Wiktionary:About Proto-Indo-Iranian

These are some guidelines for Proto-Indo-Iranian, the ancestor of all Indo-Aryan, Iranian, and Nuristani languages. See also:

Consonants

 * Labials – p, b, bʰ, m
 * Dental/Alveolar – t, d, dʰ, s (allophone: z), n, l
 * Post-alveolar – š (allophone: ž), r, r̥
 * Palatals
 * First – ć, ȷ́, ȷ́ʰ (> Sanskrit: ś, j, h, Avestan: s, z, z)
 * Second – č, ǰ, ǰʰ (> Sanskrit: c, j, h, Avestan: č, ǰ, ǰ)
 * Velar – k, g, gʰ
 * Laryngeal – H (< all PIE laryngeals)
 * Semivowels – y, w

In some transcription schemes, *y and *w are transcribed as *i̯ and *u̯. The first set of palatals is sometimes written with circumflex accents. Such entries should be moved to the correct spelling. Redirects can remain. The status of *l as a phoneme separate from *r is still being debated.

Laryngeal
If a laryngeal is reconstructed in the PIE form, then *H should also be in the PII form (e.g. *Hnā́ma, not **nā́ma). Otherwise, long vowels can be used.

The laryngeal from PIE can be vocalized sometimes, resulting in *i or zero. Read Wikipedia for a full description.

Vowels

 * Short – a, i, u
 * Long – ā, ī, ū
 * Diphthongs – ay, āy, aw, āw

Pitch accent
Marked by an acute accent. All vowels, *r̥, and *H between consonants can take the accent.

Descendants
Always include a derived terms or descendants section. There are plenty of resources available that document cognates between Indo-Aryan and Iranian languages (especially Sanskrit and Avestan), so this shouldn't be difficult for a valid reconstruction. The primary exception is Nuristani, which is still poorly described, but recent work has been done to better identify Nuristani reflexes of Proto-Indo-Iranian words.