Indo-European *y / *d^, Celtic *y / *d(y), Greek h / z

Sean Whalen
[email protected]

March 29, 2025

PIE *y > Celtic *y / *d(y) seems to exist (*d(y) as an interim notation for the sound that became
Brythonic *d > W. dd, but PI *dy, in *-yos > *-dyos > OI -de, etc.), and multiple outcomes of *y
are also found in many IE. It was not fully regular, but was much more common in Brythonic
than Irish. It created variants like :

*kH2ariyaH2- > W. caredd, OI caire ‘crime, sin / fault, reproach’

*taH2-ye- > *ta:de- > W. toddi ‘melt / dissolve’, *taH2-ye-ti > OI taḯd

*newiyo- ‘new / young’ > Gl. novio-, OI nóe; *nowid(y)o- > W. newydd, OI núide

*wiH1riyo- > S. vīryá-m ‘manliness/strength/valor/heroism’, *wiryo- > OI ferd(a)e ‘male’, MW

gwrdd ‘strong’

*wiH1ri-H2ag^-tro- ‘leader of fighting men’ > ?Ls. *wiryaxtos >> L. Viriat(h)us ‘a Ls. chief’, G.
Ouríatthos \ Uriatthos, W. Gwriad, OI Ferdiad

The accent was not a fact, and in some cases is unknown, but Morris-Jones tried to fit these into
a regular scheme based on accent (likely following Verner, and many attempts afterward to find
environmental causes have often started the same way, even when the accent was unknown or
against the proposed change, with many cases of analogy proposed for whichever examples went
against the author’s theory). I reproduce his arguments, which include other valuable data :
>
100. iii. (1) In Brit., in the diphthong ii̯ (ei̯ , ai̯ ), when accented or following the accent, i̯ became
a spirant probably like French j, which became ẟ, and appears so in W. Thus -íi̯ os > -yẟ, -íi̯ ā >
-eẟ; ´-ii̯ - > -oeẟ § 75 iv. But the change did not take place in oi̯ or īi̯ .

(2) The same change took place after l or r following the accent; thus ´li̯ > *lẟ > W. ll; and ´ri̯ >
*rẟ ≡ W. rẟ. Examples: li̯ : W. gallaf ‘I can’: Lith. galiù ‘I can’;—W. all- in all-fro ‘foreigner’,
Gaul. Allo-broges < *ali̯ o-: Lat. alius, Gk. ἄλλος < *álios;—W. gwell ‘better’: Skr. várya-ḥ
‘eligible’, várīyān ‘better’: O. E. wel, E. well, orig. ‘choice’, √u̯ el- ‘wish’.—ri̯ : W. arddaf ‘I
plough’: Lith. ariù ‘I plough’, Goth. arjan ‘to plough’;—Pr. Kelt. Iu̯ ér-i̯ on-, -iann- > W. Iwerddon
‘Ireland’, Ir. gen. Ērenn;—W. morddwyd ‘thigh’: O. H. G. muriot ‘thigh’;—W. hwrdd ‘a violent
push’ < *spuri̯ - (ur < u̯ ₑr § 63 viii (1)) √sphu̯ erē- ‘hurl, smite’ § 96 iv (1): Lith. spiriù ‘I kick’ (ir <
ₑr § 63 iii); also possibly W. g-ordd fem. ‘mallet’ (g- excrescent § 112 ii (2)), O. W. ord ox. 2,
Bret. orz < *púri̯ -ā ‘smiter’: Gk. σφῦρα ‘mallet’ < *σφυρι̯ ᾱ; in that case Ir. ordd is from British (a
Bret. orz < *púri̯ -ā ‘smiter’: Gk. σφῦρα ‘mallet’ < *σφυρι̯ ᾱ; in that case Ir. ordd is from British (a
not improbable borrowing, cf. Pedersen Gr. i 22–4).

