Mill lays down at starting(68)the postulates from which he is to reason.Here,of course,we appeal to association.
Association,he tells us,links together the thoughts of phenomena which are like each other,or which have been contiguous or successive;the link strengthens as the association is repeated,and after a time becomes 'inseparable.'Now belief in an external world means the belief that things exist when we do not think of them;that they would exist if we were annihilated;and further,that things exist which have never been perceived by us or by others.This belief is explicable by the known laws of association.For at any moment a given sensation calls up 'a countless variety of possibilities of sensation.'
They are regarded,that is,as sensations which I might experience if circumstances were altered.Again,these possibilities of sensation (which,he adds,are 'conditional certainties')are permanent,because they may be called up by any of the fleeting sensations.This permanence is one of the characteristics of the outside world;and we thus have always in the background,or as a 'kind of permanent substratum,'whole groups of 'permanent possibilities'suggested by the passing sensations.These become further consolidated when fixed orders of succession have suggested the ideas of cause and effect --themselves a product of association.Hence,we get our external world,and can define Matter to be a 'Permanent Possibility of sensation.'The phrase became famous.
This involves the metaphysical question which was reserved or evaded in the Logic.His whole purpose there is to show that thoughts should conform to things.But how things differ from thoughts was never made clear.'Attributes,'we were then told,were the same as 'sensations.'The sensations somehow cohere in clusters.But what makes them cohere in different forms?When a sensation is not accompanied by the sensation previously associated,why is not the association simply weakened or destroyed instead of suggesting a 'conditional certainty'?Ilearn that fire is hot because the sensations of brightness and heat have occurred together;but when I see the brightness without feeling the heat,why does not the association simply become fainter?Why should I interpret the experience to mean,'If I were nearer I should feel the heat'?Does not the interpretation imply that I have already some system of combining my impressions and a need of making the two experiences consistent instead of contradictory?Upon the single assumption of sensations occurring together or successively,and related in time alone,there seems to be no need for any external world whatever.The hypothesis would be exemplified in the case of an animal which,though capable of sensations,had no capacity for arranging them so as to represent space at all.And,again,the statement suggests no distinct reference to any criterion of truth or falsehood.It accounts for illusions as well as for true beliefs.What is the difference?The fact that certain sensations adhere in clusters is not the same thing as the belief in their regular recurrence;and considering the vast variety and intricacy of our sensations,the question which I have mentioned in connection with James Mill arises again:Why should any two people have the same clusters or (on this showing)the same belief --or how one association can be said to be (not real but)true,and another (not unreal but)false?
This difficulty shows itself when Mill proceeds to investigate the 'primary qualities.'They are to be simply 'attributes'co-ordinate with other attributes.With the help of Professor Bain and Mr Herbert Spencer,in whose then recent writings he saw a most encouraging development of his father's principles,Mill makes out a case to show how the perception of space may be developed.The problem discussed by those authorities and their successors is clearly a legitimate part of psychology;their investigations,though still on the threshold of a vast and difficult inquiry,are at least valuable beginnings;and when the experts have all agreed,we shall be ready to accept their conclusions.There is,however,a difficulty which exposes Mill to another criticism.(69)Briefly,it is that his so-called explanation of space-conception really presupposes space.Hamilton had pointed this out in his Kantian moods.(70)The difficulty is obvious.In a scientific theory a statement in terms of space is an ultimate statement.We do not try,nor does it appear to be possible,to get behind it.When Ihave said that a body moves in an ellipse.I do not go on to express the ellipse in terms of 'muscular sensation.'That would be to substitute for a definite measure one essentially fluctuating and uncertain.I can define a given muscular sensation as that which corresponds to a certain distance;but to reverse the definition --to express the distance in terms of the pure sensation,excluding all reference to distance,is surely impossible.Now,it may seem that Mill is here attempting just this impossible feat.Therefore he is really still on the same side of the gulf,though he supposes himself to have crossed it.
