Considering the maqaf

The maqaf is in English a hyphen. So Here is an example: Amos 8:10

Amos 8
וְהָפַכְתִּ֨י חַגֵּיכֶ֜ם לְאֵ֗בֶל וְכָל־שִֽׁירֵיכֶם֙ לְקִינָ֔ה וְהַעֲלֵיתִ֤י עַל־כָּל־מָתְנַ֙יִם֙ שָׂ֔ק וְעַל־כָּל־רֹ֖אשׁ־קָרְחָ֑ה וְשַׂמְתִּ֙יהָ֙ כְּאֵ֣בֶל יָחִ֔יד וְאַחֲרִיתָ֖הּ כְּי֥וֹם מָֽר	i vhpcti kgicm labl vcl-wiricm lqinh vhyliti yl-cl-motniim wq vyl-cl-raw-qrkh vwmtih cabl ikid vakrith ciom mr	י והפכתי חגיכם לאבל וכל־שיריכם לקינה והעליתי על־כל־מותניים שק ועל־כל־ראש־קרחה ושמתיה כאבל יחיד ואחריתה כיום מר	10 And I will change your festivities to lament and all your songs to a dirge and I will bring up over all endowments, sackcloth, and over every head, baldness. And I will set it as a unique lament and following it, as a day of bitterness.

Lambdin writes that the maqaf (maqqep in his spelling), indicates that a preposition is proclitic, i.e. has no stress of its own. 

Does it affect the music? The music determines stress much more than the maqaf. And it is clear from bars 2 to 4 that an avoidance of inner-word stress is not necessarily justified. (I am glad to see that my music program ignores the maqaf entirely. Note that in music, the hyphen joins syllables separated by note changes, whereas in Hebrew the maqaf joins separate words.)

Nonetheless, my attitude towards this jot may have been a little too dismissive in the past.

Hireq, when and when not to render it in standard spelling

This post continues my exploration of the transformations of individual vowels from a pointed text to an unpointed one. 

Here we look at hireq, when it is ignored and when it appears. There was an overall strategy. I had several rules based on prefix to suppress hireq. They almost all disappeared when I figured out the closed syllable rule.

1. Suppress hireq
1. when there is an /h/ in the prefix and the stem is in hlc, hll, clm, ixg, ixt, lvi, npl, pla, rah, rgy, wlc, wmy, wgh, wqh, xmt, yll.
2. or the stem is in hlc, irw, nwa, rbb with conditions for some affixes.
2. Suppress hireq when it is in a closed syllable, 
1. if the last character of the prefix is /n/ 
1. and stem is one of ptk, ckd, clm, csh, dmh, mlt', pla, pzr, rah, rpa, wbr, wck, wmy, and the first syllable of the word is closed with a schwa.
2. or any stem with a weak first character and the syllable is closed with a hatef-patah (only one instance of this).
2.  if the stem is not in /n/ and the stem is not in bqw, clm, csa, gll, irw, igy, lbb, nwa, rxp, qll, qvh (or it is qvh with the last character of the prefix a /t/).
1. There is an exception for words from stems hlc, wal with a final schwa.
3. Allow hireq to be come 'i*i' - (which will later translate to double i).
1. for words beginning with v||schwa||i||hireq where the stem begins with i and the word is not the name of a person, location, or people (I do have this explicit semantic information in my data),
2. for some stems formed with weak consonants (ibw, iry, ixr, n'ty, ndr, lvn, ain, ild, nwa) with conditions on the affixes,
3. for the long list of stems (see this post) that allow i with prefixes i, vi, iv, hiv (happens with stem idy),
4. Allow hireq to become i 
1. for several lists of stems with conditions on the affixes.
2. for stems with yod as part of the stem, excluding any stems, another long list, where hireq is explicitly prevented.
5. Much later in the process, the last steps, 
1. prevent double i where the stem does not allow it.
2. prevent the weak hireq from forming a syllable with a common single or double prefix unless this is overridden for a stem.

Though the prevention logic has simplified considerably, this whole process seems too complex still. 

Summary of the rules - a point of stability

Continuing to explain to myself this program I am writing in response to the transformation of a pointed Hebrew text into one that conforms to standard hebrew spelling.

Right now I have 143 explicit named 'rules' governing subsets of Hebrew stems and their various word forms. That means 143 distinct sets of transformations that apply to a set of stems.

Do they make sense? 

