173. Environmental Considerations
Two other qualifications on the distribution method must be observed, although they are sufficiently obvious to carry no great danger of oversight. The first concerns gaps or bounds due to physical environment. Metallurgy will not be practised on an isolated coral island. Snowshoes cannot be expected in equatorial lowlands. The spread of the cultivation of a tropical plant like manioc is necessarily restricted no matter how great the antiquity of its use. Limitations of diffusion, or breaks in the continuity of distribution, thus do not count as negative evidence if climate or soil suffice to explain them. This is in accord with what has been previously formulated (§ [83]) as to environment being a limiting condition rather than a cause of cultural phenomena.
Secondly, a marginal area need not be literally so. It may actually be nuclear. Thus in the Philippines, older elements of culture are best preserved in the interiors of the larger islands. The coasts show many more imported traits. Communication in the archipelago is by sea, internally as well as in foreign relations; resistance to travel, conquest, intercourse, or innovation is by land. The remote area as regards time may therefore be a mountain range fifty miles inland, while a coast a thousand miles away is near. So a rough hill tract in a level territory, a desert encircled by fertile lands, sometimes remain backward because they oppose the same obstacles to diffusion as great distances.
It is thus evident that valuable as the distribution principle is, perhaps most important of all non-excavating methods of prehistoric investigation, it can never be used mechanically. It must be applied with common sense, and with open-mindedness toward all other techniques of attack. With these provisos in mind, let us approach the problem of American culture.