[sword-devel] TR and KJV

[DM Smith]

Warning: Long preamble before asking for advice. I’m really looking for advice or concurrence.

For Strong’s numbers in the KJV we used a 2000 copy of Maurice Robinson’s TR. (It is available via the Internet Wayback Machine at byztxt.com <http://byztxt.com/>, which no longer exists).

Robinson’s work marks variations between Stephens and Scrivener.

Since 2000, the file has been updated, correcting errors in the variants and improving the mapping of the Greek to Strong’s numbers. This most recent work can be found at: https://sites.google.com/a/wmail.fi/greeknt/home/greeknt <https://sites.google.com/a/wmail.fi/greeknt/home/greeknt>. The file is TR-PRSD.zip.

I’m in the process of updating to the new version.

I have compared it to the 2000 version and the 2000 version to the TR module that was used for the KJV2003 work and they agree with regard to the Greek. There are some minor differences regarding the handling of variants.

A few years ago, I update the KJV module to handle variants properly. Earlier, we had included the Strong’s numbers for the variant that wasn’t part of the tagging. This was a manual effort of a few hundred verses.

In analyzing the differences between the 2000 version and the latest version, I was seeing odd-ball stuff that I hadn’t noticed before. It was present in the 2000 version.
A) Occasionally there is a Strong’s number of 0. I still haven’t figured out what that means. In the last round I pitched them as invalid Strong’s Numbers.
B) there are variants of Strong’s numbers and morphology that are not marked.
C) Some Greek words are compound, having two or three Strong’s numbers and one morphology.
D) In this round, I also figured out that verbs have Strong’s numbers that are part of Robinson’s morphology as links to the root verb form. (Last round I pitched them as invalid Strong’s numbers)

The basic assumption we had previously was src, lemma, and morph had values that formed parallel arrays. These findings don’t fit with that.
The question is how should we handle these 4:
A) Strong’s number 0. I’m inclined to pitch these as not having value.
B) Variants of Strong’s Numbers and Robinson’s morphology.
These are of the form: greek word, Strong’s Number, {Robinson’s Morphology}, Strong’s Number, {Robinson’s Morphology}. I’ve broken these into 3 groups:
Same word, different morphologies:
Verbs:
Book	Chap	Verse	Pos	Content
Matt	26	45	10	καθευδετε 2518 {G5719} {V-PAI-2P} 2518 {G5720} {V-PAM-2P}
Matt	26	45	14	αναπαυεσθε 373 {G5731} {V-PMI-2P} 373 {G5732} {V-PMM-2P}
Matt	27	9	11	ελαβον 2983 {G5627} {V-2AAI-3P} 2983 {G5627} {V-2AAI-1S}
Mark	14	41	8	καθευδετε 2518 {G5719} {V-PAI-2P} 2518 {G5720} {V-PAM-2P}
Mark	14	41	12	αναπαυεσθε 373 {G5731} {V-PMI-2P} 373 {G5732} {V-PMM-2P}
John	1	9	10	ερχομενον 2064 {G5740} {V-PNP-ASM} 2064 {G5740} {V-PNP-NSN}
John	5	39	1	ερευνατε 2045 {G5719} {V-PAI-2P} 2045 {G5720} {V-PAM-2P}
John	15	18	6	γινωσκετε 1097 {G5719} {V-PAI-2P} 1097 {G5720} {V-PAM-2P}
Gal	3	7	1	γινωσκετε 1097 {G5719} {V-PAI-2P} 1097 {G5720} {V-PAM-2P}
Col	3	24	14	δουλευετε 1398 {G5719} {V-PAI-2P} 1398 {G5720} {V-PAM-2P}
Heb	13	23	1	γινωσκετε 1097 {G5719} {V-PAI-2P} 1097 {G5720} {V-PAM-2P}
Jas	2	24	1	ορατε 3708 {G5719} {V-PAI-2P} 3708 {G5720} {V-PAM-2P}
1John	4	2	3	γινωσκετε 1097 {G5719} {V-PAI-2P} 1097 {G5720} {V-PAM-2P}
1John	5	1	19	αγαπα 25 {G5719} {V-PAI-3S} 25 {G5725} {V-PAS-3S}