Here for example is the first entry below: gvi, irw, mdi each allow double i /ii/ in the calculated SimHebrew word, whatever its origin. That is the only 'rule' applied to these stems. I see that each of these stems has an /i/ in it, so this collection is not contradictory. Allow double i would be a silly rule for say the stem dbr since dbr by default does not allow i. For dbr to appear with allow double i, there would have to be another rule attached to it that allowed i for some condition.

The next rule collection is allow double i, allow i and the two stems that have this rule are ml't and yl, neither of which allows i as a default, so again the rule is self-consistent. 

A little further down I saw rules I would question just as I noted above. How can words that don't allow i get to have a double i? Sure enough, when I removed those from the rule, it made no difference to the results.

A little further down, there was a rule that combined allow_i with allow_i_v. The rules that allow i as well as allowing i for prefix i are OK since i with prefix i translates into a forced double i. But to allow i with prefix v is a subset of allowing i, so the rule is redundant. Equally if i is allowed implicitly and not explicitly prevented, then rules that are subsets of allowing i can be removed. I removed about 11 such redundancies.

Remember Occam's razor.

And so on. What I should do at some point is remove all the specific references to stems as rules from the code and put them into a database as I have in these now 143 cases. Then I could invent some additional language analysis. Maybe when it's easier to do that than just adding more rules I might do it. (Programmers take the route with the fewest obstacles.)

A word about o. The default is to allow it, so if it isn't wanted, a rule must specifically prevent it. So there is a rule below entitled prevent_o. It applies to 9 stems in the test data. I have been in tension whether to name every rule. A single rule that has exceptions for every stem presents a problem. Too many rules! So I oscillate between those I can easily remember and a search for specific combinations of letters and vowels. 

The following has been updated after the inclusion of another 8 chapters. 

Rule Combination                                                      Applies to these stems

