The individual vowels, tsere

Continuing this series of exploring Hebrew vowels through software. This has been a remarkable journey for me. There are a few easy vowels (qubuts, holam, and segol) that I will pass over for the time being. 

The two remaining difficult ones that have taken the most code in my program are hireq and tsere. Tsere is the easier of these two. (38 references in the code to start compared to 99 for hireq.)

Tsere like other vowels plays a role in the double vv and several common suffixes. For tsere and i, I begin with a generic rule: for stems beginning with /i/ of length greater than 2 with prefix v||tsere||i or prefixes beginning with i||tsere with or without dagesh, the tsere generates a double i to begin the word.

That's quite a mouthful - and there are plenty of ways in which my programming thoughts and fingers might have made this more contorted than necessary. So I tested it by removing the code and the whole statement can be eliminated. I am frankly, astonished! 

If I am this illogical in programming where I can test my statements, what is my theology like?  That's worth a laugh.

Tsere begins now with the removal of the combination tsere yod. These are essentially spelling adjustments. A few of these removals have to be restricted by affix combination. It could be that this is always the case, but some word forms never occur so I can't tell if a generalization will be a worse or a better predictor of further behaviour when I have more data.

Similarly to yod, tsere followed by aleph sometimes makes the aleph disappear. When is a mater a mater?

All the following stems convert every tsere in the word to ii without exception. amn, amr, avh, blh, dah, hrs, idy, iin, iph, iqd, iqr, ird, irq, iry, ixa, iwb, kih, klx, kmm, kpz, qvm, rah, rcc, rdh, rdm, rgn, rkq, ryy, spn, ydi, yip, yni, yvr, ywh, yxb. 

That's a nice simple rule. This affects 130 words in the test data. (0.3%) 

In some stems [bar, aim, brc, krp, krw, prw, ram, yrm] and with some conditions on word forms, only the first tsere under the first letter of the stem is replaced with an i. 

In some the first tsere in the word is replaced with an i. These are +rp, +vb, ail, alm, bnh, brc, csh, hvm, irh, iwn, iwn, kth, kwb, kwc, man, ndb, nsc, pl+, pnh, rah, rgn, rpa, wcr, wpl, wrp, wrw, wyr, yrm, ywh, yxr, zrh. In all these cases there are conditions on the word form.

So far the stem - word form combination seems to allow determination of the standard spelling. Only in 10 instances, as I noted earlier, did I find conflicts where the exact same word form resulted in two different spellings. 

ryh (friend, associate) is a good example, resulting in 7 words that cannot be resolved. There are another 9 pairs of this type. You can see I resolved this one to maximize the 'right' answer. When I do my tests, I now show the different classes of discrepancy in various highlights.

Reference	stem	shape	Req.	Calc.	Hebrew
Genesis 11:3(4)	ryh	ryh\v	-ryhu	-riyhu	־רֵעֵהוּ
Jonah 1:7(4)	ryh	ryh\v	-riyhv	-riyhu	־רֵעֵהוּ
Psalms 12:3(5)	ryh	ryh\v	-riyhv	-riyhu	־רֵעֵהוּ
Jeremiah 36:16(9)	ryh	ryh\v	-riyhu	-riyhu	־רֵעֵהוּ
Ruth 3:14(10)	ryh	ryh\v	-riyhv	-riyhu	־רֵעֵהוּ
Psalms 101:5(3)	ryh	ryh\v	riyhv	riyhu	רֵעֵהוּ
Genesis 11:7(11)	ryh	ryh\v	ryhu	riyhu	רֵעֵהוּ

Enough on tsere. Now to hireq, the most complex arrangement of rules.