The Correspondence Library

[Pogostick Man]

Credit to dhok for informing me of this resource.

Proto-Algonquian to Mi'kmaq

From Dawe, Audrey Marie (1986), The Fundamentals of Micmac Historical Morphology, citing Hewson, John (1973), "Proto-Algonkian Reflexes in Micmac", and Hewson, John (1983), "Some Micmac Etymologies".

tʃ → ʃ / ! C_
n{θ,l} h{θ,ʃ} → Ø s
{ʔ,h,N} → Ø / _C
ʔ{θ,ʃ} ʔl → s Ø
x → Ø / _{p,k}
θ ʃ → s Ø / _k
ʃ → s
θ → l
k → {k,q}
o(ː) waː eː iː → u o e i
aː → a
(aw)aha → aː
{awa,iwa,iwi} → uː
{o,a}wi → oː
ehi → eː
{aja,iha,iji,ihi,ija} → iː

[Pogostick Man]

Proto-Algonquian to Piscataway

From Mackie, Lisa (2006), "Fragments of Piscataway: A Preliminary Description". This is very incomplete, partially because it seems that the only source we have on Piscataway is a single document in rather poor condition.

*#we- retained
{θ,ʃ} → ɬ ? (conjectured based on <z> in the Piscataway source and the lack of voicing in the apparent original sounds)
k → x
e → o / unclear conditioning
ʔ → h / _C

[Pogostick Man]

Information in this post is from here.

Proto-Na-Dene to Proto-Athabaskan-Eyak

kʲ kʲʼ gʲ → ts tsʼ dz
xʲ → s
{s,ʃ} → Ø / _x

Proto-Athabaskan-Eyak to Proto-Athabaskan

ɬ → ɬ~l
ʃ → ʂ~ʐ
{s,dz} → s~z
k kʼ kʷ kʷʼ g gʷ → kʲ kʲʼ ʈʂ ʈʂʼ gʲ ɖʐ
x xʷ → xʲ~j ʂ~ʐ
Qʷ → Qʷ → Q ?
χ(ʷ) → χ~ʁ

Proto-Athabaskan-Eyak to Eyak

kʲ kʲʼ gʲ xʲ → ts tsʼ dz s~ʃ
Kʷ → K
qʷ qʷʼ ɢʷ → q qʼ ɢ
s → Ø / _x
x → Ø / ʃ_
$ → xʷ → {x,s}

Proto-Na-Dene to Tlingit

Where a colon appears, forms to the left are normal and forms to the right are l-assimilated.

s → s : ɬ
ts → ts : tɬ
tsʼ → sʼ~tsʼ : ɬ~tɬʼ
ʃ → ʃ~s : ɬ
tʃ → tʃ~ts : tɬ
tʃʼ → sʼ~tʃʼ : tɬʼ
Something's going on with the velars and uvulars; apparently, both the rounded and unrounded consonants have reflexes that may or may not be rounded
kʲ kʲʼ → {k,ʃ} kʼ
xʲ → x
k(ʷ)ʼ → {x,k}(ʷ)ʼ
x(ʷ) → x
qʼ → χ(ʷ)ʼ
qʷʼ → {χʼ,q(ʷ)ʼ}
x → Ø / {s,ʃ}_
$ → χ

[----]

What is $ supposed to refer to here?

[Pogostick Man]

What is $ supposed to refer to here?

It's apparently some sort of phoneme of unknown value, possibly something like a sibilant. It seems to come up in discussions of reconstructing Athabaskan.

[Pogostick Man]

Index Diachronica v.8.1 is out.

[Pole, the]

Proto-Na-Dene to Proto-Athabaskan-Eyak

kʲ kʲʼ gʲ → ts tsʼ dz
xʲ → s
{s,ʃ} → Ø / _x

Proto-Athabaskan-Eyak to Eyak

kʲ kʲʼ gʲ xʲ → ts tsʼ dz s~ʃ
[…]
s → Ø / _x

Is this repetition on purpose?

