$From Mildred Katherine Pope's 1934 "From Latin to Modern French"
$We follow exactly the order as stipulated in Chapter II starting on page 72, with some input from other chapters where appropriate
$However, she explains in her elaboration chapters some places where clearly the order is not what it is listed as in Chapter 4,
$ ... so we implement those instead
$ ---------------------------------------------------------------------------

$Characteristics of Classical Latin as per Pope page 75
$ CORRECTION: We delete Pope's handling of Classical Latin's variant /l/ forms,
	 $as they do not seem to be diachronically relevant, 
	 	$and harm performance.
n̪ > ŋ / __ [+hi,+back,-cont]
ɡ > ŋ / __ [+nas,+cor]
[+lo] > [-front,+back]
[+lo] > [+front,-back] / __ e̯
{j;w} > [+cons,-son] / __ [+syl] $as per Pope s106 -- Pope considers w and j when not "vocalic" (i.e. part of diphthong) to be fricative
	$ for j, p55 s106c ; for w, p56 s109
$ Note that the sequence u: + w does not seem to occur in Latin 
$ If it did there would be some logical issues with Pope's rules -- i.e. s343 she states "beta from b" is effaced after u: 
	$ where of course w (by which she means gamma^w) had already shifted to beta s186 much earlier 
	$ -- of course sound change has no memory! However this issue is moot because u:+w doesn't occur in Latin it seems
	
$ CORRECTION : Pope forgets to explicitly say that w is a proper w not ɣʷ after velar consonants hence why we don't end up with tons of kv clusters.
$ To be sure she does note (s186) that "betacism doesn't occur after k" but she doesn't notice that other affects she attributes to ɣʷ are not occurring in the presence
$ of k -- such as the deletion of the rounded "fricative"/approximant before ʝ (indeed after kw, it is the ʝ that is deleted it seems...) 
ɣʷ > w / [+back,+hi,+cons] __ 

$Pope p226 s629
$In the second century, all aspiration on Greek loanwords was lost. 
[+sg,-cont] > [-sg]

$Pope p226-7 s632
$ Greek y unrounded to i 
[+hi,+front] > [-round,-lab]

$ beginning of Late Latin shifts here... 
$As per Pope's paradigm, it is considered Late Latin if it occurred before the end of the fifth century
$ we place the marker later as per the FLLAPS table. 

$p144 s359 ~ p73 s156iii --- ns > s, with compensatory lengthening of prior vowel -- must have occurred before nature of length changed
$Pope says "very early" s359
$CORRECTION: now lengthening occurs in the context of non-stressed following syllable,
	$ and the n loss is triggered by being between a specifically long vowel and s.
[+syl] > [+long] / __ n̪ s ([-syl])* [-stres] 
n̪ > ∅ / [+syl,+long] __  s

$p144 s359 ~ p73 s156iii --- rs > s after long vowels, > ss after short vowels
	$must have occurred before nature of length changed
	$ implicitly consecutive vowels are "long" and second becomes so if not-- case of deorsum
$ CORRECTION : we only implement the stipulated shifts occurring 
	$after long vowels and two vowel sequences, and not after two vowel sequences either.
	$ Because the rs > ss shift after single short vowels 
	$ seems to only be correct for two lexemes, persica and dorsum.
$ (CORR cont'd) we also now suppress the part of the rule that occurred after two non-consonantals, no targets as deorsum was attenuated)
[+cor,+son,-nas,-lat] > ∅ / [+syl,+long] __ s 
$ [+cor,+son,-nas,-lat] > ∅ / [-cons] [-cons] __ s

$ CORRECTION: stressed i + e > iː x (pietatem type lexemes affected)
	$ needs to bleed any detensing that occurs via Pope s154i
	$ may have appearance of overfitting but Pope explicitly notes case of pitie' as pixt- thus implying this (in BaseCLEFstar too)
[+hi,+front,+stres] e > [+long] x / __ [-nas]

$Page 72-73, section 154i ~ elaborated p 89 s 179 
$Classical Latin length distinction becomes a quality ("tense" for us) distinction
[+syl,-lo,-long] > [-tense]

$p73 s155i ~ p91 s185 -- h-loss word-internally Pope explains that this happened word-internally first "very early", word initially later 
h > ∅ / @ __

$p145 s361- i-prosthesis -- happens in Western Romance very early, attested second century
∅ > ˌɪ / # __ s [+cons] 

$p100 s214i -- one case of regular stress movement, attested in third and fourth centuries . 
$ movement of stress onto the penultimate if antepenultimate is immediate adjacent short vowel
$ Must come before yod formation (s220), which it feeds, as per s214i -- not attested till the 3rd century but Pope seems to be saying it actually occurred earlier. 
[+syl,-lo,-hi,-stres] > [+prim] / [+prim,-lo,+front,-long] __ ([-syl])* [+syl] ([-syl])* #
[+prim] > [-prim] / # ([-syl])* __ [+prim]
[+prim] > [-stres] /  __ [+prim]

$p100 s214ii -- the (only?) other case of regular stress movement-- occurred around the same time but not for the educated classes
$stress on antepenultimate moves forward ("ordinarily") if the second consists of short vowel + plosive + r
[-stres] > [+prim] / [+prim] ([-syl])* __ [-cont] r ([-syl])* [+syl] ([-syl])* #
[+prim] > [-prim] / # ([-syl])* __ ([-syl])* [+prim]
[+prim] > [-stres] / __ ([-syl])* [+prim]

$p73 s155iib and p91 s187 -- we know develarization of w came very early, 1st/2nd century (Reign of "Nerβa") and that it was preceded by 
$ the gliding phenomenon below

$p102 s220 (also p73 s156i)  -- yod and w formation -- gliding of unstressed short non-low vowels before other vowels
$ -- called upon to be this early by p91 s187
[+syl,-long,-stres,-lo] > [+hi,-syl,+tense] / __ [+syl]

$ CORRECTION: uː unstressed becomes sonus medius 
	$-- diachronically equivalent to ɪ 
	$ -- before a non syllable and not near word boundary 
	$ -- but does not bleed prior yod formation (see adiu:ta:re)
		$ or feed it (manu:opera)
uː > ɪ / __ [-syl] @

$ per s220 : yod is lost between e and p,r
j > ∅ / {p;r} __ [+front,-hi,-lo,+syl]

$ CORRECTION: before j is consonantalized,
	$, dark l emerges from l specifically before consonants (except l)
	$ (this is in effect a reduction of the scope of Pope's previous rule of Latin l-darkening)
l > [+back] / __ [+cons,-lat]

j > ʝ / __ [+syl] $as per Pope s105c, 106
$ w, newly from u,o in hiatus is clearly NOT effected by this as per treatment in s374

$ CORRECTION (addition) : ʝ is deleted when coming after # + velar + w (cases of quie:tum etc...)
	$ this may look like overfitting but the errors that result are of a frankly uninteresting nature and better removed.
ʝ > ∅ / # [+back,-cont] w __  

$p91 s187iic -- explicitly must come after p102 s220 
$effacement of w before jod. 
$NOTE however that in many nouns this was reversed by analogy (see p92 s188)
$affects w as per laqueum > latz > "lacs"
{ɣʷ;w} > {∅;∅} / __ ʝ

$p91 s187iia
$ CORRECTION: This seems to only reliably occur after /ɑ/ and /ɑː/ -- this is a better reason why vivum, ovum and nativum are unaffected than "analogy" (which Pope says in s188).
ɣʷ > ∅ / [+lo] __ [+syl,+round]

$ as per s186iib -- w effaced between k and a rounded vowel
	$ Note most authors interpret this differently, as the delabialization of k
	$ Note also that Pope seems to have missed that his also happend after g?
w > ∅ / k __ [+syl,+round]

$ p115 s261-264 ~ s221 ~ s156i
$ *** Note as per s156i, for Pope this is a "tendency" 
$ the slurring ... of "proparoxytones" ... i.e. where the tonic syllable was followed by two unstressed sylls before hte word coda
$ ... typically concerning those which are common suffixes here.
$ Pope attests this with the Appendix Probi, which was written in the third or fourth centuries but does not say that is when it specifically occurred. 
$ Because of words like "table", we know it probably occurred before lenition of intervocalic b (early second century) but Pope does not confirm this either. 
$ By placing it here before the beginning of betacism, we are giving Pope the benefit of the doubt. 
$ This is also inline with what she states about the -itus,-itum etc group -- she is holding that this phenomenon occurred before hte beginnings
$ 	of palatalization  (s296)
$ Further note : s262b -- Pope notes that sometimes the older forms were restored by preservative analogy via "clerkly influence" 
$ p115 s262a -- effects on -ulus,-ulum,-ula words. Attested for affecting those where preceding cons is k,g,t,p or b. There don't seem to be
	$ any words of this type ending in -dulus,-dulum, or -dula. 
$ CORRECTION : -mulum endings also effected (cumulum > comble; tremulum > tremble) 
$ CORRECTION : -rula endings also effected (berula (Gaul.) > berle not bierle) 
$ CORRECTION: -olum,-olus endings also included if not after a labial stop
{ʊ;o} > {∅;∅} / [+prim] ([-syl])* {[-cont];m;r} __ [+lat] [+syl] ([-syl]) #
ɔ > ∅ / [+prim] ( [-syl] )* {[-cont,-lab];m;r} __ [+lat] [+syl] ( [-syl] ) #

$ p115 s262b -- effects on the -idus,-idis,-itis etc group : for prior consonsants Pope lists l, g, t, s, k, and r
	$ the most faithful way to render this also in line with Occam's Razor would seem to be [+cons] 
	$ *** NOTE s353 -- seems to imply it was possible for this to NOT occur for some words
	$ note also the frigidus contradiction: s297 & s316 vs. s262ii & s296i
$ CORRECTION: ɪ does not disappear before t at this point as all the affected words (hospitem, comitem, lenditem, limitem) need to eject ə after final vowels are lost
$ CORRECTION: ɪ only disappears before d[+front,-lo] when after a sonorant (magidem must also eject ə -- so cluster must be formed in Gallo-Roman stage)
$ CORRECTION: ɪ before dʊ does not disappear if previous consonant is voiceless, as these will need to eject ə in Gallo-Roman when final vowels are lost (muccidum, nitidum, sapidum, tepidum)
	$ CORRECTION: it DOES disappear before voiced stops though (frigidum)
ɪ > ∅ / [+prim] ([-syl])* [+son] __ d̪ [+front,-lo] ([-syl]) #
ɪ > ∅ / [+prim] ([-syl])* [+cons,+voi] __ d̪ [+back,-lo] ([-syl]) #

$ p115 s262c -- effects on intertonic (non-low) vowels standing between n/l and r 
[-lo,-stres] > ∅ / [+prim] ([-syl])* n̪ __ r [+syl]
[-lo,-stres] > ∅ / [+prim] ([-syl])* [+lat] __ r [+syl] 

$ CORRECTION : ɪ deleted between k/ɡ and l if the velar stop is not preceded by an /ɑ/ variant 
ɪ > ∅ / [-lo,+syl] [-cont,+back] __ l

$s127 s296i fakere case
$ CORRECTION: targets only unstressed front vowels, does not occur after anything but ɑ, and doens't occur before l
[-stres,+front] > ∅ / [+prim,+lo] [+back,-cont] __ [+cor,-strid,-nas,-lat] [+syl]
$ CORRECTION: hits after ɡ too -- cases of digitum, fugitam. Any vowel can come before the d.
$ CORRECTION: hits before n too -- cases of plantaginem, acinum (circinum?? Not anymore.). Any vowel can come before the n.
$ CORRECTION: no k_d seqments, only ɡ_d 
[-stres,+front] > ∅ / [+syl] ɡ __ [+cor,-lat,+voi] [+syl] 

$ CORRECTION : slur short i between ɣʷ and k 
$ Pope lists avica and avicellum as having been effected by "slurring" in s187i 
	$but doesn't state that i was slurred here explicitly elsewhere
ɪ > ∅ / ɣʷ __ k

$as per p91 s187i -- ɣʷ > w when preconsonantal due to "slurring"
ɣʷ > w / __ [+cons]

$ CORRECTION: a much more effective and thorough way to describe the 
	$apparent contexts of deconsonantalization of ɣʷ here...
$ actually covering two shifts
$ it appears that this occurred in all ɑɣʷɪ sequences : navicellam, navicam, avicam, avicellam, Pictavis, etc...
$ Pope (s187i) says this is due to "slurring" but she nowhere else described the "slurring" phenomenon that would have fed the rule in this way
$ Doesn't happen when the posterior ɪ is stressed it seems -- see kˌɑwˈɪkulɑm > ʃəvij
$ Hence this is the most efficient way to represent it...: 
$ CORRECTION : abrogated for now as no effect due to correction @ line 175
$ ɣʷ > w / [+lo] __ ɪ 
$ next, to cover the ɑɣʷɑ sequences: bauam, flavam, cavannum. 
$Doesn't seem to happen for -aːre verbs (lava:re) 
	$ -- this can be represented by only allowing the posterior ɑ to be short
ɣʷ > w / [+lo] __ [+lo,-long]

$p91 s186: w>beta. Occurred during the 1st and 2nd century ("Nerβa")
$ CORRECTION : as evidenced by Pope s343 causing errors for 10 words that became β from ɣʷ,
$ they did not merge this early, and must have merged AFTER that -- instead we send ɣʷ to βʷ 
$ NOTE: delete and reinsert the <ʷ> to turn off/on the demo
ɣʷ > βʷ

$p73 s155iia and pp136-137 ss332-336 and p6 s9 : b > β intervocalically or between vowel and r (s372)
$According to s336 this occurred very early at least for the bilabial, 2nd century AD 
$Later authors group all voiced stops together for this shift, but Pope is unclear
$She states (s336) that it is attested earlier for b than the others but is unclear
$However she indicates a shift order in s307i that necessitates velar stop palatalization preceding intervocalic lenition
$So we give her the benefit of the doubt and place the lenition of the rest of the velar stops later,
$Right after the beginning of palatalization.
$ she is not clear about whether her j and w ("fricatives") are sonorant, but we will assume they are NOT
	$... given that she called them fricatives -- this is supported by her trajectory for rubeum > rouge (s676)
b > β / [+syl] __ [+son,-lat,-nas]

$ p115 s263 -- with proparoxytons with intervocalic g, both g and preceding vowel deleted "early" (s263)
	$ given the hypothesized forms listed by Pope, it is clear she thinks this happened before  ɪ > e and ʊ > o 
[-stres] ɡ > ∅ / [+prim] ([-syl])* __ [-stres] ([-syl])* #

$ per Pope s185 and s198, combination of vowels that have become adjacent and share the same quality become a single long vowel of same quality
$ in practice this only happens for e'e and o'o.
$ CORRECTION: o'o made tense.
[+front,-hi,-lo] [+front,-hi,-lo,+prim] > ˈeː
[+round,-hi] [+round,-hi,+prim] > ˈoː
[+front,-hi,-lo,+stres] [+front,-hi,-lo] > ˌeː

$ CORRECTION: shift occurring in Gallian (Gallo-Romance, Occitano-Romance, Gallo-Italic, Rhaeto-Romance), Venetian and Sardinian
	$ ɑ ɡ ˈɪ n sequences become ˈɑ n 
$[+lo] ɡ ˈɪ n̪ > ˈɑ n̪ 
$ KS: Removing this change as it is not regular; see changes in lexicon to multiples of ten (trigintam>trintam, quadragintam>quadrantam, etc)

$Page 72-73, section 154i ~ elaborated p 89 s 179 , s222
$Classical Latin length distinction becomes a quality ("tense" for us) distinction
[+syl,-lo,-long] > [-tense]
[+syl,+long] > [-long]

$p73 s154iii ~ p190 s504 -- seems to have first occurred in the first and second centuries s504
$giving her the benefit of the doubt, we place this after the length > tense shift, as otherwise oe would become ɛ
$ 	given that she doesn't specify it as long and writes it like a short vowel 
{a e̯;ˌa e̯;ˈa e̯} > {ɛ;ˌɛ;ˈɛ}
[+round,-hi,+syl] e̯ > [+front,-round,-back,+tense,-lab] ∅

$p144 s359 ~ p73 s156iii : simplification in ŋks, ŋkt clusters
k > ∅ / ŋ __ [+cor,-voi]

$p144 s359 ~ p73 s156iii : kst > st rule 
k > ∅ / __ s [+cons]

$p144 s359 ~ p134 s324-325 ~ p73 s156iii ~ p64 s130 ~ p6 s9i
$ frication of k before t,s ; attributed to Gaulish substratum p6 s9 (RECTUM > REXTUM -- see also discussion of attestations like paraxsidi in other sources.) 
$ Pope doesn't discuss the specific ordering of this vis-a-vis the kst>st rule, but this one would bleed that one and she never discusses an xst>st rule,
$ so we favorably assume this one comes second.
$ BaseCLEF* CORRECTION -- Pope fails to explicitly state that this only occurs after nonconsonantals
	$ hence nkt and etc clusters are not effected. But this is not an interesting error. 
k > x / [-cons] __ [+cor,-voi] 

$CORRECTION: countertonic /ɛ/ closes back to /e/ before ssʝ (not sj -- see prehensionem > prison)
ˌɛ > ˌe / __ s s ʝ

$p144 s359 ~ p73 s156iii : g > w before m, Pope says this is first attested in teh 3rd/4th century Appendix Prodi 
$"opened and vocalized"-- Pope is explicit that it passes through a fricative stage p65 s135b
ɡ > [+cont] / __ m 
ɣ > w / __ m 

$p73 s156vi ~ p191 s505 ~ p64 s131: au > a under influence of tonic u in next syllable (irregardless of tens e or length)
$ ... NOTE: Pope is not clear about when exactly this occurred, but we place it before ʊ > o for convenience
[-syl,-cons,-lo,+round,+back] > ∅ / ˌɑ __ ([-syl])* [+stres,+hi,+round,+back]

$p73 s154i ~ elaborated p89 s180 : ɪ > e and ʊ > o 
$according to Pope the lowering of ɪ was earlier than taht of ʊ "which retained its value in parts of the Roman Empire up to the present"... but doesn't specify when exactly
$best to just place them together here. 
[+hi,-tense] > [-hi,+tense] 

$p328 s854 ~ p64 s131 : dissimilation of tonic vowel juxtaposed to final flexional vowel. 
$ *** this is a "tendency" as per Pope s854
$ Other authors have other explanations for this, like tensing of vowels before other vowels or even much earlier lengthening of them 
ˈo > ˈɔ / __ [+hi,+syl,+back]
ˈɛ > ˈe / __ [+lo]
ˈe > ˈi / __ e

$ p73 s156v ~ s262 ~ s360 
$ tl > kl -- attested first in the Appendix Probi as per Pope p145 s360
$ as noted by Pope s262, sometimes reversed under "clerkly influence" by preservative analogy
t̪ > k / __ [+lat]

$ p97-98 s205i ~ p73 s156ii : final m deletion  -- 
$ ... unlike some later authors Pope does not recognize a nasalization in Late Latin, instead holding that simple nasal consonant deletion occurred
$ Pope doesn't explicitly date this but considers it first attested in teh 3rd/4th century Appendix Probi
$ Pope notes it did not occur in monosyllabic words 
$ NOTE: as per p97-98 s205i -- unlike Italy and also South Gaul, North Gaul did not see effacement of s and t word-finally,
$ ... because "pronunciation was influenced by the teaching in schools" in North Gaul -- though this mainly concerns final t and s 
$ CORRECTION : clear also final -n in words like materiamen
$ CORRECTION : change trigger -- we are deleting final m specifically after vowels that lack stress, regardless of number of syllables
	$ monosyllables -- quem is totally stressless; suum, tuum, meum, in -- countertonic and nasal preserved at this stage. 
[+nas,+cons] > ∅ / [-stres] __ #
$ CORRECTION : also deleted after countertonic ɑ
[+nas,+cons] > ∅ / [+lo,-prim] __ #

$ p185 s488 : o > ɔ before labial consonants 
$ *** NOTE : This is called a "TENDENCY" by Pope
$ she does not give a date to it other than saying it Late Latin
$ --- we give her the benefit and place it after final m deletion 
$[+round,-hi,+syl] > [-tense] / __ [+lab,-son] 
$ CORRECTION: we are discarding this rule : it seems to be just wrong, per 40 effected words. 

