This,says Mill following Malthus,would be fatal,because the individual would have no motive to support himself.He must only have such a right as implies personal responsibility.But then,as facts also show,many individuals may be unable to support themselves even if they wish it,and their responsibility becomes a mockery.If we enforce duties on all,must we not make the duty possible?Must not every one be so trained and so placed that work will be sure of reward?There is the problem,which he sees and feels,though his answer seems to imply a doubtful shifting between antagonistic theories.
VIII.LOGICAL METHOD
I must glance finally at the relation of Mill's method to his general principles.In an early essay (162)he declares that the method must be 'a priori,'that is,as he explains,'reasoning from an assumed hypothesis.'(163)In the Logic it is treated as a case of the 'direct deductive method.'This involves an important point in his system.He had derived from Comte,as he tells us,(164)only one 'leading conception'of a purely logical kind,the conception,namely,of the 'historical'or 'inverse deductive method.'This method,implied in Comte's sociology,starts,as Mill says,from the 'collation of specific experience.'Now Mill agrees that this 'historical'method was appropriate to sociology in general.He agrees,too,with Comte that it was not the method used by economists.But,whereas Comte had inferred that political economy must for that reason be a sham science,(165)Mill holds that economists were justified in using a different method.Comte,he thought,had failed to see that in certain cases the method of 'direct deduction'was applicable to sociological inquiry.One such case,though he will not undertake to decide what other instances there may be,is political economy.(166)He decides that the difficulties,regarded by Comte as insuperable,may be overcome.His early account is still valid;and he therefore explicitly rejects the 'historical'method.
I confess that the use of these technical phrases appears to me to be rather magniloquent,and to lead to some confusion.
Setting them aside,Mill's view may be briefly stated.He argues,in the first place,that we cannot apply the ordinary method of experiment to economic problems.To settle by experience whether protection was good or bad,we should have to find two nations agreeing in everything except their tariffs;and that,of course,if not impossible,is exceedingly difficult.(167)It follows that if there be a true science of political economy,it must have a different method.We might indeed adopt Comte's answer:
'There is no such science';a view for which there is much to be said.Mill,however,being confident that the science existed had to justify its methods.Political economy,he says,considers man solely as a wealth-desiring being;it predicts the 'phenomena of the social state'which take place in consequence;and makes abstraction of every other motive except the laziness or the desire of present enjoyment which 'antagonise'the desire of wealth.Hence it deduces various laws,though,as a fact,there is scarcely any action of a man's life in which other desires are not operative.Political economy still holds true wherever the desire of wealth is the main end.'Other cases may be regarded as affected by disturbing causes'--comparable,of course,to the inevitable 'friction'--and it is only on account of them that we have an 'element of uncertainty'in political economy.
Otherwise it is a demonstrable science,presupposing an 'arbitrary definition'of a man as geometry presupposes an 'arbitrary definition 'of a straight line.'(168)The relation of this doctrine to Mill's general views on logic is clear,but suggests some obvious criticisms.'Desire for wealth,'for example,is not a simple but a highly complex desire,involving in different ways every human passion.(169)To argue from it,as though its definition were as unequivocal as that of a straight line,is at least audacious.Mill,no doubt,means to express an undeniable truth.Industry,in general,implies desire for wealth,and the whole mechanism supposes that men prefer a guinea to a pound.The fact is clear enough,and if proof be required can be proved by observation.We must again admit that whatever psychological theorem is implied in the fact must be assumed as true.But it does not follow that because we assume the 'desire for wealth'we can deduce the phenomena from that assumption.That inference would confound different things.
If we were accounting for the actions of an individual,we might adopt the method.In some actions a man is guided by love of money,and in others by love of his neighbour.We may 'deduce'his action in his counting-house from his love of money,and consider an occasional fit of benevolence as a mere 'disturbing cause'to be neglected in general or treated as mere 'friction.'
A similar principle might be applied to political economy if we could regard it as the theory of particular classes of actions.
But we have to consider other circumstances to reach any general and tenable theory.We have to consider the whole social structure,the existence of a market and all that it implies,and the division of society into classes and their complex relations:
