In all these cases (concludes Sokrates) the characteristic property is essentially relative, implying something distinguishable from, yet correlating with, itself. May we generalise the proposition, and affirm, That all properties are relative, and that every thing in nature has its characteristic property with reference, not to itself, but to something else? Or is this true only of some things and not of all--so that cognition may be something in the latter category? This is an embarrassing question, which I do not feel qualified to decide: neither the general question, whether there be any cases of characteristic properties having no reference to any thing beyond themselves, and therefore not relative, but absolute--nor the particular question, whether cognition be one of those cases, implying no separate _cognitum_, but being itself both _relatum_ and _correlatum_--cognition of cognition.[18] [Footnote 18: Plato, Charm. 168-169. 169 A: [Greek: mega/lou dê/ tinos a)ndro\s dei=, o(/stis tou=to kata\ pa/ntôn i(kanô=s diairê/setai, po/teron ou)de\n tô=n o)/ntôn tê\n au(tou= du/namin au)to\ pro\s e(auto\ pe/phuken e)/chein, a)lla\ pro\s a)llo\--ê)\ ta\ me\n, ta\ d' ou)/; kai\ ei) e)/stin au(= a(/tina au)ta\ pro\s e(auta\ e)/chei, a)=r' e)n tou/tois e)sti\n e)pistê/mê, ê(\n dê\ ê(mei=s sôphrosu/nên phame\n ei)=nai. e)gô\ me\n ou) pisteu/ô e)mautô=| i(kano\s ei)=nai tau=ta diele/sthtai.]] But even if cognition of cognition be possible, I shall not admit it as an explanation of what temperance is, until I have satisfied myself that it is beneficial. For I have a presentiment that temperance must be something beneficial and good.[19] [Footnote 19: Plato, Charm. 169 B. [Greek: ô)phelimo/n ti ka)gatho\n manteu/omai ei)=nai.]] [Side-note: Even if cognition of cognition were possible, cognition of non-cognition would be impossible. A man may know what he knows, but he cannot know what he is ignorant of. He knows the fact _that_ he knows: but he does not know how much he knows, and how much he does not know.]