Nyāya Sūtras 94

The status of perception as inference is maintained by relying on the grasping of a single part. And that—

[It is] not [so], because there is apprehension by perception of whatever amount [is perceived].

Perception is not inference. Why? Because apprehension [occurs] by perception itself. That which is relied upon as the grasping of a single part is apprehension by perception. And an apprehension is not without an object. And whatever collection of things is its object, that much, being acknowledged, establishes perception. What distinct collection of things is there other than that—a whole, or an aggregate? Nor can the grasping of a single part be established as an inference, owing to the absence of a logical reason.

And on other grounds as well, there is no consequence of perception being inference, because [inference] is preceded by it. Inference is preceded by perception: for one who has seen fire and smoke connected through perception, an inference of fire occurs through the perceptual sighting of smoke. And both the perception of the connected inferential sign and signified, and the perceptual grasping of the inferential sign alone—without these, the operation of inference does not occur; but this [latter] is not an inference, because it arises from contact between a sense organ and an object. Nor does an inference occur through contact of the inferable object with a sense organ. This is the great distinction between the defining characteristics of perception and inference that must be maintained.

Nor is there apprehension of only a part, because of the real existence of the whole. And it is not merely apprehension of a part. What then? Apprehension of a part and apprehension of the whole accompanying it. Why? Because of the real existence of the whole. For this whole, which is distinct from the part, exists; it is untenable that there should be non-apprehension, when a part is apprehended, of that [whole] which resides in the parts and has attained the causes of apprehension. If [it is argued that this is so] because of non-grasping of the entire thing—[we reply:] no, because there is no other part than the causes [themselves]. The parts are not entirely grasped because other parts are screened off by parts themselves;

[and if you say] the whole is not grasped entirely because it is not exhausted in the grasped parts, and thus this apprehension of a single part does not cease at all—[we reply:] indeed, 'entire' occurs when there is nothing remaining; 'not entire' occurs when there is a remainder. And that applies to the many parts, because of grasping when there is no screening and non-grasping when there is screening. Come now, being asked, please explain: of the whole being grasped, what do you consider to be ungrasped,

by which there could be apprehension of [only] a part? For it has no parts other than its causes [the parts]; therefore, the condition of being a part does not apply there.

This is its condition: with those parts whose grasping occurs through contact with a sense organ, it is grasped along with them; with those parts whose non-grasping occurs through screening, it is not grasped along with them. And there is no difference caused by this.

Either the tree would be an aggregate having no remainder of the collection, or it would be their contact; in either case there is an absence of grasping. Either the tree would be an aggregate having no remainder of roots, trunk, branches, leaves, and so on, or it would be the contact of the aggregated constituents; in either case, grasping of the tree consisting of an aggregate is not tenable. First, grasping without remainder is not tenable because other parts are screened off by parts. Grasping of contact is also not tenable, because of the non-grasping of the possessors of contact. This cognition of a tree accompanied by the grasping of a single part is plausible on the origination of a distinct substance, not on a mere aggregate.