Throughout much of his work our author's object is to invalidate the
anthropological method. That method sets side by side the customs,
ideas, fables, myths, proverbs, riddles, rites, of different races. Of
their _languages_ it does not necessarily take account in this process.
Nobody (as we shall see) knows the languages of all, or of most, of the
races whose ideas he compares. Now the learned professor establishes the
'harm done' by our method in a given instance. He seems to think that,
if a method has been misapplied, therefore the method itself is
necessarily erroneous. The case stands thus: De Brosses {117a} first
compared 'the so-called fetishes' of the Gold Coast with Greek and Roman
amulets and other material objects of old religions. But he did this, we
learn, without trying to find out _why_ a negro made a fetish of a
pebble, shell, or tiger's tail, and without endeavouring to discover
whether the negro's motives really were the motives of his 'postulated
fetish worship' in Greece, Rome, or Palestine.