SECTION 1. ABSOLUTISM AND RELATIVISM IN METHODOLOGY. 19 



John Stuart Mill, though an empiricist in philosophy, was 

 nevertheless, like his distinguished predecessors, an absolutist 

 in logic. He set little store by approximate generalisations, 

 and looked on them as definite though incomplete; his identi- 

 fication of inductive with causal investigations was apparently 

 due to his desire of disposing of something once for all; his 

 canons demanded proofs as unerring as those of the syllogism; 

 and his repeated use of letters of the alphabet to symbolise 

 the various unknown factors in a problem, illustrated how over- 

 simple was his conception of the Universe. The methodological 

 guidance he proposes is consequently only applicable in the 

 main to the concluding stages of an enquiry when bewilder- 

 ment has ceased and the principal facts are established and 

 classified. 



Sheer indefinable probability, a groping one's way in the 

 dark, a chaos growing gradually less confused, a thinker feebly 

 illuminating a humble corner here and there or slightly intensify- 

 ing the light; in other words, the plastic form of the actual 

 process of concrete enquiry had not impressed itself upon the 

 older logicians. They were concerned with final products, not 

 with complicated and elusive facts; nor did they treat of hypo- 

 theses, generalisations, and certainties of an unfolding and 

 progressive character. Even where, as in Laplace's theory, pro- 

 bability was postulated, it was of a calculable character, and 

 not of the undefined quality which almost invariably attaches 

 to investigations as they develop under the hands of genera- 

 tions of men of science, as say in the progressive discovery of 

 the nature of flame, in the slow determination of the principal 

 causes of meteorological changes, in the gradual localisation of 

 the sensory and motor areas in the brain, in the involved 

 unravelling of the problem of heredity, or in the step-by-step 

 ascertainment of the nature of a perfect diet. Likewise in our 

 new century there are few savants who adequately recognise 

 that the most learned treatise written on any subject to-day is 

 bound to be comparatively crude because of its dependence on 

 other treatises which are being or will be written, e.g., a trea- 

 tise on what education should be depends, among other develop- 

 ments, on a perfected science of hygiene, psychology, ethics, 

 aesthetics, and technology, and on something like unexcep- 

 tionable physical, economic, intellectual, political, and moral 

 conditions in society as a whole. This interdependence is notice- 

 able throughout the groups of sciences, beginning with the 

 least dependent and terminating with the most dependent- 

 elementary mathematics, mechanics, ethereology, chemistry, 

 crystallography, biology, psychology, and the cultural or specio- 

 psychical sciences, 1 there being "scarcely any natural pheno- 

 menon which can be fully and completely explained in all its 



1 For a classification of the sciences, see Conclusion 33. 