[Pogostick Man]

No, I think that was an oversight. Hopefully I will fix it in time for the next update.

[Zju]

Do you happen to know based on which word pair are the sound changes for proto languages posited? I'm specifically interested in the Altaic branch, but also in other proto Kartvelian, Northeast Caucaisan and Northwest Caucasian.

[Pogostick Man]

Do you happen to know based on which word pair are the sound changes for proto languages posited? I'm specifically interested in the Altaic branch, but also in other proto Kartvelian, Northeast Caucaisan and Northwest Caucasian.

I'm afraid I don't, sorry.

Proto-Indo-European to Sanskrit, from Edgerton, Franklin (1946), Sanskrit Historical Phonology: A Simplified Outline for the Use of Beginners in Sanskrit; and Whitney, William Dwight (1879), A Sanskrit Grammar, Including Both the Classical Language, and the Older Dialects, of Veda and Brahmana

The following is likely at least somewhat incorrect due to my main source, which is from the '40s, and my probably somewhat-incorrect understanding/methods. The author held that PIE had voiceless aspirates in addition to voiced ones (apparently corresponding to *SH sequences under laryngeal theory), but seeing as nobody else has posted sound changes from PIE to Sanskrit it's probably better than nothing. For the purposes of ease of reading PIE, assume <y>, when you see it, refers to /j/. It gets kind of confusing due to how <y j> get used in the various relevant notations.

{ḱ,ǵ}T ǵʱT → ʂʈ ɖʱ
sḱ → cʰ (the voiced palatal aspirate is apparently a loan phoneme?)
ḱ ǵ ǵʱ → ʃ ɟ h
k(ʷ) g(ʷ) g(ʷ)ʱ → c ɟ h / _{E,y}
kʷ gʷ gʷʱ → k g h
s → ʂ / {B,w,E,y,k,r}_ ! _{r,#}
{ʃ,ks} → ʂ / _T
n → ɳ / {ʂ,r,r̥,r̥r}…_ (this change is blocked if, per Whitney, "there intervene[s]. . .a palatal (except. . .y), a lingual, or a dental")
{o,a,e}y → eː
{o,e}(ː) → a(ː)
w → v
{o,a,e}v {o,a,e}ːv → o aːw / _{C,#}
{r̥,l̥} → r̩ (this is probably actually retroflex but I'm leaving it as <r> for the sake of the font)
Non-syllabic *r and *l get mixed up; it seems that an outcome of /r/ is more typical
r̥r l̥l → {u,i}r ~ {u,i}l
r̥ː l̥ː → {u,i}ːr ~ {u,i}ːl
N̥(ː) → a(ː)
m̥m n̥n → am an (probably redundant, but just making sure)

[Tropylium]

Proto-Indo-European to Sanskrit

That can be done in much more detail. Here is a chronologically rearranged version, including some laryngeal theory addenda (a few interpretations are to taste), and several further developments in particular from Masato Kobayashi (2004), Historical Phonology of Old Indo-Aryan Consonants:

PIE to "Proto-Satem"
e > a / _h₂_
e > o₂ / _h₃_
p > b / _h₃
h₁ h₂ h₃ > ə / syllabic
ə > ∅ / #_
h₁ h₃ > H
H > ∅ / [plosive]_
kʷ gʷ gʰʷ > k g gʰ
s > ʃ / {i u r l k ḱ g ǵ gʰ ǵʰ}_ (RUKI)
ḱ ǵ ǵʰ > tʃ dʒ dʒʰ

"Proto-Satem" to Proto-Indo-Iranian
Bʰ₁P₂ > B₁Bʰ₂ (Bartholomae's Law; includes s ʃ > zʰ ʒʰ)
Bʰ-Bʰ > B-Bʰ (Grassmann's Law)
ptʃ > pʃ
ttʃ ddʒʰ > t.ʃ d.ʒʰ
tʃt dʒdʰ > ʃt ʒdʰ
tʃʃ > ʃʃ
k g gʰ > c ɟ ɟʰ / _{y i e} (Law of Palatals)
o > aː / _CV (Brugmann's Law - does not affect o₂ < eh₃)
e o o₂ m̥ n̥ > a