(3) The change of i̯ to *ẟ in the above cases took place before the Roman period, for there is no
example of it in any word borrowed from Lat. The alteration was therefore earlier than the period
of vowel affection, and the *ẟ could not affect; hence arẟaf, not *eirẟaf, etc.The fact that the
change does not take place initially corroborates the view that it did not happen before an
accented vowel. All forms that occur can be explained under this supposition; thus all- < *áli̯ o-,
but ail ‘second’ < *ali̯ ós, etc.; see § 165 vi.
>

Based on some Celtic showing *d > *dz, written with the double-axe (also called san) also used
for *ts (Stifter 2024), it is likely that a related explanation exists for *d^ > *dz^ > *d(y) > *d /
*dy, with 2 elements needed. I propose that there was free variation between *y / *d^ at an old
stage, likely *d^ / *dz^ in some dia., which would violate no laws of regularity. Only later, when
*y > y but *d^ > *d in Brythonic, *d^ > *dz^ > *dy in OI (or similar), would it appear to split
without cause. Since *d(h) > d & some *y > d would cause a partial merger, the older stage
where *y / *d^ did not encroach on any other phoneme is not visible without analysis. In the
Italo-Celtic hypothesis, some Latin *y > d, *gWHen-yo- ‘drive away, beat (off), hit’ > of-fendō ,
G. theínō, Li. geniù ‘lop off’ (Sihler 1995), would be strong support if this variation were
recognized.

This can not be easily separated from *y > dz \ zd = G. z (zeta). This also is matched by the
great variation in Ar. *y > y / ǰ / ž / l. Some say *y > h vs. *y > z in Greek was fully regular &
came from different Hy-, etc. This is shown to be false from Ar. also having multiple outcome of
*y but the cognates not matching in terms of which appeared where (*ye(H1)kWrt ‘liver’ > S.
yakŕ̥t, gen. yaknás, Ar. leard, G. hêpar; luc, *yugo-m > S. yugá-, E. yoke, G. zugón). Some kind
of optionality probably also led to variants like G. biáō \ biázō ‘constrain/force/etc.’ (more
below). In the same way, since Ph. is known to have *d > t, Ph. *y > *d > t (below) would
match nearby Lydian *y > d (likely pronounced /ð/) in mariwda-, etc. If *y > *y / *d^ at an early
stage in many IE branches, it could explain G. z, Ar. y vs. ǰ, Celt. y / d. Maybe 1st *ð^ if Lydian
*y > d is related, since a path like *y > *γ^ > *ð^ might be simpler, and languages with d / dð / ð
are hardly odd, and *d > l in Lamẽtrus ‘Demeter’ and Lefś ‘Zeus’.

I do not think methodology that shows 2 outcomes is lacking in its strictness. In just such a way,
*t > th / d in English was the proper analysis, it just lacked Verner’s Law to explain the
distribution. A first glance can be true, yet not find all there is to be found. Finding all outcomes
is the 1st step, details about the distribution being caused by regular sound changes is often
found later (much, much later in cases like Winter’s Law). In the same way, *y > y / dd in Welsh
still has no regular explanation that can pick out which words would definitely have y or dd, yet
knowing that *y produced either, not any other C, is very important in its own right. It would be
foolish to ignore the obvious reality because there is (yet) no analog of Verner’s Law.

In these, just as in Celtic, free variation between *y / *d^ at an old stage would violate no laws of
regularity. Only later, when *y > h but *d^ > z would it appear to split (since *s > h, *gy > z,
etc., would mean h / z were not simple variants of only one underlying sound). This simply is a
wider range of outcomes and mergers than in Celtic, since G. had more *C that changed into the
same h, etc. This led to variants like biáō \ biázō Since verbs in -zō have derivatives in -smos,
etc., this also shows that *gWhermn-ye- > G. thermaínō ‘heat’ >> *thermanz^-tro- > thermástrā
‘furnace’ had optional y > z^ ( > s before t , etc.). Others like meidiáō ‘smile’ >> meidiasmós
show the principle. This is not restricted to G., since closely related Phrygian also had it
(*g^h(e)r(i)ye- > *g^hariðe- > G. kharízō ‘show favor / oblige’, kekharisménos ‘agreeable’,
*gharid- > Ph. gegaritmenos ‘*found innocent > judged > judged innocent/guilty’, with the
spread of the bad outcome like E. doom). It would be foolish to ignore kekharisménos :
gegaritmenos, and knowing the source of z / s in G. allows t in Ph. to be compared. This verb
had -y- in other IE, with no ev. for *-t- or *-d- :

