THE ART OF SCIENTIFIC INVESTIGATION 



Practically all reasoning is influenced by feelings, prejudice and 

 past experience, albeit often subconsciously. Trotter wrote : 



" The dispassionate intellect, the open mind, the unprejudiced 

 observer, exist in an exact sense only in a sort of intellectualist 

 folk-lore; states even approaching them cannot be reached with- 

 out a moral and emotional effort most of us cannot or will not 

 make." 



A trick of the mind well known to psychologists is to " rational- 

 ise ", that is, to justify by reasoned ai^ument a view which in 

 reality is determined by preconceived judgment in the sub- 

 conscious mind, the latter being governed by self-interest, 

 emotional considerations, instinct, prejudice and similar factors 

 which the person usually does not realise or admit even to him- 

 self In somewhat similar vein is W. H. George's warning against 

 believing that things in nature ought to conform to certain 

 patterns or standards and regarding all exceptions as abnormal. 

 He says that the " should-ought mechanism " has no place what- 

 ever in research, and its complete abandonment is one of the 

 foundation stones of science. It is premature, he considers, to 

 worry about the technique of experimentation until a man has 

 become dissatisfied with the " should-ought " way of thinking. 



It has been said by some that scientists should train them- 

 selves to adopt a disinterested attitude to their work. I cannot 

 agree with this view and think the investigator should try to 

 exercise sufficient self-control to consider fairly the evidence 

 against a certain outcome for which he fervently hopes, rather 

 than to try to be disinterested. It is better to recognise and face 

 the danger that our reasoning may be influenced by our wishes. 

 Also it is unwise to deny ourselves the pleasure of associating 

 ourselves whole-heartedly with our ideas, for to do so would be 

 to undermine one of the chief incentives in science. 



It is important to distinguish between interpolation and extra- 

 polation. Interpolating means filling in a gap between estabUshed 

 facts which form a series. When one draws a curve on a graph by 

 connecting the points one interpolates. Extrapolating is going 

 beyond a series of observations on the assumption that the same 

 trend continues. Interpolation is considered permissible for most 

 purposes provided one has a good series of data to work from, 

 but extrapolation is much more hazardous. Apparently obvious 



90 