$ -- kwinkwe > kinkwe dissimilation
$ Pope s823 iii. 
w > ∅ / k __ [+syl] ([-syl]) k w [+syl]

$ CORRECTION : ts goes to f before a sonorant consonant (tsrokna, etc.) 
$ can be seen as overfitting, but necessary to assimilate loanwords and we are more interested in French sound shifts not Latin ones. 
t͡s > f / __ [+cons,+son]

$FIRST PALATALIZATION: Palatalization begins in the fourth century as per Pope p129 s304
$t and d preceding yod become palatalized affricates
$p130 s308-9
{t̪;d̪} > {t͡sʲ;ɟ} / __ [+hi,+front,-syl,+cont] 

$ as per s306ii : k+yod > c c intervocalically only 
k > c c / [+syl] __  ʝ [+syl]

$p 124 s283-284 : palatalization of velar stops before e and i, s306-307 for before yod
[+hi,+back,-cont] > [+front] / __ [+cont,-lo,+front]
[+back,+front] > [-back] 

$ CORRECTION : regressive palatalization of geminate k
k > c / __ c

$p127 s296ii : palatalization of k blocked if coming after secondarily stressed velar vowel
$ ... and before intertonic unstressed e (<ɪ) 
$ CORRECTION : we broaden this to all unstressed e,ɛ before t, r or l, to hit words like iacet 
$ CORRECTION : occurs specifically after stressed velar vowels, and not when t is word final
c > k / [+back,+stres] __ {e;ɛ} {t̪;r;l} @

$ CORRECTION : l deleted when immediately before ɲ 
$ she actually has the effects of this in her citation forms for Late Latin (s686) 
 $ -- but she doesn't explicitly state it. 
 $ (included in BaseCLEFstar) 
[+lat] > ∅ / __ [+nas] ʝ

$CORRECTION: opening of voiced velar and palatal stops intervocalically occurs early enough to get into Pope's sound tables.
[-cont,+hi,+voi] > [+cont] / [-cons] __ {[+son,-lat,-nas];ʝ}

~Late Latin

$absorption of yod into prev, now palatalized, consonant 
ʝ > ∅ / [-cont,+hi,+front] __

$p131 s311, assigned to Late Latin in p73 s155-5iv :
$ as per s293 and s156 however this did not YET affect ng + pal-stop -- i.
$ trigger again in Gallo-Roman stage.
$CORRECTION: This is bled by an early shift of mnj to mmj
$ does not appear in Pope's s660-680ish tables. 
n̪ > m / m __ ʝ
[+nas,+cor] [+hi,+front,-syl,+voi,+cont] > ɲ

$p131 s312, assigned to Late Latin p73 s155-5iv -- but not evident in Pope's tables in 660-680 for that stage
$ as per s293 and s156 however this did not YET affect l + pla-stop -- i.e. not words like culcita 
$trigger again in Gallo-Roman stage
l l ʝ > ʎ
l ʝ > ʎ

$ CORRECTION: unstressed vowels deleted between [+syl] w and {t;d}
	$ even when not penultimate (Pope's rule s353 applies only for penultimates
		$... where Late Latin is concerned) 
	$ 2 examples -- may need to delete as overfitting? But it is consistent with other patterns of unstressed /ɪ/ deletion 
[-stres] > ∅ / [+syl] βʷ __ [+cor,-cont]

$per Pope s353: unstressed penultimate ɪ in -item, itum, idum, ita endings is deleted before any d > ð
$ we can use /e/ here because if it was e:, it would be stressed. 
$ CORRECTION: include also the syncope of vowels before n̪ 
$ CORRECTION: when before a vowel in coda position it instead goes to "evanescent" ə̯ 
 $ (which will be completely effaced instead only in the Gallo-Roman stage) 
 	$ before t when after a consonantal continuant then a stop
e > ə̯ / [+cont,+cons] [-delrel] __ t̪ [+syl] #
e > ∅ / __ [+distr] [+syl] ([-syl]) #

$ CORRECTION: unstressed ɛ after palatal deleted 
	$ when before t̪ then some a-variant
$ SUPPRESSED -- overfitting. 
$ ɛ > ∅ / [+front,+cons] __ t̪ [+lo]

$CORRECTION ə̯ inserted between p and d (rapidum, sapidum, tepidum > rade, sade, tiede | etc) 
∅ > ə̯ / p __ d̪

$ CORRECTION : n̪ > m / p __  -- case of carpinum > charme etc
$ Pope does cover it but doesn't attribute it to Late Latin.
n̪ > m / p __ 

$ CORRECTION : earliest flapping with n comes into contact with d due to falls of vowels here: 
	$ cases of Londres, ordre, desordre, etc. 
$ We can just send it straight to the trill here
$ bleeds s365
$ Pope attributes this to a variety of non-phonological factors
	$ but it can be modeled regularly. 
n̪ > r / d̪ __ 

$ CORRECTION : earliest implementation of s365 (Pope attributes these to Gallo-Roman)
$ necessary to correctly bleed some of the following lenition shifts.
$ s365 : deletion of all plosive consonants internal to a cluster not ending in l or r 
$ CORRECTION: also does not target consonants before ʝ
[-cont] > ∅ / [+cons] __ [-son,-hi]
[-cont] > ∅ / [+cons] __ [-cont,+hi]

$ CORRECTION : before nasals, effacement does not occur at this stage
$ [-cont] > ∅ / [+cons] __ [+nas,+cons]

$ CORRECTION: s365i inserted here too, earlier than Pope explicitly states it is. 
$ s365i : additional deletion of velars in rgl, skl
{s k;r ɡ} > {s;r} / __ l

$p73 s155iia and pp136-137 ss332-336 and p6 s9 : intervocalic lenition between vowel and (vowel or r)
$ valid targets before r included plus d before l -- s372
$According to s336 this occurred very early at least for the bilabial, 2nd century AD 
$Later authors group all voiced stops together for this shift, but Pope is unclear
$She states (s336) that it is attested earlier for b than the others but is unclear
$Her tables show a lenition of g to gamma for Late Latin but not d to delta
$However she indicates a shift order in s307i that necessitates velar stop palatalization preceding intervocalic lenition
$So we give her the benefit of the doubt and place the lenition of the rest of the velar stops later,
$Right after the beginning of palatalization. This should be roughly 4th century
$ inclusion of palatal voiced stop in this shift -- as per Pope p127 s297
[-delrel,+voi] > [+cont] / [+syl] __ [+son,-lat,-nas]
d̪ > ð / __ [+lat]

$ s353 : devoicing of labial and dental consonants that are now before t 
$ since aidier is excluded, we can dodge fixing palatals too. 
[+ant,+cons,-son] > [-voi] / __ t̪

$p73 s155i ~ p91 s185 -- as Pope explained, h-effacement happened word-internally first. 
$It was apparently still being resisted among the educated classes at the time of St. Augustine, 4th century and early 5th century 
h > ∅ 

$p73 s156vi -- influence of tonic vowels on countertonic vowels
$***NOTE Pope calls these "TENDENCIES"
$ she does not seem to give a date to these
$ s127 for i > e
$ s128 for e >a 
$ s505 for aw > a under influence of u specifically 
ˌi > ˌe / __ ([-syl])* [+hi,-round,+prim,+tense] $i.e. by i
${ˌe;ˌɛ} > {ˌɑ;ˌɑ} / __ ([-syl])* [+lo,+prim]
$ CORRECTION: e mutation deleted -- does much more harm than good. 
$s505 ~ ˌɑ w > ˌɑ before round stressed sylls -- already handled elsewhere, don't need to redundantly do it here. 

$p73 s154ii : length returns for stressed vowels only, conditioned the open/closed context condition:
$ as per s197-198 -- not "blocked", including word final, followed by vowel, followed by cons + word coda only in a monosyllable,
$ followed by cons + vowel, or followed by plosive + l/r
	$ CORRECTION: before β/v + r + V + # vowels are also lengthened -- chevre, couleuvre, levre, fievre, etc... 
		$ Pope only stipulates that plosive + r is free -- neglecting that b has become fricative here.
			$ (the same does not occur for ðr clusters which are indeed blocked) 
$ note as per p53 s100, note that Pope does not consider glides to be consonants 
	$ but to be fair she specifies  "in hiatus with a following vowel" -- and all her examples feature
		$ hiatus with syllabic vowels, not semivowels
$ -- give her the benefit of the doubt and specify +syl not -cons
[+prim] > [+long] / __ #
[+prim] > [+long] / __ [+syl]
[+prim] > [+long] / __ [+cons] [-cons] 
[+prim] > [+long] / __ [-cont] [+cons,+son,-nas]
[+prim] > [+long] / __ [+lab,+voi,+ant] r [+syl] #
[+prim] > [+long] / # ([-syl])* __ [+cons] #

$ ~Early Gallo-Roman
$ Sixth and seventh centuries, as per Pope s163 

$ CORRECTION: deletion of unstressed -we- after velars and before a cons, in syllables internal to a tri-or-more syllabic word (cases of aquila > aille, sanguisuga>sangsue) 
w e > ∅ / [+syl] ([-syl])* [+hi,+back,+cons] __ [+cons] ([-syl])* [+syl]

$ Pope s325ii intervocalic sk > ks before u,o  -- precedes s325i
$ Logically, should be placed before breaking of long lax vowels -- otherwise it will be triggered by lax o: > uo  '''/
$ CORRECTION : only before boundary. Which is the soft way of deleting this. 
s k > k s / [+syl] __ [+syl,+tense,+round] #

$ Pope s325i ks > xs again (s359) for words like kresko (s325ii) , then to > ç (s378,s324) ultimately to j (s325i) 
$ for convenience we are also placing here k>x before t (s325, again after the first time in Late Latin: plakitum as per s296i)   
	$ and the opening of k and ɡ to x and ɣ before l,r -- conveniently due to sound distributions
$ we can just refer to these as the coronals (since k and ɡ don't occur before n, d, z and we place the assibilation of palatals after this for convenience)
$ CORRECTION: Pope failed to explicitly specify that this occurs after nonconsonantal sounds
[+hi,+back,-cont] > [+cont] / [-cons] __ [+cor,+cons]
$ CORRECTION : also open before ʎ -- s407, Pope implies that in cuilliere that the former k has disappeared somehow, but she doesn't ever say how. 
[+hi,+back,-cont] > [+cont] / [-cons] __ ʎ

$ intervocalic -qu- 
$ k > x before Pope's fricative w as per s327-330
$ note that Pope is fairly inconsistent as to the identity of the <u> in qu 
$ she seems to label it as a semivowel (u with glide mark) in s327-330
$ but marks it as her "fricative" w in s186-188
$ to reflect this we don't question it and just use w to reflect her usage, without questioning it...
k > x / [+syl] __ w [+syl]

$ s374ii : intervocalic -lw- > ll and -sw- > ss 
$ CORRECTION: this happens for sw everywhere it occurs, not just intervocalically 
$ CORRECTION: for l only after rounded elements. 
[+lat] [-syl,+cont,+round] > l l / [-cons,+round] __ [-cons] $only necessary for verb conjugations really though...
s w > s s
s > ∅ / # __ s

$ Pope p138-139 s341 loss of early Gallo-Roman ɣ:
$ Pope places this in the Gallo-Roman stage which she says starts at the end of the fifth century
$ Attested in mid-sixth century as per Greg of Tours (p138-139 s341) 
$ and was attested in Gaulish in the fourth century or even earlier (Pope s9)
$ Precedes (s341) loss of unstressed final o,u 
$ According to symbols used, this should precede ɑ > a (s341)
ɣ > ∅ / __ [+tense,+round,+syl]
ɣ > ∅ / [+hi,+round,+syl] __ [+lo]
ɣ > βʷ / [+round,-hi,+back,+syl] __ [+lo]

$also, as per s323 ɣr > r pretonic after e
$ she says only /e/ but given the inclusion of peregrinum in her examples, 
	$ we can give her the benefit of the doubt and let it affect /ɛ/ too.
ɣ > ∅ / [+front,-hi,-lo,+syl] __ r [+prim]

$ Pope s374ii first round of β > w before another w -- will recur
$ as per comparing s484 and s481, clearly βw had a different fate from ðw or ɣw (including < -qu-) -- so this one happened first
$ note that this one will have no targets unless some verb conjugations are added to the dataset. 
β > w / __ w 

$ Pope ~ p78 s163.7 p139 s343 first effacement of β 
$ "when it was derived from classic Latin intervocalic b and stood before the vowels u or o or after tonic u" 
	$ s163:7 says "before labial vowels", as does 343 (though 343 says u or o too) 
$ (PSEUDO-)CORRECTION : Pope neglects to note that this only happens AFTER other vowels -- i.e. not word-initially. 
	$ ( w (>βʷ) now not merged into β yet when this rule applies, so it counterfeeds)
β > ∅ / __ [+lab,+syl]
β > ∅ / [+prim,+hi,+round,+back] __ $i.e. after tonic u 

$ CORRECTION: ɸ, ɸʷ deleted if in between consonants.
{ɸ;ɸʷ} > {∅;∅} / [+cons] __ [+cons] 

$ Pope s352: -ika endings -- atonic e is deleted BEFORE k becomes voiced
$ CORRECTION: occurs in all -ika- instances including inside the word (i.e. praedicatorem >precheur)
$ CORRECTION: occurs only after voicing if the e is preceded by nasal+voiced stop, or by r
e k > ɡ / r __ [+lo]
e k > ɡ / [+nas] [-cont,+voi] __ [+lo]
e > ∅ / __ k [+lo]

$ Arteaga Mazzola version of this: voicing is the regular phenomenon (but later cluster assimilation plus non-phonetic influence of the "Frankish" voiceless palatal final change this for many words)
$ e k > ɡ / [+cons,-round] __ [+lo]

$ Pope p137 s335-336, also p77 s163-5 : voicing of k to ɡ 
$ Began in fifth century, "became general" in the sixth century as per s336
$ Explicitly occurred BEFORE slurring of intertonic vowels s349 -- unlike same slightly later shift for t(>d) (s350)
$ Explicitly occurs BEFORE beginning of second palatalization and before au > o s299
	$ potentially RECURRENT as per Pope s162 
$ this incorporates ɡw > ɣw as per Pope s329 
k > ɡ / [-cons] __ [+syl]

$ CORRECTION : syncope of unstressed ɛ,e before countertonic ɛ+r 
	$ if followed by consonant then primary stressed back vowel (verecundiam, caerefolium)
[+front,+syl,-hi,-lo,-stres] > ∅ / [+front,-hi,-lo,-prim,+stres] r __ [+cons] [+back,+prim]
 
$ Recurrence of Pope s333i
ɡ > ɣ / [-cons] __ [-cons] 

$ CORRECTION: fʝ sees metathesis which blocks voicing to vʝ
$ SUPPRESSED: overfitting
$ f ʝ > ʝ f

$Pope s334ii voicing of breathed fricatives 
$ must happen before first palatalization of s and z 
$ not specified in Pope's chronology -- but "recurrent" as per s162
$ and happened before first formation of palatal sibilant fricatives...
[+cont,-son] > [+voi] / [-cons] __ [+syl] $CORRECTION: fixed to exclude w. We are not hitting xw clusters with this yet. 
[+cont,-son] > [+voi] / [-cons] __ ʝ [-cons] $CORRECTION: Pope doesn't consider yod at this point a consonant, but lenition clearly occurs before it... 

$ Pope p77 s163-1 ~ p164-165 s419 : Influence of final (atonic) /i/ on preceding tonic /e/ and /o/. 
$ Dated to 6th or 7th centuries (s163-1) 
$ Pope states this happened before the "weakening" of final atonic vowels. 
$ e > i and o > y 
$ Based on the way she has ordered the rules on page 77, 
$ it would appear she is asserting that this is the first appearance of /y/ in French. 
[+prim,+tense,-hi,-lo] > [+hi,+front] / __ ([-syl])* i # 

$ CORRECTION : l > ɫ can go here . Will need to be recurrent. 
l > ɫ / __ [+cons,-lat]

$ as per Pope p126 s293
$ palatalization of ŋ, l and s before palatal stops followed by short unstressed paroxytone e,ɛ before r,t
$ potentially RECURRENT as per s162 
$ but definitely started before effacement of ɛ pre r/l and cluster simplification as per s293
{ŋ;ɫ;s} > [+hi,+front,-back] / __ [-cont,+hi,+front] [-stres,+front] [+cor,+cons,-lat] ([-syl])* [+syl] ([-syl])* #

$ Pope p77 s163-3 "effacement of penultimate ɛ before r and l"
$ comes before early Gallo-Roman cluster modification s293 
$ as per symbols used should come AFTER regressive palatalization assimilation of nasals, s and laterals
$ but BEFORE regressive palatalization of r ! 
ɛ > ∅ / __ [+cor,+son,-nas] ([-syl])* [+syl] ([-syl])* #

$ s189 : nβl > mbl -- only relevant for inwolare>embler, an obselete word, it seems though. 
$ (PSEUDO-)CORRECTION: modified to hit also βʷ
n [+cont,+voi,-son,+lab] > m b / __ [+lat] 

$ CORRECTION : β and w are deleted elsewhere when in middle of cluster not following a labial or velar 
$ (we can use -hi since we assume a palatal cons would not emerge here - -back won't work because of ɫ) 
$ would appear to be overfitting but this is a mere replication of existing phenomena with other interobstruent consonants
[+cont,+voi,-son,+lab] > ∅ / [+cons,-lab,-hi] __ r

$ action resulting on clusters as formed from this previous rule (s163-3) and from proparoxyton deletion in Late Latin (s262) 
$ Denasalization -- s369, and glide development s370
$ b inserted between m + r or l 
$ d inserted between n,l,dark l,z,ʎ + r 
$ t inserted between s and pal. s + r
$ should RECUR whenever these consonants are brought into contact.  
∅ > b / m __ [+cor,-nas,+son,-syl]
∅ > d̪ / [+lat] __ r
{z r;n̪ r;s r} > {z d̪ r;n̪ d̪ r;s t̪ r}

$ CORRECTION : c,ɟ spirantized between vowel and true dentals t̪d̪n̪ to bleed following rule, s293
$ might have slight issue with culcita > coute. 
{c;ɟ} > {ç;ʝ} / __ [+distr]

$ as per Pope p126 s293: depalatalization of palatal stops
	$ that come into contact with following non palatal consonants
$ explicitly BEFORE assibilation (bleeds it)
[-cont,+hi,+front] > [-hi,-front,+cor,+ant,+distr] / __ [+cons,-front]

$p73 s154ii : length returns for stressed vowels only, conditioned the open/closed context condition:
$ as per s197-198 -- not "blocked", including word final, followed by vowel, followed by cons + word coda only in a monosyllable,
$ followed by cons + vowel, or followed by plosive + l/r
	$ CORRECTION: b/β/v + r + V + # also allows lengthening
$ note as per p53 s100, note that Pope does not consider glides to be consonants 
	$ but to be fair she specifies  "in hiatus with a following vowel" -- and all her examples feature
		$ hiatus with syllabic vowels, not semivowels$ CORRECTION : shortening function operates before [-syl] rather [+cons] (and then # or non-sonorant [+cons] )
[+syl] > [-long] / __ [-syl] [+cons]
[+prim] > [+long] / __ # 
[+prim] > [+long] / __ ([+cons]) [+syl]
[+prim] > [+long] / __ [-cont] [+cons,+son,-nas]
[+prim] > [+long] / __ [+lab,+ant,+voi] r [+syl] #
[+prim] > [+long] / # ([-syl])* __ [+cons] #

$ CORRECTION: ˈɔ becomes long before a rounded continuant cons then a stop
$ SUPPRESSED: overfitting
$ ˈɔ > ˈɔː / __ [+round,+cont,+cons] [-cont]

$ CORRECTION: now we can finally merge βʷ to β (and ɸʷ to ɸ) 
{βʷ;ɸʷ} > {β;ɸ}

$p127 s297 : developments of ʝ < ɟ : effacement intervocalically except between countertonic a and tonic e 
$ CORRECTION: moved much earlier so it can bleed e > i raising in the case of ˌeʝˈe. 
	$ We are no longer currently sending ʝ to j at this point now
$ CORRECTION: only happens after countertonic ˌe 
ʝ > ∅ / ˌe __ [+syl]

$ CORRECTION: geminate ʝʝ cluster degeminates, 
	$ but not before raising/tensing any prior eh or oh to e and o if countertonic
{ˌɛ;ˌɔ} > [+tense] / __ ʝ ʝ
ʝ > ∅ / __ ʝ

$ CORRECTION: tonic ˈe after a palatal becomes long if closed by coronal + oral coronal or coronal + ʝ 
ˈe > ˈeː / [+hi,+front,+cons] __ [+cor] {ʝ;[+cor,-nas]}

$ Pope p164 s418 : raising of "tonic free e" to i (tonic free defined in s197-198)
$ when preceded by a palatal 
$ Sixth/Seventh centuries as per Pope s163-11
$ Conveniently all such vowels will be long as per s154ii -- we use this for our purposes
$ Must come before assibilation of first palatalization results, which would bleed this for the voiced pal. stop
	$ (Pope does not specify the voiced one retained any palatal element after assibiliation-- she does for c though as per p )
ˈeː > ˈiː / [+cons,+hi,+front] __

$ CORRECTION : x n becomes ŋ ɲ ; may appear overfitt-y but necessary for assimilation of loanwords
x n̪ > ŋ ɲ 

$ s426 : ɔ tonic blocked, ɔ countertonic free or blocked, 
	$ and any o free or blocked with any level of stress raise to u 
	$ ... if before nasals, before diphthongization can occur
$ CORRECTION : these go to already nasalized ũ, so that it won't feed subsequent u-fronting for these segments
$ CORRECTION: free tonic ɔ is also effected... so essentially ANY variant of o or ɔ is affected -- cases of frogne and son
[+round,-hi] > [+tense,+hi,+nas] / __ [+nas]

$ s163-10 ~ s225, 227 -- mid vowel breaking --  first completed in Gallo Roman period I i.e. 6th-7th century as per s163-10
$ RECURRENT throughout period as per s162
$ ˈɛː and ˈɔː tonic free diphthongized 
$ because of action of previous rules we don't have to call the context, just call the long versions
$ which only occur in that context as per our previous rule. 
$ in practice to make this recur we probably have to make them long whenever they become free 
{ˈɛː;ˈɔː} > {ˈi e̯;ˈu o̯}

$s357 Opening of final k 
k > x / [+syl] __ # 

$ s311 and s312 cont'd : (can bleed assibilation) 
$ n palatalizes again before voiced palatals 
$ potentially RECURRENT as per Pope s162 
$ note earlier cases like li:neam have already been hit by an earlier rule. 
[+nas,-lab] > ɲ / __ [+cons,+hi,+front,+voi,-nas] 

$ (Pope s311 and s312 cont'd)
$ l palatalizes again before all palatals, period
[+lat] > ʎ / __ [+cons,+hi,+front]

$ CORRECTION: ɲ back to ŋ if before ɟ then tonic vowel -- case of ingenios,ingenium to engin(s0
$ SUPPRESSED: overfitting
$ ɲ > ŋ / __ ɟ [+prim]

$ (Pope s311 and s312 cont'd)
$... and they absorb the following palatal 
$ CORRECTION : ʎc not effected : dulkiɑm > doʎca > dus etc
[-lat,-nas,-syl,+hi,+front] > ∅  / [+hi,+front,+son,+cons,+nas] __

$ s320-321, palatalization of ŋ before n,s,t (can also include d,z)
$ feeds dental palatalization
$ CORRECTION: moved earlier so that we can simplify this rule 
	$(it does not include velars which will become palatalized...but does include tsj from tj)
ŋ > ɲ / __ [+cor]
$ŋ > ɲ / __ [+ant,+distr]
$ŋ > ɲ / __ [+strid,+cont]

$ CORRECTION : in the case of r r ʝ , ʝ does not merge into a prior rʲ, but rather closes to ɟ
$ would appear to be overfitting, but fits in with what we know of syllable structure factors that caused
	$ ... the emergence of ɟ
ʝ > ɟ / r r __

$ s313-315 : r,s,z don't lose their primary place of articulation but DO palatalize under influence of succeeding palatals
$ potential RECURRENT as per Pope s162
$ s absorbs following: sc, stsj incldued
r ʝ > rʲ
z ʝ > zʲ
s [+hi,+front,+cons] > sʲ sʲ
s [+hi,+front,+cons] > sʲ sʲ  $ RECURRENCE in the form of s sʲ > sʲ 
sʲ > ∅ / __ sʲ sʲ $for sanity : prevents s s j from becoming triple sʲsʲsʲ

$ s306ii -- assibilation of cc done slightly differently than other contexts (see below) 
c c > t̪ t͡sʲ

$ Pope p77 s163-4 ~ p92-93 s191 ~ p125-126 s291-296 : first assibiliation of palatals
$ Attributed (s163) to sixth and seventh centuries
$ as per s291 c retains a palatal element in its articulation as t͡sʲ, but ɟ does not 
$ This stage (but NOT the depalatalizatoin of t͡sʲ explicitly (s296) happens BEFORE unstressed vowel loss
{c;ɟ} > {t͡sʲ;d͡ʒ}

$ Pope p139 s342 : second effacement of ɣ, "Later Gallo-Roman ɣ" -- but it seems this actually belongs to the early stage
$ should come BEFORE any raising of a 
$ this clearly BLEEDS the second palatalization where ɣ is concerned
$ as per s481 this seems to come BEFORE low vowel fronting 

$ s342i : before tonic u or o 
ɣ > ∅ / __ [+tense,+round,+prim]

$ s342iii : between o, u, aw and a 
ɣ > ∅ / [+round,-cons] __ [+lo] $ o|u|au __ a

