3.4. Indo-Iranian

3.4.1. Indo-Iranian evolution

Important phonological developments from Pre-Proto-Indo-Iranian up to Proto-Indo-Aryan (PIAr.) and Proto-Iranian (PIr.) include:

·       Pre-Proto-Indo-Iranian:

o   Late DIE laryngeal developments:

§  Lengthening of colourised vowels before and after merged laryngeal and progressive laryngeal loss (see below §II.2. Laryngeal evolution).

§  Interconsonantal *h → *i, although Iranian shows some exceptions in initial syllables. For other evolutions of the merged laryngeal *h specific to Pre-PIIr, see e.g. Lubotsky (2018).

o   Early satemisation trend:

§  Velars become palatovelars, with exceptions (e.g. before *r): *k→*kj, *g→*gj, *gh→*gjh.

§  Labiovelars become (and merge with) plain velars: *kw→*k, *gw→*g, *gwh→*gh.

o   Brugmann’s Law: *o in an open syllable lengthens to *ō; *dohtór-m̥ → Pre-PIIr. *dōtṓr-m̥.

o   Merge of syllabic * → *. Trend to the merge *l → *r too, with exceptions found in PIAr. and PIr.

·       Proto-Indo-Iranian:

o   ruKi-rule: phonetic retraction of sibilant to *š (or *ž) after *r, u, K, i; it becomes phonemicised only in the separate branches. *š remains a marginal phoneme in Proto-Indo-Iranian.

o   Bartholomae’s Law (continues in PIAr. and PIr): an aspirate immediately followed by a voiceless consonant becomes voiced stop + voiced aspirate, cf. DIE *ubhto- → Pre-PIIr. *ubdho- ‘woven, made of woven material’; *augh-tá- → Pre-PIIr. augdhó- ‘said’. In addition, DIE *dh+t → Pre-PIIr. *dzdh; as, DIE *r̥dhto- → Pre-PIIr. *r̥dzdho-, ‘complete, mature’.

o   Satemisation (law of palatals):

§  Palatalisation of palatovelars: *kj →*ć, *gj →*j́, *gjh →*h.

§  Plain velars, when before a front vowel (*i or *e) or the glide *, are palatalised to affricates: *k→*č, *g→*ǰ, *gh→*ǰh; as, *ke → *če ‘and’, *eti → *ǰīeti ‘lives’, *ghéntiǰhanti ‘slays’.

o   Merge of Pre-PIIr. *e, o → PIIr. *a, Pre-PIIr. *ē, ō PIIr. .

o   General * → *a. Also, * → *a, except in word initial position before resonants *mnV-, *mrV, etc.

o   Trend *→ *v (continues in PIAr. and PIr.).

o   Loss of laryngeal remains, likely the glottal stop (continues in PIAr. and PIr.?).

·       Proto-Indo-Aryan:

o   *ć→*ś, *→*ź, *h→*źh. Merge of *-źdh- < *-źh-t- with *ždh < PIE *sdh after *i, *u or *.

o   *s and *z between stops are lost, including “intrusive” *s inserted between two heteromorphemic dental stops, e.g. DIE *id-s-tó- ‘seen’ PIAr. *vit-tá-.

o   PIIr. *h and *ǰh, and sporadically other voiced aspirates, were debuccalised and became /h/.

o   PIIr. *š (or *ž) from the ruKi rule (and from *ć, * before voiceless dental stops) phonemicises and becomes retroreflex *.

o   PIIr. *n →* when immediately followed by a retroreflex stop.

·       Proto-Iranian:

o   Voiced aspirated plosives PIIr. bh →*b, dh →*d, h→*, gh →*g.

o   PIr. voiceless fricatives *f, *ϑ, and *x as a result of spirantisation of consonant clusters, of voiceless aspirated stops ph, th, kh, assimilation of aspiration, or PIIr. *s before * in initial position.

o   PIIr. *s →*h in most positions. A source of PIr. *s < PIIr. *sć.

o   Double dental law: the “intrusive” *s is kept, the first dental is lost (ruKi-rule does not affect it, cf. Av. cisti- ‘insight’, OInd. cittí-).

o   Phonemic *š is found as a result of PIIr. *ćs or the groups PIE *ć/ + t.

o   After a labial *s →*š (or *ž after a voiced aspirate), e.g. PIIr. *drapsa ‘banner’ in Av. drafša, MP drafš.