Others:
Book	Chap	Verse	Pos	Content
Matt	4	15	1	γη 1093 {N-NSF} 1093 {N-VSF}
Matt	4	15	4	γη 1093 {N-NSF} 1093 {N-VSF}
Matt	4	15	11	γαλιλαια 1056 {N-NSF} 1056 {N-VSF}
John	21	15	15	τουτων 3778 {D-GPM} 3778 {D-GPN}
Acts	4	 9	10	τινι 5101 {I-DSM} 5101 {I-DSN}
Rom	8	28	8	παντα 3956 {A-APN} 3956 {A-NPN}
1Cor	2	13	15	πνευματικοις 4152 {A-DPN} 4152 {A-DPM}
Jas	4	 5	11	το 3588 {T-NSN} 3588 {T-ASN}
Jas	4	 5	12	πνευμα 4151 {N-NSN} 4151 {N-ASN}
1Pet	1	 6	2	ω 3739 {R-DSM} 3739 {R-DSN}
2Pet	2	19	9	ω 3739 {R-DSM} 3739 {R-DSN}
2Pet	2	19	13	τουτω 3778 {D-DSM} 3778 {D-DSN}
Rev	3	17	9	ουδενος 3762 {A-GSN-N} 3762 {A-GSM-N}

Different Strong's numbers, same morphology:
Book	Chap	Verse	Pos	Content
John	6	19	5	εικοσιπεντε 1501 {A-NUI} 4002 {A-NUI}
1Cor	10	 8	12	εικοσιτρεις 1501 {A-NUI} 5140 {A-NPF}
2Tim	2	 6	5	πρωτον 4412 {ADV-S} 4413 {A-ASN-S}
Rev	5	 8	11	εικοσιτεσσαρες 1501 {A-NUI} 5064 {A-NPM}
Rev	5	14	9	εικοσιτεσσαρες 1501 {A-NUI} 5064 {A-NPM}

Should we pitch the variant or include it? We’ve not marked up variants of the attributes before. How would we represent that? I’m of the opinion that we should pitch the second variant. I’ve analyzed the verbs and it seems that the first matches the translation.

Note, that Jas 4:5 has a matter of article, noun agreement being identified. το πνευμα is either 3588 {T-NSN} 4151 {N-NSN} or 3588 {T-ASN} 4151 {N-ASN}. Similar thing in a few other verses. The choice of a variant in one verse may affect the selection of another variant in the same verse. Can’t mix and match.

C) Compound words
Book	Chap	Verse	Pos	Content
Matt	7	20	1	αραγε 686 1065 {PRT}
Matt	17	26	12	αραγε 686 1065 {PRT}
Luke	2	37	6	ογδοηκοντατεσσαρων 3589 5064 {A-GPN}
John	21	11	15	πεντηκοντατριων 4004 5140 {A-GPM}
Acts	11	18	10	αραγε 686 1065 {PRT}
Acts	17	27	7	αραγε 686 1065 {PRT}
Heb	3	 6	13	εανπερ 1437 4007 {COND}
Heb	3	14	6	εανπερ 1437 4007 {COND}
Heb	6	 3	4	εανπερ 1437 4007 {COND}
Rev	7	 4	7	ρμδ 1540 5062 5064 {A-NUI-ABB}

Should we represent these as a single word having multiple Strong’s numbers, breaking the notion of parallel arrays? Or repeat the Greek word and the morphology for each of the Strong’s numbers? That is, is src= indicating the word or the Strong’s number?

I think the most interesting one is the last, which is 144, but in the KJV is "an hundred and forty and four”. This is badly marked up as:
        hundred1540 <gdef:G1540>, ρμδ A-NUI-ABB <gmorph:A-NUI-ABB> and forty1540 <gdef:G1540>, ρμδ A-NUI-ABB <gmorph:A-NUI-ABB> and four1540 <gdef:G1540>, ρμδ A-NUI-ABB <gmorph:A-NUI-ABB> 
This is implemented via x-split because of the added word and.

Visually, it should probably be:
        hundred1540 <gdef:G1540>, ρμδ A-NUI-ABB <gmorph:A-NUI-ABB> and forty5062 <gdef:G1540>, ρμδ A-NUI-ABB <gmorph:A-NUI-ABB> and four5064 <gdef:G1540>, ρμδ A-NUI-ABB <gmorph:A-NUI-ABB> 
And not (since it is one underlying word):
        hundred1540 <gdef:G1540>, ρ A-NUI-ABB <gmorph:A-NUI-ABB> and forty5062 <gdef:G1540>, μ A-NUI-ABB <gmorph:A-NUI-ABB> and four5064 <gdef:G1540>, δ A-NUI-ABB <gmorph:A-NUI-ABB>

D) Strong’s numbers used for Morphology codes.
In Robinson’s morphology Strong’s numbers that are > 5624 are used for morphology. These are the old TVM codes. These are links to the real entry.
In each case, it is a reference to the root morphology of the verb.
For example (I’ve added G as a prefix and {} to surround for clarity):
καθευδετε 2518 {G5724} {V-PAI-2P}
The {G5724} has the following:
$$$G5724
@LINK V-PAI

What should we do with them? I don’t think these are of value.

Many thanks!

In Him,
	DM


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.crosswire.org/pipermail/sword-devel/attachments/20160201/79cdb202/attachment-0001.html>