$ s342ii : between a velar vowel and atonic u (short u > o , so atonic u or o)
$CORRECTION: This one does not occur after sequences of [+syl] [+son] [+back,-cons] not followed by w 
	$-- i.e. cases of Sabigniacum > Savigny, Albiniacum > Aubigny, veracum > v(e)rai etc
$CORRECTION also includes after uo̯ -- case of locum > lieu 
ɣ > ∅ / [+round,-cons,-syl] __ [+tense,-stres,+round,+syl]
ɣ > ∅ / # [+back,-cons] __ [+tense,-stres,+round,+syl]
ɣ > ∅ / [-son] [+back,-cons] __ [+tense,-stres,+round,+syl]
ɣ > ∅ / {[-syl];#} [-syl] [+back,-cons] __ [+tense,-stres,+round,+syl]

$ CORRECTION : ɣ disappears also between ie̯ and o (a: has not yet been raised -- ie, so Savigny, Aubigny etc won't be effected)
ɣ > ∅ / ˈi e̯ __ o

$ CORRECTION : ɣ (possibly actually palatal at this point) also disappears between e and ˈɑː -- loyal, royal etc. 
ɣ > ∅ / [+front,-hi,-lo,+tense] __ ˈɑː 

$Pope p113 s254.
$Final unstressed i, o, u become semi-vocalic and combine into diphthong with previous vowel. 
o > u / [+prim] __ # 
[+hi,-stres] > [-syl] / [+prim] __ # 

$ CORRECTION: deletion of e (< {ɪ or eː} unstressed) before t then  ɑː . 
$ Pope's forms in s350 show results of this but she does not separate it from general unstressed vowel "slurring"
$ CORRECTION: include /i/ too -- ki:wi:tatem
$ CORRECTION: in fact its context includes not only before t but also all coronals
[+front,+tense,-lo,-stres] > ∅ / __ [+cor] [+lo,+long]

$ s350: this time where a b or β has come into contact with t due to the prior rule, the t becomes voiced to d
$ CORRECTION: suppressed. Not diachronically useful (semitam > sente, not sende, dubitare > douter, although subitanum > soudain) 
$ t̪ > d̪ / [+lab,+cons,+ant,+voi] __ [+lo]

$ Pope p160-161 s404:7 -- she calls this one a "glide" -- NOT a fricative (but uses the same symbol) 
∅ > j / ˌe __ [+lo]

$ CORRECTION : ð before a lateral can now assimilate to ɫ here-- just early enough for espɑtulɑ to become espɑɫlɑ
$ CORRECTION: Pope s372 moved here, date was rather unclear in Pope's prose previously ("imply a velarized pronounciation in these words"-- when?)
$ CORRECTION: only becomes dark l after a back or lo vowel, else to normal l 
ð > ɫ / [+back] __ [+lat]
ð > l / __ [+lat]

$ CORRECTION: t coming into contact with sj or zj becomes tsj 
$ essentially an earlier iteration of the existing t s > ts rule -- though it would look like overfitting otherwise
t̪ [+ant,+cont,+strid,+front] > t͡sʲ

$ Pope p77 s163-2 ~ p90 s182 : Palatalization of ɑ > a
$ Pope confirms this also occurred in the diphthong ɑu (> au) -- p191 s505
$ Attributed (s163) to sixth and seventh centuries
[+lo] > [+front,-back]

$ per s182 : temporarily blocked in monosyllables
[+lo] > [-front,+back] / # ([-syl])* __ #

$ s373 assimilation of oral labials b,p,β (and v) to following dental positions 
$ RECURRENT as per s162
$ appears to resolve in EARLY OLD FRENCH, concurrent with degemination 
$ CORRECTION: at this stage, β/ɸ does not assimilate to t or ts ( noce not noice, doute not dotte)
[+lab,-nas,-round,-cont] > t̪ / __ [-cont,+cor,-voi]
[+lab,-nas,-round,+cons] > d̪ / __ [-cont,+cor,+voi] 
[+lab,-nas,-round,+cons] > s / __ s

$s371 nasal cluster assimilation 
$ CORRECTION: m b > m m deleted as it is completely unnecessary and mostly causes errors
$ CORRECTION : only mn, nm doesn't occur and is inconsistent with what appears to be the actual regular pattern elsewhere
n̪ > m / m __

$s374ii : (n)nw > nβ intervocalically 
w > β / [+syl] (n̪) n̪ __ [+syl]

$removal of geminates resulting from assimilations in 3+ consonant clusters
$Pope never explicitly says this but her trajectories for words make this clear
	$ and it is probably better to give her the benefit of the doubt on this rather boring error
{n̪ n̪;s s;d̪ d̪;t̪ t̪;w w;m m}>{n̪;s;d̪;t̪;w;m} / [+cons] __  

$s374i : w from hiatus u,o effaced if coming after two consonants 
$CORRECTION: Pope includes post-velar w (uses placui) as an example, but we exclude post-velar environments
	$ more examples on the other side since at this point former hiatus u,o are totally merged with w
	$ i.e. lingua, linguaticum, quinque, quinquaginta, sanguem, omquam...
w > ∅ / [+cons] [+cons,-back] __

$ s322-326, palatalization of velar frics pre-coronal
[+back,+hi,+cont,-son] > [-back,+front] / __ [+cor]
$ CORRECTION: also pre-ʎ
[+back,+hi,+cont,-son] > [-back,+front] / __ ʎ

$ CORRECTION: ç assimilates to t̪ (no palatal coarticulation) before t͡sʲ
	$ (frisson, façon, sucer, soupçon, ...) 
ç > t̪ / __ t͡sʲ

$ CORRECTION : çs > ss after word initial ˌɛ
ç > s / # ˌɛ __ s

$ CORRECTION : ɛç > eç when secondarily stressed, or after tsj
ˌɛ > ˌe / __ ç

$ CORRECTION: s320-321 moved much earlier. 

$ s321, absorption of following nasal by ɲ
[+nas,+cons] > ∅ / ɲ __ 

$ CORRECTION : ŋ ɲ > ɲ ɲ 
ŋ > ɲ / __ ɲ

$ s322 
[+lat] > ʎ / [+hi,+front,+cons] __ 

$ s316 palatalization of d and t following palatal cons.
$ occurs also for n and r when in position  
$ z and s do not need to be filtered out as they will already be palatalized, edh has not emerged yet. 
$ RECURRENT as per s162
$ must have been active AFTER assibilation (amicitatem > amitstie)
$ and active AFTER intertonic vowel loss (amicitatem > amistie, medietatem > moitie . adiutare >aidier etc) 
$ and active BEFORE depalatalization of tsj (amicitatem > amitstate as per pope > amistie) 
$ and active BEFORE influence on a by preceding palatal
$ and active AFTER palatalization of velar fricatives
[+cor,+ant,-lat] > [+hi,+front] / [+hi,+front,+cons] __ 

$ CORRECTION: p127 s297 moved much earlier. 

$ s167 ~ s181 ~ s42 ~ s9 : beginnings of u > y -- u goes to ʉ
$ s42 -- "a very slow process" -- s167 begins in Early Gallo-Roman (but must come after the beginning of the second palatalization)
$ s9 -- due to Gaulish (also happened in Welsh,Breton,Cornish)
$ CORRECTION: moved earlier because it bleeds the following rule of initial t voicing after ʝ, bleeding specifically for words that had etymological u:
	$ this is also more consistent with Pope's view that it arose due to Gaulish
[+hi,+tense,+round,+syl,-nas] > [-front,-back]

$ s350 -- ... and t "immediately" devoices if it comes into contact (by fall of unstressed vowels) 
$ with a PRECEDING b, β or ʝ
$ CORRECTION : this also effects t͡s
$ CORRECTION : voicing influence of ʝ does not occur if target is right before word coda (iacet >git, pejus > pis etc) or right before another consonant (Tricassae > Troyes, ...) 
$ CORRECTION : same as above for the labials too except that it does occur before voiced consonants (derbitarum > dardre, etc..) 
$ CORRECTION : moved before schwa reinsertion.
$ CORRECTION : ʝ only has this effect at this point if it is coming after a velar or low vowel (ʉ is excluded)
[+cor,+ant] > [+voi] / [+back,+syl] ʝ __ [-cons]
[+cor,+ant] > [+voi] / [+lo] ʝ __ [-cons]

$p73 s154ii : length returns for stressed vowels only, conditioned the open/closed context condition:
$ as per s197-198 -- not "blocked", including word final, followed by vowel, followed by cons + word coda only in a monosyllable,
$ followed by cons + vowel, or followed by plosive + l/r
	$ CORRECTION: v/β + r + V + # also allows lengthening
$ note as per p53 s100, note that Pope does not consider glides to be consonants 
	$ but to be fair she specifies  "in hiatus with a following vowel" -- and all her examples feature
		$ hiatus with syllabic vowels, not semivowels
$ CORRECTION : shortening function operates before [-syl] rather [+cons] (and then # or non-sonorant [+cons] )	
	$; multiple sonorants at the end of the word will cause shortening 
[+syl] > [-long] / __ [-syl] ([-syl])+ #
[+syl] > [-long] / __ [-syl] [+cons]
[-prim] > [-long]
[+prim] > [+long] / __ #
[+prim] > [+long] / __ ([+cons]) [+syl] 
[+prim] > [+long] / __ [-cont] [+cons,+son,-nas] 
[+prim] > [+long] / __ [+lab,+voi,+ant] r [+syl]
[+prim] > [+long] / # ([-syl])* __ [+cons] #

$ s198 checked for /e o a/ before enya and elya
[-hi,+tense,+syl] > [-long] / __ {ʎ;ɲ}

$ CORRECTION: ˈɔ becomes long if it comes before rounded continuant then a stop (mowitam > meute)
$ SUPPRESSED: overfit. 
$ ˈɔ > ˈɔː / __ [+round,+cont,+cons] [-cont]

$ s414 recurrence of diphthongization (s225, 227 ~ s163-10)
{ˈɛː;ˈɔː} > {ˈi e̯;ˈu o̯}

$ s410, s414-415
$ Breaking of mid lax vowels by following palatal 
$ CORRECTION: only happens to primarily stressed vowels though! 
{ˈɛ;ˈɔ} > {ˈi e̯;ˈu o̯} / __ [+cons,+hi,+front]

$CORRECTION: countertonic ɔ goes to ʉ before two palatal consonants consecutive.
ˌɔ > ˌu / __ [+cons,+hi,+front] [+cons,+hi,+front]

$ as per s410 niece, tuertre cases 
$ CORRECTION: This is much more restricted than Pope describes.
$  --- It does not ultimatedly diachronically matter at all for ɔ it seems 
		$ --- (see noptiam > noce, skorteam > ecorce etc)
$  --- It only occurs for primarily stressed ˈɛː (messionem > moisson, mentionica > mensonge etc) 
$ It also only occurs before ttsʲ and rtsʲ segments (walentiam>valence not valience, Arewernia>Auvergne not Auviergne etc)
$ CORRECTION: only occurs triggered by noncontinuant palatals (not ɲ -- Auvergne not *Auviergne) 
ˈɛ > ˈi e̯ / __ [+cons,+cor,-nas] [+cons,+hi,+front,-cont] 

$ s298-302, s413-415 -- Beginning of second palatalization 
$ Note: although Pope lists the second palatalization for the next period, her notes make it clear that she holds it to have preceded the a-raising therefore e/o-breaking phenomena
$ so caused. 
[+hi,+back,+cons] > [+front] / __ [+front,+syl] 

$ CORRECTION : Formation ˈeː o > j ˈoː of certain diphthongs for unstressed tense final vowels before becoming glides or effaced
$ SUPPRESSED: overfit
$ ˈeː o > j ˈoː

$ s254 recurrence -- final unstressed i , o , u in hiatus with tonic vowel (in Late Latin or)
	$ brought into hiatus in Early Gallo Roman become semi-vocalic. 
	$ bleeds effacement
$ CORRECTION: effects all ɛei in hiatus with prior stressed vowels (aerem, lui, jeudi, lai, etc...) 
$ CORRECTION : same for unstressed o and ɔ in hiatus wiht stressed vowel
[-lo,-stres] > [+hi,-syl,+tense] / [+prim] ([-cons,-hi]) __

$ CORRECTION: any vowel before w shortens 
[+syl] > [-long] / __ w

$ Pope p183 s481 : Gallo Roman a-mutation under influence of following w or u>o 
$ must bleed s413 but not second palatalization
{a;ˌa;ˈa} > {ɔ;ˌɔ;ˈɔ} / __ w

$ s417 Raising of free countertonic ˌa to ˌɛ if preceded by a palatal consonant
$ unless it is followed by tonic in next syllable or if l/r come next
$ CORRECTION: actually whether the next syllable is tonic e does not matter
$ instead the constraining factors are that it doesn't occur before coronals
$ and that it doesn't occur before nasals unless followed by a long vowel
ˌa > ˌɛ / [+cons,+front] __ [-cor,-nas] [-cons]
ˌa > ˌɛ / [+cons,+front] __ [+nas] [+syl,+long]

$ s413
$ Raising of free tonic a to ɛ if preceded by palatal consonants 
$ CORRECTION: unless at beginning of word, this does not happen before nasal consonants and also doesn't happen before ʝ
ˈaː > ˈɛː / # [+cons,+front] __
ˈaː > ˈɛː / [+cons,+front] __ [+cons,-front,-nas]
ˈaː > ˈɛː / [+cons,+front] __ rʲ $rʲ is the only one except ʝ that would occur in the relevant spot in the dataset. 

$ s414 recurrence of diphthongization (s163-10)
$ hits former tonic a 
{ˈɛː;ˈɔː} > {ˈi e̯;ˈu o̯}

$ CORRECTION: beginning of u > y fronting moved earlier. 

$MIDDLE GALLO-ROMAN
$ Seventh century as per Pope s164

$Absorption of prior velar (now palatal) frics by lambda, resolving s322
$also by rj -- resolving s323
$ note : this technically will incorrectly make cases like
	$ varium etc
	$ equivalent to facere, disfacere etc
	$ in reality the latter ejected a final schwa but the former did not
		$ but there is no way yet to formalize this without massive overfitting to small data. 
		$ this only occurred word-finally: see lacrima > lairme (>larme), not laireme
		$ also it did not effect gr clusters : agrum > air, nigrum > noir (however tragere> traire ... analogy?)
		$ TODO work on this case may be necessary
[+hi,+front,-lat,-syl] > ∅ / __ {ʎ;rʲ}

$ 164-1 Second palatalization resolves s298-302
[+back,+front] > [-back] 

$ s164-4 ~ s333 -- opening of d intervocalic
$ as per s372 -- also happens before r,l,m,n 
$ RECURRENT as per s162
$ since g and b have already opened and htey recur-- and ɟ doesn't occur in the right context -- we can include them too. 
$ however b explicitly doesn't not open before l or nasals as per s372
[-delrel,+voi] > [+cont] / [-cons] __ {[-cons];r}
[-delrel,-lab,+voi] > [+cont] / [-cons] __ [+lat] 

$ s372 : d̪ > ð before w 
d̪ > ð / __ [+cont,+lab,+cons]

$ s355-356 first opening of final t,d
$does not however occur after slurred vowel ə
$ would seem to happen BEFORE the palatal fricatives are lost as it doesn't happen it words like lit, nuit (<lectum, noctem)
{t̪;d̪} > [+cont] / [+syl] __ #

$ s305-- yod in initial position closes, as does Pope's w 
$ also -- s203 (for yod > gj) ; yod is closing after consonants 
$ w ~ ɣʷ closes word-initially : s636 ~ s192ii
$ s203 holds this begins in the "early Gallo-Roman" though its later stages occur after the influx of Frankish vocab
$ CORRECTION : Yod-closing does NOT occur in the context of # [+cons] __
[+hi,-syl,+cont] > [-delrel] / # __ [+syl]
{j;ʝ} > {ɟ;ɟ} / @ ([+cons])+ __ [+syl]

$ s305 -- closed yod devoices if after a voiceless element (sapia > sache) 
ɟ > c / [-voi] __ 

$ s164.2 : palatal stops c,ɟ become postalveolar affricates 
$ Dated to seventh century in summary by Pope (s164)
$ s299 "must have taken place before the end of the 8th century" 
[+hi,+front,-cont,-ant] > [+delrel,+cor,-hi,+strid,+distr]

$ per s685 -- never explicated elsewhere -- m > n before d͡ʒ
m > n̪ / __ d͡ʒ

$ CORRECTION : former geminate velars assimilate
{k t͡ʃ;ɡ d͡ʒ} > {t̪ t͡ʃ;d̪ d͡ʒ}

$ CORRECTION: tense ˈeː becomes lax ˈɛː if before a d/ð e ɣ sequence $ in future there may be a more graceful way to handle this...
ˈeː > ˈɛː / __ [+voi,+cor,+ant,+distr] e ɣ

$ CORRECTION: β deleted after countertonic ˌɔ or ˌo and before a tense front vowel, when not after a front element
	$ not a lax one : novɛllum > nouveau etc -- ouaille, ou, rouille etc.
	$ (essentially a generalization of the existing β effacement after round back vowels) 
β > ∅ / {#;[-front]} [+stres,-prim,+round,+back] __ [+syl,+front,+tense]

$s164-10 ~s250 complete effacement of unstressed penultimate
$as per s260: occurs in 7th century; but either does not occur or is reversed where voyelle d'appui is required as per s257-258
$the voyelle d'appui is inevitable ə
$ as per s259 : this resolves after the reduction of non-stressed non-low vowels everywhere, as there is a different outcome for
$ paroxytones as opposed to proparoxytones. 
$ CORRECTION: cannot come immediately before final vowel
$ CORRECTION: a actually doesn't go to schwa yet if before t/ð then a vowel
a > ə / __ ([-syl])+ [+syl] ([-syl])* # 
ə > a / __ [+distr] [+syl]
[-stres,-lo] > ə / __ ([-syl])+ [+syl] ([-syl])* # 

$ s261 ~ s350 : as per word trajectories seen in s350, under justification of s261
$ non-penultimate unstressed vowels except for a are effaced "early" when next syllˌs stressed vowel is a
$ this happens before the voicing of t
$ this will resolve after s257-259 does. 
$ CORRECTION: requires posterior context of non-vowel
[-stres,-lo] > ə / __ ([-syl])+ [+lo,+prim]  

$ s26: purportedly under Frankish influence (Germanic "mutation"), the suffixes -arium and -aria
$ end up rendering -ier and iere, by means of changing the a to an open/lax e
$ r depalatalizes when this happens
$ only happens in the "suffix" -- i.e. when a prior syllable exists
$ element exists in the word, not cases like varium or aria
$ thanks to Latin stress rules, we can safely assume we are talking about a long and stressed a
$ placement before 164.11 is decided by the presence of non-reduced unstressed final vowels in her forms
	$ but it must be after a-fronting, as she uses a rather than alpha. 
$ Attested in the Glossary of Reichenau and, in a name, in the Oaths of Strasbourg
$ CORRECTION: when not preceded by a consonant, this instead goes to jˈɛː 
ˈaː rʲ > j ˈɛː r / [+syl] __ [+syl] #
ˈaː rʲ > ˈɛː r / [+syl] ([-syl])* __ [+syl] # 

$ s164-11 ~ s256 Reduction of non-stressed non-low vowels to schwa 
$ NOTE: this actually occurred a bit later than the ones above, 
$ but diachronically it is convenient to place it here.
$ ultimate effacement is "before the ninth century".
$ CORRECTION: requires posterior context of non-vowel (or word final) 
	$unless it is lax (advocatum >avoue but deunde > dunt in GR (>dont), manuopera > manœuvre etc 
[-stres,-lo,+syl] > ə / __ {[-syl];#} 
[-stres,-lo,+syl,+tense] > ə

$ CORRECTION : specification of added rule -- s365iii bleeds some n flapping
$ this is going to be recurrent...
[+cons,+lab] > ∅ / [+cons] __ [-son]
[+cons,+lab] > ∅ / [+cons] __ [+nas,+cons]

$ Pope dismisses n-flapping phenomena as characteristic only of loan words (s642) 
	$ but it can be modeled regularly. 
$ CORRECTION: n,r,l become flap when after velar or dental (NOT alveolar) occlusives 
	$ and between two schwas then (s)#
[+son,+cor] > ɾ / [-son,-lab,-strid,-front] ə __ ə (s) # 

$ CORRECTION: specifically after /r (ə)/, /n/ will flap (and in no case be renasalized)
$ Dating as per TLFi : appears to complete around 1100-- 1050 jorn > 1100 jor
	$ 1100 sujurn > 1150 sejor 
$ However it is essentially impossible to get a correct rule cascade if we do that
	$ perhaps there was some distinct sequence for a time,
		$ or the written form was just conservative? 
	$ or perhaps the written -rn (wiht no e) ending indicated -rr in fact...?
n̪ > ɾ / r (ə) __ ə

$ CORRECTION: nə,ra,rə (əna never will be generated)
	$  all become ɾə before bilabials
	$before intervocalic/pre-rhotic continuant, 
	$ as long as neither of hte next two segs are hi
[+cor,+son,-hi] [+syl,-stres] > ɾ ə / [+syl] __ [+cons,+cont,-cor] [+son,-lat,-hi]

$ CORRECTION: recurrence of tap whistling as dissimilation from prior r at prior syllable onset
$ PRUNED: appears to have only one example at the moment, Jero^me. 
$ ɾ ə > z / r [+syl] __ 

$ s256 : ə effaced before unstressed a becomes the new ə, but retained in specific places (s256-259):  
ə > ə̯

$ CORRECTION: ɾə immediately collapses to trill if after non-low front vowel (derree, merme, cerveau, cervelle). Allow a /r/ in between which can coalesce with ɾ > r
ɾ ə̯ > r / [+front,-lo,+syl] (r) __

$ CORRECTION: lambdacism of remaining tap before ə̯ then voiced consonantal (palefroi, aume, aumaille, pele(gh)rin...)
	$ as long as followed by vanishing schwa then a voiced cons
ɾ > l / __ ə̯ [+cons,+voi]

$ CORRECTION : final ə̯ r > r ə̯
{ə̯ r;ə̯ ɾ} > {r ə̯;ɾ ə̯} / __ #

$ CORRECTION: schwa retention after l to block formation of new clusters of shape l (obstruent) r (pelerin, palefroi...) 
$ may be overfitting (two examples, pelerin and palefroi) but does align with what we know of what clusters are likely impermissible (i.e. lβr, lɣr...̠
ə̯ > ə / [+lat] __ ([-son]) (ə̯) [+son,+cor,-nas]

