APPLIED TO HISTORY 243 



the human element. This is one main reason why there 

 is so little room for prediction in dealing with it: we 

 cannot tell what men must have done, we can only learn 

 indirectly what they have done. While we deal with 

 documents in ways hinted at above we use methods 

 which follow almost mechanical rules. We analyse, 

 compare, and classify their statements : we fix their dates 

 by various processes of more or less certain validity : and 

 we obtain, as a result, a sort of provisional view of the 

 state of things in the period. Many of the documents 

 before us will probably be formal state-papers, records 

 made in the way of business, and so on. If we are 

 persuaded that these genuinely belong to the period they 

 claim they have a high degree of authority. But a large 

 portion of the documents under discussion will be due to 

 the voluntary enterprise of individuals, and it is in these 

 that we shall find the most intimate and detailed account 

 of things. It seems to me that, whereas it is laborious 

 but comparatively easy to estimate formal documents, it 

 is difficult to estimate the value of independent voluntary 

 records, because much will depend on the character, 

 opportunities, and capacities of the writer. We cannot 

 do without chroniclers and historians if we are to have 

 anything like a full account of the past, but it is of vital 

 importance to us to know whether they are trustworthy 

 witnesses or not. 



This judgement of our authorities must be of gradual 

 growth, and is made up of many considerations. We 

 must, for instance, note the conditions, so far as they are 

 revealed, in which our author claims to write, and, even 

 more important than that, the indications which he 

 supplies (unconsciously) of his way of working and his 

 conception of a fact. So far as it is possible, we must 

 endeavour to distinguish between the real record of 