Proto-Indo-Iranian to Proto-Indo-Aryan
ə > i

Laryngeal simplification
ph₂ th₂ kh₂ bh₂ dh₂ gh₂ > pʰ tʰ kʰ bʰ dʰ gʰ
h₂ > H
VH > Vː / _{C #} (includes a < N̥)
R̥H > uR, iR / _V (irregularly, but R̥Hu > uru usually)
R̥H > uːR, iːR / _C

Introduction of retroflexion
ʃ ʒ > ʂ ʐ
R(V)n > R(V)ɳ
t d dʰ n > ʈ ɖ ɖʰ ɳ / [retroflex]_, except _r
s t d dʰ n > ʂ ʈ ɖ ɖʰ ɳ / _[retroflex]
ls lt ld ldʰ ln > ʂ ʈ ɖ ɖʰ ɳ (Fortunatov's Law; disputed)
s > ʂ / _V{ʂ ʈ ɖ ɖʰ}

Demise of affricates and voiced sibilants
∅ > a / #_z
aZ iʐ uʐ > əː iː uː
əː > oː / w_
əː > eː / otherwise
tʃ dʒ dʒʰ > tɕ dʑ dʑʰ > ɕ ɟ ɟʰ
tst dzdʰ > tː dʰː

Other consonant changes
kt pst > t st / #_
ɟʰ > ɦ
dʰ > ɦ / in about a dozen cases, conditions remain disputed
bʰ > ɦ / in about two cases
Vm > Ṽ / _{s ɕ}, also between words (there was no -mʂ-)
n > ɲ / ɟ_ (but not in ɕn, ɦn < cn, ɟʰn)

Proto-Indo-Aryan to Vedic Sanskrit
s ʂ > h / _#[voiceless]
s ʂ > r / _#[voiced]
l > r
ɖ ɖʰ > ɭ ɭʰ / V_V
sɕ ɕʂ > ɕɕ ʂʂ
ss ʂʂ ɕɕ > t.s t.ʂ t.ɕ
t.ʂ t.ɕ > kʂ cʰː
cʰː > cʰ / C_
bzʰ > ps
pʂ cʂ ɟʐʰ gʐʰ > kʂ
(Prakrits generally retain ss (and ɕɕ > ss), some retain sɕ; most have kʂ cʂ gʐʰ ɟʐʰ > kʰː cʰː gʰː ɟʰː)

Vedic Sanskrit to Classical Sanskrit
retroflex laterals reverted to stops
H > ∅
ai au aːi aːu > eː oː ai au / _¬V
w > v

Various changes are probably still missing from this.

[KathTheDragon]

p > b / h₃_

Did you mean _h₃?

[Pogostick Man]

Thank you, Tropylium and Kath!

_¬V

What does this mean?

[Tropylium]

"Except before a vowel".

You're correct, Kath; fixed.

[Pogostick Man]

Proto-Athabaskan to Hupa

From Sapir, Edward (1936), "Reflexes of Proto-Athabaskan in Several Languages (Hupa, Navaho, Chipewyan, Sarcee)"

NB: First, part of the list of correspondences was cut off; second, it is sometimes difficult to read Sapir's handwriting; and third, I'm hoping I made the correct inferences about his notation.

h → {h,Ø}
q ʁ → x w
qʷ qʷʼ ɢʷ xʷ ɣʷ → x(ʷ) q(ʷ)ʼ ɢ(ʷ) x(ʷ) w
z → s
{ʃ,ʒ} → w
xʲ → w