$ s257ia : ə intertonic retained when after cons + r or l but not rr or ll 
$ CORRECTION: extended to include word final scenarios
ə̯ > ə / [+stres] ([-syl])* {[+cons,-ant];[+distr]} [-distr,+son] __
$ s257ib : ə intertonic retained when before ʎ,ɲ,ts or "any particularly heavy group of consonants"  
$ CORRECTION: only retained before ʎ or ɲ -- not tsʲ or ç (with regards to palatal sounds 
	$-- i.e. we add +son specification)
$ CORRECTION: retained before ŋ t͡sʲ 
	$but it doesn't seem like this is true of any other non-sonorant palatal ending cons clusters 
	$ (second rule here thus changed away from just marking "heavy" palatal ended clusters to just ŋ t͡sʲ) 
$ CORRECTION: ə also preserved between m and ts followed by a tonic
$ CORRECTION: ə is NOT preserved before rʲ
ə̯ > ə / __ [+front,+cons,+son]
ə > ə̯ / __ rʲ
ə̯ > ə / m __ t͡sʲ [+prim]
ə̯ > ə / __ ŋ t͡sʲ
ə̯ > ə / __ [+cons] [+cons] [+cons]
$s258a : ə retained before n̪ t̪ #
$ CORRECTION: retained before all -nt, not just finally
ə̯ > ə / __ [+nas] t̪
$ CORRECTION ə temporarily retained between t/d/ð/r/l and ɣ, just long enough for t to lenit and handling cases like pelerin
ə̯ > ə / [+cor,-strid] __ ɣ
$ CORRECTION : ə also retained after dental ə ɣ for now 
ə̯ > ə / [+cor,+ant,+distr] ə ɣ __
$ CORRECTION : ə also retained when after [+cons] w and before [+cons] 
ə̯ > ə / [+cons] w __ [+cons] 
$ CORRECTION : ə also preserved when after [+cons] p __ followed by -cont, and the next syllable is tonic 
ə̯ > ə / [+cons] p __ [-cont] ([-syl])* [+prim]
$ CORRECTION : ə always retained after dʒ or tʃ\
ə̯ > ə / [+delrel,-ant] __

$ CORRECTION: ə̯ > e before final s when there is no other syllabic in the syllable
$ overfitting within this set possibly (mes, les) but consider also: ses, tes, etc. 
ə̯ > e / # ([-syl])+ __ s #

$s259: ə retained in final syllable of proparoxytones that were not deleted in Late Latin or Early Gallo-Roman 
ə̯ > ə / ə̯ ([-syl])* __ # 

$ CORRECTION: ə retained if there are no syllabic elements BEFORE it 
$ we know this operated from the end rather than the beginning of the word because of cases like illo >>>> lə (le) 
$ examples: que, ne; could be seen as overfitting but not having this would imply a major violation of syllable restraints as French requires a phonemic coda 
	$ (note however that both ne and que tend to have no coda when the next word begins in a vowel...) 
		$ (but their citation forms do contain ə)  
ə̯ > ə / # ([-syl])* __ #

$ CORRECTION : ə̯ retained after STOP r and before a cons
ə̯ > ə / [-cont] r __ [+cons]

$ CORRECTION : ə̯ retained after non-lateral cons then l, and before cons
ə̯ > ə / [-lat,+cons] l __ [+cons]

$ Completion of fall of the schwas 
$Delete all the rest which were not specifically retained: 
ə̯ > ∅ 

$ Pope s351 : affricate tsj voices and depalatalizes when it is about to comes into contact with d through schwa fall, 
	$ and is followed by /ə/
	$ Pope attributes it technically to the "consonant group"
d̪ t͡sʲ > d͡z / __ ə 

$ CORRECTION : ð that comes into contact with a prior lateral sonorant cons becomes a stop again
$ exemplified but not explained in forms in s669
ð > d̪ / [+lat] __

$ CORRECTION : ð is also stopped after labials (sade, tiede, rade) but only labial oral stops, not nasals (comme < quomodo)
ð > d̪ / [-cont,+lab] __

$ CORRECTION : ð that has come into contact with any other prior consonant effaces 
	$ Autun may have retained the  ð longer -- see cited "Ostedun" form in s257
	$ forms cited by Pope show this shift but she never explains it anywhere
ð > ∅ / [+cons] __

$ CORRECTION : β that has come into contact with l becomes b
β > b / __ l

$ CORRECTION : ŋ ɡ w at end of word to ŋ ɡ (sanguem > sang etc)
	$ this looks like overcorrecting, but given French syllable structure it is so obvious this needs to be here, it is ok. 
w > ∅ / ŋ ɡ __ #

$ CORRECTION: simplify word initial k t͡sʲ clusters to just t͡sʲ
$ definitively after first fall of schwas 
k > ∅  / __ t͡sʲ

$ RECURSE
$ s350 -- ...  t "immediately" voices if it comes into contact (by fall of the schwas) 
$ with a PRECEDING b, β or ʝ
$ CORRECTION : this also effects t͡s -- this is exemplified in Pope s250 but she never explains it
$ CORRECTION : voicing influence of ʝ does not occur if target is right before word coda (iacet >git, pejus > pis etc) or right before another consonant (Tricassae > Troyes, ...) 
$ CORRECTION : same as above for the labials too except that it does occur before voiced consonants (derbitarum > dardre, etc..) 
$ *CORRECTION* (this one not covered by earlier recursion)
	$ : influence of yod only effects palatalized consonants -- voidier (βˌɔʝd̪ʲˈie̯r) and aise (aːʝd͡zʲə) but moitie' (mˌɛʝt̪ˈaːt̪)
$ CORRECTION : moved before schwa reinsertion.
$ CORRECTION : ʝ only has this effect at this point if it is coming after a velar or low vowel (ʉ is excluded)
[+cor,+ant] > [+voi] / [+lab,+voi,-hi,+cons,-nas] __ [+voi]
[+cor,+ant,+front] > [+voi] / [+back,+syl] ʝ __ [-cons]
[+cor,+ant,+front] > [+voi] / [+lo] ʝ __ [-cons]

$ CORRECTION : ʝ > j if no syllabic units left afterward
	$ -- ə will not be reinserted just because ʝ is "not sonorant" 
ʝ > j / __ ([-syl])* #

$ CORRECTION: r ə t͡s(ʲ) > r t t͡s(ʲ)
$ SUPPRESSED: overfit
$ə > t̪ / r __ [+delrel,-voi]
$ CORRECTION : r t t͡s > r t t͡s
$ SUPPRESSED: overfit
$t̪ > ∅ / __ r t̪ [+delrel]

$s258b: ... but ə retained finally after tʃ,dʒ, cons+r/l but not rr/ll/lr, and after [+lat][+nas] 
$ CORRECTION : Pope only has ə retained after ɫ [+nas] -- 
	$ but per her notes (s390) ɫ hadn't developed yet! Hence we only use [+lat] 
$ CORRECTION : expand [+lat] [+nas] __ context to any non-nasal consonantal then a nasal cons
$ CORRECTION : ə inserted after cons + son before word codas in general
$ CORRECTION : retained after +cons ɲ __ # (Auvergne, peigne, etc) 
$ CORRECTION: retained after double m
∅ > ə / [+cons,-son] [+son,-syl] __ #
∅ > ə / [+cons,-nas] [+nas] __ #
∅ > ə / [+cons] ɲ __ #
∅ > ə / [-ant,+delrel] __ #
∅ > ə / [-lat,+cons] l __ #
∅ > ə / m m __ #

$ CORRECTION : endings of [+lab] [dental,-cont] (NOT alveolar!), 
	$ td/cd and dt also eject ə. td also assimilates at this time
∅ > ə / [+cons,+lab] [+cor,-delrel] __ #
[+cor,+ant,+distr,+voi,-cont] > t̪ ə / [-voi] __ #
∅ > ə / d̪ t̪ __ #

$ s355 -- final spirantization reversed when come into contact via slurring to consonantal. 
[+ant,-son,+distr] > [-delrel] / [+cons] __ #

$ s316 palatalization of d and t following palatal cons.
$ occurs also for n and r when in position
	$ CORRECTION: s palatalization (s318) and z merged into this. 
$ RECURRENT as per s162
[+cor,+ant,-lat] > [+hi,+front] / [+hi,+front,+cons] __ 

$ s367 d,t + s becomes the affricate when brought into contact first fall of the schwas. 
$ must come before degemination of t t s (adsatis > assez etc but plattos > plats etc) 
$ CORRECTION: this also occurs for ð,θ,s
$ CORRECTION: but this only occurs word finally
[+ant,+distr,-nas] s > t͡s / __ #

$ CORRECTION: z deleted before s 
	$ would appear overfit but consistent with syllable structure laws which certainly do not permit codas of -zs. 
z > ∅ / __ s

$ CORRECTION: geminate stops are degeminated if after consonant, or before consonant then non-non-consonantal (i.e. either consonantal or coda)
$ deal with labials here -- others later, given that labials will mutate to form new geminate clusters with all the coronals after this
$ CORRECTION: will shorten also for a single consonant that is not r,l
$ would appear overfit, but this is just a generalization of the phonological rules Pope already agrees with. 
{b b;p p}>{b;p} / [+cons] __
{b b;p p}>{b;p} / __ [+cons] {[+cons];#}
{b b;p p}>{b;p} / __ {[+cons,-son];[+cons,+nas]}

$ s373 assimilation of labials b,p,β (and v) to following dental positions 
$ RECURRENT as per s162
$ appears to resolve in EARLY OLD FRENCH, concurrent with degemination 
[+lab,-nas,-round,+cons] > t̪ / __ [-cont,+cor,-voi]
[+lab,-nas,-round,+cons] > d̪ / __ [-cont,+cor,+voi] 
[+lab,-nas,-round,+cons] > s / __ s

$s371 nasal cluster assimilation (recurrence)
$ CORRECTION: m b > m m deleted as it is completely unnecessary and mostly causes errors
$ CORRECTION : only mn, nm doesn't occur and is inconsistent with what appears to be the actual regular pattern elsewhere
n̪ > m / m __

$removal of geminates resulting from assimilations in 3+ consonant clusters
$Pope never explicitly says this but her trajectories for words make this clear
$ recurrent since the triggers are! 
$ CORRECTION: geminate stops are degeminated if after consonant, 
	$ or before consonant then non-non-consonantal (i.e. either consonantal or coda)
$ --don't have to deal with the affricate clusters or palatals in the latter case as they never occur 
$ CORRECTION: will shorten also for a single consonant that is not r,l
$ CORRECTION : s,m,l also shortened when double after another consonant or word initially - include also ɫl for good measure
{d̪ d̪;t̪ t̪;t̪ t͡s;d̪ d͡ʒ;t̪ t͡ʃ;ɟ ɟ;c c;ɡ ɡ;k k;s s;m m;l l;ɫ l} > {d̪;t̪;t͡s;d͡ʒ;t͡ʃ;ɟ;c;ɡ;k;s;m;l;l} / {[+cons];#} __
{d̪ d̪;t̪ t̪;ɡ ɡ;k k} > {d̪;t̪;ɡ;k} / __ [+cons] {[+cons];#}
{d̪ d̪;t̪ t̪;ɡ ɡ;k k} > {d̪;t̪;ɡ;k} / __ {[+cons,-son];[+cons,+nas]}
{s s;m m}>{s;m} / __ [+cons] 

$ CORRECTION : t̪ t͡sʲ in word initial position also eliminated. 
$ For good measure, handle also any word initial stop followed by a non-continuant
[-cont] > ∅ / # __ [-cont]

$ CORRECTION : t͡sʲ sʲ > t͡sʲ everywhere 
	$ would appear as overfitting but this is consistent with known syllable structure rules (otherwise we end up with malvats.s) 
t͡sʲ sʲ > t͡sʲ

$ action resulting on clusters formed form vowel losses 
$ Denasalization -- s369, and glide development s370
$ b inserted between m + r or l 
$ d inserted between n,l,dark l,z,elya + r 
$ t inserted between s and pal. s + r 
$ CORRECTION : denasalization of nr > ndr comes later -- after length is determined for the relevant words, determining diphthongization
$ note: minor > meindre > moindre, but simulare > sembler (not *soimbler) and camera > chambre (not *chaimbre) 
	$ -- note however criembre -- so the diphthongization of ɛ > ie came before) and 
∅ > b / m __ [+cor,-nas,+son,-syl]
∅ > d̪ / [+lat] __ r
{z r;s r} > {z d̪ r;s t̪ r}

$ s189 : nβl > mbl 
n β > m b / __ [+lat] 

$ CORRECTION: in cases of [-cons] ([+cons])+ rʲ, rʲ depalatalizes 
 $after ejecting a /j/ at the beginning of the cons cluster
$ -- evrium, ostrea, cupreum, materiamen... 
$ in the case of ostrea>huitre and materiamen>merrain, the effects are apparent in cited forms
	$ in s313, but never explained. 
$ this is in effect a reordering as the depalatalization is moved earlier for these segments.  
∅ > j / [-cons] __ ([+cons,-front])+ rʲ

$ CORRECTION: rʲ emits j immediately backward when preceded by a vowel... but... 
∅ > j / [-cons] __ rʲ

$ note that the a: rʲ have already been handled above by s26 . 
$ CORRECTION: in the case of ˈi e̯ it's going to be deleted soon so we can just delete it now... 
j > ∅ / [+syl] ([-syl])* ˈi e̯ __ rʲ 

$ s164-5 ~ ss334
$ as per S350 -- t was voiced AFTER INTERTONIC VOWEL LOSS 
$ voicing of p, t, ts intervocalically ss334-335 [Second Major Lenition]
$ RECURRENT as per s162
$ since k,s,f have already voiced intervocalically, and the new results of palatalization will never be intervocalic
$ and because the voicing of k,s,f is also recurrent (s162)
$ we can just have this as a universal intervocalic voicing rule
$ CORRECTION: moved to before post-palatal a-raising
[-voi] > [+voi] / [-cons] __ [-cons]
$ as per s372 p, t also voice before l,r and t voices before n 
[-voi,-delrel,+ant] > [+voi] / [-cons] __ [+cor,+son,-nas]
t̪ > d̪ / __ [+nas,+cons]

$p73 s154ii : length returns for stressed vowels only, conditioned the open/closed context condition:
$ as per s197-198 -- not "blocked", including word final, followed by vowel, followed by cons + word coda only in a monosyllable,
$ followed by cons + vowel, or followed by plosive + l/r
	$ CORRECTION: also [+lab,+voi]rV# allows lengthening
$ note as per p53 s100, note that Pope does not consider glides to be consonants 
	$ but to be fair she specifies  "in hiatus with a following vowel" -- and all her examples feature
		$ hiatus with syllabic vowels, not semivowels$ CORRECTION : shortening function operates before [-syl] rather [+cons] (and then # or non-sonorant [+cons] )	
$ CORRECTION : no more shortening at end of word if there is only one consonant between the vowel and the coda 
$ CORRECTION: shortening doesn't occur before fric + r/l 
[+syl] > [-long] / __ [-syl] [+cons,-son]
[+syl] > [-long] / __ [-syl] [+cons,+nas]
[-prim] > [-long]
[+prim] > [+long] / __ #
[+prim] > [+long] / __ [+syl]
[+prim] > [+long] / __ [+cons] [+syl] 
[+prim] > [+long] / __ [-cont] [+cons,+son,-nas] 
[+prim] > [+long] / __ [+lab,+ant,+voi] r [+syl] #
[+prim] > [+long] / # ([-syl])* __ [+cons] #
$ s198 checked for /e o a/ before enya and elya
[-hi,+tense,+syl] > [-long] / __ {ʎ;ɲ}

$ CORRECTION: length also gained by primary syllables when followed by an ejected j
[+prim] > [+long] / __ j ([+cons])* [+cor,+hi,+front]

$ CORRECTION: length retained in stressed syllables in multisyllabic words followed by rs#
[+prim] > [+long] / [+syl] ([-syl])* __ r s #

$ s352 : in deɣo and remaining deɣa (-tikum,-tika) suffixes, e (>ə) deleted ɑnd ɣ palatalizes and closes to make dʒ
$ CORRECTION: This happens anywhere in the word where this sequence exists, not just at the end
$ CORRECTION: where d has remained t, it becomes a voiceless tʃ
$ CORRECTION: t and d are actually preserved before the new affricates at first. 
$ CORRECTION: this happens in the midst of the process of the fall of the first schwa. 
$ CORRECTION: this bleeds post-palatal a-raising (moved earlier)
{t̪ ə ɣ;d̪ ə ɣ;ð ə ɣ} > {t̪ t͡ʃ;d̪ d͡ʒ;d̪ d͡ʒ}

$ s413-4 -- recurrence as per medietatem, adiutare
$ Raising of free tonic a to ɛ if preceded by palatal consonants
$ recurses as per cases of adiutare and medietatem (s414) 
$ CORRECTION: also before rʲ , r s 
$ CORRECTION: also after pure palatal continuants (j,ʝ)  before nasals
$ CORRECTION: does not occur immediately word finally *this time*
[+lo,+prim] > [-lo,+front,-tense] / [+front,-syl] __ [-cons]
[+lo,+prim] > [-lo,+front,-tense] / [+front,-syl] __ [+cons] [-cons]
[+lo,+prim] > [-lo,+front,-tense] / [+front,-syl] __ [-cont] [+cons,+son,-nas] 
[+lo,+prim] > [-lo,+front,-tense] / [+front,-syl] __ rʲ #
[+lo,+prim] > [-lo,+front,-tense] / [+front,-syl] __ r s #
[+lo,+prim] > [-lo,+front,-tense] / [+front,-cor,-syl,+cont] __ [+son,-lat,+cons]

$ s414 recurrence of diphthongization (s163-10) that hits former tonic a 
$ CORRECTION: includes tonic ɔ
{ˈɛː;ˈɔː} > {ˈi e̯;ˈu o̯}

$ CORRECTION: j ˈi e̯ then becomes ˈi e̯ when it came from rʲ and would not create a hiatus. 
j > ∅ / [+syl] ([-syl])+ __ ˈi e̯ rʲ

$ s167 ~ s181 ~ s42 ~ s9 : continuation of u > y -- u goes to ʉ
[+hi,+tense,+round,+syl,-nas] > [-front,-back]

$ CORRECTION: s164.5/s334 moved earlier, as is s352.

$ s164-7 ~ s335e : another bout of spirantization (recurrent anyways as per s162) 
$ before l,r included as per s372 (but not b before l as per 372iii)
[-delrel,+voi] > [+cont] / [-cons] __ {[-cons];[+son,-nas,-lat]}
[-delrel,-lab,+voi] > [+cont] / [-cons] __ [+lat] 
$ s372 d > ð before nasals , and then ð > z before n "sometimes" 
	$(excluding the latter, does not seem to have diachronic use, examples are Roðnə >Rosne, boðne>bozne, but it ended as Rhone and borne.)
d̪ > ð / [+syl] __ [+nas] 

$ CORRECTION: flap nasalizes if after ð (pro(z)ne, Rho(z)ne, borne)
ɾ > n̪ / ð __

$ CORRECTION: ɾ goes to r everywhere else 
ɾ > r 

$CORRECTION : remaining ʝ becomes j sometime at this point 
$ CORRECTION : moved earlier as bleeds first voicing lenition of t : medietatem > moitie -- moved further down past this point. 
$ CORRECTION : also counterfeeds final spirantization of voiced consonants -- rigidam > raide not raie, wokitare > voidier not voyer etc. 
ʝ > j 

$p93 s192i kw > k before e and i at beginning of word
$ must come after second palatalization resolves
w > ∅ / k __ [+front,-lo]

$ s164-12 : closing of tense and lax o to u by following nasal consonants ~s426
$ according to Pope (s426), attested in the Glossary of Reichenau (?)
$ s426 -- this explicitly comes BEFORE and BLEEDS the diphthongization of o > ou 
$ Pope does NOT specify that it becomes nasal at this point -- actually she says the "first nasal vowels in Frnech were a and ai in the tenth century" (s434)
$ but in all fairness it probably SHOULD be to prevent it from getting swept up in u>ʉ>y, the first stages of which are still active at this point
$ CORRECTION : we now have this happening much earlier, but we maintain this here in case /o/ variants got reintroduced before this point 
	$-- allowing for later possibility of including loanwords into paradigm.
[+syl,+back,+round] > [+hi,+tense] / __ [+nas,+cons]

$ CORRECTION : early degimination of geminate palatal s before depalatalization of coronals resolves
$ 	(this is a reordering)
sʲ sʲ > sʲ

$ s164-3 ~ s190 ~ s279 ~ s311,312,313,314,315 
$ ... palatalized dentals resolve into j + simple dental; lj not effected
$ CORRECTION: from palatalized coronals except tsj, backward ejection of j after something not palatally articulated itself 
$ CORRECTION: any remaining d͡zʲ also ejects j backwards and depalatalizes at this stage
$ CORRECTION: /rʲ/'s emission of j backward moved earlier, away from this point, so that it can trigger breaking of /a/
rʲ > r
$ for maintenance -- Pope doesn't account for the fact that this could end up with a /j/ + j + Cons cluster
$ but this is not an interesting error, just a clumsy one -- obviously no such cluster can exist
$ so we bleed it with a special rule... 
{t̪ʲ;d̪ʲ;sʲ;zʲ;n̪ʲ;d͡zʲ} > {t̪;d̪;s;z;n̪;d͡z} / [+hi,+front,-syl] __ 
{t̪ʲ;d̪ʲ;sʲ;zʲ;n̪ʲ;d͡zʲ} > {j t̪;j d̪;j s;j z;j n̪;j d͡z} 


$ CORRECTION: case of disiunare, prehensionem, etc -- ˌej goes to ɛj, and thus ultimately ˌi, before zʲ
$ deaffircation of dz counterfeeds this (lice:re > loisir, ve:ci:num > voisin, etc) 
ˌe > ˌɛ / __ j z

$ CORRECTION: ɛj > i when countertonic and before coronal (priser, prison, di^ner...) before depalatalization of coronals resolves
$ (NOTE: may make sense to merge this with the broader ˈɛj > i shifts)  
ˌɛ j > ˌi / __ [+cor]

$ CORRECTION: palatalized coronals, excluding tsʲ are actually preserved into Old French in a specific position between a palatal continuant cons and i
$ ... tsʲ is handled separately
$ There is evidence that the deletion of j in -ier endings centuries later, which Pope and others attributed to analogy, can instead be explained regularly
	$ as the absorption of yod by a persistently palatalized coronal
	$ that it was absorbed by palatal segments that uncontroversially remained -- ʃ,ʒ,ʎ and ɲ at that point -- is the consensus view
		$ to this list, we add persistently palatalized coronals, which persisted between two palatal segments (consonantal or not)
			$ except in cases where the prior palatal consonant was preceded by countertonic i or e 
		$ persistently palatalized coronals will disappear after they can condition the loss of j in those endings
	$ this may be more likely than analogy (we argue for palatal persistence 
		$because otherwise there would have to be a significant change in the chronology of diphthong developments 
			$so that 'ie > j'e happened much earlier
$  NOTE : needs to be suppressed for demonstrative purposes for case traiter
$ may need to reanalyze this -- unsure if analogy (see extended discussion in palatalization remarks...) 
[+cor] > [+hi,+front] / [+front,+hi] __ [+hi,+front]
{t̪ʲ;d̪ʲ;sʲ;zʲ;n̪ʲ;d͡zʲ;lʲ;rʲ} > {t̪;d̪;s;z;n̪;d͡z;l;r} / [+front,+tense,-lo,+syl] [+hi] __

$ we trust Pope on her word (s318) that s "palatalizes" into t͡sʲ word finally after palatal lateral and nasal elements rather than a simple spirant
	$ and also [CORRECTION] expand it to include all palatal sonorant consonantals
s > t͡sʲ / [+cons,+son,+hi,+front] __ #

