The question however arises that if all apprehensions are valid, how are we to account for illusory perceptions which cannot be regarded as valid? The problem of illusory perception and its psychology is a very favourite topic of discussion in Indian philosophy. Omitting the theory of illusion of the Jains called _satkhyati_ which we have described before, and of the Vedantists, which we shall describe in the next chapter, there are three different theories of illusion, viz. (1) _atmakhyati_, (2) _viparitakhyati_ or _anyathakhyati_, and (3) _akhyati_ of the Mima@msa school. Theviparitakhyati or anyathakhyati theory of illusion is accepted by the Nyaya, Vais'e@sika and the Yoga, the akhyati theory by Mima@msa and Sa@mkhya and the atmakhyati by the Buddhists.
The commonest example of illusion in Indian philosophy is the illusory appearance of a piece of broken conch-shell as a piece of silver. That such an illusion occurs is a fact which is experienced by all and agreed to by all. The differences of view are with regard to its cause or its psychology. The idealistic Buddhists who deny the existence of the external world and think that there are only the forms of knowledge, generated by the accumulated karma of past lives, hold that just as in the case of a correct perception, so also in the case of illusory perception it is the flow of knowledge which must be held responsible. The flow of knowledge on account of the peculiarities of its own collocating conditions generates sometimes what we call right perception and sometimes wrong perception or illusion. On this view nothing depends upon the so-called external data. For they do not exist, and even if they did exist, why should the same data sometimes bring about the right perception and sometimes the illusion? The flow of knowledge creates both the percept and the perceiver and unites them. This is true both in the case of correct perception and illusory perception. Nyaya objects to the above view, and says that, if knowledge irrespective of any external condition imposes upon itself the knower and the illusory percept, then the perception ought to be of the form "I am silver" and not "this is silver." Moreover this theory stands refuted, as it is based upon a false hypothesis that it is the inner knowledge which appears as coming from outside and that the external as such does not exist.
The viparitakhyati or the anyathakhyati theory supposes that the illusion takes place because on account of malobservation we do not note the peculiar traits of the conch-shell as distinguished from the silver, and at the same time by the glow etc. of the conch-shell unconsciously the silver which I had seen elsewhere is remembered and the object before me is taken as silver. In illusion the object before us with which our eye is associated is not conch-shell, for the traits peculiar to it not being grasped, it is merely an object. The silver is not utterly non-existent, for it exists elsewhere and it is the memory of it as experienced before that creates confusion and leads us to think of the conch-shell as silver. This school agrees with the akhyati school that the fact that I remember silver is not taken note of at the time of illusion. But it holds that the mere non-distinction is not enough to account for the phenomenon of illusion, for there is a definite positive aspect associated with it, viz. the false identification of silver (seen elsewhere) with the conch-shell before us.
The akhyati theory of Mima@msa holds that since the special peculiarities of the conch-shell are not noticed, it is erroneous to say that we identify or cognize positively the conch-shell as the silver (perceived elsewhere), for the conch-shell is not cognized at all. What happens here is simply this, that only the features common to conch-shell and silver being noticed, the perceiver fails to apprehend the difference between these two things, and this gives rise to the cognition of silver. Owing to a certain weakness of the mind the remembrance of silver roused by the common features of the conch-shell and silver is not apprehended, and the fact that it is only a memory of silver seen in some past time that has appeared before him is not perceived; and it is as a result of this non-apprehension of the difference between the silver remembered and the present conch-shell that the illusion takes place. Thus, though the illusory perception partakes of a dual character of remembrance and apprehension, and as such is different from the ordinary valid perception (which is wholly a matter of direct apprehension) of real silver before us, yet as the difference between the remembrance of silver and the sight of the present object is not apprehended, the illusory perception appears at the moment of its production to be as valid as a real valid perception. Both give rise to the same kind of activity on the part of the agent, for in illusory perception the perceiver would be as eager to stoop and pick up the thing as in the case of a real perception. Kumarila agrees with this view as expounded by Prabhakara, and further says that the illusory judgment is as valid to the cognizor at the time that he has the cognition as any real judgment could be. If subsequent experience rejects it, that does not matter, for it is admitted in Mima@msa that when later experience finds out the defects of any perception it can invalidate the original perception which was self-valid at the time of its production [Footnote ref 542]. It is easy to see that the Mima@msa had to adopt this view of illusion to maintain the doctrine that all cognition at the moment of its production is valid. The akhyati theorytries to establish the view that the illusion is not due to any positive wrong knowledge, but to a mere negative factor of non-apprehension due to certain weakness of mind. So it is that though illusion is the result, yet the cognition so far as it is cognition, is made up of two elements, the present perception and memory, both of which are true so far as they are individually present to us, and the cognition itself has all the characteristics of any other valid knowledge, for the mark of the validity of a cognition is its power to prompt us to action. In doubtful cognitions also, as in the case "Is this a post or a man?" what is actually perceived is some tall object and thus far it is valid too. But when this perception gives rise to two different kinds of remembrance (of the pillar and the man), doubt comes in. So the element of apprehension involved in doubtful cognitions should be regarded as self-valid as any other cognition.