Proto-Athabaskan to Proto-Southern Athabaskan

From Hoijer, Harry (1938), "The Southern Athapaskan Languages". American Anthropologist 40:75 – 87

kʲ kʲʼ gʲ → ts tsʼ dz
m → {b,m} (→ b seems more common)
ɣ → h / in derivational prefixes
VnC → V[+ nas]C / _# (unless C = ʔ)
t n x → d n h / in derivational prefixes

Proto-Southern Athabaskan to Proto-Eastern Southern Athabaskan

From Hoijer, Harry (1938), "The Southern Athapaskan Languages". American Anthropologist 40:75 – 87

t → k
{s,z}(ʔ) {ʃ,ʒ}(ʔ) {ɬ,ɮ}(ʔ) → s ʃ ɬ / _#

Proto-Eastern Southern Athabaskan to Kiowa Apache

From Hoijer, Harry (1938), "The Southern Athapaskan Languages". American Anthropologist 40:75 – 87

n → d
d → Ø / _#
k → tʃ / _E
{xʲ,j}ʔ {x,ɣ}ʔ → Ø h(ʔ) / _#
Vnʔ VnC → V[+ nas] V[+ nas]C / _#
x ɣ → h Ø / _#
{d,j} → Ø / _#
xʲ j → ʃ ʒ
{s,z}(ʔ) {ʃ,ʒ}(ʔ) {ɬ,ɮ}(ʔ) → s ʃ ɬ / _#

Proto-Eastern Southern Athabaskan to Jicarilla

From Hoijer, Harry (1938), "The Southern Athapaskan Languages". American Anthropologist 40:75 – 87

d → ʔ / _#
xʲ xʔ j ɣʔ → h ʔ Ø ʔ / E_#
xʲ xʔ j ɣʔ → ih iʔ i iʔ / _#
xʲʔ jʔ → h ʔ / _#
x ɣ → h Ø / _#
n → ⁿd
xʲ → s
j → ɣ / _E
Vnʔ VnC → V[+ nas] V[+ nas]C / _#
{s,z}(ʔ) {ʃ,ʒ}(ʔ) {ɬ,ɮ}(ʔ) → s ʃ ɬ / _#

Proto-Eastern Southern Athabaskan to Lipan

From Hoijer, Harry (1938), "The Southern Athapaskan Languages". American Anthropologist 40:75 – 87

d → Ø / _#
xʲ j → ʃ Ø / _#
{xʲ,j}ʔ → Ø / _#
{x,ɣ} → Ø / _(ʔ)#
Vnʔ VnC → V[+ nas] V[+ nas]C / _#
n → ⁿd
xʲ → s
j → ɣ / _E
{s,z}(ʔ) {ʃ,ʒ}(ʔ) {ɬ,ɮ}(ʔ) → s ʃ ɬ / _#

Proto-Southern Athabaskan to Proto-Western Southern Athabaskan

From Hoijer, Harry (1938), "The Southern Athapaskan Languages". American Anthropologist 40:75 – 87

j → ɣ / _E

Proto-Western Southern Athabaskan to Chiricahua

From Hoijer, Harry (1938), "The Southern Athapaskan Languages". American Anthropologist 40:75 – 87

d → Ø / _#
xʲ j → ʃ Ø / _#
{xʲ,j}ʔ → Ø / _#
{x,ɣ} → Ø / _(ʔ)#
Vnʔ → V[+ nas] / _#
n → ⁿd
{s,z}(ʔ) {ʃ,ʒ}(ʔ) {ɬ,ɮ}(ʔ) → s ʃ ɬ / _#

Proto-Western Southern Athabaskan to Mescalero

From Hoijer, Harry (1938), "The Southern Athapaskan Languages". American Anthropologist 40:75 – 87

d → Ø
xʲ j → ʃ Ø / _#
{xʲ,j}ʔ → Ø / _#
{x,ɣ} → Ø / _(ʔ)#
Vnʔ → V[+ nas] / _#
n → ⁿd
s(ʔ) ʃ(ʔ) ɬ(ʔ) → s ʃ ɬ / _#
z(ʔ) ʒ(ʔ) ɮ(ʔ) → dz dʒ dɮ / _#