$ s294 dz opens to z intervocalically due to lenition (Pope implies this by citing s334) before it resolves to jz
d͡z > z / [-cons] __ [-cons]

$p73 s154ii : length returns for stressed vowels only, conditioned the open/closed context condition:
$ as per s197-198 -- not "blocked", including word final, followed by vowel, followed by cons + word coda only in a monosyllable,
$ followed by cons + vowel, or followed by plosive + l/r
	$ CORRECTION: also allow lengthening before vrV#
$ note as per p53 s100, note that Pope does not consider glides to be consonants 
	$ but to be fair she specifies  "in hiatus with a following vowel" -- and all her examples feature
		$ hiatus with syllabic vowels, not semivowels$ CORRECTION : shortening function operates before [-syl] rather [+cons] (and then # or non-sonorant [+cons] )	
$ CORRECTION : no more shortening at end of word if there is only one consonant between the vowel and the coda 
$ CORRECTION: multiple final consonants, sonorant or not, will always shorten a vowel in the final syllable
$ CORRECTION: shortening doesn't occur here before fric + r/l : frere, mere, poele, moindre, pere, sourcil
[+syl] > [-long] / __ [-syl] ([-syl])+ #
[+syl] > [-long] / __ [-syl] [+cons,-son]
[+syl] > [-long] / __ [-syl] [+cons,+nas]
[-prim] > [-long]
[+prim] > [+long] / __ #
[+prim] > [+long] / __ [+syl]  
[+prim] > [+long] / __ [+cons] [+syl] 
[+prim] > [+long] / __ [-cont] [+cons,+son,-nas] 
[+prim] > [+long] / __ [+lab,+ant,+voi] r [+syl] #
[+prim] > [+long] / # ([-syl])* __ [+cons] #
$ s198 checked for /e o a/ before enya and elya
[-hi,+tense,+syl] > [-long] / __ {ʎ;ɲ}

$ CORRECTION: vowels not longer before [-cont] l except for round vowels
[+syl,-round] > [-long] / __ [-cont] l

$CORRECTION: not long before ɲ 
[+syl] > [-long] / __ ɲ

$ CORRECTION : nr > ndr denasalization occurs AFTER length is determined,
	$  not blocking diphthongization behavior (minor>moindre, etc)
	$ (this is a reordering)
∅ > d̪ / n̪ __ r

$ CORRECTION: schwa inserted after stop + sonorant clusters if they are followed by nonsyllabic
∅ > ə / [-cont] [+son,-syl] __ [-syl]

$ CORRECTION: long a shortens when between a sonorant non-syllabic and final l
	$ cases: chenal, royal, loyal, mal
	$ not when not by a sonorant (ð is not sonorant) : sel, tel, hotel, Noel < Noðel, pieu etc
	$ also not kw -- must be continuant two phones before 
ˈaː > ˈa / [+cont] [-syl,+son] __ l #

$ s164-13 : diphthongization of tense e, o > ei, ou when tonic free(s225),
$ per Pope also tonic free a to ae "?" (s233)
$ and also s164-14 : "second bout of diphthongization of lax e tonic free
$ CORRECTION : if eː is already coming before j, that is absorbed. Same for ʝ 
	$ (this one is in BaseCLEFstar too)
{ˈeː j;ˈeː ʝ} > {ˈe j;ˈe j} 
{ˈaː;ˈeː;ˈoː;ˈɛː} > {ˈa e̯;ˈe j;ˈo w;ˈi e̯}

$s182 : finally remaining velar a (only in monosyllables) is fronted, late enough so that it counterfeeds breaking and diphthongization
[+lo] > [+front,-back]

$ s420 : tsj ejects a /j/ across a prior n if the vowel before is e 
$ CORRECTION: only countertonic e is affected.
∅ > j / ˌe __ n̪ t͡sʲ

$ s308: tsj ejects j into vowel groups that are not e before # or vowel then # (-itia and -itium suffixes). 
$ CORRECTION: no, only under certain conditions: 
	$ (1) prior is ˈie̯ (decimum > dime, pretium > prix
	$ CORRECTION (2) prior is ˈow (nukem > noix, crucem > croix) 
	$ CORRECTION : this must happen before degemination of ttsj (laetitia > liesse , not lisse; neptiam> niece, not nisse; noptiam > noce, not noice; pettiam > piece not pisse) 
	$ CORRECTION: moved this to right after palatization of aː and oː 
	$ CORRECTION: this also changes any tonic a-variant to aj and blocks any diphthongization to ae̯ (palais, paix, etc...) 
	$ noix/croix results are shown in Pope when demonstrating separate phenomena, but never explained. 
∅ > j / e̯ __ t͡sʲ
w > j / [+round,+syl] __ t͡sʲ

$ CORRECTION : we also delete e̯ before ʎ, ɲ (as well as the newly ejected j -- and also j from other sources 
	$ ( i.e. braie, rai, maie, orfraie, verai(e), baie, aine, lairme etc...) 
e̯ > ∅ / ˈa __ [+front,+son,-syl]

$ s343ii second effacement of beta between diphthongs ending in w and unstressed final u (or ʉ) 
$ as per Pope's example in s343 this includes Latin lupum -- so must come AFTER beginning of 2nd diphthongization
β > ∅ / w __ [+tense,+round,-stres]

$ s164-15 ~ s223 ~ s234: raising of lax countertonic ɛ,ɔ to tense e,o except in ɛr[+cons] ɛɫ[+cons] and ɔɫ[+cons] , see also s223, 121
$it is not as simple as Pope argues...
	$ CORRECTION 1 : shift is NOT blocked before geminate r (terrain, etc)
	$ CORRECTION 2 : ð then cons will block shift of ɛ
		$ must be non-nasal -- see semaine. 
	$ CORRECTION 3 : s (not z! -- temoin etc) then cons will block shift of ɛ usually...
		$ ... but when it's not without prior onset (and before s)
		$ (clearcut cases: prêcher, prêcheur, prêter, vêtir, vêtement) 
		$ debatable cases as we don't know which pronounciation /e/ or /ɛ/ is regular:
			$ essai, essaim, presser
	$ we group together all three of these as oral coronal continuants. 
	$ CORRECTION 4:  also does happen before interdental frics always if word initial
	$ CORRECTION 5: but, unless covered by a rule above, does not happen for C #
{ˌɛ;ˌɔ} > {ˌɨ;ˌɯ} / __ [+lat] [+cons]
ˌɛ > ˌe / __ r r
ˌɛ > ˌe / # __ [+cor,+cont]
ˌɛ > ˌɨ / __ [+cor,-nas,+cont] [+cons,-nas]
ˌɛ > ˌe / __ @ @

$CORRECTION: closing of ɔ (s164-15~s223~s234) is bled before voiceless coronals (prochain, fosse', côte', ...)
ˌɔ > ˌɯ / __ [+cor,-voi]
	$ this cannot be done later as it would overgeneralize -- see coutume, nous, vous, soupcon etc... 
$CORRECTION: closing of ɔ doesn't occur right before tonic o
$CORRECTION: blocked in ɔw dipth
ˌɔ > ˌɯ / __ ˈɔ
ˌɔ > ˌɯ / __ w

$(resolving 164-15 ~ s223 ~ s234
ˌɔ > ˌo
{ˌɨ;ˌɯ} > {ˌɛ;ˌɔ}

$ per Pope s481 : original aw ("au̯") becomes ɔ -- we do not differentiate "au̯" from a + w sequences
	$ but we could reliably capture it as being whenever aw precedes a consonantal. 
$ CORRECTION: moved to after closing of countertonic ɔ  (s164-15~s234)
	$ because original AU segments are not effected
$ CORRECTION: doesn't happen before DENTAL CONSONANTS
	$ this is far too early by anyone's account for the interdentals to have been lost,
		$ and cases like gauta > joue, aut > ou, nauda > noue, alauda > aloue etc show a clear pattern
w > ∅ / [+round,-tense] __ {[-syl,-cor];[+strid];[-syl,+son]}

$ feed nasalized u formation if necessary
[+syl,+back,+round] > [+nas,+hi,+tense] / __ [+nas,+cons]

$ s359 : β and ɣʷ brought into contact with other consonants vocalize to w (ex: fabrica > faurga > forge) 
$ CORRECTION: β does not do so if before r then a non-consonantal
ɣʷ > w / __ (ə) [+cons]
β > w / __ [-son] 
β > w / __ [+cons,+nas]
β > w / __ r (ə) [+cons]

$ s164.6 ~ s 189 
β > v 

$ CORRECTION: s258 moved much earlier. 

$ Later Gallo-Roman 
$ 8th and 9th centuries as per Pope s165

$ s427 tonic free a ... i.e. ae̯  ... becomes aj before n, m 
$ note : this does not seem to occur before ɲ -- with the exception of a single word châtaigne, and the possibly dialectally derived varaigne
$ also does not seem to occur before ŋ -- note sangwem > sɑ̃ (sang), not sɛ̃
e̯ > j / [+lo,+stres] __ [+nas,+cons,-hi]

$CORRECTION: s256 moved much earlier. 

$s355-357 open word finals again if unsupported 
[-delrel] > [+cont] / [-cons] __ # 

$ CORRECTION : ɸ > f (also in BaseCLEFstar)
ɸ > f

$s326 Later Gallo-Roman ks formed from vowel deletions > s, not to xs
k > ∅ / __ s #

$ s165-2 ~ s251 : reduction of unstressed a to ə final and intertonic
a > ə / [+syl] ([-syl])* __

$ CORRECTION: This is no longer necessary
$ s318 ~ s165-4 palatalization of final s by enya and elya preceding 
$s > t͡sʲ / {ɲ;ʎ} __ # 

$ CORRECTION : ɲ ɲ degeminated at this point 
ɲ > ∅ / __ ɲ 

$ s165.8 ~ s279iv resolution of prae-consontal enya to jn when praeconsonantal
$ CORRECTION: ejects nasalized j̃ not purely oral j
ɲ > j̃ n̪ / __ [+cons]

$ s165-3 ~ s206 : devoicing of all final consonants that have become final -- operates after s165-1.
$ she says "all consonants but doesn't list the sonorants -- so we say non-sonorants
[-son] > [-voi] / __ #

$ CORRECTION: Least overfit interpretation possible for how to get Pope's akwa>iaue>eaue thing to work with the data
$ First:  ˈa e̯ ɣ > ˈa j / __ w (case of aqua >>> aiue >>> eaue, aquas >>> aiues >>> eaues >>> eaux) 
$ (this also handles vera:cum > vrai (not vrec)
e̯ [+back,-round,-son] > j / ˈa __
$ Second: w, possibly with following schwa, is nullified before on an obstruent and after gamma
	$ (suis, suit)
{w ə;w} > {∅;∅} / ɣ __ [-son]
$ Third: 'i (probably via ɯ) becomes ʉ before e̯ɣC, under assimilative influence of ɣ (> ɯ, then to nearest phoneme at that time, ʉ) 
$ this one may have to be suppressed as although it is highly predictive, it could also be explained by analogy
	$ cases: suivre, suis, suie, suit...
ˈi > ˈʉ / __ e̯ ɣ [+cons] 
$ Fourth: remaining ɣw > jv (e'vier, ive, antive, antif, suivre, sive)
ɣ w > j v 
$ (rule for e'vier not e`vier deleted for overfitting) 

$ CORRECTION: Old length distinction lost at this point, to be replaced by new one where
	$ length arises from compensatory lengthening due to the deletion of praeconsonantal fricatives
$ TODO: may want to discuss placement of this (in absence of any explication by Pope)
[+syl] > [-long]

$ s231-3 : Pope is very unsure of the fate of a e̯ when it did not become aj before nasals, but what she settles on is this:
$ a e̯ >"e e̯ " (we interpret as long eː) except various places it was bled (after palatal cons, before palatal cons, before nasals, before /w/) 
$ syllable length rules of Gallo-Roman (open and stressed =long, else =short) must no longer be effective after this point. 
$ Buckley criticizes this paradigm from Pope.
$ CORRECTION: moved much later as it counterfeeds other phenomena like aqua > eau after eː becomes diphthongized and thus is no longer productive, 
	$ else a e̯ becomes oj > wa 
$ CORRECTION: but it needs to be before ae > aj / __ N
{a e̯;ˌa e̯;ˈa e̯} > {eː;ˌeː;ˈeː}
 
$ CORRECTION : deletion of w internal in remaining cons clusters (fits with general cluster rules)
w > ∅ / [+cons] __ [+cons] 
 
$ CORRECTION: s327 deleted from here as entire set of rules pertaining to 
	$ intervocalic ɣw, ðw etc reformulated
 
$ CORRECTION s328 unnecessary -- deleted
$  {ɛ;ˌɛ;ˈɛ;ˈɛː} > {i e̯;ˌi e̯;ˈi e̯;ˈi e̯} / __ ʝ w
 
$ CORRECTION: s329 deleted as entire set of rules handling  ɣʷ moved earlier.

$ s 192ii :  ɣʷ at beginning of word to ɡ if this is still around.. 
ɣʷ > ɡ / # __  

$s355-357 close if became supported
$ CORRECTION: this is moved later to come after the -xw- handling phenomena
[+cont,-son,-strid] > [-delrel] / [+cons] __ #

$ s357 -- loss of final x after unstressed sylls 
x > ∅ / [-prim] __ # 

$ s357 x palatalizes after tonic a, i. However, neither is raised. 
$ CORRECTION: x also palatalizes after ae̯, aj and ie̯ \
	$ -- so allow one front glide in between
x > ç / [+lo,+front,+prim] ([-cons,-syl,+front]) __ #
x > ç / [+hi,+front,+prim] ([-cons,-syl,+front]) __ #

$ s357, elsewhere, x closes back to k 
x > k / __ #

$ s357, s359, s324, s325: ç merges into ʝ
ç > ʝ

$ s372ii t,d become edh before r,l,n,m
$ CORRECTED -- Pope missed that this only happens after vowels (so restrict context)
$ CORRECTED -- moved later so that this can be blocked in former post velar enviroments (i.e. pictorina > poitrine not poirine) 
[-delrel,+cor] > ð / [-cons] __ [+son,+cor] 
[-delrel,+cor] > ð / [-cons] __ [+nas] 

$ s372iii p,b become beta then v before r, halts at b before l
$ CORRECTION: happens specifically after non-consonantals. 
p > b / [-cons] __ [+son,+cont,-nas]
b > β / [-cons] __ r
β > v

$ should be after resolution of palatal consonants. 
$ s404 : ʝ > j ("i-glide" for Pope) after vowels; j is simply absorbed by preceding i
$ CORRECTION: also absorbed by preceding j (incl nasal j) or some variant of /i/ (varying by stress/length/nasality)
ʝ > j / [-cons] __ 
j > ∅ / [+hi,+front,+tense] __

$ s165.5 leveling of au to o (s505)
$ note that the "original" a w have now become ɔ w -- this rule is handling when it was once β 
{a w;ˌa w;ˈa w} > {ɔː;ˌɔː;ˈɔː} 

$ s165.7 : another round of cluster modifications follows
$ CLUSTER REDUCTIONS pp 145-150 ish... 

$ 365ii early denasalization nns, rns . Utlimately becomes the alveolar affricate, see two rules down. 
n̪ > t̪ / [-cons] [+cor,-lat,+son] __ s

$ s365ii s t s > ts
s t̪ s > t͡s / [-cons] __ [-cons]
s > ∅ / __ t͡s

$ s367 d,t + s becomes the affricate; 
[+cor,-delrel] s > t͡s

$ s365 : effacement of middle consonant in groups that don't end in l or r 
[+cons] > ∅ / [+cons] __ [-son]
[+cons] > ∅ / [+cons] __ [+nas,+cons]

$ s374ii: fricative + w > w w 
[+cont,+cons,+voi,-son] > w / __ w 

$ recurrence of s374i 
w > ∅ / [-cont] [-cont] __

$ CORRECTION: delete w after stops that are not velar
w > ∅ / [-cont,-back] __

$ s377 s > z before voiced and/or "breathed" consonants (excluding s of course, but including f)
$ CORRECTION: only after nonconsonantals. 
s > z / [-cons] __ [+voi,+cons]
s > z / __ [+cont,+cons,-cor]

$ CORRECTION: s372ii, 372iii moved elsewhere from here

$ s365i : elimination of middle velars in three consonant intervocalic clusters
{s k l;r ɡ l} > {z l;r l} / [-cons] __ [-cons]

$ s368-370 denasalization and glide development
$ b inserted between m + r or l 
$ d inserted between n,l,dark l,z,elya + r 
$ t inserted between s and pal. s + r 
∅ > b / m __ [+cor,-nas,+son,-syl]
∅ > d̪ / [+lat] __ r
{z r;n̪ r;s r} > {z d̪ r;n̪ d̪ r;s t̪ r}

$s371 nasal cluster assimilation 
$ CORRECTION: m b > m m deleted as it is completely unnecessary and mostly causes errors
$ CORRECTION : only mn, nm doesn't occur and is inconsistent with what appears to be the actual regular pattern elsewhere
n̪ > m / m __

$ s373 assimilation of oral labials b,p,β (and v) to following dental positions 
$ as well as "d,t" to following labials or d, dzh, ts
$ RECURRENT as per s162
$ appears to resolve in EARLY OLD FRENCH, concurrent with degemination 
[+lab,-nas,-round,+cons] > t̪ / __ [-cont,+cor,-voi]
[+lab,-nas,-round,+cons] > d̪ / __ [-cont,+cor,+voi] 
[+lab,-nas,-round,+cons] > s / __ s
[+lab,-nas,-son] > ∅ / __ [+delrel] 
[+cor,-delrel] > ∅ / __ [+lab,-cont] 
[+cor,-delrel] > ∅ / __ [+cor,-cont] 
[+cor,-delrel] > ∅ / __ [+delrel] 

$ CORRECTION: dental fricative edh is also deleted before non sonorants
	$ not just t and d as Pope says. 
	$ (adfactare, pedicum, peditalia, radicinam, iudicem...)
	$ CORRECTION: does so before all non-continuants
	$ CORRECTION: if it does so after onset, single consonant, then countertonic e or eh,
		$ then the latter is replaced with 'ie 
[+stres,+front,-hi,-lo] > ˈi e̯ / # [+cons] __ ð [-cont]
ð > ∅ / __ [-cont]

$ CORRECTION: ɣ devoices next to voiceless segments (technically overfitting to amicus > amis, but consistent with syllable structure laws) 
ɣ > x / __ [-voi]

$ s366 degemination 
$ CORRECTION: extended to also hit rr when word final 
{p p;b b;f f;m m;t̪ t̪;t̪ t͡s;t̪ t͡sʲ;t̪ t͡ʃ;d̪ d̪;d̪ d͡ʒ;n̪ n̪;s s;z z;l l;k k;ɡ ɡ;n̪ n̪} > {p;b;f;m;t̪;t͡s;t͡sʲ;t͡ʃ;d̪;d͡ʒ;n̪;s;z;l;k;ɡ;n̪}
r r >  r / __ # 

$ CORRECTION: s378 moved earlier. 

$s378 z > ð before voiced dentals 
$ CORRECTION : deleted -- not diachronically useful
$z > ð / __ [+ant,+cor,+son]

$ s258 ə inserted after final CONS + r,l clusters 
$ CORRECTION where jr is concerned only a and ie̯ before the j are contexts where emission is triggered. 
	$ it does not happen after oj (dortoir, heur, etc) or ʉo̯ (cuir) or ej (avoir, chaloir, voir, hoir, savoir, noir etc...)  
$ in the case of the palatal j r clusters, Pope shows the effects in citation forms when documenting other shifts
		$ but never specifically explains this specific phenomenon 
$CORRECTION: this only happens for actual consonants followed by l, not glides.
∅ > ə / [-son] r __ #
∅ > ə / e̯ j r __ #
∅ > ə / [+lo] j r __ #
∅ > ə / [-lat,+cons] l __ #

$|~Early Old French 
$ as per Pope p9 s16, this spans from the 9th century to the end of the 11th. "Formative period of the French language.”
$ however it is NOT the stage set we use for FLLAPs, in which "Old French I" comes a bit later. 

$ Pope s257 : unstressed ə > e before yod. She implies this is also happening before ʎ and ɲ in her examples
$ s252 : this feeds ej > oj phenomena.
ə > e / __ [+front,+son,+hi,-syl]

$ s329 and s484 countertonic e becomes o before w and tonic u -- doesn't state when explicitly other than "in Early Old French" 
$she doesn't seem to mention this being effected by former o,
	$ so we assume it comes before s166.2 ~ s184 
ˌe w > ˌo
ˌe > ˌo / __ [+hi,+round,+prim] 

$ s422 e "intertonic" (i.e. stress-less) moves up to i before palatal consonants (ɲ,ʎ)
$ CORRECTION : this is moved earlier, originally in LOF
e > i / __ [+hi,+son,+front,+cons]

$s166.9 ~ s383-391 -- "beginnings" of l-vocalization praeconsonantal. 
$As per Pope s167, "unclear" how much and if the vocalization aspect was present during this period.
$First attested in 9th century -- s390. 
$semi-correction: for simplicity sake, let's not deal wiht lambda- gamma, and just use /ɫ/
[+lat] > ɫ / __ [+cons]

$s382iv and s384 : but back l becomes ʎ after i (originally Pope has this as lʲ, but Occam's razor as
	$ diachronically there is no difference -- Pope is inconsistent, using ʎ for fils but lʲ for ficelle)
ɫ > ʎ / [+hi,+front,-round] __

$ s166.1 ~ s183 ~ s9 ~ s36 ~ s42 completion of u fronting. 
$"Spread slowly from the South" (s183, s36), continuing in the tenth century (s42), with some Eastern and Northern (esp Walloon, Picard) dialects remaining unaffected.
$ precedes u>o (s166.2~s184) as per s183. 
[+hi,+round,-back] > [+front]

$ s166.6 ~ s434 ~ s441.1,441.3: first nasalization, effects a and aj, in the 10th century
j > [+nas] / [+lo] __ [+nas,-syl]
[+lo] > [+nas] / __ [+nas,-syl]

$ CORRECTION: countertonic ˌe to ˌə can bleed second nasalization 
	$ -- moved earlier to before it. 
$ s166.5 ~ ~s223 ~ s234-235 reduction of ˌe > schwa when free. 
$ Due to increasing weight of stressed syllables. 
$ Attributed in the late 11th century via a letter of Anne of Russia, see s235
ˌe > ˌə / __ ([+cons]) [+syl]

$s665 (stated in sound table but nowhere else): ˌə reverts to ˌe 
	$ if before palatal consonants (ʎ,ɲ...) 
ˌə > ˌe / __ [+hi,+front,+cons]

$ CORRECTION: reversals include ə > ˌɛ before lə (pelerin, celle, celles... bleeds following shift (s166.5~s223 ~s234-5)
	$ s684 -- Pope definitely does not accomodate this, and has an explicitly different view of the development of pelerin 
		$ which she cites as pˌələri:n in Old French and only getting /ˌɛ/ in Middle French 
			$ -- yet she never explains anywhere this purported ˌə > ˌɛ shift. 
		$ CORRECTION: expand post-post-context to all non-round syllabics
ˌə > ˌɛ / __ l [+syl,-round]

$ CORRECTION: ˌə > ˌe before r then anything but a : peril, derriere, Jerome, paresse
ˌə > ˌe / __ r [-lo]

$ CORRECTION: initial countertonic schwa to e 
	$ ... if before any consonant except a dental
		$ due to the rules of how intervocalic clusters developed
			$ this can be summarized with [+strid]
	$ cases: eveque(s), essai(s), essaim(s), etc.
ˌə > ˌe / # __ [+strid]

$ NOTE: other cases not handled here-
	$ pedo:nem > pion -- form of "pëon" attested in 1380 as per TLFi
	$ denara:ta > denree -- should be handled by earlier schwa DELETION rules
	$ desordre, prevot, plenier(e), trem(u)i(e) -- too isolated. 