*g^h(e)r(i)ye- > S. háryati, G. khaírō ‘rejoice / be glad, L. *xoriye- > horīrī

*g^h(e)r(i)ye- > *g^hariðe- > G. kharízō ‘show favor / oblige’
G. kekharisménos ‘agreeable’, Ph. gegaritmenos ‘*found innocent > judged (innocent/guilty)’

In this way, *maH2n(i)- ‘good’ (L. mānus, im-mānis, i-stem pl. Mānēs) seems to be the source of
Ph. Manes, gen. *Maniyos > Manitos. When several words with *y shown by cognates have -t-,
looking for a common origin is a reasonable step. That so many IE also show *y > d, etc., when
*d > t is already known for Ph., makes it a simple explanation. That G. dia. *dz > z \ d(d) was
reflected in Ph., and also later reg. d > t, shows its age and optional nature (a dia. change in G.
appearing in Ph., of such unusual type, would not be chance; not later areal influence, since d > t
is old and shared with Ar., etc.). This also applied to clusters like *-n(C)y- :

*ny > nǰ
*gWhen-ye- ‘drive away, beat (off), hit’ > L. of-fendō, G. theínō, Li. geniù ‘lop off’, Ar. ǰnǰem
‘destroy/wipe clean’, -ǰinǰ \ -ǰunǰ ‘destroyed’

*ny > yn
*g^hwano\i-? > OCS zvonъ ‘sound’, o-stem, & Ar. jayn ‘voice / sound’, i-stem

*nty > *yny

*g^erH2ont-yo- > Gl. Gerontios, Ar. *ćeroynyo > ceruni ‘old person’
? > Ar. *arkhon-yo- > *arkhoynyo > arkhuni ‘royal’

*nty > *ny

*H1dntyo- > Ar. *zyantyo > *ǰantyo > *žanyo > žani ‘tusk’ [d(h) > z by H (Whalen 2025a), H1 >
y (Whalen 2025b) ]

The change of *zy- > ž- also seen in [with *sm- > *zm- like G.] :

*sm(e)id- ‘smile / laugh’ > G. meidiáō, Ar. *zmid-ye- > *zyimde- > žptim / žmtim ‘I smile’, žpit
‘smile’

Seeing *y > *y / *dy in OI nóe, núide; *y > y / ž in Ar. yolov ‘much/numerous / many people’,
but *žolov- in žołovurd ‘multitude’ makes any attempt at regularity based on folowing vowels,
etc., impossible with the data that exists. This does not mean that each stage was wholly
irregular. I encourage looking for more regular aspects, even if I don’t think all parts will ever be
found regular. I even think I can build on Hrach Martirosyan’s ideas in :
>
2.1.7 PIE *i̯ - > Arm. l-

Examples: leard ‘liver’ vs. Skt. yákr̥ t [sic] etc.; luc ‘yoke’ vs. Skt. yugá-, Lat. iugum, etc.
Different explanations have been offered for these words (see s.vv.). Hamp (1982: 191) assumes
l < [λ] < *[j] < *[i̯ ], “an unspectacular phonetic sequenceknown from current attestation in
dialects of a number of languages”.

The alternation *i̯ - : *l- is reminiscent of the possible correlation seen in designations of
‘elephant’ (see Gamkrelidze/Ivanov 1984, 2: 524-525;Mallory/Adams 1997: 176-177).