3.4.2. Influence from Uralic

There are some obvious phonetic similarities between Uralic and Proto-Indo-Iranian, in contrast with Late PIE. The eastern LPIE dialect seems to have undergone multiple processes of phonetic uralisms, since Uralic languages had, in general, plenty of palatalised sounds, but only voiceless stops (Kallio 2001):

·       Satemisation, i.e. affrication and assibilation of stops by front vowels: LPIE *K pronounced through PU *ć, , *j.

·       Delabialisation of labiovelars: LPIE *Kw through PU *k.

·       Loss of sonorisation and aspiration: LPIE *T, *D, *Dh through PU *T.

·       ruKi-rule, as part of the adoption of affricates and sibilants typical of Uralic languages.

·       The vocalic merge from Pre-PIIr. to PIIr, i.e. *e/o→*a, *ē/ō→*ā, seems to have a parallel in the evolution from PFU to Proto-Ugric: *ō → *a, *ē → *ä,  *o → *a (in most cases, with some cases of *a → *o), and also *e → *i under certain circumstances (Häkkinen 2009).

Uralic lexical influence on Indo-Iranian is underresearched, with continued contacts with Proto-Iranic[xv]. The direction of the borrowing is sometimes unclear (Kümmel 2019):

·       PU *weŋćä ‘knife’ ~ PIIr. *wā́ćī- ‘axe’.

·       PU *pëŋka ‘mushroom’ ~ PIIr. *bhanga- ‘narcotic’.

·       Substitution of Indo-European reflexive *s(- (which loses its nature as strict reflexive marker in Indo-Iranian languages) for, among other possibilities, the reflexive pronoun PIIr. *tanū́- ‘self’ < ‘body’ (Orqueda 2018), comparable with the use in Samoyedic and Finno-Ugric languages of ‘body’ as reflexive pronoun (see below §3.5.1. Disintegrating Uralic evolution and §4.19.2. Samoyedic evolution).

·       Isolated Indo-Aryan form *aika- ‘one’ (with ancient PIA vocalism, as in Mitanni Aryan aika-), different from Proto-Iranian *aia- (cf. Gk. *oio-), shows a striking parallel with Ugric forms in ek-, uk-, ak- derived from PFU *ükti ‘one’ (Fournet 2010). Since a borrowing into Finno-Ugric in *-t- is not likely, and the form cannot be reconstructed for an older Indo-European stage (suffixes like LPIE *-ko- are added to the stem, not the root), a contamination from Finno-Ugric (or Ugric) into Late PIIr. or Early PIA is a plausible explanation.

·       P(F)U *kota ‘hut’ ~ PIr. *kata- ‘house’.

·       PU *kaða ‘female animal’ ~ PIr. *gadā- ‘bitch’, PIr. *kaϑā- ‘female donkey’.

·       PFU *katV- ‘steal, thief’ ~ PIr. *gada- ‘robber, thief’

·       PU *kala ‘fish’ ~ PIr. *kara- ‘big fish’

3.4.3. Asian agricultural substratum

Loanwords of a non-Indo-European language, attributed to the close contact of Proto-Indo-Iranian speakers with the Bactria-Margiana Archaeological Complex (BMAC), whose language may have been related to the modern Burushaski language isolate, are dated to the period between the Pre-Proto-Indo-Aryan/Proto-Finno-Ugric contacts and the appearance of Indo-Aryan words in Mitanni (Witzel 2003; Pinault 2003, 2006; Parpola 2015).

The following are examples of the adopted words (Lubotsky 2001, 2010):

·       Construction: *ʰarmii̯a- ‘(permanent) building’; *sikatā ‘sand, gravel’, *išta- ‘brick’, although for a potential IE etymology see Mayrhofer (2005) and (Adams 2013); *maūkha ‘wooden pin’.

