Calculate a ``sector population index'' for the sector

A population index for the sector is obtained by multiplying the number of households for the sector's local maximum (obtained in section 4.1.4) by an average number of persons per household. The same set of informal observations used to establish an average residential compound area also yielded an average of 6 persons per compound. This concurs approximately with ethnohistoric information. For example, at the time of Viceroy Toledo's visit to Tiwanaku in 1574, a total population of 4329 was recorded for the repartimiento ([Choque Canqui 1993]: 81). Among these were 868 tributarios (heads of household). This yields an approximate ratio of 4.99 persons/household.

Therefore:

  $P_{s}$ = $H*P_{h}$ 




where $P_{s}$ = sector population index


		 		 $H$ = number of households in sector


		 		 $P_{h}$ = persons per household

This sequence of steps eventually returns a sector population index of 42.67 for a one hectare sherd scatter, and of 66.67 for a hectare of occupied area. These numbers are definitely on the low end of the range of numbers typically used to derive population estimates from site area. Adams, for example, used a figure of 200 persons/ha in his study of Mesopotamian settlement ([Adams 1965]), This was based on modern population densities from the old quarters of Baghdad and from other local examples. Kramer's ethnoarchaeological study of modern settlements in Southwest Asia yielded an average of 120 persons/ha (cited in [Hassan 1981]: 66-67). Similarly, the classical city of Melos has been estimated to have contained between 130 and 200 persons/ha within its walls ([Whitelaw and Davis 1991]: 280). While my numbers therefore seem low compared to figures employed in the analysis of Mediterranean and Near Eastern urban settlements, they fall within the 25-50 persons/ha range of Basin of Mexico ``High Density Compact Villages'' ([Parsons 1976]: 72). My numbers are also considerably lower than the 100 persons/ha used by Parsons to estimate the population of Tiwanaku ([Parsons 1968]). Finally, 42-66 persona/ha falls into the middle of the range of Drennan's ``compact settlement'' category of Mesoamerican archaeological sites ([Drennan 1986]: Table 13.1).

The sector population index, while not a direct measure of population, can be used as a proxy measure. This is so because the index should vary directly with actual population. In the remainder of this analysis I will use this index in place of actual population figures. At no point should it be forgotten, however, that the measure I am using here is a provisional one, and that it could stand a great deal of improvement. Finding better and more phase-specific values for all of the variables employed is the only way in which this index will ever come to be trusted as a direct population estimate.