Sriharsha (d. circa 1225), a philosopher belonging to the non-dualist Vedanta school of Hindu philosophy, rejected, contra the evident facts of experience, real multiplicity or differences in the world. One reason he did so, apparently, was on following grounds (tr. Ganganatha Jha):
Here the Logician interposes the question—“But what is your proof for Non-duality?” This very question, the Vedantin replies, cannot be asked by one who does not admit Non-duality. Unless one has an idea or conception of Non-duality, how could the question as to its proof be asked at all? […]. If then you admit that you have a cognition of that Non-duality regarding which you ask a question, we further ask you—Is this cognition of yours a true or a false one? If you hold it to be a true cognition, then the very same means of proof (or true cognition) on which that cognition is based is, at the same time, the means of proof for Non-duality and as thus the means of proof is already known to you, the question is idle. […]. If, on the other hand, you declare the cognition you have of Non-duality not to be true, then your question amounts to this “what is the proof for that which is the object of wrong cognition?”— and does not this question clearly imply a self-contradiction?
As you’ve probably already estimated, I don’t find the proof convincing in the least. The problem with it seems to me to be two-fold: first, it trades on an ambiguity as a result of failing to heed a crucial distinction between our intellectual operations; second, it can be used to prove its contrary (i.e., dualism). Consequently, it lacks probative force and so one should not be tempted by it. I’ll explain what I have in mind below.
Sri Harsha, if I have him right, argues something like the following:
To ask for a proof of non-dualism is (in some sense) to have a conception or cognition of it. Now, he further says, is this conception of non-dualism (about which proof is being asked) a true cognition or not. If it is true, then that by which it is a true conception (i.e., what makes it a true conception, as opposed to a false one) will also be the reason for non-dualism being true. Therefore, non-dualism is true. If, on the other hand, one’s conception (of non-dualism) is false, then asking a proof of non-dualism would amount to asking a proof for something that is false, which is clearly absurd.
From an Avicennian point of view though, the first problem with the above argument is that it badly confuses two distinct intellectual operations. These are (1) the act of ‘conceptualizing’ (tasawwur), i.e., the meaning of some x and (2) that of ‘judging’ or ‘assenting’ (tasdiq) to whether or not x is true, i.e., corresponds to how reality actually is. As applied in the present context, when Sriharsha asks the opponent of non-dualism whether his ‘conception of non-dualism (a proof of which is being asked by the opponent) is true or not’, the question needs to be disambiguated – either Sriharsha is asking his opponent whether his conception is a true conception in that he has correctly conceived of what non-dualism simply means, whether he has understood the intelligible content of the concept – which concerns tasawwur proper (i.e., act (1)) – or Sriharsha is asking his opponent whether his conception is true in that that content corresponds to external reality – which concerns tasdiq proper (i.e., act (2)). If it is the former that is being asked, then Sriharsha cannot, on that basis alone, validly infer that non-dualism is a true, i.e., in sense (2), cognition. Otherwise, it would amount to inferring the existence (in external reality) of a ‘phoenix’ simply on the basis of conceiving of what the term ‘phoenix’ means, which is clearly fallacious. If it is the latter though, then Sriharsha is in effect asking his opponent what his opponent is demanding of him i.e., a proof of non-duality! For whether or not non-dualism is a true conception qua tasdiq is exactly what the opponent of Sriharsha is asking an argument for. So from this point of view then, Sriharsha has, rather sophistically, avoided responding to the question altogether, to say nothing of having answered it.
The second problem with the argument is simply that even if, arguendo, one grants it as sound, the same method can be used to prove that dualism (or any other view for that matter) about reality is true in sense (2). All one has to do is reverse the roles of Sriharsha and his opponent and substitute ‘dualism’ as the focus of the argument in for ‘non-dualism’ in order to derive the relevant conclusion. Therefore, etc.