Proto-Western Southern Athabaskan to Navajo

From Hoijer, Harry (1938), "The Southern Athapaskan Languages". American Anthropologist 40:75 – 87; and Sapir, Edward (1936), "Reflexes of Proto-Athabaskan in Several Languages (Hupa, Navaho, Chipewyan, Sarcee)"

{ɣ,h} → j / in derivational prefixes
xʲ(ʔ) j(ʔ) → h ʔ / _#
x ɣ → h ʔ / _#
{x,ɣ}ʔ → {ʔ,g} (Hoijer notes a reflex "-g-")
xʲ → s
j → ɣ / _E
{s,z}(ʔ) {ʃ,ʒ}(ʔ) {ɬ,ɮ}(ʔ) → s ʃ ɬ / _#

Proto-Western Southern Athabaskan to San Carlos

From Hoijer, Harry (1938), "The Southern Athapaskan Languages". American Anthropologist 40:75 – 87

x(ʲ) {j,ɣ} → h Ø / _#
xʲʔ jʔ → h ʔ / _#
{x,ɣ}ʔ → g
Vnʔ → V[+ nas] / _#
n → ⁿd
{s,z}(ʔ) {ʃ,ʒ}(ʔ) {ɬ,ɮ}(ʔ) → s ʃ ɬ / _#

[Pogostick Man]

Information in this post is from Brown, Cecil H., David Beck, Grzegorz Kondrak, James K. Watters, and Søren Wichmann, "Linking proto-Totonacan and proto-Mixe-Zoquean". <http://www.ualberta.ca/~dbeck/TzEILNXI.pdf>

Proto-Totozoquean to Proto-Mixe-Zoquean

l → j
q → ʔ
nʲ tʲ ʃ tʃ kʲ → n t s ts k
x {ɬ,tɬ} → h j
V̰ → V
ɨ ɔ → ə o

Proto-Totozoquean to Proto-Totonacan

ʔ → Ø
nʲ tʲ kʲ k → l tʃ k q
h → Ø / ! #_
j → t
o o̰ → u ṵ
{ə,ɔ} {ə̰,ɔ̰} → a a̰
{e,ɨ} {ḛ,ɨ̰} → i ḭ

[Pogostick Man]

Information in this post is from Dwyer, David J. (1987/1988), "Towards Proto Mande Morphology". Mandekan: Bulletin semestriel d'études linguistiques 14/15:139 – 152. He has a lot of changes involving the velar nasal going to various palatal sounds; I'd be inclined to think it was an original palatal nasal but I'm keeping it as it was in the paper.

Proto-Mande to Bobo

p → f
tʼ {kʼ,g} → t k
l → d
z → {s,j}
w → g
ŋ → ɲ
kp gb → k gw

Proto-Mande to Busa

tʼ kʼ → t k
d → l
ŋ → ɲ
kp kpʼ → {kp,k} gb
j → i

Proto-Mande to Dan

tʼ kʼ → t k
d → l
ŋ → ɲ
kpʼ → gb

Proto-Mande to Guro

p f → f v
tʼ kʼ → t k
d → l
ŋ → ɲ
kp {kpʼ,gb} → p b

Proto-Mande to Kono-Vai

p → f
l → d
z → s
{g,w,kp} kpʼ gb → k kp b
tʼ kʼ → t k
ŋ → ɲ

Proto-Mande to Southwest Mande

f → p
{tʼ,d} {kʼ,g} → l k
z → s
w → g
ŋ → ɲ
kp kpʼ gb → {k,B} kp B (not sure whether this is supposed to be /β/ or /ʙ/)