In some Armenian dialectal words, we see an initial l- instead of y-, cf. ystak‘pure’ > Muš listag,
hiwsem ‘to weave’ (q.v.) > Łarabaɫ lüsil, yesan ‘whetstone’ > Alaškert, Muš, Sasun lɛsan. In
some cases, contamination is possible. For Łarabaɫ lüsil, Ačaṙyan (HAB 3: 101b) assumes
contamination with PIE *plek̂ - ‘to weave’. Muš listag may be due to the influence of loys ‘light’.
On the whole, however, a phonetic explanation seems more reasonable. It is remarkable that, in
all cases, the first following consonant is the sibilant -s-. Thus, we may be dealing with a sound
change of the type y...s > l...s, which is younger and is hardly related to the cases seen in leard
and luc.

With this hypothetical sound development in mind, one can consider the following possible
example: dial. *liz ‘female buffalo’, in Van [Ačaṙean 1913: 423a] and Moks [HayLezBrbBaṙ 1,
2001: 225b]. NPl liz-n-ir is attested in a Moks version of the famous folk-song “Camt‘el” (see
Šahpazean 1913: 26L-6 and footnote 3). The plural ending -ner (Van and Šatax) : -nir (Moks)
presupposes an older NSg form with -n (see Ačaṙyan 1952: 108; M. Muradyan 1962: 85; M.
Muradyan 1982: 139); cf. Van/Šatax yezner, Moks iznir, the plural of yez (Moks iz) < ClArm.
ezn ‘bullock’. This implies that the older nominative form of the word under discussion would
have been *lezn. One wonders, then, if *lez-n ‘buffalo’ is identical with the synonymous by-form
*ye/iz < ClArm. ezn ‘bullock’. Typologically, compare the above-mentioned ystak, which is
represented in Muš by two forms next to each other: h’istag and listag (see Baɫdasaryan-
T‘ap‘alc‘yan 1958: 266a). Note that here, too, the following consonant is a sibilant, although in
this case it is a voiced one.
>

With these insights, *ye(H1)kWrt > S. yakŕ̥t, Ar. *liard > leard ‘liver’, G. hêpar; *yugo-m > Ar.
luc, S. yugá-, E. yoke, G. zugón might also be caused by following *g^ ( < *g after u) & *H1 (if
= x^ or R^). This would be dms. of 2 palatals, maybe at a stage *d^-K^ > *l^-K^. In a similar
way, *p > *f > *xW > h / w in most, but sometimes > y- / ž- has never been explained.
Martirosyan’s idea that y- is an affix from *en does not explain why an affix that added no
meaning was added so many times specifically to words in *p-. In fact, all ex. are of *p-p/kW/
w/u, which implies that at stage *xW there was *xW-W > *x^-W, with *x^ > h / y / ž later :

*pH2trwyo- > Gr. patruiós, Av. tūirya-, *xWatharwyo > *x^atharwyo > *x^aharwy ? > Ar.
yawray ‘stepfather’ [since *tr > wr, it is hard to know if *w-w existed with *wy > y later, or other
changes related to “fixing” *-rwy-, etc.]

*pltH2u- > Av. pǝrǝθu-, S. pṛthú-, G. platús ‘broad / flat’, *xWalthxu- > *x^althxu- > Ar. yałt`
‘wide / big / broad / spacious / mighty’, yałt`(k)u ‘mighty / victorious’ [*H2 > *x / *k]

*penkWe > OI cóic, Ar. hing ‘5’

*penkWe-dk^omt()- > *xWenxWi:s^ond- > *x^enxWi:s^ond- > *yinxisund- > *yihisun > Ar.
yisun ‘50’

*piH1won- > S. pīvan-, fem. pīvarī-, *xWīwerī > *x^īwerī > *yiweri > *yewri > Ar. yoyr -i-
‘fat’

*plH1u- > G. polús, stem *plH1ew- > *xWolew- > *x^olew- > Ar. yolov ‘many (people)’,
žołovurd ‘multitude’

*p(e)lH1- > Li. pilti , *pelH1neu- > *xW-u > *x^-u > Ar. hełum ‘pour/fill’, _-yełc’ ‘full of _’ (in
compounds)

*yi-pl(e)H1- > S. píprati ‘fill’, G. pímplēmi, *xWif- > *x^if- > Ar. yłp’anam ‘be filled to
repletion / be overfilled’ [*f > *xW likely prevented near *xW]

This does not account for when *x^ > h / y / ž, but it should be clear that it’s better to narrow
down the range of *p > y before attempting to find additional causes, if they exist. By
attempting to disregard variants that are only unexplained at the first stage, the second stage can
never be reached, let alone any third. This manner of thinking would have left Verner without a
job, and no one knew Winter was needed for 200 years. Do not make the mistake of letting clear
variants with environmental causes go unrecognized for so long because we can’t find all
explanations at the first examination.

