[7] Owen, Philosophical Transactions, Feb. 22, 1848.
[8] From protos = first, and plastos = formed.
[9] From the Greek telos = an end.
[CHAPTER III.]
Methods—the science one of observation and deduction rather than experiment—classification—on mineralogical, on zoological principles—the first for Anthropology, the second for Ethnology—value of Language as a test—instances of its loss—of its retention—when it proves original relation, when intercourse—the grammatical and glossarial tests—classifications must be real—the distribution of Man—size of area—ethnological contrasts in close geographical contact—discontinuity and isolation of areas—oceanic migrations.
In the Natural History of Man we must keep almost exclusively to the methods of deduction and observation; and in observation we are limited to one sort only, i. e. that simple and spontaneous kind where the object can be found if sought for, but cannot be artificially produced. In other words, there is no great room for experiment. The corpus is not vile enough for the purpose. Besides which, “even if we suppose unlimited power of varying the experiment (which is abstractedly possible), though no one but an oriental despot either has the power, or if he had would be disposed to exercise it, a still more essential condition is wanting—the power of performing any of the experiments with scientific accuracy[10].” Experiment is nearly as much out of place in Ethnology and Anthropology as it is in Astronomy.
Psammetichus, to be sure, according to Herodotus, did as follows. He took children of a poor man, put them in the charge of a shepherd who was forbidden to speak in their presence, suckled them in a lone hut through a she-goat, waited for the age at which boys begin to talk, and then took down the first word they uttered. This was bekos, which when it was shown to mean in the Phrygian language bread, the Egyptians yielded the palm of antiquity to that rival.
Now this was an ethnological experiment; but then Psammetichus was an oriental despot; and the instance itself is, probably, the only one of its class—the only one, or nearly so—the only one which is a true experiment; since in order to be such there must be a definite and specific end or object in view.
We know the tradition about Newton and the apple. This, if true, was no experiment, but an observation. To have been the former, the tree should have been shaken for the purpose of seeing the fruit descend. There would then have been an end and aim—malice prepense, so to say.