1 88 Appendix 



A population movement, looked at as an effort to satisfy 

 desires, renders this task somewhat easier than it is in most 

 social problems. The movement itself is a very tangible, mea- 

 surable trait, and, furthermore, two situations are involved, the 

 one from which population shifts and the one into which it 

 shifts. Certainly, if groups of men are so profoundly affected 

 that they leave their residences and acquaintances and chance 

 their future among strangers there must be some powerful mo- 

 tive or complex of motives back of the move — some condition 

 which is odious, some desire which cannot be satisfied in their 

 old home. Given a sufficient number of more or less homo- 

 geneous areas which are losing population, an observer can de- 

 termine certain elements common to the situations in each, 

 from which the movement seems to rise. Whether these con- 

 ditions are fundamental and permanent causes of the movement 

 or not, can be verified by observing the migrant in his new sur- 

 roundings, and finding out if he escapes the conditions which 

 were odious in his old surroundings — if he satisfies the desire 

 which he could not satisfy before moving. This gives a double 

 check on the causes of the movement which is based on con- 

 crete, observable facts. 



But, in the midst of such complexity as organized social 

 groups present, how is definite assurance to be obtained that 

 observations are accurately made, or relationships correctly 

 tested. The scientific method for obtaining this assurance was 

 outlined by Durkheim in 1895 in his "Les Regies de la Methode 

 Sociologique," as follows: 



The old logicians' methods are of very little value in social 

 reasoning because they assume a science already advanced to 

 such a stage as to offer incontestable laws from which logical 

 reasoning may proceed by comparison of cases which agree 

 or differ in one point only. Social groups are too complex to 

 ever agree or differ in only one respect. The real social method 

 is, therefore, a statistical method. The groups studied may be 

 compared with respect to the phenomenon under investigation 

 and a phenomenon which is thought to be its cause. When 

 the extent to which the two are present or absent in the same 

 group fluctuates, uniformly in the same direction, this simple 

 parallelism of values constitutes, in itself, a proof of a relation- 

 ship which may often be stated quantitatively, provided a suf- 

 ficient number of cases are studied. 



This process of accurately measuring the quantitive relation- 

 ship between a social force and a change in society is a great 



