Assume it’s possible for S, a maximally perfect person, to think of himself qua maximally perfect person. What would happen, metaphysically speaking? Well, the object of S’s thought must be a perfect representation of a perfect person. But what, exactly, does ‘perfect representation’ amount to?
One understanding would threaten to eliminate any distinction between S and S’s object of thought. In Parmenides 130-134d, Plato argues that no object in the sensible world can perfectly resemble the Form of which it partakes, for if it did, it would just be the Form itself. Similarly, Frege argued against the correspondence theory of truth on the grounds that perfect correspondence between thought and reality—a relation he thought essential to the theory—would require thought and reality to be identical. The reasoning seems straightforward: compare any two things, x and y, where y is a representation of x; if y differs from x in any respect, then y cannot be a perfect representation of x, for y does not represent x at least in that respect. The application is clear: if S’s thought of himself were truly a prefect representation, then S’s thought of himself would be identical to S himself. But that seems either incoherent (what sense can be given to it? I cannot find any) or false (oneself and oneself qua object of thought are not identical).
Maybe this difficulty can be overcome by qualifying “perfect—in relevant respects.” But what are the relevant respects? If S is thinking of himself qua maximally perfect person, the relevant respects would be the those which make S perfect. S and the object of S’s thought would then be individuated by properties that are not perfections (e.g., “being the thinker” and “being the object of thought” etc.). But now we run into another problem: being able to think of oneself qua oneself is a surely perfection. So, the object of S’s thought of himself would inherit the perfection of being able to think of oneself qua oneself. But any entity able to think of oneself qua oneself is a person. We have come to a conclusion familiar to us from psychological analogies of the Trinity. But if S’s thought of himself is also a perfect person (S*, say), would not S* also think of himself qua perfect person, and so generate another perfect person, S**, who does the same? It would seem that if a perfect person generates another person in virtue of thinking of himself then an infinite number of perfect persons would be generated, like a mirror reflecting itself.
What we have, then, is a double dilemma facing those who think it’s possible for a maximally perfect person to think of himself qua maximally perfect person. Either
(A) S’s thought of himself is not a perfect representation
which I’m assuming is a non-starter, because S is ex hypothesi a maximally perfect person, and so would upon thinking of himself do so perfectly, or
(B) S’s thought of himself is a perfect representation
But then it seems that either of the two proposed ways of understanding ‘perfect representation’ is problematic, for the one
(C) Eliminates any distinction between S and S’s object of thought (a la Plato and Frege)
and the other
(D) Generates an infinite number of perfect persons
Maybe there is another understanding of ‘perfect representation’ that isn’t problematic in the way (C) and (D) are. Or maybe the original assumption—that being able to think of oneself qua oneself is a perfection—is false. If it isn’t a perfection, or at least a divine perfection, then one could also escape similar worries raised here.