TRANSACTIONS OF SECTION F. 651 



the isolation of their study, this is the most serious of all — is that of exag- 

 gerating immensely the office of deduction in their investigations. 



Deduction has indisputably a real and not inconsiderable place in Sociology. 

 We can sometimes follow the method which Mill calls the direct deductive, that 

 is, we can, from what we know of the nature of man and the laws of the external 

 world, see beforehand what social phenomena will result from their joint action. 

 But, though the economists of the so-called orthodox school recognize no other 

 method, we cannot really proceed far in this way, which is available only in simple 

 cases. Social phenomena are in general too complex, and depend on too manifold 

 conditions, to be capable of such a priori determination. In so far as the method can 

 be used, the vital condition of its legitimate employment is the ascertainment of 

 the consilience of the results of deduction with those of observation ; and yet such 

 verification from fact of the conclusions of theory, though essential to the admissi- 

 bility of this process of inquiry, is too often entirely overlooked. 



Much more commonly the function of deduction is different from what has just 

 been described, and its relation to observation is inverted. The laws of the 

 economic constitution and movement of society are first obtained by observation, 

 whether directed to contemporary life or to the history of the past. The office of 

 deduction is then to verify and control the inductions which have been arrived 

 at, using for this purpose considerations founded on the qualities of human nature 

 and the external conditions to which society is subjected. Results which could 

 not have been elicited by a priori reasoning from the latter data, may, when in- 

 ductively obtained, be in this way checked and rationalized. The pretension of 

 the economists, formally set forth in Senior's treatise, to deduce all the phenomena 

 of the industrial life of communities from four propositions, is one that cannot be 

 sustained. But conclusions derived from observation may be placed in relation 

 with the laws of the world and of human nature, so far at least as to show that 

 they contradict nothing we know respecting those laws. This method, in which 

 inductive research preponderates, and deduction takes a secondary place as means 

 of verification, is the really normal and fruitful method of sociological inquiry. 



But the method of Sociology must be not only inductive, but historical ; and by 

 the latter name it may best be characterised. By this is meant, not merely that it 

 finds the materials for its studies in the general field of human history : . we mean 

 further that it institutes a comparison of the successive states of society in order 

 to discover the laws of social filiation — a process similar in principle to the biological 

 comparison of organisms of different degrees of development. If we followed exclu- 

 sively the a priori method, in (for example) economic research, and sought to infer 

 the economic facts of life from the nature of the world and man, we could arrive 

 only at one determinate order of things, whilst we know that in reality the economic 

 organization and functions of society vary in time according to definite laws of suc- 

 cession. Mr. Lowe, indeed, will have it that " political economy is founded, on the 

 attributes of the human mind, and nothing can change it ;" which means, I suppose,, 

 that its formulas must always correspond with the phenomena. But how can this 

 view be reconciled with the now ascertained fact, that society has passed through 

 states in which the modern economic constitution was so far from existing, that 

 property did not belong to the individual, but to the community ? The a priori 

 method, in fact, overlooks what is the main agency in the social movement — namely, 

 the accumulated influence of anterior on subsequent generations of mankind ; an 

 influence too complex to be estimated deductively. Every department of social life, 

 and amongst the rest the industrial system, undergoes transformation — not arbi- 

 trarily indeed, but in accordance with law ; and if we wish to understand any of 

 those departments, we must study its transformations, considering each successive 

 form in relation to all the preceding and contemporary conditions. 



There is, indeed, no more important philosophical theorem than this : that the 

 nature, of a social fact of any degree of complexity cannot be understood apart from 

 its history. " Only when its genesis has been traced," says Mr. Herbert Spencer, 

 " only when its antecedents of all orders have been observed in their co-operation, 

 generation after generation, through past social states — is there reached that inter- 

 pretation of a fact which makes it a part of sociological science." To understand, 



