[For] linguistic matters [pertaining to an yanfaʿil], see the beginning of “action”.

Terminologically, Āmidi defined [affection] as a “state that obtains in a body because of its being influenced by something other than itself, for as long as this [state of] being influenced is taking place, like “being cooled” and “being heated” (al-Mubīn 113).

Both with respect to existence and nonexistence, the categories of action and affection mututally entail one another. They are counted amongst the relative accidents, the positive existence of which the kalām theologians negated. [An example of] affection is [something] being heated (for as long as it is being heated). The [state of] affection obtains incrementally, which is to say that its state is unfixed. This “being heated” is distinct from heat [itself], because [the latter may] continue to exist in the wake of the [state of] being heated [having passed away]; after this, [a given particular instantiation of] the category of affection will endure no longer. Heat itself is a stable entity from the category of quality, and is distinct from having the capacity to be heated, which is to say that it is distinct from the capacity the thing [capable of] being heated has to receive heat. This is because heat already exists, before the state of being heated [obtains]. The capacity [itself] is, rather, from the category of quality. Because both the categories of action and affection constitute renewable phenomena, which are unfixed, the names “to act” and “to be affected” were chosen for them, rather than “action” and “affection” (Sharḥ al-Mawāqif 5:19-20, Maṭāliʿ al-Anẓār, 72).

The kalām theologians rejected [the notion that] this category is an accident with a positive existence, which, according to the philosophers, [this] as well as the other relational accidents all have. [The kalām theologians] reduced them to pure perspectival entities. They deduced [in the following manner]: “Were [relational accidents] accidents with a positive existence, they would depend upon a locus wherein to subsist, and a relation between them and their loci would have to obtain. The same [argument] would apply to these relations [themselves], just as it did to the [original relations], and an invalid regress ad infinitum would be entailed. These relations, then, have no existence, being rather purely perspectival entities” (Al-Mabāḥith al-Mashriqiyya 1:583, Sharḥ al-Maqāṣid 1:285).