Proto-Mande to Mandekan

p → f
l → d
z → s
{k,g,kp} {kp',gb} → s b
t' k' → t k
ŋ → ɲ
w j → k dʒ

Proto-Mande to Mano

f → v
tʼ kʼ → t k
d → l
ŋ → ɲ
kpʼ → gb

Proto-Mande to Mwa

f → v
tʼ kʼ → t k
d → l
ŋ → ɲ
kpʼ → gb

Proto-Mande to San

tʼ kʼ → t k
d → l
ŋ → ɲ
kp kpʼ → k b

Proto-Mande to Sembla

p → f
f → d (yes, really)
tʼ {kʼ,g,w} → {t,d} k
l → d
z → s
gb → b
j → dʒ
ŋ → ɲ

Proto-Mande to Soninka

p → f
tʼ {kʼ,g} → {t,d} k
z → j ?
{w,ŋ} → j
kp gb → k b

Proto-Mande to Susu

p → f
tʼ → t
l → d
z → s
w → x
ŋ → j
kp gb → k b

[----]

Blust, Robert: The Austronesian Languages, Revised Edition (2013)

Proto-Polynesian to Luangiua

h > 0
q > 0
r > l
f > h
k > ʔ
t > k
n > ŋ
w > v
(the resultant consonant inventory is a little strange: /p k ʔ m ŋ l v s h/)
It also seems that ʔ was lost word initially but this could possibly be an artifact of the transcription system.

[dhok]

By the way, for those Micmac changes, note that <q> is actually /x/. This is mentioned in passing in most sources, and I didn't notice it when I sent the paper in question to Pogostick Man.

[Chengjiang]

[The main changes linking PIE to Sanskrit]

This is pretty awesome, but one thing's confusing me. Classical Sanskrit clearly has /l/. Where did it come from, since the original /l/ seems to have rhotacized?

[KathTheDragon]

The general theory is that it's of dialectal origin, seeing as there's no obvious regular conditioning (and, iirc, you can even get variation between r and l in the same word, in a few cases)

[Pogostick Man]

Index Diachronica v.10.0 is out.

Anybody got a better set of changes for Norse? I've been told mine are wrong.

[Ser]

? and Serafín, the former citing Penny, Ralph (2002), A History of the Spanish Language, 2nd Ed. Cambridge University Press; and Lipski, John (1994), Latin American Spanish. Longman Pub Group

The other mysterious editor was Whimemsz. And the one who cited Lipski was me, not him.

[Vijay]

Sorry, I'm not sure whether I'm doing this right, but here's some sound changes from Old Indo-Aryan (OIA) to Romani anyway (not an exhaustive list at all; most of these are from Turner (1926)):

r̥ > i, u
śm, ṣm, sm > mh > m
y > j / #_
m > v / V_V
kṣ > kh
t, d > l / V_V
ś, ṣ > ś
st(h), ṣṭ(h) > Romani št

[2+3 clusivity]

Since I have Romani: a Linguistic Introduction (Matras, 2002), I will add to Vijay's post. Matras draws on Vijay's source and a few others. I also added his comparative notes to Domari and Lomavren. The changes are in no particular order unless noted otherwise.

Shared with Central Middle Indo-Aryan (MIA) Languages

(1) /r̥/ > /i, u/; but /er/ in /kerd/ < /r̥/, c.f. Domari /kard/ "done")
(2) /sm/ > /mʱ/ > /m/
(3) /y/ > /j/ / #_
(4) /p/ > /v/ / V_V
(5) dropping of various singleton medial stops; e.g. OIA /bʱagini:/, MIA /bʱahiɳi:/ v. Romani /pʰen/, Domari /be:n/
(6) right to left assimilation of consonant clusters into geminates followed by shortening; e.g., OIA /rakta/, MIA /ratta/, Romani /rata/, Hindi /ra:t/ "Blood." Unlike NIA languages, like Hindi, compensatory lengthening does not (currently?) appear.''
(7) /tm, tv/ > p. E.g. OIA /a:tman/, MIA /appa:/, Romani /pe(s)/ "reflexive pronoun." The later has a few exceptions perhaps protected by larger clusters. E.g. /catvari/ > /ʃtar/ "four."
(8) /kʂ/ > /kh/, E.g. OIA /akʂi/, MIA /akkʰi/, Romani /dʒ-akʰ/ "eye," omari /iki/, Lomavren /aki/; but > /tʃʰ/ on occasionally or perhaps as a loan: /kʂurika:/, MIA /cʰuri:/, Romani /tʃʰuri/, Domari /tʃuri/ "Knife."
(9) /stʰ/ > tʰ / #_. E.g. OIA /stʰu:la/, MIA /tʰulla/, Romani /tʰulo/ "fat"

