Abstract

In Tarkasaṁgraha (TS), Annaṁbhaṭṭa defines inferential cognition (anumiti) as that cognition which results from parāmarśa. In his own commentary on TS, namely Tarkasaṁgraha-dīpikā (TSD), Annaṁbhaṭṭa shows an over-coverage of the TS definition of anumiti in ‘perception after doubt.’ And then in TSD, he moves on to remove the over-coverage by qualifying the definition (of anumiti) with pakṣatā. In this article, we will raise an objection against the strategy Annaṁbhaṭṭa chooses to remove the above-mentioned over-coverage. We will argue that qualifying the definition (of anumiti) with pakṣatā does not remove the over-coverage in question. Following that an alleged solution to the over-coverage will be presented, and subsequently, we will show why the alleged solution is also flawed. Next, we will return again to defend Annaṁbhaṭṭa’s strategy of eliminating the over-coverage. Despite the defense, eventually, it will be suggested that qualifying the definition of anumiti with pakṣatā does not remove the over-coverage. Hopefully, this article will bring to light one controversial aspect of the definition of anumiti (as rendered in TS and TSD) which perhaps remains unnoticed.