Proto-Indo-Iranian is the reconstructed proto-language of the Indo-Iranian branch of Indo-European. Its speakers, the hypothetical Proto-Indo-Iranians, are assumed to have lived in the late 3rd millennium BC, and are often connected with the early Andronovo archaeological horizon.

Proto-Indo-Iranian was a Satem language, likely removed less than a millennium from the late Proto-Indo-European language, and in turn removed less than a millennium from the Vedic Sanskrit of the Rigveda. It is the ancestor of the Indo-Aryan languages, the Iranian languages, the Dardic languages and the Nuristani languages.

• 1 Descriptive phonology
  • 1.1 Two palatal series
  • 1.2 Laryngeal
  • 1.3 Accent
• 2 Historical phonology
• 3 Subsequent sound changes
• 4 See also
• 5 Notes
• 6 Bibliography

alveolar

alveolar

voiceless

voiceless

In addition to the vowels, *H, and *r̥ could function as the syllabic core.

Proto-Indo-Iranian is hypothesized to contain two series of stops or affricates in the palatal to postalveolar region. The following table shows the most common reflexes and origins of the two series (Proto-Iranian is the hypothetical ancestor to the Iranian languages, including Avestan and Old Persian):

Proto-Indo-European is usually hypothesized to have three to four laryngeal consonants, each of which could occur in either syllabic or non-syllabic position. In Proto-Indo-Iranian, the laryngeals merged as one phoneme /*H/. Beekes suggests that some instances of this /*H/ survived into Avestan as unwritten glottal stops.

Like Proto-Indo-European and Vedic Sanskrit (and also Avestan, though it was not written down), Proto-Indo-Iranian had a pitch accent, indicated by an acute accent over the accented vowel.

The most distinctive phonological change separating Proto-Indo-Iranian from Proto-Indo-European is the collapse of the ablauting vowels *e, *o, *a into a single vowel, Proto-Indo-Iranian *a (but see Brugmann''s law). Grassmann''s law, Bartholomae''s law, and the Ruki sound law were also complete in Proto-Indo-Iranian.

A fuller list of some of the hypothesized sound changes from Proto-Indo-European to Proto-Indo-Iranian follows:

• The Satem shift, consisting of two sets of related changes. The PIE palatals *k̂ *ĝ *ĝʰ are fronted or affricated, eventually resulting in PII *ĉ, *ĵ, *ĵʰ, while the PIE labiovelars *kʷ *gʷ *gʷʰ merge with the velars *k *g *gʰ.

• The PIE syllabic liquids *l̥, *r̥ merge as *r̥.

• The PIE syllabic nasals *m̥ *n̥ merge with *a.

• Bartholomae''s law: an aspirate immediately followed by a voiceless consonant becomes voiced stop + voiced aspirate. In addition, dʰ + t > dzdʰ.

• The Ruki rule: *s is retracted to *š when immediately following *r *r̥ *u *k or *i. Its allophone *z likewise becomes *ž.

• Before a dental occlusive, *ĉ becomes *š and *ĵ becomes *ž. *ĵʰ also becomes *ž, with aspiration of the occlusive.

• The sequence *ĉs was simplified to *šš.

• The "second palatalization" or "law of palatals": *k *g *gʰ develop palatal allophones *č *ǰ *ǰʰ before the front vowels *i, *e.

• Brugmann''s law: *o in an open syllable lengthens to *ō.

• The vowels *e *o merge with *a. Similarly, *ē, *ō merge with *ā. This has the effect of giving full phonemic status to the second palatal series *č *ǰ *ǰʰ.

• In certain positions, laryngeals were vocalized to *i. This preceded the second palatalization.
  • Following a consonant, and preceding a consonant cluster

• Following a consonant and word-final

• The Indo-European laryngeals all merged into one phoneme *H, which may have been a glottal stop. This was probably contemporary with the merging of *e and *o with *a.

• According to Lubotsky''s Law, *H disappeared when followed by a voiced nonaspirated stop and another consonant:

Among the sound changes from Proto-Indo-Iranian to Indo-Aryan is the loss of the voiced sibilant *z, among those to Iranian is the de-aspiration of the PIE voiced aspirates.