$ CORRECTION : ə from previous period (restoration after schwa falls) becomes e 
	$ if in closed nasal syllable. (parente', jouvencelle, etc.) 
$ CORRECTION : but only if it is not the final syllable 
	$ (hence all the verb -ent endings remain non-nasalized schwas)
ə > e / __ [+nas] [-nas,+cons] ([-syl])* [+syl]

$ CORRECTION: rounding of ˈe to ˈo before j is *early* when preceded by a voiced labial consonant
ˈe > ˈo / [+lab,+voi] __ j

$s405-406 ɲ final > j ɲ when after a e (not ɛ or ie) u (<o) or y.
$ called "Early Old French" by Pope (s406)
∅ > j̃ / [+tense,+syl] __ ɲ # 

$s407i : glide develops between e tonic or countertonic and ɲ intervocalic  
$ CORRECTION: /ɛ/ as prior context also sees this happen
$ that it is here can be evidenced by Pope's tables. 
∅ > j̃ / [-hi,-lo,-round,+stres] __ ɲ [+syl]

$ s166.6 ~ s434, s439 ~s441.2,441.4: second nasalation, effects syllabic e variants and ej, in the 11th century. Bleeds later ei > oi shift.
$ should NOT effect ie̯ , which was nasalized in the third nasalization, in the 13th century. (s441.5 etc -- and then lowered later).
j > [+nas] / [+front,-hi,+syl] __ [+nas,-syl]
[+front,-hi,+syl] > [+nas] / __ [+nas,-syl]

$ s166.3 ~ s226-7 : uo > ue; see also ss550-554 for later development. 
$As per s226, this happened "before the end of the [EOF] period". 
$Attested in 1084 in Domesday book 1081
$The mutations of ei and ou seem to have occurred later.
o̯ > e̯ 

$ s411: ˈue̯j >"yi̯j" > yj and ˈie̯j >(ii̯j >) i
$ see also s410~s514-5
$ CORRECTION: also includes j̃
ˈu > ˈy / __ e̯ j
e̯ > ∅ / [+hi,+front,+tense,+prim] __ [-syl,-cons,+hi,-round]
[-syl,-cons,+hi,-round] > ∅ / ˈi __ 

$ s330 : early leveling of aj before w -- case of aqua -- 11th century. 
ˈa j > ˈɛː / __ w 

$s489 : ou >eu bled in Early Old French by ou >u before labials. 
$Date not exactly known. 
$ CORRECTION: this doesn't happen when ow is also preceded by a voiced labial consonant though (meuble)
	$ CORRECTION: but DOES separately happen after p. 
[+tense,+syl,+round,-hi,-lo] w > [+hi] ∅ / [-lab,+cons] __ [+lab,-syl]
[+tense,+syl,+round,-hi,-lo] w > [+hi] ∅ / p __

$s490: countertonic o > ə under influence of (following?) tonic u>y or o>u  
$ note, called a "tendency" by Pope.
$ CORRECTED -- seems to happen immediately before (u>y) [speaking in terms of context, not date]
	$ and for u>y and o>u seems to happen before alveopalatal affricates, and nasals.
ˌo >ˌə / __ [+prim,+hi,+round,-back]
ˌo >ˌə / __ [-syl,+nas,+cons] [+tense,+prim,+round]
ˌo >ˌə / __ [+delrel,-ant] [+tense,+prim,+round]

$ CORRECTION: ˌo opens before r t or ts , throw in d and dz for good measure, then a syllabic (not tourtre < torturella) 
ˌo > ˌɔ / __ r [+ant,-lab,-cont] [+syl]

$s435: dentalization of final m after vowels. 
$s435 -- accepted by late 11th century 
m > n̪ / [-cons] __ #

$CORRECTION: ŋ > n when before a non-velar 
$ Pope misses this although her use of "luns" for OF for longs in the table implies it
$ attestation through TLFI:
$ LONGS~LONS: lonɡoːs > GR lũŋɡəs > lũns as "lons" attested in late 1100s
$ SANGS~SANS: sɑnɡʷeːs > GR saŋɡəs > sãns -- "sans" attested in 1176 
$ BANCS~BANS: bɑnkoːs > GR baŋkəs > bãns -- "bans" attested in 1000s in Alexis
$ BLANCS~BLANS: blans attested in 1174 Chretien de Troyes and again throughout 1210-1260...
ŋ > n̪ / __ [-back]

~Old French I

$ CORRECTION: later shifts on countertonic ˌo and tonic ˈo going to u 
	$ ... are BLED by effects of ð : 
	$ except when the next vowel is tonic and tense, ˌo and ˈo go to ˌɔ before ð 
		$ i.e. affected: borne, boyau, boyaux, prone, Rhone; 
			$ unaffected: nouer, noueux, nourrice, nourrisson, vouer
{ˌo;ˈo} > {ˌɔ;ˈɔ} / __ ð ([-syl])* {#;[-stres];[+syl,-tense]} 

$ CORRECTION: final l WILL be labialized if it is preceded by a mid vowel that is short: 
	$ ɛl# > ɛa̯w# : agneau, anneau, arceau, oiseau, beau, boyau, faisceau, fleau, jouvenceau, chameau, chapeau, chateau, cerveau, niveau, manteau, marteau, nouveau, panneau, pinceau, poteau, rameau,
		$ rateau, ruisseau, sceau, seau, e'tourneau, vanneau, vaisseau, veau, peau
	$ el# > ew# > øw#: cheveu, ce
	$ ɔl# > ɔw# :  cou, fou, mou
	$ NOT al (mal, chenal, loyal, cheval, val...) 
	$ NOT e:l (quel, Noel, sel, tel...) -- single exception of pieu < palum
	$ NOT iel :  ciel
	$ uel is inconsistent : deuil, moyeu, linceul/filleul each show diff behavior
	$ ol is inconsistent: coule but saoul
	$ not owl: aieul, seul
	$ NOT il: mille, il, vil, avril, fil -- one exception of fusil 
	$ yl inconsitent: cul but nul
	$ NOT awl: Paul
	$ NOT ajl: frele, grele
l > ɫ / [+syl,-hi,-lo,-long] __ #

$ Pope s372 : ð occasionally assimilated to r but when it did and didnˌt was
	$undetermined at the time of writing (assimilated: butina >borne but Rodanum >Rhone, etc...)
	$helps show how Pope can be improved upon using DiaSym.
	$note the z >ð >r shift may have actually been a Picard characteristic (s378i)
	$ includes also ð > z before m : rhythme > rîme , consuetudinem > coutûme, semaine, dimanche 
$ CORRECTION: moved from Later Old French to here so it is before effacement of z 
$ CORRECTION: ð goes to z before n specifically when r is on otherside of syllable, and is followed by n 
$ CORRECTION: ð is simply deleted without going to z if between non-primary stressed vowel and m
$ (borne one looks like overfitting -- but is unchanged from the baseline.)
ð > ∅ / [-prim] __ m
ð > z / __ m
ð > z / r [+syl] __ n̪
ð > r / __ n̪

$ CORRECTION: various laxing effects moved to here (earlier) 
$ all should feed s541
$ CORRECTION: only r, not l, for s492 context
$ CORRECTION: broaden to include ˌe for rC postcontext (s492-3)
[+stres,-hi,-lo,-round] > [-tense,+front] / __ r [+cons] $Pope s492-3
ˈe > ˈɛ / __ [+cons] [+syl] $Pope s575
ˈe > ˈɛ / __ [-son] [+son,+cor,-nas] [+syl]
$ CORRECTION: doesn't occur before non-lateral coronal cons then #, except for s when not after a palatla
$ note that this covers ʎ# codas, which Pope asserts were only effected much later
ˈe > ˈɛ / __ [+cons,-cor] #
ˈe > ˈɛ / __ s #

$ CORRECTION: previous distinctive length (in ew>øː, ae̯>eː and aj>ɛː) is leveled
[+syl] > [-long]

$ s166.10 effacement of preconsonantal z, accompanied by compensatory lengthening of prior vowel. 
$ Complete before the Norman Conquest of England (1066) as per s377
$ CORRECTION: lengthening doesn't happen before d -- it seems z is simply absorbed
$ CORRECTION: does not occur word-initially
$ CORRECTION: for countertonics, lengthening only happens when z is after only ONE consonant (not two)
z > ∅ / __ d̪
[+prim] > [+long] / @ __ z [+cons]
[+stres] > [+long] / @ __ z [+cons] [+syl]
z > ∅ / __ [+cons]

$ s166.4 ~ s226-7 :  ei > oi, ou > eu. As per s226, occurred "before the middle of the twelfth century".
$ She places it in EOF-- so we place it there too, though clearly at the end. 
$ Any e before yod effected.
$ per s226 : effects ei whether tonic, intertonic or countertonic, but only tonic ou 
ˈo > ˈe / __ w 
[+front,+tense,-hi,-lo,+syl,-long] > [+back,-front,+round,+lab] / __ j 

$ CORRECTION : s291-s292, tsj depalatalization, is DELETED. 
	$Already had to be moved after s226-7 ou > eu, because there is different behavior for palatal t͡sʲ , as opposed to simple t͡s
	$ see krukem, nukem > croix, noix, whereas prodis > prowts > preux. 
 $ other unsalvageable issues included words like suctiare > sucer not sucier, percer not percier, blesser not blessier, chasser not chassier etc
 $ as for the two cases of aciarium and sortiarium -- this are much more likely cases for analogy than the verbs, as it is a very restricted but recognizable noun class
 	$ that is much more "learned" and knonw to correspond to Latin -iarium in both cases. 
$ NOTE: needs to be reinstated if doing demonstration with traiter
$ t͡sʲ > t͡s / {[-nas];#} __

$ CORRECTION: ˌo and ˈo open before pre-coda t or k 
{ˌo;ˈo} > {ˌɔ;ˈɔ} / __ [-cont,-lab,-delrel] #

$ CORRECTION : ˌo opens before r,l then stressed a or ɛ or ə  (volage, foret, soleil)
$ CORRECTION : first one is counterfed by ae > eː -- moved down past s231-3
ˌo > ˌɔ / __ [+cons,+cor,+son] {ˈa;ˈɛ} 

$ s520: early oj(uj?),ɔj >oɛ̯,ɔɛ̯ before r 
j > ɛ̯ / [+syl,+back,+round] __ r 

$ CORRECTION: w [+lat] ejects an /ə/ at the end of the word -and also after velar l then normal l, etc...
∅ > ə / [+back,-syl] [+lat] __ #
$ CORRECTION : for moule < musculum and coule < cucullum : long u in monosyllable or only other syllable being u
	$ also causes this ejection
∅ > ə / # ([+cons])* ([+tense,+round,+back])* [+lat] __ #

$ s674 : countertonic ɔw > uw bled before z and v where it goes to ɔ,
	$ as well as when blocked by one cons then free after.
$ (Pope doesn't realize that at this point ð may not have been deleted yet...
	$ she gives laudare as an example of "in hiatus" 
	$ (but she notes herself the loss of θ and ð was not finalized 
		$"until the middle of the twelfth century") 
$ s673 : this also happens for tonic ɔw
$ Pope only calls this for ɔw, not ow, but she had shifted countertonic ɔ to o earlier... 
$ CORRECTION : s also causes automatic simplification (Autun) like v and z do. 
	$With s,v, and z we can just say +strid will cover it (no examples either way of f, tʃ, dʒ, ts or dz)
$ CORRECTION : correct the "blocked and free" second realization of this shift 
	$ to be before any coronal sonorant, further context doesn't matter
$ CORRECTION : this does not happen word finally 
$ CORRECTION : also targets countertonic ˌow
{ˌɔ w;ˌo w;ˈɔ w} > {ˌɔ;ˌɔ;ˈɔ} / __ [+strid] @
{ˌɔ w;ˌo w;ˈɔ w} > {ˌɔ;ˌɔ;ˈɔ} / __ [+cor,+son]

$CORRECTION: tonic o opens ˈo > ˈɔ in context of r [-cont,+voi] r
	$ (may be analogy...)
ˈo > ˈɔ / __ r [-cont,+voi] r

$ s166.2 ~ s184 o > u. As per s184, occurs in the eleventh and twelth centuries.
$ s518: this also means oi > ui
$ except for early shift before labial consonants (see s489), this is bled by 166.4. 
$Pope places it in EOF so we place it there too-- but clearly at the end of the period.
[+round,+tense,+back] > [+hi] 

$s407i : glide develops between e tonic or countertonic 
	$  and ʎ intervocalic after e TONIC
$ CORRECTION: /ɛ/ to also be included in prior contex, and 
	$ include e or ɛ in prior context whether tonic or countertonic
∅ > j / [-hi,-lo,-round,+stres] __ ʎ [+syl]
∅ > j̃ / [-hi,-lo,-round,+stres] __ ɲ [+syl]

$s407ii : same for y countertonic or tonic 
∅ > j / [+hi,+front,+round,+stres] __ ʎ [+syl]
∅ > j̃ / [+hi,+front,+round,+stres] __ ɲ [+syl]

$ CORRECTION: tonic iej and iej̃ move to (long) iː after a palatal
$ SUPPRESSED: no longer seems to have an effect...? 
$ ˈi e̯ [-cons,-syl,+hi,-round] > ˈiː / [+cons,+hi,+front] __ 

$s166.6~s439 ; beginning of first lowering of nasalized e-varieties -- late eleventh century.
[+front,-hi,-lo,+syl,+nas] > [-tense]

$~Later Old French
$AKA Old French II. As per Pope s16, from end of eleventh century to beginning of the 14th. 
$ However we are just using this to mark the time boundary -- this is not where OFII is for Pope's periodic table
	$ which we used for FLLAPS. 

$ s555 ue̯w > ye̯w : Pope seems to imply this happens in the late 11th century
$ CORRECTION: suppressed, redundant. 
$[+hi,+round] > [+front,-back] / __ e̯ w 

$ s556 ye̯w > yø̯w 
	$ s556 "in the course of the eleventh and twelfth centuries the middle element was rounded to ø and stressed" 
	$ BASECLEF* CORRECTION: allow this to also happen before ɫ, which is not fully vocalized at this point yet. 
e̯ > ø̯ / [+hi,+round] __ [+son,+back]

$s171.3 : voicing of intervocalic stridents, recurrence of s334
$She says it occurs for s and f both, and specifically says it occurs in Later Old French
	$ in s171.3 : "... in intervocalic were voiced, s beoming z and f, v..." 
$ CORRECTION: s > z here demonstrably only causes problems 
$ CORRECTION: f > v has no effect here deleted  currently this rule has no effect for f > v
$s > z / [-cons] __ [-cons]
$f > v / [-cons] __ [-cons]

$Pope s441.2~s447-8 : ɛ̃>ã in the 11th to 12th centuries.
$ BASECLEF* CORRECTION: doesn't happen before j̃
[+nas,-hi,+syl] > [+lo,+tense] / __ {#;[+cons]}

$ CORRECTION: word-initial /k/ voices when before r,l and then followed by a (regardless of stress)
	$ Pope does notice this but she (likely incorrectly) attributes this to a "tendency" in Gallo-Roman 
		$ and only conditions it on the following l or r
	$ However if we condition it on a following a AFTER aw becomes round, and open a > ae̯ > eː 
	$ then we hit every occurrence except irregular caveola > geole
		$ -- i.e. craticula > gr(a>e>null)ille,  crassum > gras, crassia > graisse, classum > glas. 
	$ by placing it here, we also do not accidentally hit as that become o later, and do not trigger the shift:
		$ i.e. clavem > clef, clavum > clou
	$ ... or those that become ae̯ : clarum > cler, etc. 
k > ɡ / # __ [+cor,+son] [+lo]

$ Pope s500 -- ir > ie̯r when stressed and before consonants. 
$ feeds ie̯ >je shift in 12th century, must precede it.
∅ > e̯ / [+front,+hi,-round,+prim] __ r [+cons]

$ CORRECTION: efface ʎ between i variants and consonant
ʎ > ∅ / [+front,+hi,-round,+syl] __ [+cons]

$s385~s390 vocalization of dark l variants 
$ per Pope s390: attested in rhymes by middle of 12th century: 
[+lat,+back] > w

$ Pope s550 - 551 : ˈue̯ is monophthongized to ø
$ Attested in 1137 -- "over the course of the 12th century"
$ "Stages not fully determined but appear to be ˈue̯ > ˈuø̯ > w'ø > ˈø"
$ BASECLEF CORRECTION: Pope notates this as ue̯ undergoing this shift, but the u should have become fronted to y by this point...
	$ -- it is actually ye̯ -- make notation sufficiently broad
$ She also explicitly limits it to hte stressed diphthong 
[+hi,+round,+tense,+prim] e̯ > ˈø

$ BASECLEF CORRECTION: w after u or y is immediately deleted
w > ∅ / [+syl,+hi,+round] __

$ s382iv ~ s384 meanwhile lʲ which has been limited to praeconsonantal position after i-variants is effaced
lʲ > ∅

$s330 ~ s388 ~ s538-540 : a glide develops between ɛ and w
$ can occur for either stressed or secondarily stressed ɛ (bellitatem >beaute -- a case of secondary stressed ɛ for this)
∅ > a̯ / [-hi,-lo,+front,-tense,+stres] __ w
[+syl] > [-long] / __ a̯ w

$s388 ~ s544 : glide of /e/ after eː (from Gallo Roman a in open sylls)
$ note this can only come from primarily stressed a (>ˈeː) 
$ no effect for this as of right now so suppressed -- I believe it is for certain cases where word final l vocalized, and things ending in former aːlVs
	$ likely that in case of pa:lo:s the a stayed long even after the o was deleted for some reason -- but unclear why this would be
		$ another relevant target of s388 would be taːleːs > tieus, but this is not in our dataset at the moment as it was not in Pope's index
ˈeː > ˈe e̯ / __ ɫ

$ Pope s466-467 : ãj > ɛ̃j, everywhere, twelfth century.
$ CORRECTION: moved earlier, formerly was right after 528-9
[+lo,+nas] > [-lo,-tense] / __ j̃

$ s388 ~ s544: ee̯w > ie̯w -- must be in the 12th century
$ CORRECTION: no targets, suppressed (only fed by s388~s544 from earlier e:>e schwa / __ w?)
ˈe > ˈi / __ e̯

$ CORRECTION: s172a (th-effacement) moved much later so that it can condition some coming phenomena 

$ CORRECTION: s422 moved much ealrier. 

$ CORRECTION ˌe > ˌɛ, blocking reduction to schwa, 
	$ in open syllables before voiced continuants that are not lateral or nasal
	$ (r ,z ,v, ð) -- in many of these cases it will almost immediately return to ˌe.
	$ CORRECTION: just r 
ˌe > ˌɛ / __ r [+syl]

$ CORRECTION : t͡s goes to s before consonants that aren't coronal -- and feeds the next rule (case of decimum > dime) 
t͡s > s / __ [-cor,+cons]

$ CORRECTION recurrence of countertonic e to schwa reduction 
	$ moved down four or so rules, so that it comes after
	$ s544 (iew > jew etc etc ) 

$ Pope s378ii : attested by the Orthographica Gallica, written in 1300
$ Pope noted that the final s-effacement that was engendered by this was first reported 
	$ in the late 12th century in the Southwest -- so we place this before the late 12th
$ Pope argues s > x before dentals specifically. 
$ This bleeds normal s-effacement as per Pope -- debatable. 
$ Walloon wrote this as 'h' (Pope s378)
$ CORRECTION: s > x before ALL consonants
$ CORRECTION: moved here from later. 
$ CORRECTION: include also s^j as source
[-distr,+strid,-voi] > x / [-cons] __ [+cons]

$ CORRECTION : unrounding and lowering of u in ˈuj before ə then # in late 12th century
$ Pope argues that oi > ai shifting only occured under Italian influence in Middle French
$ However, there is textual evidence of it beginning in this context much earlier, and no exceptions
$ # Example: rika (<Gaulish) appeared became roie under the rounding ei by 1140, but in ** 1170 we see it attested as raie
	$ and the formation of rayure (raie + ure)  www.cnrtl.fr/definition/raie
$ It is likely the sound used was a rare allophonic sound mostly absent elsewhere in the language
	$ which was neither round enough to be o 
	$ nor low or front enough to be a 
	$ therefore ʌ makes a good guess.
$ the two exceptions to this -- via and korrigia -- are handled by an later exception rule 
$ CORRECTION: unrounding and lowering of u in uj before j d ə , j ð ə and j ð r ə in late 12th century
$ Pope argues that oi > ai shifting only occurred under Italian influence, and specifically centuries later in Middle French
$ She also argues that ð was lost during the 12th century, not hte end of it. 
$ However, there is textual evidence pointing toward some sort of oj-unroudning occurring where ð was about to leave a hiatus
	$ but NOT where θ would create the necessary context 
$ witrum: voirre in 1174-1187, but becoming "verre" and "warre" in 1200
		$ cre:ta : attested as croie in 1175, but then craie in the root for craiere in 1290
		$ cle:ta : attested as claie in 1306
		$ mone:tam : monoie in 1170 then monnoie in 1245, but attested as monae 1296
$ There are other interpretations of this that are tempting, but do not seem to work 
	$One is that ð and θ disappeared early (as Pope argues), and fed the previous added shift where oj went to ʌj before ə then coda #
$		 However there is an issue with this: Pope and all the relevant literature argue that ð and θ disappeared at the same time
		$ and where it is θ disappearing and
			$n not ð to create the environment of ojə#, the shift does not occur
		$ example : ficatum > fojəθ > fwa (foie), not *fɛ (*faie)
		$ example 2 : plicat > plojəθ > plwa (ploie), not *plɛ (*plaie) 
			$ this also would not explain why we see the same pattern in oide words (rigidam > roide >raide) 
				$ except those preceded by labial elements which rerounded ʌ (via > voie, frigida > froide, lampreda > lamproie)
	$Another is that ð and θ disappeared early and there was a global shift where this would have occurred
		$ this works for the ojðə# segments, but not the ojðrə# segments (voirre > verre, attested in 1200; tonitrum > tonerre, etc...)
		$ It appears unrounding before rə only occurs where ð once was, meaning ð really may have survived longer than we thought
	$one way to explain this is that the shift originally manifested in a conservative lower class dialect 
		$ that did not make it into the written form into a bit after it happened 
		$ i.e. it occurred in the late 12th century,
$			early enough to be conditioned on ð just as it was about to vanish when we first see
$			roie become raie and voirre became vairre [1200], but did not become widespread in the writing until around 1300 
				$ when we begin to see craie 1290, monae 1296, claie 1306, etc.)
			$ this also would not explain why there is similar behavior before d (rigidam > roide > raide) 
		$ all exceptions except se:tam can be explained by a rule by which teh former uj is restored under the influence of preceding
			$ labial elements 
$ Like above, it is likely this sound, which saw delayed entrance into the written language except for a few words (raie, verre)
	$ was neither really an a or an o for about a century-- hence ʌ is a good guess, with Canadian raising being a good analog
$NOTE: in later review with FranTEXT, this looks very dubious -- cited in paper as case where PATCH with DiaSim may have led us astray.
ˈu > ˈʌ / __ j [+ant,+voi,+distr] (r) ə 