·       Land cultivation: *i̯awīi̯a- ‘irrigation channel’; *čāt ‘well’; *kʰā- ‘wellspring’.

·       Religion: e.g. gods *Indra- (also attested in Mitanni), *Ćaru̯a, *Ghandaru̯a; *atʰaru̯an- ‘priest’; *magha ‘ritual offering, sacrifice’; *r̥ši- ‘seer’; *anću- ‘Soma plant’.

·       Local fauna: *uštra- ‘Bactrian camel’; *kʰara- ‘donkey’; *kaći̯apa- ‘tortoise’; *kapauta ‘dove’, aj́ʰuka ‘hedgehog’, *matsa ‘fish’, etc.

·       Social life: hairstyles (*kaića-/*gaića-, *stuka), dress (*atka- ‘cloak’, *paastā ‘cloth’), utensils (*kapāra ‘dish’, *naii(s) ‘skewer’), etc.

The phonological and morphological features of the dozens of proposed Indo-Iranian loanwords are strikingly similar to those of loanwords found only in Sanskrit (i.e. after Indo-Ayrans had crossed the Hindukush), which suggests that continued contacts can only be traced back to peoples speaking similar dialects in Central Asia (Lubotsky 1999).

3.4.4. Mitanni Indic

Mitanni Indic shows features of dialectal Indo-Iranian or Old Indo-Aryan, such as diphthongs merged in Vedic Indic, e.g. aika- ‘one’ instead of Skt. éka-. Pre-Vedic Indo-Aryan is therefore supposed to have been spoken in the Middle East, strongly linked to the Mitanni state (16th–14th c. BC). Evidence include (Kroonen, Barjamovic, and Peyrot 2018):

·       The text in Hittite CTH 284 dating to the 15th-14th centuries BC gives detailed instructions by “Kikkuli, master horse trainer of the land of Mitanni.” Indo-Iranian (possibly Indo-Aryan) terms include wa-ša-anna- ‘training area’, and a-i-ka, ti-e-ra-, pa-an-za-, ša-at-ta-, na-a-wa-ar-tan-na- ‘one, three, five, seven, nine rounds’.

·       Names of Indo-Aryan derivation among the ruling class of the (mainly Hurrian-speaking) Mitanni population (Mayrhofer 1982; Witzel 2001).

·       Indo-Aryan adjectives denoting horse colours are from the texts of the provincial town of Nuzi on the eastern frontier of Mitanni, including pabru-nnu- ‘reddish brown’, parita-nnu- ‘gray’, pinkara-nnu- ‘reddish brown’ (Mayrhofer 1982).

·       Divine witnesses of Mitanni in the treaty CTH 51 between its ruler Šatiwazza and Šuppiluliumas of the Land of Hatti include “the Mitra-gods, the Varuna-gods, Indra, and the Nāsatya-gods” (Beckman 2016)

·       Personal names with apparent Indo-Aryan etymologies survived accross a large territory, as far as Nuzi in the east and Palestine in the west.

An early letter from Tell Leilān in Northern Syria dating shortly before the end of Zimri-Lim’s reign in 1761 BCE (Eidem 2014) makes reference to mariannu, which could extend the Indo-Iranian linguistic presence in Syria back two centuries prior to the formation of the Mitanni state. The word is generally seen as a Hurrianised form of the Indo-Aryan word *mara- ‘man/youth’ and taken to refer to a type of military personnel associated with chariot warfare across the Near East (Dassow 2008).

It is believed that the rise to power of an Early Indo-Aryan-speaking elite among a heterogeneous population might have given a mark of elite warrior-class identity to the language and names for dynastic succesors, which survived among certain groups during the Late Bronze Age.

Based on the presence of Mitanni Indic in the 15th century, and on the archaic language inferred from the Rig Veda and the Avesta, it is often assumed that Proto-Indo-Iranian may be dated to the centuries around 2000 BC.


3.4.5. Schleicher’s fable in Proto-Indo-Iranian



óu̯is ékjōs-ke

ou̯is osiās ŕ̥ˀnā ne ēst

ekjons dédorke;

tom gˀúm u̯ógjhom u̯égjhontm̥,

tom megjám bhórom,

tom īróm ōkjú bhérontm̥.

ou̯is ekjobhi̯os éu̯euket:

kjd mō rišeti,

nérm̥ kékkji ékju̯ons ágjontm̥.”

