Reconstruction:Proto-Indo-Iranian/-wants

From Proto-Indo-European *-wénts.

*-wants

1. Forms possessive adjectives from nominal stems.

The declension of this suffix does not straightforwardly descend from the declension of *-wénts. Most saliently, the nominative singular in Sanskrit has a long vowel, and the suffix's nominative singular also manifests quite erratically in Avestan. The following schools of thought exist for how this situation evolved:

• Hoffmann and Lipp believe that regular sound changes can yield *-wénts > *-wénss > *-wḗns > *-wḗs > *-wā́s in the nominative, specifically via a combination of Szemerényi's law and Stang's law.^[1] This theory, however, cannot explain why participles like सत् (sat) underwent no such development.
• It is alternatively possible to start from original *-wants that was only later contaminated by reflexes of the perfect active participle suffix *-wās.^[2]
  • De Vaan states that "in *-mant- and *u̯ant-stems, analogy to the stems in -uuāh/-uš has induced the substitution of the [nominative singular] in *-u̯ant-s (e.g., cuuąs 'how much') by *-u̯ās and *-mās (e.g., astuuā̊ 'bone (→ material)', xratumā̊ 'intelligent')."^[3] It is unclear whether De Vaan also meant to encompass the possessive *want-stems in his explanation.
  • A contamination-based explanation has the advantage of allowing for a reason to why present participles do not behave similarly in Sanskrit (the possessive suffix and perfect active participle begin in *-w- while the participle suffix does not).

• Proto-Indo-Aryan:
  • Sanskrit: -वत् (-vat), -मत् (-mat)
    • Pali: -vant
    • → Bengali: -বান (-ban) (from the masculine singular nominative -वान् (-vān))
    • → Hindi: -वान (-vān) (from the masculine singular nominative -वान् (-vān))
• Proto-Iranian: *-wāh (from either contamination or the original inherited outcome), *-wānh (following Lipp, with strong stem restored)
  • Avestan: -𐬎𐬎𐬀𐬥 (-uuan) (for example: 𐬀𐬱𐬀𐬎𐬎𐬀𐬥 (ašauuan, “follower of order”)
  • Persian: وند (vand)