allow_dbl_i,	gvi, irw, mdi
allow_dbl_i, allow_i,	ml+, yl
allow_dbl_i, allow_i, hatef_qamats_vav,	yvp
allow_dbl_i, allow_i_c, qamats_1_vav,	zvh
allow_dbl_i, allow_i_h,	abh
allow_dbl_i, prevent_i, qamats_yod_ii,	hih
allow_dbl_i, prevent_i_h,	ixt
allow_i,	+vb, acr, amx, at, bqw, brk, btq, bwr, bzr, ckd, clh, cn, cnr, csa, csh, ctr, cys, dbh, dlh, dvd, gbr, gll, gn, hnm, k+h, kbl, kc, kdql, kmd, kmw, knm, kth, ktt, kx, lb, lpd, lvn, lyg, m+h, mdd, mgn, mvab, nax, nba, ngr, nkl, npx, nqb, nqh, nsh, nsy, ntc, nym, pnnh, psl, pss, ptk, ptt, pzr, q+r, qnm, qxx, rbb, rmvn, rpa, rxp, rxx, sll, snsn, tcn, tmr, w+k, wn, wqx, wvb, wvh, ww, wyy, xhr, xma, xmk, xmq, xnh, xnr, xph, y+r, yqb, ywr, yxm, yzz, zch, zmh, zmm, zmr
allow_i, allow_i_i,	bxr, glh, qvh, wbr, wck
allow_i, allow_i_i, closed_qamats,	nwq
allow_i, allow_i_i, hatef_qamats_vav,	dmh
allow_i, closed_qamats,	crm, lbb
allow_i, hatef_qamats_vav,	wbl, xpr
allow_i, hatef_qamats_vav, qamats_1_vav,	klh
allow_i, i_tsere_ii,	yxb
allow_i, i_tsere_ii, qamats_1_vav,	amn
allow_i, prevent_i_h,	hlc, hll, rgy, wgh, xmt, yll
allow_i, prevent_i_h, allow_i_i,	pla, wmy
allow_i, prevent_i_h, allow_i_i, qamats_u,	clm
allow_i, prevent_i_h, i_tsere_ii, prevent_o,	rah
allow_i, prevent_o,	abd, azn
allow_i, prevent_pfx_hi,	nxb, pgy
allow_i, qamats_1_vav,	rnn
allow_i, vavqamats_vv,	xvh
allow_i_0afx,	gdy, kzq, lmd, nar, qxr, slh, tqn, wcn
allow_i_0afx, allow_i_c, allow_i_l,	wcr
allow_i_0afx, allow_i_h,	psk
allow_i_0afx, allow_i_h, allow_i_l, allow_i_v,	awh
allow_i_0afx, allow_i_l, allow_i_v,	pnh
allow_i_0afx, allow_i_sfx, allow_i_h, allow_i_i, allow_i_t, allow_i_v, remove_h,	ntn
allow_i_0afx, allow_i_sfx, allow_i_i, allow_i_t,	mla
allow_i_0afx, allow_i_sfx, allow_i_t,	bly
allow_i_0afx, allow_i_sfx, allow_i_v,	ynh
allow_i_0afx, allow_i_sfx, i_tsere_ii, hatef_patah_vav,	avh
allow_i_0afx, allow_i_v,	wlk
allow_i_0afx, allow_i_v, closed_qamats,	rkm
allow_i_0afx, hatef_qamats_vav,	qdw
allow_i_0afx, hatef_segol_vav, vavtsere_vv,	yvh
allow_i_0afx, i_tsere_ii,	blh, yvr
allow_i_0pfx,	bxy, dca, gdp, yqw
allow_i_0pfx, allow_i_b, allow_i_h, allow_i_t,	kll
allow_i_0pfx, allow_i_b, prevent_pfx_hi, allow_i_t,	pll
allow_i_0pfx, allow_i_i, allow_i_t,	wkt
allow_i_a,	drw, wvm
allow_i_a, allow_i_i,	nxl, wp+
allow_i_b,	abib, mn, zq
allow_i_b, allow_i_i,	crt
allow_i_b, allow_i_i, allow_i_m, allow_i_t,	sbb
allow_i_b, allow_i_i, allow_i_t,	ngp
allow_i_b, allow_i_i, closed_qamats,	bra
allow_i_c,	wtl, xpkt
allow_i_c, allow_i_l, allow_i_v,	cpr
allow_i_h,	hgh, mss, nsc, nvk, nxx, rby, twy, wpc, zhr
allow_i_h, allow_i_i,	wby
allow_i_h, allow_i_v,	nqm
allow_i_h, prevent_pfx_hi,	rym
allow_i_h, prevent_pfx_hi, allow_i_v,	wal
allow_i_h, qamats_vav_iv,	wlv
allow_i_i,	arc, crh, cwl, gml, lpt, mas, mcc, mvl, mvr, n+r, nbl, npk, nsk, ntq, ntx, nvn, nwc, nzl, prd, qhl, qnh, qvx, rbh, scn, scr, svg, wkk, wrp, xmd, zry
allow_i_i, allow_i_l,	ndd
allow_i_i, allow_i_l, allow_i_t,	lkm
allow_i_i, allow_i_t,	bhl, ctb, dmm, lcd, mkh, mv+, n+w, ngw, ngy, nxr, tmm
allow_i_i, allow_i_t, allow_i_v,	tpw
allow_i_i, allow_i_v,	str
allow_i_i, hatef_qamats_vav,	krh
allow_i_i, i_patah_ii_exc1,	ild
allow_i_i, i_tsere_ii,	idy
allow_i_i, init_i_ii, i_patah_ii_exc1,	ikl
allow_i_l,	wyn
allow_i_m,	qrn
