An Appraisal of Techniques with a New Hemispheric Estimate

Social scientists often consider population size as an independent variable major importance. The author analyzes, therefore, methodological reasons why most prior estimates of aboriginal America population imply small scale preconquest societies and concludes that the population was far large than has been thought. The range of previous estimation is so great as to indicate that some methods or data must have been faulty. Skeptical anthropologists and historians have regarded historic population figures reported by contemporary observers as larger than reality. Ethnohistorical estimations based on careful cross-checking of direct and indirect sources of population data demonstrate that contemporary observers underreported the true magnitude of native American populations. Estimates obtained by projecting backward through time population data from modern Indian census counts are seriously defective, as are past applications of the the dead reckoning mode of estimation. Ethnohistorical methods are less deficient. Crosschecking sources serves to increase the accuracy of estimation, and informant knowledge is evaluated in terms of opportunities for accurate observation and especially for accurate memory, since it is more difficult to retain quanlitative than qualitative data. Additive reconstruction of Indian population from historical records is handicapped because the geographic spread of literate Europeans lagged behind the diffusion of new disease agents which decimated aboriginal populations. Direct observation of Indian population trends by anthropologists suffers the same limitation, even though several have recorded extremely rapid depopulation of Amazon basin peoples within the century. Calculation of aboriginal population from preconquest social structure is possible only in 1 area governed by an imperial ideal of administrative units consistent in population. Aproximately accurate estimates of aboriginal American population may be achieved by comparing the population of a given area at two or more times in order to establish population trends expressed as ratios of the size of the population at 1 time to its size at another. Early historic depopulation was great in the Americas: a well-documented instance of recovery following depopulation from 50 to 1 over one century indicates limits to such a trend. Greater population loss probably results in extinction. A hemisphere-wide historic depopulation ratio of 20 to 1 is postulated. Applying it to more or less well-established historic Indian nadir populations suggests that the New World was inhabited by approximately 90,000,000 persons immediately prior to discovery.