2^ Science Re/igion and Reality 



tries to use them both for his benefit. Whenever he has been 

 taught by experience that effort guided by knowledge is of some 

 avail, he never spares the one or ignores the other. He knows 

 that a plant cannot grow by magic alone, or a canoe sail or float 

 without being properly constructed and managed, or a fight be won 

 without skill and daring. He never relies on magic alone, while, 

 on the contrary, he sometimes dispenses with it completely, as in 

 fire-making and in a number of crafts and pursuits. But he clings 

 to it, whenever he has to recognise the impotence of his knowledge 

 and of his rational technique. 



I have given my reasons why in this argument I had to rely 

 principally on the material collected in the classical land of magic, 

 Melanesia. But the facts discussed are so fundamental, the 

 conclusions drawn of such a general nature, that it will be easy to 

 check them on any modern detailed ethnographic record. Com- 

 paring agricultural work and magic, the building of canoes, the art 

 of healing by magic and by natural remedies, the ideas about the 

 causes of death in other regions, the universal validity of what has 

 been established here could easily be proved. Only, since no 

 observations have methodically been made with reference to the 

 problem of primitive knowledge, the data from other writers could 

 be gleaned only piecemeal and their testimony though clear would 

 be indirect. 



I have chosen to face the question of primitive man's rational 

 knowledge directly : watching him at his principal occupations, 

 seeing him pass from work to magic and back again, entering into 

 his mind, listening to his opinions. The whole problem might 

 have been approached through the avenue of language, but this 

 would have led us too far into questions of logic, semasiology, and 

 theory of primitive languages. Words which serve to express 

 general ideas such as existence, substance, and attribute, cause and 

 effect, the fundamental and the secondary \ words and expressions 

 used in complicated pursuits like sailing, construction, measuring 

 and checking ; numerals and quantitative descriptions, correct and 

 detailed classifications of natural phenomena, plants and animals — 

 all this would lead us exactly to the same conclusion : that primi- 

 tive man can observe and think, and that he possesses, embodied 

 in his language, systems of methodical though rudimentary 

 knowledge. 



Similar conclusions could be drawn from an examination of 