Kloekhorst, Alwin (2012) The origin of the Lydian dat. sg. ending -λ
https://www.academia.edu/3204833

Martirosyan, Hrach (2009) Etymological Dictionary of the Armenian Inherited Lexicon

https://www.academia.edu/46614724
Morris-Jones, John (1913) A Welsh Grammar, Historical and Comparative: phonology and
accidence
https://en.wikisource.org/wiki/A_Welsh_Grammar,_Historical_and_Comparative/Phonology

Oreshko, Rostyslav (2019) Phonetic value of Lydian letter <d> revisited and development of PIE
dentals in Lydian, Wekwos 4, 2019: 191-262
https://www.academia.edu/39978695

Sihler, Andrew (1995) New Comparative Grammar of Greek and Latin

Stifter, David (2024) More on san in Cisalpine Celtic

https://www.academia.edu/128037763

Whalen, Sean (2025a) Laryngeals and Metathesis in Greek as a Part of Widespread Indo-
European Changes (Draft 5)
https://www.academia.edu/127283240

Whalen, Sean (2025b) Indo-European Roots Reconsidered 9: *H1ek^wo-s ‘horse’

https://www.academia.edu/128170887

Abbreviations

(all others as standard or as given in references)

1s first person singular
3p third person plural
a., acc accusative
abl. ablative
aj. adjective
an. animate
ana. analogy
asm. assimilation
av adverb
bc because (of)
brd borrowed
cau causative
cl clitic
com comparative (degree)
con connected (with)
cont continous
contm contaminate, -ed, -ion]
cp. compound
cpm compound member
d(at). dative
def definite
der derive, derivation
ds desiderative
dsm. dissimilation
ep epigraphic
ety. etymology
ev. evidence
ex. example
exc except
exl exclamation
f. feminine
fu. future
fv. free variation
g(en). genitive
gda. gen+dat+abl
i. instrumental
ij. interjection
imf imperfect (past)
imp imperative
inan inanimate
inf infinitive
ins inscription
int intensive
intr intransitive
irr irregular
l(oc) locative
lw. loanword
m. masculine
md. middle
met. metathesis
n., nom. nominative
na. nom+voc+acc (used of base form of standard IE n.nu.)
nu. neuter
no. noun
ono onomatopoeia
op optative
opt optional(ly)
p plural
pas passive
pc positional consonant
pf perfect
pfp perfect participle
pl pluperfect
pn pronoun
po postposition
pos positive (degree)
poss possible, -ly
pp past participle
pr preposition
pre prefix
ps present
pt past, preterite
red. reduplicated
reg. regular, -ly, -ity
rel. related (to)
s singular
sb subject
sem. semantics
sim. similar (to)
sj. subjunctive
suf. suffix
sup. superlative (degree)
supl. suppletive
tr transitive
trans. transcribed (in)
v. verb
vo. vocative
w with
w/o without
~ a kind of (in def.), akin to, related in some (indirect?) way (in roots)
> to (descends to with sound change(s))
>> loan to
-> derivation to

Sounds (all others as standard or as given in references)

Consonants
C^ palatalized C
N uvular n
R uvular r
X uvular x
! lateral fricative
Vowels
E open e
O open o
U open u
Others
y~ nasalized y
v~ nasalized v
V~ nasalized V
V´ stressed V (or with high tone when appropriate to system)
V. retroflex V
etc

A Atshareetaá / Ashrit (older Palola < *Paaloolaá)