allow_i_sfx,	+pk, amt, aw, axl, b+k, bwl, dwn, kbq, kch, klq, mhr, mll, my+, ncr, qdm, qll, wcl, wdp, wvy, ym
allow_i_sfx, allow_i_b,	xl, yt
allow_i_sfx, allow_i_b, allow_i_c, prevent_pfx_hi,	gdl
allow_i_sfx, allow_i_b, allow_i_i, allow_i_v,	qbx
allow_i_sfx, allow_i_h, allow_i_i, allow_i_v,	nkm
allow_i_sfx, allow_i_i, allow_i_l,	mnh
allow_i_sfx, allow_i_i, prevent_o,	bnh
allow_i_sfx, allow_i_i, remove_h,	spr
allow_i_sfx, allow_i_l,	am
allow_i_sfx, allow_i_v,	rmh
allow_i_sfx, hatef_qamats_vav,	lq+
allow_i_sfx, i_tsere_ii,	klx
allow_i_sfx, prevent_i_h, allow_i_i, allow_i_t,	npl
allow_i_sfx, prevent_pfx_hi, qamats_u,	pqd
allow_i_t,	+ma, mxa, n+p, nvd, qrb, rhb, war
allow_i_t, allow_i_v,	ndr
allow_i_t, allow_i_v, qamats_vav_iv,	alm
allow_i_v,	+hr, +vl, al, awr, kqr, krw, lh+, ngd, nkt, nqp, nvy, nyr, rb, rxh, wbk, wcm, wkr, wlm, wqh, zqq, zvd
allow_i_v, closed_qamats,	kwc
allow_i_v, i_tsere_ii, remove_yod,	qvm
allow_i_v, qamats_u,	wlc
closed_qamats,	anknv, bkn, brc, hdr, krp, mhh, mlc, prx, scc, slk, wcb, wrt, yml
cmn_final_i_exc,	az, dvi, iwi, wi
cmn_final_i_exc, allow_dbl_i, prevent_i, i_tsere_ii, remove_h, qamats_yod_ii,	kih
cmn_masc_pl_exc,	abivn, ivnh, mim
cmn_masc_pl_exc, allow_i_a, allow_i_i, allow_i_t,	wmm
cmn_masc_pl_exc, i_patah_ii,	igy
cmn_sfx_exc,	id, ira, ixr, wna, wph
cmn_sfx_exc, allow_i,	byt
cmn_sfx_exc, create_final_ha, allow_i_i,	qra
create_final_ha,	kgr, qry
create_final_ha, allow_i,	kwb
create_final_ha, allow_i_i,	qrh
create_final_ha, allow_i_i, allow_i_t,	n+y
create_final_ha, allow_i_i, allow_i_t, allow_i_v,	lqk
create_final_ha, allow_i_i, allow_i_t, remove_h,	cvn
create_final_ha, allow_i_sfx,	zcr
create_final_ha, qamats_1_vav,	yzr
create_final_ha, remove_h,	ymd
create_final_ha, restore_a, allow_i_i, allow_i_v,	bva
create_final_ha, vavqamats_vv,	wvq
hatef_patah_vav,	gah, kvl, wkh
hatef_qamats_vav,	kmr, krb, qdqd, sbl, wkd, xri, ymr, ypr
hatef_qamats_vav, qamats_1_vav,	anih
hatef_segol_vav,	avil
i_patah_ii,	bliyl, din, iaw, ixb, yit
i_tsere_ii,	dah, hrg, hrs, iph, iqd, ird, irq, iry, iwb, kmm, kpz, rcc, rdh, rdm, rgn, rkq, ryy, spn, ydi, yip
i_tsere_ii, hatef_qamats_vav,	yni
i_tsere_ii, i_patah_ii, qamats_yod_ii, vavtsere_vv,	iin
i_tsere_ii, prevent_o,	amr, ywh
init_i_ii, i_patah_ii_exc1,	isr
init_i_ii, i_tsere_ii,	iqr
prevent_i,	bin, mmi
prevent_i, prevent_o,	ihvh
prevent_i_h,	ixg, lvi
prevent_o,	acl, aiph, aph, bch, bl, ch, dkh, iawih, la, mwh, nad, ph, pry, raw, wilh, wla, wlmh, wmal, wth, xan, zat, zh
prevent_o, hatef_qamats_vav,	akz
prevent_o, vavqamats_vv,	lvh
prevent_pfx_hi,	ink
prevent_pfx_hi, allow_i_m, allow_i_t, allow_i_v, qamats_1_vav,	knn
prevent_pfx_hi, allow_i_m, i_tsere_ii, remove_yod,	ixa
qamats_1_vav,	ark, avn, bvw, ctl, kq, wr, ybd
qamats_1_vav, closed_qamats,	ahl
qamats_u,	+by, ryl
qamats_vav_iv,	ikd
qub_a_exc,	+la, kla, wvah
qub_a_exc, allow_i_sfx,	cla
qub_a_exc, closed_qamats,	gal
remove_h,	bgd, ryw
remove_h, prevent_o,	pdh
remove_yod,	kbr
restore_a,	nva
restore_a, qub_a_exc, allow_i_h, allow_i_t, allow_i_v,	nwa
sp_exc, allow_dbl_i, allow_i_v,	cny
sp_exc, allow_i_0afx, allow_i_sfx, allow_i_m, allow_i_v, closed_qamats,	dbr
vavqamats_vv,	gvy, xvk, yrh
vavqamats_vv, vavpatah_vv,	rvh
vavqamats_vv, vavtsere_vv, vavpatah_vv,	wlh, ynv
vavtsere_vv,	kgv, mzv, ninvh, qxh
vavtsere_vv, vavpatah_vv,	nvh