$ CORRECTION: rerounding of ʌ when after a labial (possibly separated by r) #
	$ cases: via, korrigia, froide
ˈʌ > ˈu / [+lab] (r)* __

$ s544 : ie̯w >jew as part of stress shift. Occurred during 12th century (s538 -- later 12th century) 
$per s509-510 ~ s470-472 (see also s512-513) this can be input as part of a larger shift,
	$ dated to the same time, that moved ie̯ to je everywhere it occurred. 
$ per s470 -- predates a second nasalization of e, in the 13th century. 
	$This second nasalization then saw ẽ >ɛ̃ lowering, but it wasnˌt accepted till the 17th century
$ Fed by s388. Feeds e>ø / __ w shift. 
$ Seems to only occur for stressed ˈi.
$ also s538 : ɛa̯w > e̯aw for both stressed and secondarily stressed ɛ and 
	$per s538 "Later 12th century"
	$Pope doesn't give relative dating for the closure (ɛ>e) and stress shift so we group
		$them together. 
	$Pope notes -- s508 ~ s472-- the leveling of diphthongs started in the south and west. 
	$ / __ w context isn't necessary since ɛa̯ only occurs specifically in that context...
$include also here s555-556 : yø̯ > ɥø
$ CORRECTION: some of the final ˌɛ transformations moved to after this (like reduction to schwa) 
	$ as it seems to block at least one (derriere)
{ˈi e̯;ˌɛ a̯;ˈɛ a̯;ˈy ø̯} > {j ˈe;e̯ ˌa;e̯ ˈa;ɥ ˈø} 

$ CORRECTION: Pope does note a few of the effects of this shift in s496 but she attributes the main effect of the shift to 
	$ the Middle French period, and holds that hte effects in Early Old French were sporadic
$ However, she missed that there is a consistent context here : exclusively countertonic ˌɛ
	$  is regularly shifting after bilabial (NOT labiodental) consonants in closed syllables followed by voiceless p or tʃ -- both voiceless and slightly rounded
$ CORRECTION: moved earlier 
$ CORRECTION: countertonic ˌɛ goes to ˌa 
	$... when after a labial consonant and before r in a closed syllable
ˌɛ > ˌa / [+lab,-strid,+cons] __ r {#;p;t͡ʃ}

$ CORRECTION: before ð is effaced, countertonic ˌu to ˌə if followed by a tonic a (rable)
$ PRUNED: lexically specific, would lead to overfitting.
$ ˌu > ˌə / __ ð [+lo,+prim]

$ s172a ~s346-7~s372 : th-effacement  – finalized in middle of twelth century. 
	$began dialectally in North and East in 9th century. 
	$ CORRECTION: moved to during changes to countertonic e, because it appears that some of these are being driven by ð, 
$ CORRECTION: moved to late twelth century because of data Pope missed. There are attestations of the beginning of a oj > aj 
	$ that seems clearly conditioned on a following ð (cle:tam,cre:tam,tonitrum,vitrum,mone:tam; exceptions are lampre:dam and se:tam) 
	$ the first attestations of this appear in 1200 -- not in Middle French as Pope argues. See shift inserted above.
{θ;ð} > {∅;∅}		

$Pope s541-545 :  12th century -- ew >øw . Is fed by ie̯>je (12th cent); feeds øw>ø (late 12th cent)
[-hi,-lo,+tense,+syl] > [+round,+lab] / __ w

$ s421 countertonic a brought into contact with tonic y raises to e ... (eventually to ə...) 
$ happens after the loss of the interdental frics. 
$ CORRECTION: ˌa is similarly reduced before ˈø -- pavorem becoming pëor by 1176 definition (not 1119 : "pour") 
	$ i.e. ^meaning we broaden posterior context 
$ CORRECTION: moved to after ew > øw
ˌa > ˌe / __ [+front,+round]

$ Recurrence of s166.5 ~ ~s223 ~ s234-235 reduction of ˌe > schwa when free. See s421. 
$ CORRECTION -- only sonorant consonants can come between this and next syllabic next time	
	$ -- hence must be "over open" syllable. 
$ CORRECTION: moved to after ew > øw
ˌe > ˌə / __ ([+cons,+son]) [+syl] 

$ CORRECTION : degemination of rr happens many centuries before Pope says it is
$ this is surprising because Pope has much backing in historical texts describing the language -- however after making this change
$ accuracy improved by a whole point and it appears that a lot of her French historical text-based evidence
$ led her in the wrong direction -- such as also asserting that any a before rr would become long (and thus /ɑ/ ultimately -- but this did not
	$ happen in every single one of the available cases: barram > /bar/, karru:kam > ʃary, karrum > ʃar ...) 
r r > r 

$ CORRECTION: ˌɛ to ˌe before r,z,v then vowel
ˌɛ > ˌe / __ [+cont,-lat,-nas,+voi] [+syl]

$ CORRECTION: case of terre:num > terrain -- ˌe is persistently reduced to ə 
	$ if before continuant then tonic e or ẽ or ɛ or ɛ̃
$ PRUNED: likely overfit. 
$ ˌe > ˌə / __ [+cont] [+prim,+front,-hi,-lo]

$ CORRECTION : ʌ lowers and fronts to a, completing early oi>ai shift attested before middle French for a number of words
	$ cle:tam, cre:tam, mone:tam, the:ca, rica, vitrum, tonitrum > claie, craie,  e, taie, raie, verre, tonnerre 
$ Place this before Pope s528-529 because of cases of verre and tonnerre (aj > ɛj > ɛ praeconsonantally)
ˈʌ > ˈa

$Pope s528-529: aj > ɛj when praeconsonantal -- aj monopthongization
$Pope does not give a specific date. Makes sense to be here, given spellings ˌai" for wɛ >ɛ later...
$ note post facto-- Rheinfelder 1976 (:s272- 279, pp 105-108) also attributes this to 12th-13th cents.  
$ CORRECTION: include also cases of ˌej 
[+syl,-round,-hi] > [-lo,-tense] / __ j [+cons] 

$Pope s528-529 : aj > ɛj word finally
[+lo] > [-lo,-tense] / __ j #

$Pope s541~545: øw >ø in later 12th century.
w > ∅ / [+round,+front,-hi,+syl] __

$ (Pope s541-5 cont'd) yj >ɥi early when it will be deleted after a velar (s503 ~ s514 ~ s516) 
$ CORRECTION : this only occurs word finally. 
$ currently suppressed as has no targets anymore. 
$ {y j;ˌy j;ˈy j} > {ɥ i;ɥ ˌi;ɥ ˈi} / [+cons,+back] __ #

$ s166.8 ~ s192 : deletion of w in kw,gw. 
$ ɥ also deleted after k,g at this time (s516) 
$ As per s192, late 12th century.
[-syl,+round] > ∅ / [+back,-cont] __ 

$ s556  ɥø > jø is bled after labials and velars. Velars have already been covered by the rule above.
$ CORRECTION: moved from between third nasalization and s670
ɥ > ∅ / [+lab,+cons] __ [+front,+round]

$ s547-548 : ɔw > u in the thirteenth century. 
$ s548: late 12th and 13th centuries. 
[-hi,-lo,+back,+round] w > [+hi,+tense] ∅ 

$THIRD NASALIZATION : late 12th and early 13th 
$ must be after the original ɛ̃ > ã
$ all old targets effected, in addition to: 
$ je : s434 ~ s441.5 ~ s470
$ u : s434 ~ s441.7 ~ s459 ~s464  (s464: started early 12th century)
$ ø : because we know it comes before i and y, convenient to incldue here [s434 ~ 441.11 ~ 477]
$ CORRECTION: include also nasalization of j to j̃ again -- feeding the other nasalizations here.
j > j̃ / __ [+nas,+cons]
[+syl,-front] > [+nas] / __ [+nas,-syl]
[+syl,-hi] > [+nas] / __ [+nas,-syl]

$ as suggested by s670, normal development of oj>uj,ɔj } wɛ is interrupted 
	$when before a nasal 
	$Pope does not state this explicitly except that her trajectory in s670 suggests it
	$some inference done to make this work (and efficiently so)
	$... giving her the benefit of the doubt
{ũ j̃;ˌũ j̃;ˈũ j̃} > {w ĩ;w ˌĩ;w ˈĩ}  

$ s409i ~ s239 : oi (> wɛ) emits j between itself and other tonic palatal vowels
$ for reasons of organization we realize this as having uj,ɔj go to wɛj before tonic palatals
$ s409i : in cases like botellum (bˌɔe̯ˈaw at this point) it is the e̯ that becomes j 
$ CORRECTION : diachronically this is not helpful but harmful in after countertonic ˌa 
	$ -- including for Pope's examples cathedra > chaire (>chaise) etc
	$ -- ˌa removed from consideration. 
e̯ > j / ˌɔ __ [+front,+prim]
{ˌu;ˌɔ} > {w ˌɛ;w ˌɛ} / __ j [+front,+prim]

$ s519-521 : oj(>uj), ɔj > oɛ̯(uɛ̯),ɔɛ̯ everywhere 
	$ except (s521) when before another yod or countertonic in hiatus with the tonic vowel. 
$ TODO/NOTE: A bit of a contradiction is present here
	$ s519 Pope attributes this shift to the "late 12th and 13th centuries"
	$ but she describes the first stage, lowering of j to ɛ̯ as follows: 
		$ "the less stressed high element of both diphthongs, under the influence of the lower first element, was lowered to ɛ..."
		$ however, if this is after o > u, which occurs in the 11th-12th centuries as per Pope s184, then it that logic does not obtain
		$ because u is not lower than i. 
	$We give Pope the benefit of the doubt and assume she meant uj>uɛ̯ by analogy
		$ and so place it in the date she mentioned in LOF.
j > ɛ̯ / [+syl,+back,+round] __ 

$s521: reverse for now when in the conditions specified. Will be completed in the 16th century in Middle French section. 
$ CORRECTION : this doesn't seem to be useful and is actually harmful. Removed. 
$ ɛ̯ > j / [+syl,+back,+round] __ j
$  ɛ̯ > j / [+syl,+back,+round] __ [+prim]

$ s518-521 : uɛ̯,ɔɛ̯ > wɛ 
$ s519: late 12th and 13th centuries. 
{u ɛ̯;ˌu ɛ̯;ˈu ɛ̯;ɔ ɛ̯;ˌɔ ɛ̯;ˈɔ ɛ̯} > {w ɛ;w ˌɛ;w ˈɛ;w ɛ;w ˌɛ;w ˈɛ} 

$ CORRECTION: depalatalization of tj after a syllabic element
t̪ʲ > t̪ / [+syl] __

$s409ii ~ s239 : y emits j before non-high unrounded palatal vowels with at least some stress (stress part inferred from s239, which is cited by Pope)
$ CORRECTION: this does not appear to be useful and in fact seems harmful
	$ Pope's only example is the verb exsucare going to essuyer -- but the j there could have been from ɣ
$ ∅ > j / [+hi,+front,+round,+syl] __ [+stres,+front,-hi,-round]

$s503 ~ s514-517: yj >ɥi everywhere not already mutated.
$s514: 13th century
$ as per s671 -- nasal variant included. 
[+front,+round,+syl,+hi,βstres,ɣprim] [-round,-syl,-cons] > [-syl] [+syl,βstres,ɣprim]

$ s555-556 ɥø > jø
ɥ > j / __ [+front,+round,-hi]

$ s556 jø > ø after dʒ 
j > ∅ / d͡ʒ __ [+front,+round,-hi]

$ FOURTH NASALIZATION: finally effects all vowels, extended to i and y
$ 13th century
$ s434 ~ s439~s441.6-441.10~s451-457
[+syl] > [+nas] / __ [+nas,+cons]
$handling of unstressed ə̃ as per s434
ə̃ > ə 
ə > ə̃ / __ n̪ t̪ #

$ as per s669-670 : wɛ nasalized takes form of wĩ for now. 
$ Exact order unknown but this seems most in line with what Pope has said elsewhere
$ CORRECTION: currently unnecesssary -- suppressed
$ [+front,-tense,+nas] > [+hi,+tense] / w __ 

$ Pope s528-529 : ɛj > ɛ praeconsonantally. 
$ ɛ may have become long ɛː compensatorily as per s560 ("some writers say..")
	$  but this does not appear diachronicallly useful. 
j >  ∅ / [-round,-tense,+stres] __ [+cons]

$ CORRECTION : s378ii moved earlier. 

$s437: n̪ is absorbed by ə̃ in ə̃nt ending actually goes back to the 13th century. 
n̪ > ∅ / ə̃ __ t̪ #

$ s570ii : lengthening in a:sə,ɛ:sə,i:sə endings
$ CORRECTION: moved this centuries earlier, to before merger of ts and s (<ss) 
	$ as there are clearly different results
	$ as there is no length where the affricate formerly followed. 
$ NOTE: suppress for ts/s demo
[+front,-round,+syl] > [+long] / __ s ə #

$ CORRECTION: tonic a become long before a pre-coda s but NOT a pre-coda ts 
	$ (thus this must come before deaffrication -- moved earlier) 
	$ CORRECTION: actually we move all of s565-7 (earlier to) here 
$ NOTE: suppress for ts/s demo, move s565-7 back to where it is in baseclef.
[+prim] > [+long] / __ s #

$CORRECTION: t͡s effaced before consonants where it remains (pre-coronal) WITHOUT lengthening.
$ NOTE: suppress for ts/s demo. 
$ t͡s > ∅ / __ [+cons] 

$ s172b ~ s194 : before end of 13th century, all affricates become sibilants
[+delrel] > [+cont]

$Pope s247 : ə effaced if after fricative or plosive and before r/l.
$Pope does not give an exact date here...
$ CORRECTION: the inclusion stipulated by Pope of also stops being contexts here 
	$ (though not affricates according to Pope)
	$ is removed so that we are only dealing with continuants for prior context
ˌə > ∅ / [+cont,-son] __ [+son,+ant,-nas] [+prim]

$ CORRECTION: final j after vowel just deleted except after high vowel
$ this is s532 but Pope has it attributed to hte 17th century.
	$ i.e. veracum > vrai, also mai, bai, essai -- Pope seems to think that ʝ remained different from j much later than she does elsewhere (s532)
	$ only case that did go to /e/ seems to be habeo > ai... which is irregular anyways.
	$ thus this is bleeding s529ii. 
j > ∅ / [-hi,+syl] __ #

$ s464 : continuation of nasalizing tendency into Middle French -- still productive except for i and y (i.e. promener, fromage...) 
[+syl,-hi] > [+nas] / __ [+nas,+cons]
[+syl,-front] > [+nas] / __ [+nas,+cons]

$ i.e. where it is in Pope's table -- even though this is effectively the beginning of Middle French. 

$handling of unstressed ə̃ as per s434 -- recurse
	$ but no ə̃nt anymore as it was absorbed in the thirteenth century as per s437
{ə̃;ˌə̃} > {ə;ˌə} 

$ Pope s172d ~ s375-379 ~ s564 ~ s567 : praeconsonantal s effaced after vowels...
$ s377 : had not yet happened in mid 13th century -- early 14th century
$ Not reported in the 1300 Orthographica Gallica, which may be conservative, but implies 
	$ praeconsonantal s as being pronounced /x/
$ CORRECTION: now this is done to x too for fitting with other changes in rules (not a meaningful correction) 
[+syl] x > [+long] ∅ /  __ [+cons]
x > ∅

$ CORRECTION: long countertonic tense e (not ɛ) is shortened.
ˌeː > ˌe

$ s173a ~ s580 ɔː > oː  when free and long -- Pope says 13th century in s173a 
$ but must come after s effacement and resultant lengthening as per s580ii
$ will repeat after effacement of s creates more targets for it. 
$ CORRECTION: moved earlier, was previously a few rules later. 
$ TODO may need to further work on this -- call only once and after preconsonantal s effacement. 
[-lo,-hi,+back,-tense,+long] > [+tense] / __ ([-syl]) [+syl]
[-lo,-hi,+back,-tense,+long] > [+tense] / __ ([-syl]) #

$ Pope s685: implied elsewhere but only stated in the table
$ m praeconsonantal goes to n
m > n̪ / __ [+cons] 

~Old French II

$ s435 : final ɲ > n̪ : likely in 13th century based on rhymes, 
	$ but "ng","gn" spellings retained till the 16th
	$ Pope has this completed only by Middle French in her tables, not as part of Later Old French. 
ɲ > n̪ / __ #

$CORRECTION: single final consonants effaced after ə
[+cons] > ∅ / ə __ #

$ s268-9 : stressless ə effaced before another vowel too
	$ s269 began in 13th century, not accepted till the 16th
$ lengthening per s562
 $ Pope consistently does not have this for the attestment based tables:
	$ s664 MF form "bəos" (beaux), s665 "bəote" for beaute', s667 agneaux, s674 oiseaux
	$ our compromise is having the intermediate stage of ə̯V in FLLAPS
	$ realizing the lengthening aspect afterward as well. 
$ per examples in s269, 271 etc : when she says vowels she must mean "nonconsonantals"
ə > ə̯ / __ [-cons]

$ s244 : lengthening of following a is preceded by assimilation of ə > a : actual formation of lengthened vowel by "contraction" handled by another rule, s242a
$ lengthening implemented in later rule, via s242a ~ s561 
$ per s666 -- this happens before Pope places the marker for Middle French with her table
	$ which we interpret to be circa 1550
$ unlike other effects on ˌə (243-245 etc) adjacent to another vowel, which are placed after it
ˌə > ˌa / __ [+lo,+stres] 

$ s243-245 : countertonic ˌə effaced / __ V 
	$ s243 began in 13th, 
	$ "became very usual" in 16th... but traditional spellings are maintained a bit longer. 
	$ s245 -- some exceptions regarding əy > y, some cases it instead went to ø
$ lengthening per s562
$ consistent with treatment of s268-9, we have reduction to ə̯ realized before MF marker
	$ and realize the lengthening and total effacement afterward
$ per examples in s269 and s271 : when she says vowels she must mean "nonconsonantals" 
ˌə > ə̯ / __ [-cons]

$ s270 ə coming after another vowel is effaced 
$ lengthening per s562
$ s270 : began in the 14th century but not accepted until the 16th, 
	$ when it was "general and accepted by most grammarians" 
$ $ consistent with treatment of s268-9, we have reduction to ə̯ realized before MF marker
	$ and realize the lengthening and total effacement afterward
$ per examples in s269, 271 etc : when she says vowels she must mean "nonconsonantals"
ə > ə̯ / [-cons] __

$ s529ii: ɛ j > e word-finally
$ note s560 says "earlier" (i.e. EOF) aj(>ɛj)>ɛː renders ɛ (possibly long ɛː per poets cited) 
	$  but does not (explicitly) state this for ɛj going to e word finally -- so we do not implement it as such. 
$ s529: accepted in the 15th century (Villon accepts it)...
$ this is for -ai final words not the -aie final words, see below. 
$ DISTINCT from the later shift of final aj# formed Middle French which resolved in the 17th
		$ i.e. s532. ; these are fed by the final loss of e̯ per s562
$ CORRECTION: suppressed -- bled by earlier implementation of "Middle French" s532 -ai effect > /ɛ/. 
	$ all cases that would be effected by this seem to be /ɛ/ except for "ai" which is already irregular. 
$ [-lo,-hi,-tense,+front,+syl] j > [+tense] ∅ / __ #

$ s486 : rounding of e-sounds to ø between two labial consonants or ʒ __ [+lab]
$ based on Pope's evidence, say it was accepted in early 15th century 
	$ first reported by Guiart, who wrote in the early 1300s. 
	$ Next reports of it are usage by Villon (lived 1431-1463) and then Palsgrave (born 1485) 
$ CORRECTION: in case of f (and, we assume p) only happens before v and m (and we assume, b, also voiced),
{ə;ˌə;ˈɛ} > {ø;ˌø;ˈø} / {ʒ;[+lab,+cons,+voi]} __ [+lab,+cons]
{ə;ˌə;ˈɛ} > {ø;ˌø;ˈø} / [+lab,+cons,-voi] __ [+lab,+cons,+voi]

$ CORRECTION: the following denasalization of ĩ  (s452) ...
	$ ... is bled in the case of wĩ by the movement of ĩ to ẽ in this case
	$ Pope says that wĩ regularly denasalizes to wɛ *ultimately* but never explains
		$ why it wouldn't be effected by the earlier denasalization of i when it was wĩ at the time 
			$ and thus a valid target. 
[+front,+nas,+syl] > [-hi] / w __

$ denasalization of ĩ before it is lowered as per s452
$ occurs "before intervocalic nasal" (s452)
$ as per s440 - ỹ can be handled at the same time. 
$ as per 441.10 :  so can ɥ̃ĩ 
$ must be before when ĩ is lowered, which is 15th century.
$ CORRECTION: can also happen when next element is in fact a glide.  
[+nas,+front,+hi,+syl] > [-nas] / __ [+nas,+cons] [-cons]
[-syl,-cons,+round,+front] > [-nas] / __ [-nas]

$ s451 : ĩ > ẽ  -- 15th century (>ɛ̃ not until later sixteenth century as per s452 
	$ -- H Estienne (1582) still says there is a difference,
		$  but lowered pronounciation is accepted by Tabourot (1587) and Lyttelton(1566)
[+nas,+front,-round,-lo,+syl] > [-hi] 

$ s441.3 ~ s466-467 :  ɛ̃j̃ > ɛ̃ before denasalization which was 15th and 16th centuries -- so 15th century
	$ Pope later says s467 that denasalization was in 16th century
$ this specific shift is accepted by Villon (15th century) 
j̃ > ∅ / [+nas,+front,-tense] __ 

$ s496 : ɛ > a / __ r in 15th and 16th centuries (later reversed for many words)
$ reported by Villon (15th) and Henri Estienne (16th) 
$ CORRECTION: may very well have been synchronically valid but not diachronically useful
	$ -->suppressed. 
$[-round,-tense] > [+lo,+tense] / __ r

$CORRECTION: initial ˈɔː shortened before a consonantal cluster.
ˈɔː > ˈɔ / __ [+cons] [+cons]

$ s534-538 ~ s560 : aw > o: -- originally said to be 15th-16th centuries
$ long as per s560
$ s535 -- attested in 1537-1548, and in Ramus 1561 -- so seems to be complete by early 16th
{ˌa w; ˈa w} > {ˌoː;ˈoː}