ékjōs éu̯eukont: “krudhí hou̯i!

kd nōs rišeti kékkji,

nēr, pótis, ou̯ii̯ōm ŕ̥ˀnām

su̯ebhi ghermóm u̯estrom kr̥néuti,

ou̯ii̯ōm-ke u̯ŕ̥ˀnā ne hesti.”

Tod kjekjruu̯ós ou̯is ágjrom ébhuget.



áis áćās-ka

áis i̯asiās ŕ̥ˀnā na āst

áću̯ans dádarća;

tam gurúm u̯aham u̯ahantam,

tam ma(h)ā́ntam bharam,

tam īrám āćú bhárantam.

áis áću̯abhi̯as áu̯aučat:

hd mā rišati,

náram čši áću̯ans áantam.”

áćās áu̯aučánt: “ćrudhí ái!

hd nās rišati čši,

nār, pátis, áīnām ŕ̥ˀnām

gharmám u̯ástram tanáu̯ai kr̥náuti,

aīnām-ka u̯ŕ̥ˀnā na asti.”

Tad ćaćruu̯ás áis aj́rám ábhuat.





ávis áśvās-ka

ávis i̯asiās ū́rnā na āst

áśvans dádarśa;

tom gurúm ham váhantam,

tom mahā́ntam bháram,

tom vīrám āśú bhárantam.

ávis áśvabhi̯as ávaučat:

źhd mā riati,

náram čáṭṣati áśvans áźantam.”

áśvās ávaučan: “śrudhí ávi!

źhd nās riati čáṭṣati,

nār, pátis, ávīnām ū́rnām

gharmám vástram tanávai kr̥náuti,

ávīnām-ka ū́rnā na asti.”

Tad śuśruvás ávis aźrám ábhuźat.



ávih áćvāh-ka

ávih i̯ahiāh rnā na āht

áćvanh dádarśa;

tom gurúm am váantam,

tom maā́ntam báram,

tom vīrám āćú bárantam.

ávih áćvabi̯ah ávaučat:

d mā rišati,

náram čašati áćvanh áantam.”

áćvāh ávaučan: “ćrudí ávi!

d nāh rišati čašati,

nār, pátih, ávīnām rnām

garmám váhtram tanávai kr̥náuti,

ávīnām-ka várnā na ahti.”

Tad ćaćruváh ávih aj́rám ábuat.


·       The evolution of * into *v- and the laryngeal loss in *ŕ̥ˀnā must have happened at roughly the same time, given the differing outputs in both PIAr. and PIr. In the PIIr. version of the fable, both traits are therefore left intact.

·       For PIIr. *ma(h)āntam- ‘great, large, big’, cf. Ved. mahā́nt-, YAv. mazā́ṇt-, an enlargement of inherited *ma(h)ā- after *bhrh-ant- ‘high’ (Schmitt 2018).

·       For PIIr. *raiš-, pres. *riš-a- ‘suffer, be hurt’, cf. OInd. riati, Av. raēš-, iriš-.

·       For PIIr. perf. part. *čákšuš < *kwékwkus- ‘looking at something’, cf. Ved. cákšuš-, with a shift in meaning to ‘eye’, similar to derivative in *-men-. More appropriate for the Post-PIIr. stage seems to be the reduplicated desiderative forms in -s- PIIr. *čsati < *kwéks-ti, PIAr. *čáṭṣati, cf. OInd. cákṣate.

PIIr. gen. pl. *-nām comes probably from the extension of feminine gen. pl. **-ā-ām (to distinguish it from the accusative), hence a Proto-Indo-Iranian innovation after the merge of vowels. Stems in -i- and -u- share a long vowel before the new ending, probably due to the generalised (previous allophonic) IE endings in *-i-, -u-, which is not found in genitives made in *-ām. In this precise case of *áu̯i-, the root shares an acrostatic paradigm with other LPIE languages.