pipeline, reversing [2] and [3] would produce different results.

 

But, given your point about how Accordance interprets this as a set intersection, I'm guessing that it's executing two separate searches and then finding the intersection between the two. That is, search [1] is all hapax in the NT, search [2] is all words in Colossians, and the result is the intersection of [1] and [2], which is indeed commutative. That would seem to make the best sense of what I'm observing. Thank you!

[A McPherson]

On 11/24/2023 at 8:23 PM, Nathan Parker said:

Are you needing this in Greek? If so, I may move this to Original Languages for you. You're likely to get an answer there quicker.

Yes, thank you. I appreciate the quick response and learning points.

[A McPherson]

On 11/24/2023 at 11:32 PM, mgvh said:

To get hapax in Colossians that are hapax only in Colossians, add a Range:

 

To quickly get hapax in Colossians that are hapax in the NT as well, use RANGE in the command line.

[COUNT 1] [RANGE Col]

Thank you very much, mgvh! With a little tweaking, I got it!

[A McPherson]

On 11/25/2023 at 8:19 AM, Lawrence said:

@Steven S

 

I don't know whether a canonical order is documented somewhere, but "[x] [y]" is equivalent to "[x] <AND> [y]" in Accordance.

That's effectively set intersection, which would explain why order doesn't matter in that case.

 

To impose an ordering, do mgvh's trick (which is cool!).

Another idea is to do a search in one tab, then a second search in another tab that references the first tab. You may be able to daisy-chain the tabs to do deeper sequential searches. With these more funky searches, you'll need to know what you're doing - or just have fun playing with it.

Thank you very much, Lawrence. This is most helpful.

[A McPherson]

On 11/24/2023 at 11:00 PM, Lawrence said:

@A McPherson

 

If you're interested in just a subset of the NT (e.g. just Colossians), be sure to do mgvh's search on the whole NT first, then choose what you want from (say) Colossians. Otherwise, you'll get the words that appear once in Colossians but there's no guarantee that they don't also appear in (say) Philippians.

Thank you for this 'tip', Lawrence. It makes sense also to do a [COUNT 1] search on the entire NT first, then choose Colossians.