In fact, the more I think about it, the more Russell's paradox appears to be much ado about nothing. The smallest set that contains the set  is either the set  (if sets are defined by their members) or the set [] (if they are not). So either sets by definition contain themselves, or they by definition do not. If sets are defined by their members, the notion that the set of all squares does not contain itself because it's not itself a square is bullshit: the only reason it's not a square is that it's multiple squares---supposing that there's more than one square. If the set of all squares cannot contain multiple squares, because multiple squares are not a square (i.e., one square), then the set of all squares can only exist if there's no more than one square!
"In man, creature and creator are united: in man there is matter, fragment, excess, clay, mud, nonsense, chaos; but in man there is also creator, sculptor, hammer-hardness, spectator's-divinity and seventh day:—do you understand this antithesis? And that your compassion is for the 'creature in man', for that which must be formed, broken, forged, torn, burnt, made incandescent, purified,—that which must suffer and shall suffer? And our compassion—do you not grasp whom our reverse compassion is for when it defends itself against your compassion as against the worst of all pamperings and weaknesses?" (Nietzsche, Beyond Good and Evil, aphorism 225. Cf. The Will to Power, Kaufmann edition, section 367.)