180 SECTIONAL ADDRESSES. 



of individuals unaffected by the factor in question, whose behaviour is 

 known and observed. By splitting the problem or the community up 

 into its smaller significant or fractional sections, and making an estimate 

 for each section, the possibility of error in the aggregated estimate is much 

 reduced. If the resultant economic system which the economist deduces, 

 following the subtraction or the addition of the particular custom or law, 

 differs widely from the actual state, then the efiect of that custom or law 

 is obviously large and important. But if much the same state of affairs 

 is hypothetically evolved, then the explanation of such a state must be 

 elsewhere, if the explanation that is being sought is a true diflferential. 



Everywhere we observe that men are not born equal ; stations oi 

 fortunes in life are influenced by the fact that A and B were their parents, 

 and not C and D. Something that A and B did or had, that C and D did 

 not or had not, lives after them, and influences the economic position of 

 X, the son of A and B, so that he is essentially different from Y, the son 

 of C and D. The fact that men ' inherit ' seems to be a fact that prima 

 facie should have real economic significance. What would the economic 

 world be like, as compared with the present economic world, if men really 

 started equal ? Or what would the economic world be like if men started 

 with great inequalities, but these inequalities were quite fortuitous and 

 had no relation to the circumstauces or qualities of parents ? In either 

 case we postulate a world in which inheritance is absent as an economic 

 factor. 



It may well be that such an analysis would be inconclusive or indeter- 

 minate at the last, that at certain points we find we need close or exact 

 statistical data that are absent, that at others the balance of probability 

 as to economic psychology in the mass is in doubt, and that at a critical 

 point unbiassed scientific estimates differ widely. At the worst we should 

 know the area of scientific uncertainty ; we should have exposed the points 

 on which exact observation ought in future to be focussed ; we should 

 have given an estimated result with an idea of the probability of error. All 

 of these stages are some way towards truth, at least further on than no 

 analysis at all. In practical matters we may, after all, like others who have 

 not joined in our analysis, have to ' jump ' the gap and flagrantly guess, or 

 act empirically by instinct. This the world has been doing on the widest 

 scale for centuries while knowledge has been growing. But it is some- 

 thing to know that we are voting or deciding not indeed unscientifically 

 but «on-scientifically, which we have no business to do, ssivefautedemieux. 



III. The General Heritage of an Environment formed under Certain 

 Conditions of Inheritance. 



I am not referring particularly to what we call our social heritage, i.e. to 

 what the whole community A enjoys by reason of all that the preceding 

 whole community B has left, either produced and evolved by B itself, or 

 received and perpetuated by the whole community C that preceded B. 

 I am dealing with the principles and fact of individual heritage. But the 

 two cannot be wholly dissociated. As Prof. Pigou has said, ' environments 

 have children as well as individuals.' And if the social heritage which 

 A received from B was one in which individual heritage played an 

 important part, it may well be that it is an entirely different social 



