=145.= The chief argument for the certainty of knowledge is that we assume as much in the practical affairs of life[11]; and (as we have already seen) Aristo found it ridiculous that his Academic neighbour should not even know who he was[12]. Against it is the fact that men frequently disagree even as to what they see, and commonly distinguish between what is known to them and what ‘seems’ to be this or that. Hence Epictetus well defines the function of dialectic as ‘a perception of the disagreement of men with one another, and an inquiry into the cause of this disagreement; a condemnation and distrust of that which only seems, and some kind of investigation of that which seems, as to whether it rightly seems: and the discovery of some rule (κανών)[13].’ Of all kinds of knowledge that which comes through the senses appears to the ordinary man most worthy of confidence, and of the five senses that of sight seems to the philosopher the most divine[14]. In consequence, the whole controversy hinges on the question whether the eyes can be trusted. The positivist argues that the evidence of sight is so plain and unmistakeable that man, if he had the choice, could wish for no better informant. The sceptic replies that nevertheless, if a straight oar be placed partly in the water, it appears to the eyes to be bent; and that the feathers on a dove’s neck, though really alike, appear to the eyes as many-coloured[15]. To deal with such questions we must examine closely the nature of sensation. [Sidenote: Process of sensation.]