Ab Arabic
Ac Aramaic; JBAc Jewish Babylonian Aramaic
Ak Akkadian
Al Albanian
Ala Alashai dialect of Pashai
Ap Apabhraṁśa (Northern Indic dialects)
Ar Armenian
Aro Aromanian
As Assyrian form of Akkadian
Asm Assamese
Av Avestan
Awn Awāṇkārī dialect of Lahndā
B Bangani
Barg Bargromaṭal dialect of Kati
Bc Bactrian
Be Bengali
Bhal Bhalesī dialect of West Pahāṛī
BHS Buddhist Hybrid Sanskrit
Bi Bithynian
Bih Bihari
Bl Balochi \ Baluchi
Br Breton
Bs bHaṭé-sa zíb \ Bhaṭeri
BS
Bt
Bu Burushaski
C Cassite / Kassite
Ca Caucasian
Car Carian
Ch Chinese (Mandarin)
Co Cornish
CI Celtiberian
Ct Celtic
Cur Curāhī dialect of West Pahāṛī
D Degaanó \ Degano
Dar Darrai-i Nūr language of Pashayi
Dc Dacian
Deo
Dk Domaaki \ Domaá \ Ḍumaki
Dm Dameli
Dv Domari \ Dōmvārī
E English
Et Etruscan
Ete Eteocretan, dia. of or rel. Sardinian
Fc French
Fr Frisian
G Greek
Gae Gaelic
Gau Gauro / Gawro
Gh Garhwali
Gi Gultari
Gj Gujarati
Gl Gaulish
Gmc Germanic
Go Gothic
Gur Gurani \ Gorani; .h = Hawrami dia., .g = Gahwāraī \ Gahvareh
Gw Gawar-Bati \ Gubber \ Narsati
Gy Gypsy (if not specified)
H Hittite
Hi Hindi
Hu Hurrian
I Irish
Id Indus Kohistani
IIr Indo-Iranian
In Indic
Ir Iranian
Is Ishkashimi
It Italic
Itn Italian
J Japanese
Jaun Jaunsārī dialect of West Pahāṛī
K Korean
Ka Kalam Kohistani / Kalami / Gawri / Bashkarik / Daraaki
Kd Kurdish
Kh Khowàr
Khet Khetrānī dialect of Lahndā
Kho Khotanese
Ki Kati
Kkb Kok Borok \ Tripura
Km Kashmiri
Kmd Kāmdesh dialect of Kati
Ks Kalasha / Kaláṣa [Dardic]; Ks.b. = Birir dialect, Ks.u. = Urtsun dia., Ks.r. = Rumbūr dia.
KS Kundal Shahi
Kt Ktívi Kâtá Vari / Kâtá-vari
Ktg Koṭgaṛhī dialect of West Pahāṛī
Ku Kusunda
Kum Kumaoni
Kv Kâmvíri
Kva Kvari
Kx Karakhanid
KxM [Dybo’s MK; by Mahmud al-Kashgari, for Turkic in city of Kashgar]
L Latin
Laur Laurowani, NE language of Pashayi
Lc Lycian
Ld Lydian
Lep Lepontic
Lh Lahnda
Li Lithuanian
Lt Latvian
Ls Lusitanian
Lv Lomavren
Lw Luwian
M- Middle (added to others here)
M Mitanni
Ma Marsian
Mj Munji
MHG Middle High German
MI Middle Irish
Mld Maldive
MP Middle Persian
Mrr Marrucinian
Mth Marathi
Mthl Maithili
Mult Multānī dialect of Lahndā
Mw Marwari
Mz Mazanderani
Ni Nišei-alâ
Nir Nirlāmī dialect of Pashai
Np Nepali
NP (New) Persian (Farsi)
NPc North Picene/Picenian
Nur Nuristani / Khafir Group
O Oscan
O- Old (added to others here)
OCS Old Church Slavonic
OE Old English
OHG Old High German
OI Old Irish
OIc Old Icelandic
ON Old Norse
OPr Old Prussian
OP Old Persian
Or Oṛiyā / Oriya / Odia (of Orissa / Odisha)
Orm Ormuri / Bargistā / Baraki
Os Ossetian; Os.d = Digor, Os.i = Iron
OSx Old Saxon
P- Proto-
Pach Pachaghani dialect of Pashai
Pae Paeonian
Pal Palaic
Pc Parachi
Pg Paelignian
Ph Phrygian
Pi Pisidian
Pj Punjabi
Pk Prakrit
Pl Paaluulaá
Po Polish
Pr Prasun
Ps Pashto
Psh Pashai \ Pashayi (see various languages or dialect of Pashai)
R Russian
Rn Runic (Gmc., many ~ ON ?)
Ro Rošanī \ Rushani
Rom Romani; .b = Bohemian dia. (Czech); .bg Bulgarian, .e English, .g Greek, .gm
German, .hu Hungarian, .i Italian, .s Southeast European
Ru Rumanian \ Romanian
S Sanskrit
S BH / BHS Buddhist Hybrid Sanskrit
Sa Saňu-vīri
Sar Sarikoli
Sard Sardinian dia. of Italian
Sc Sicel
SC Serbo-Croatian
Sdh Sindhi
Sem Semnani (NW Iran.)
Sg Sogdian
Sh Shina
Shewa dialect of Pashai
Shm Shumashti
Shu Shughni
Si Sinhalese
Siv Sivand(i) dia. of NP
Sj Sawi \ Savi \ Sauji
Sk Slovak
Sl Slavic
Snj Sanjan dialect of Pashai
SPc South Picene/Picenian
Sr Sardinian (ancient Italic language)
Srb Sorbian: USrb = Upper Sorbian, LSrb = Lower Sorbian
Su Sumerian
Sv Slovene
Sy Scythian
T Tocharian
TA Tocharian A
Taj Tajrish(i) dia. of NP
Tal Talysh \ Taleshi (NW Iran.)
Th Thracian
Tq Tumshuqese
U Umbrian
V Venetic
Vo Volscian
W Welsh
Wg Waigali \ Kalas.a-alâ
Wx Wakhi
Xw Khwarezmian \ Choresmian \ Chorasmian
Y Yidgha
Yg Yaghnobi
Yv Yatvingian \ Yotvingian \ Sudovian