$ s535~s538-539 : e̯oː >ə̯oː
$ s539 -- reported by Erasmus, who lived around the turn of the 16th century, clearly complete by the writing of Beze on 1584. 
$ fed by aw > o
e̯ > ə̯ / __ [+round,+back,-hi]

$ s510 je > e after ʎ,ɲ,ʃ,ʒ . Late 15th - 16th centuries
$ still not accepted by Palsgrave and Meigret in 1545 but accepted by Estienne 1582
	$ Pope doesn't say this explicitly but give her the benefit of the doubt -- this does NOT occur for the nasal jẽ
$ s512 -- shift was earlier in the West.
$ note as per prior corrections, palatal sʲ, tʲ, zʲ, dʲ have now been preserved to this point in restricted
	$ set of words, and are also triggers for this change
j > ∅ / [+front,+cons] __ [+front,-hi,-lo,-round,+syl,-nas]

$ CORRECTED: now we finally lose the remaining anterior palatals
$ after changing their developments in Gallo-Roman so they are preserved until now in
[+ant,+front] > [-front,-hi]

$ s516 : ɥi to i in "less stressed syllables' and "after the group consonant + r" 
$ "still hesitation in the 16th century" -- say early 16th cent. 
$ CORRECTION: ɥ is effaced exclusively in non-tonic syllables, after a labiodental, possibly with r in between
ɥ > ∅ / [+lab,+cont,+cons] (r) __ [+hi,+front,-prim]

$ s521 : oɛ̯ (uɛ̯, ͏ɔɛ̯?) > wɛj before tonics, o > wɛ before j #
$ final completion of process started much earlier.
$ Sixteenth century -- accepted by Meigret and Baif
[+back,+round,-stres] ɛ̯ > w ɛ
[+back,+round,+prim] ɛ̯ > w ˈɛ
[+back,+round,+syl] ɛ̯ > w ˌɛ

$ CORRECTION : include /we/ resulting from denasalization in this (this is in BaseCLEF* too)
[+front,-round,+syl,-lo] > [-hi,-tense] / w __

$ s436-437 ~ s563 : effacement of nasal consonant praeconsonantal after nasal vowels
$ s437 : before middle of 16th century
	$ mentioned by Peletier (1517-1582)
	$ absorption of consonant in unstressed ə̃n̪t̪ ending goes back to 13th century per Hebrew glosses 
$ s563 effacement of praeconsonantal nasals causes lengthening of prev nasal vowels... 
$ s436, 438 : at the same time n̪ word final disappears except when the next word starts with a vowel -- in keeping with our dataset we delete these now as well.
[+nas,+syl] [+nas,+cons] > [+long] ∅ / __ {[+cons];#}
[+nas,+syl] j̃ [+nas,+cons] > [+long] j̃ ∅ / __ {[+cons];#}

$Pope s247 : ə effaced if after fricative or plosive and before r/l.
$Accepted in 16th century. 
$CORRECTION: suppressed entirely, with restricted use earlier in Middle French.
$ˌə > ∅ / [-delrel] __ [+son,-hi,-nas] [+prim] 
$ˌə > ∅ / [+cont,-son] __ [+son,-hi,-nas] [+prim]

$ s392 : final -l effacement after e:, i, y, u
$ regular by Later Middle French
$ restoration for many cases for e:, y and u 
	$-- but this is not a phonologically mediated process (rather lexically and socially driven)
$ accepted by Henri Estienne (late 16th)
$ also by Palsgrave, Tabourot (late 16th) 
$ say mid 16th 
$ CORRECTION: does not happen to /e/, /eː/ etc. May have occurred synchronically (as Pope demonstrates) but did 
	$ not become established diachronically. 
$ CORRECTION : after i, only happens after sibilant (in all known cases for /i/, they were palatalized meaning 
	$ a long-distance assimilation of lʲ is not unlikely -- but more data necessary) 
	$ always happens after u 
[+lat] > ∅ / [+strid,+cor] [+syl,+hi,+front] __ #
[+lat] > ∅ / [+syl,+hi,+back] __ #

$ s530 ~ s532: s531 final ajə become ɛː right after final ə is effaced next to it...
$ moved here for agreement with FLLAPS which has these words with final ɛː more in line with Peletier
$ that it is long ɛː not short ɛ is inferred because of s562...
$ obv comes after s271
$ s532 : "accepted only in the 17th century"
$ Peletier (1549) accepts it. 
$  -- Lanoue (1596) accepts only in some cases
$ "hesitation in Early Modern French" 
[+lo,+front] j > [-lo,-tense,+long] ∅ / __ (ə̯) #

$ s562 -- per consistency on FLLAPS Pope tables, ə̯ after another vowel effaced with prior lengthening at this point.
$ but only before a stressed syllabic
[+stres] ə̯ > [+long] ∅
ə̯ > ∅ / [-cons] __

$ s272 : effacement of intertonic ə between "n, r or l" and other consonants
$ intertonic -- most have stress level on either side 
$ but there do not seem to be any examples of where it occurred with n coming after ə
$ as per s272: begins twelfth and thirteenth centuries but only becomes regular in the 16th, and even then hesitation leads to doublets
ə > ∅ / [+stres] [+cor,+son,+cons] __ [+cons] ([-syl])* [+stres]
ə > ∅ / [+stres] ([-syl])* [+cons] __ {r;l} ([-syl])* [+stres]

$ s242a ~ s561 merging of adjacent and similar vowels into long vowels.
$ Accepted in the sixteenth century (but started in the thirteenth) 
$ accepted by Regnier -- very late 16th, early 17th -- s242
[+lo] [+lo] > ∅ [+long]
[-lo,-hi,+front,-round,+syl] [-lo,-hi,+front,-round,+syl] > ∅ [+long]
[+hi,+back,+round,+syl] [+hi,+back,+round,+syl] > ∅ [+long]
[-hi,+back,+round,+syl] [-hi,+back,+round,+syl] > ˈuː

$ s242b specifications on coalescence of countertonic a
$ s561 : length of aũ destination
$ CORRECTION: aũ only becomes ã word finally 
$ CORRECTION: occurs regardless of the length of the second element
$ CORRECTION: ˌa ˈɛ > ˈɛː 
$ CORRECTION: delete ˌa ˈi > ɛː -- way more harm than good. Instead have no effect. 
$ CORRECTION: restrict remaining a contraction to when some variant of u follows.
ˌa [+prim,+back,+nas] > ˈãː / __ # 
ˌa > ∅ / __ [+prim,+back,+hi]
$ [+lo,+stres] [+prim,+front,+hi,-round,-nas] > ˈɛː 
ˌa [+prim,-round,-tense,+front] > ˈɛː 
[+lo,+stres] [+prim,+front,+hi,-round,+nas] > ˈɛ̃ː

$s579: ɔ lengthens before z,v ... and then any long ɔː becomes oː 
[-hi,-lo,+back,+syl,-tense] > [+long] / __ {z;v} 
[-hi,-lo,+back,+syl,+long] > [+tense]

~Middle French 
$ Middle French marker for us is at 1550 

$ s451-452 ẽ > ɛ̃ in later sixteenth century
$ However this does not occur in the clusters of jẽ and wẽ until the early 17th century, in Early Modern, rather than Middle, French 
	$ -- see s473-474, 669-670 for wẽ; s470-471 for jẽ; 476 for ɥɛ̃
$ Accepted by Tabourot (1587) and Lyttelton (1566) 
$ and attributed early to Parisians by Bovelles (1533) --
	$ -- though ĩ retained late in Northwest (while Burgundy has lowering early)
[+front,-lo,+nas,+syl,-round] > [-hi,-tense] / # __ 
[+front,-lo,+nas,+syl,-round] > [-hi,-tense] / [+cons] __ 
[+front,-lo,+nas,+syl,-round] > [-hi,-tense] / [+syl] __ 

$ s467-468 : denasalization of ɛ̃ (from ɛ̃j̃, ĩ, ãj̃, and ẽj̃) in mid to late 16th century
$ attested in Lyttelton 1566 (s468)
$ we can, for convenience, run this with the denasalization of jɛ̃ from jẽ (<je + n <ɛː + n)
$ but note technically this was a few decades later, as Beze does not accept it, nor do Meigret or Peletier
	$ with denasalization as jɛ finally being attested in Lanoue 1596
$ CORRECTION: we also merge into this the denasalization of œ̃ (œ,ø + n originally NOT y + n which has already denasalized to y)
	$ -- as per s477
$ also merge into this the denasalization of wɛ̃, even though Pope implies that was in the 17th century (s474-476)
	$ -- simpler this way. 
[+front,-lo,+nas,+syl] > [-nas] / __ [+nas,+cons] [-cons]

$ s494: e (from ae̯, ie̯...) > ɛ before l, ʎ in 16th century
$ Peletier (1549) still has /e/, Meigret gives both before both l and ʎ,
$ while Meigret (also 1500s) fluctuates
$ while Beze has /ɛ/ -- so we place it in the second half of the 16th. 
[-hi,-lo,+syl,+front,-round] >[-tense] / __ [+lat]

$ s495 : e > ɛ when r CLOSES the syllable in 16th century (will repeat again when final ə before an r is deleted-- in 17th century)
$ for word-final r, only applies in monosyllables
$ not accepted yet by Meigret (1500s), Lanoue,
$ H. Estienne appears to be unsure (?)
$ seems accepted by the time of Hindret (1687) though some hesitation continues
$ -- very late Middle French to early Modern, possibly early 17th century
$ CORRECTION: ø also lowers to œ before r in these same positions
	$ Pope does mention y and u lowering to œ and ɔ but not exactly what happens here, and she calls it a tendency
		$ which for the most part, she argues was reversed-- i.e. not diachronically relevant. 
$ CORRECTION: does not happen finally with rs#
$ CORRECTION: ø also lowers in non monosyllables before r#
[-hi,-lo,+syl,+front] > [-tense] / __ r [+cons,-strid]
[-hi,-lo,+syl,+front,-round] > [-tense] / # ([-syl])* __ r #
[-hi,-lo,+syl,+front,+round] > [-tense] / __ r #

$ s486 : these new ø variants (from e/ɛ/ə under labial influence) all close to y except in stressed syllables
$ appears to have been accepted in the late 16th century based on Pope's presented evidence
$ Ramus (lived 1515-1572) reports that beuvons and buvons are used by different people
$ Vaugelas (1585-1650) also reports on variation still occurring in his lifetime. 
$ only happens to ø that was once e/ɛ/ə -- must be before a labial.
[+front,+round,+tense,-prim] > [+hi] / __ [+lab,+cons]

$ s562, s243-245, s268-9, s270: 
$ completion of loss of schwa adjacent to vowels
$ and lengthening of adjacent stressed vowels (s562)
$ s562: accepted widely by sixteenth century, spelling changes a bit later, 
	$ we broke the shift up because of the way Pope did tables not recognizing it (based on orthography no doubt)
		 $as a "compromise"
$ Və# finals are simplified before MF marker per consistency.
$ CORRECTION: effacement adjacent to j,w too
[+stres] > [+long] / ə̯ __
ə̯ > ∅ / __ [-cons]

$ CORRECTION: s477 merged into earlier rule. 

$ s570 i : a lengthens before ʎ in final ending aʎə
$ accepted by Beze -- 17th century
[+lo] > [+long] / __ ʎ ə #

$ CORRECTION: s570ii a:se, i:se etc lengthening moved earlier so that it is not fed by deaffrication. 
$ CORRECTION: lengthening before final s moved earlier. 

$NOTE: original rule of Pope here following used only for ts/s demo, 
	$otherwise suppress. 
$ s570ii : lengthening in a:sə,ɛ:sə,i:sə endings (Pope says this happens when /s/ was originally a geminate "ss" --- but ts and ss had merged into s centuries ago as per her chronology)
$[+front,-round,+syl] > [+long] / __ s ə #

$ s173.1b ~ 576 e:, je: >ɛ:, jɛ: before single consonant then ə, or single cons then word coda
$ not accepted by Peletier, Lanoue oscilates, Baif oscilates
$ CORRECTION: not before coronals then coda -- two exceptions to this counterrule are amer and tres
	$ but amer is likely contamination from amere, and tres has syntactic considerations
$ CORRECTION: include also ˌe as in from pietaille etc.
$ CORRECTION: include also ˈe before two consonants (-evre cases, tiers, tierce...)
[+tense,-hi,-lo,-round,+stres] > [-tense] / __ [+cons] [-cons,-back,-front]
ˈe > ˈɛ / __ [+cons,-cor] #
ˈe > ˈɛ / __ [+cons] [+cons]

$ s273 final ə postconsonantal becomes an "off-glide" before it disappears forever... 
$ reported by Beze 1564 but does not bleed some of the lengthening and quantity changes conditioned on final ə
ə > ə̯ / __ #

$ s366: late degemination of r, with lengthening of any prior vowels
$ s366 only says "before the end of Middle French" 
$ CORRECTION: moved much earlier -- contradicts Pope's philological evidence but works much better result wise. 
$ [+syl] > [+long] / __ r r 
$ r r > r / __ {ə;ə̯}
$ r r > r / ə __ 

$173.1c ~ s586 : long a becomes back ɑ (still long) -- 
$ s586 -- still was palatal in 16th century as per Ronsard and interchange of e/ɛ and a before r 
$ "but possible that some differentiation had begun earlier" 
$ first remark seems to be 1567 by Plantin 
$ "toward the end of the period" --  late 16th century 
[+lo,+long,+syl] > [+back,-front]

$s511 establishment of je>ie "diaeresis" in cases of C + l/r + je
$ CORRECTION: moved a century earlier, because it bleeds s241
$ CORRECTION : goes to i j not just i.e
∅ > i / [+cons] [+ant,+son] __ j [+front,-hi,-lo,+tense,-round,+syl]

$ CORRECTION j also is inserted between i and a following vowel if the ˌi is after [+back] ɥ
∅ > j / [+back] ɥ [+hi,+front] __ [+syl]

$s241 : consonantalization of countertonic high vowels before lower vowels
$ 16th century, "Later Middle French" 
$ accepted by early 17th -- attested in Vaugelas
$ CORRECTION: does not occur before low vowels either (truant) 
[+hi,+stres,-prim] [-hi,-lo,+syl] > [-syl] [+stres]

$s569: Pope citing Beze, intervocalic z engenders lengthening of prior consonants
$giving Pope benefit of the doubt, we place this after long a velarization, because of examples like masure /mazyʁ/
$ early 17th?? 
[+syl] > [+long] / __ z [+syl]

$s570iii ending aʒə sees lengthening of a 
$ giving Pope benefit of the doubt, we place this after long a velarization, because of examples like voyage /vwajaʒ/
[+lo] > [+long] / __ ʒ ə̯ #

$ CORRECTION: ... but o opens if it is followed by two or more consonants (autre, votre, pauvre)
[-hi,-lo,+back,+syl] > [-tense] / __ [+cons] [+cons]

$ recurrence of s239 : wˌɛ emits j if before tonic vowel -- late 16th century
$ CORRECTION: removed. Unnecessary.
$ ∅ > j / w ˌɛ __ [+prim]

$|Early Modern French

$ CORRECTION: pre-coda t will be preserved in pre-coda cons loss series, if it is only preceded by
	$  monosyllables with no onset consonantals (ɥ, w, j -- not consonantal)
∅ > ə̯ / # ([-cons,-nas])* t̪ __ #

$ CORRECTION denasalization shifts 474-477 moved back into prefvious time period. 

$ s395 : r becomes a glide in certain positions... 
$ r "losing its trill" (in the index, Pope explicitly attributes the destination as ɹ)
r > ɹ / [-cons] __ {#;[+cons]}

$ s580 open o lengthens and raises before s or z that is about to be deleted
[+prim,-tense,+back,+round,-hi] > ˈoː / __ [+strid,-cont] #

$ s611-620 : final fricative and plosive consonants deleted as a result of liaison phenomena
$ s611 : began in 13th century, established by 16th, s620 accepted (grudgingly) after the early 17th 
$ s618-619 : counteraction led to preservation quasiregularly where it was f or k in a monosyllable. 

$for preserving f and k as per s618-169
$ CORRECTION: f is ALWAYS preserved (regularlY)
$ CORRECTION: only applies for k when the single syllabic vowel is not nasal
$ CORRECTION: final s supported in monosyllables starting with a sibilant
∅ > ə̯ / # [+syl,-nas] [-syl] k __ #
∅ > ə̯ / # [-syl] [+syl,-nas] k __ #
∅ > ə̯ / f __ #
∅ > ə̯ / # [+strid] [-round,+hi,+syl] s __ #

$ə̯ will be deleted by rule after this one ... 
$now implement final cons loss as per s611-620: 
[-son] > ∅ / __ #

$ s273 : final "off-glide" ə is deleted.
$ for convenience we group this with:
	$ s539 ə̯o: > o: finally accepted in the 17th century
	$ CORRECTION: but deletion of ə̯ does not occur in monosyllables. 
ə̯ > ə / # ([-syl])* __ ([-syl])* #
ə̯ > ∅

$ RECURRENCE
$ s495 : e > ɛ when r CLOSES the syllable in 16th century (will repeat again when final ə before an r is deleted-- in 17th century, ...)
$ for word-final r, only applies in monosyllables
$ not accepted yet by Meigret (1500s), Lanoue,
$ H. Estienne appears to be unsure (?)
$ seems accepted by the time of Hindret (1687) though some hesitation continues
$ -- very late Middle French to early Modern, possibly early 17th century
$ CORRECTION: ø also lowers to œ before r in these same positions
	$ Pope does mention y and u lowering to œ and ɔ but not exactly what happens here, and she calls it a tendency
		$ which for the most part, she argues was reversed-- i.e. not diachronically relevant. 
$ CORRECTION: does not happen finally with rs#
$ CORRECTION: ø also lowers in non monosyllables before r#
[-hi,-lo,+syl,+front] > [-tense] / __ r [+cons,-strid]
[-hi,-lo,+syl,+front,-round] > [-tense] / # ([-syl])* __ r #
[-hi,-lo,+syl,+front,+round] > [-tense] / __ r #

$ CORRECTION: ø opens before consonants in final syll. Any final /ə/ does not change this.
	$ Pope mentions the emergence of the new open /œ/ sound but does not explicate this specific way it emerged
[+front,+round,-hi,+syl] > [-tense] / __ ([+cons])+ (ə) #

$ s470-474 :  final acceptance of wẽ,jẽ opening to wɛ̃,jɛ̃
$ accepted in the early 17th century
$ for convenience ø̃ > œ̃ handled also in here. 
[+nas,+front,+syl,-lo] > [-hi,-tense]

$ s457 : ỹ >ø̃ begins in late 16th century only accepted in the 17th century. 
[+nas,+syl,+front,+round] > [-hi]

$ s444 : denasalization of ã  -- 16th century beginning, accepted seventeenth century 
$ s440 : accepted in "Early Modern French"
$ for convenience we also group the denasalization of wɛ̃,jɛ̃ here...
[+front,-round,+syl,+nas] > [-nas] / __ [+cons,+nas] [-cons]

$ CORRECTION: s511 diarhesis moved much earlier

$ s531.ii : aj followed by tonic vowel becomes ɛj, accepted in late 17th century
$ CORRECTION: only before tense tonic vowels, and it goes to e, not ɛj
[+lo,+front] j > [-lo,-tense] ∅ / __  [+prim,+tense]

$ CORRECTION: mid vowels to lax before i,j
[+syl,-lo,-hi] > [-tense] / __ [-cons,+hi,+front]

$ CORRECTION: countertonic ə, ɛ after [+lab] l become ˌe
[-round,-hi,-back,-lo,+stres,-prim] > ˌe / [+lab] l __

$ s400-401 : final r deletion 
$ began in sixteenth century but successfully reversed by sociolinguistic interference (prestigious restoration) 
	$ for all except for er and jer terminations in polysyllabic words.
$ CORRECTION: ɹ does also efface in monosyllables of pattern we(ː)ɹ#
ɹ > ∅ / [+syl] ([-syl])* [-hi,+front,+tense,-lo] __ #
ɹ > ∅ / w [-hi,-lo,+syl,+tense] __ #
ɹ > r $reassertion of trill elsewhere. 

$s196 Second h-deletion. Starts much earlier but accepted in late 17th century
	$ has no effect unless Frankish and other Germanic words are in dataset
h > ∅ 

$ s394 ~ s498 : r becomes uvular 
$ seventeenth century (s394)
r > ʀ

$ s121~s170~s223~s508~s602: leveling of stress -- stress no longer distinctive for vowels.
[+stres] > [-stres]

$ CORRECTION: ũː denasalizes before short ũ, as o rather than ɔ (Saone, Jerome...) 
ũː > o / __ [+nas]

$ CORRECTION : (Pope didn't mention this oddly) loss of distinctive vowel length entirely by 1700
$ Though it may have started in the 16th century -- see Gess 2001 for more info
$ Length is in fact preserved in conservative French dialects including those of Belgium and Quebec -- but not the Paris dialect. 
[+syl,+long] > [-long]

$ s460 : ũ > ɔ̃ in 17th century, but still not complete (s462) in 1696 (!)  
ũ > ɔ̃
$[+back,-lo,+nas,+syl] > ɔ̃

$ s463: ɔ̃ denasalizes in open syllables when followed by nasal cons. 
$ s463: first attested late 17th century, not accepted until 18th.
$ for convenience œ̃ also handled in here. 
[+round,+nas] > [-nas] / __ [+nas,+cons] [+syl]

$ CORRECTION: denasalization of nasal vowels before nasal consonants that have become final. 
$ Can be handled by doing this anytime the nasal cons has not been deleted.
[+syl,+nas] > [-nas] / __ [+nas,+cons]

$ CORRECTION : word final ɔ to o
ɔ > o / __ #

$ s525 wɛ > wa -- first attested late 16th century
$ as per s525, not really accepted except before r "until the Revolution destroyed the old tradition"
ɛ > a / w __ 

$s441.1~s443 ã > ɑ̃ in Modern French
$Pope does not give an exact date for this; she merely says "in the modern period"
$ CORRECTION: ABROGATED -- this is handled correctly earlier, based purely on length, so that cases like ane, Bretagne etc are unaffected.
$ {ã;ãː} >[-front,+back]

$ CORRECTION: uvular ʀ trill becomes  a fricative 
ʀ > ʁ 

$ Pope s381 ~ s687 : ʎ > j "in the course of Modern French" 
ʎ > j

$ CORRECTION: j and i after other vowels absorbed into j
[+hi,+front,-round] > ∅ / [+syl] __ j

$ Pope s177.2: ə effaced anywhere it does not bring three consonants into a cluster. 
$ CORRECTION: last one of these is surpressed as it caused inconsistencies with Pope's later notations
$ CORRECTION: final schwa deletion does not occur for lexical citation forms in monosyllables
	$ despite phenomena like l', m', n', qu' etc
ə > ∅ / [+syl] ([-syl])* __ #
ə > ∅ / __ [+syl]
ə > ∅ / [+syl] __
$ ə > ∅ / {#;[+syl]} [-syl] __ [-syl] {#;[+syl]}