Sound Changes Shared with Central NIA Dialects

(1) /u/ > /o/ in some cases which the source is not clear on.
(2) /v/ > /b/ / #_. This is clearly an independent innovation as it is not shared with Domari (> /w/) and Lomavren (/v/ retained).
(3) /m/ > /v/ / V_V. Likely independent, Domari appears to retain medial /m/.

Changes diverging from Central MIA/NIA Dialects

(1) dental stop + /r/ clusters retained. Similar to NW NIA languages including Dardic languages. Shared with Domari and no comparable material to prove if shared with Lomavren.
(2) labial stop + /r/ clusters retained. Similar to NW NIA languages BUT not shared (currently) with Domari or Lomavren.
(3) /st/ > /st/ / V_V. E.g. OIA /hasta/, Romani /vast/, but Hindi /ha:tʰ/ "hand."
(4) /ʂʈ/ > /ʃt/ / V_V. E.g. OIA /miʂʈa/, Romani /miʃto/, but MIA /miʈʈʰa/, Hindi /mi:ʈʰa:/ "good." Comparable to Domari and Kashmiri.
(5) final dentals preserved as /l/. E.g. /pʰral/ "brother." Domari and Lomavren instead preserve as /r/; at least a few Dardic languages also preserve as /l/
(6) /ʃ, ʂ/ > /ʃ/
(7) /rs, rʃ/ > /s, ʃ/

Various other changes (the whole dental and retroflex stop thing is a major mess and reflexes shared/different with Lomavren and Domari suggest shifts outside of the MIA/NIA context but nevertheless along similar lines)

(1) /d/ > /d/ #_.
(2) /ɖ, d>ɖ/ > <ř> #_ . A rhotic or retroflex contrasting with /r/, i.e. [rr, ʀ, etc.]
(3) /t, tʰ, d, dʱ/ > /l/ / V_V (also note the rule above re final dentals.)
(4) /ɖ/ > /l/ / V_V and where reflected as /l/ in MIA languages. ??? Personally, given the rules below and above, this suggests some /ɖ/ resolved to /*d/ then /l/ or some kind of loan word situation. Shared with Domari and Lomavren.
(5) /ʈ>?ɖ, ʈʰ>?ɖʱ, ɖ, ɖʱ/ > /r/ / V_V
(6) /ʈ>ʈʈ, ʈʈ, ɖɖʱ/ > <ř> / V_V. A rhotic or retroflex contrasting with /r/, i.e. [rr, ʀ, etc.]
(7) ɳɖ > <ndř, nř, ř, ř>r, řn>, /nd, ngl, n/. C.f. /nd/ > /n/
(8) /ʈʈʰ/ > /ʃt/ in a few OIA sandhi contexts; e.g. /*ud+stʰa/ gives OIA /uttʰa/.
(9) /c/ > /ʃ/ / before voiceless stops. E.g. /catvari/ > /ʃtar/ "four."
(10) /gn/ > /g/
(11) non-retroflex, velar groups of Nasal + Stop[unvoiced] > Nasal+ stop[voiced]
(12) /ʂk/ > /k/ / internally only?
(13) Stop[voiced] > Stop[breathy voiced] / #_VC[breathy voiced]. This rule feeds into the next rule.
(14) Stop[breathy voiced] > Stop[voiceless, aspirated]. This later rule is shared with Domari and Lomavren BUT those languages do not show the effect of the rule immediately above.