Greek dialects
Aeo Aeolic
Arc Arcadian
Att Attic
Boe Boeotian
Co Coan, Heracleian (Kōs)
Corc Corcyrean
Cr Cretan
Cyp Cypriot
Ion Ionic
LA Linear A
LB Linear B
Les Lesbic
Meg Megarian
Pamp Pamphylian
Pol Polyrrhenian
Sal Salamis
Thes Thessalian
>
Mac Macedonian
Ms Messapic

SCc \ Kartvelian
Gr Gruzhian = Georgian
Lz Laz = Chanuri = č'anuri
Mg Megrel = Margal = Mingrelian
Sn Svan

Uralic

PU Proto-Uralic

FU Finno-Ugric >

FL Finno-Lapp >>
Sm Saami \ Lapp
Finnic >>>
F Finnish
Ingrian / Izhorian
Kl Karelian
Liv Livonian
Vp Veps
Vod Votic \ vadʹdʹa \ vodskij
Es Estonian

Volgaic >>
Mr Mari / Cheremis
Mv Mordvin / Erzya
Mh Moksha

Permic >>
Ud Udmurt / Votyak
Z Zyrian / Komi / Komi-Zyrian
Py Permyak / Komi-Permyak
Ugric >>
Hn Hungarian / Magyar
X Khanty / Ostyak
Mi Mansi / Vogul

Other Uralic

Samoyedic >
Nga Nganasan / Tavgi
En Enets
Nen Nenets
Skp Selkup

?>

Yr Yukaghir

Tungusic
Ne (Negidal); Evk (Evenki); Evn (Even); Na (Nanai); Orc (Oroch); Ok (Orok); Ord (Ordos); Ul
(Ulcha); Ud (Udighe); Sol (Solon); Man (Manchu); L-Man (Lit. Manchu); Jur (Jurchen);