14.7 Evidence from Field Archaeology
245
Table 14.3 Data comparing living area to total area of pa
Living area
Total area
Living area
Total area
360
445
700
800
1,667
1,780
5,932
7,860
545
750
2,500
3,900
2,135
2,318
750
1,490
383
500
1,875
4,775
1,463
1,628
550
660
915
1,400
924
1,114
2,469
3,121
1,470
2,510
2,255
2,530
800
1,120
4,756
5,819
725
2,420
5450
10,300
575
1205
2,133
2,770
3,835
4,430
2,460
3,045
815
1,300
2,200
2,325
1,090
1,800
1,582
2,032
2,516
5,670
4,720
6,500
645
1,800
675
715
677
1,810
1,195
1,900
140
150
6,150
6,900
1,274
1,365
Third, to obtain population estimates from pa data, we need to consider the proportion of living area to total area for pa. As can be seen from Table 14.3, the data
is limited.
Table 14.3 presents a list of the 38 pa, drawn from Walton’s sample of 931
pa, for which information about both defended living space (y i ) and total area
(x i ) are provided. A basic model to link these is y ~ a + bx. A regression indicated that the constant a was largely irrelevant, and that the correlation between
x and y was 0.9281. The variation inflation factor was 1, so there was no evidence
of multicollinearity. The Cameron and Trivedi decomposition of the IM test and
the Breusch–Pagan/Cook–Weisburg test both indicated heteroscedasticity. Working
with robust standard errors gave y i ~ b x i for coefficient b = 0.676, with robust
standard error 0.049 for the coefficient b. Thus a 95% confidence interval for b is
(0.576, 0.775).
The proportion of defended living space to total pa area can thus be estimated as
0.676 × total area, with 95% confidence interval (0.576, 0.775) for the multiplier.
If the total area of 90% of all pa of 2,100 m 2 or more is 23,094,000 m 2 , their total
defended living space in 1445 will be 15,607,000 m 2 .
Raoul Naroll [33] has suggested “that the population of a prehistoric settlement
can be roughly estimated by archaeologists as of the order of one-tenth of the floor
area in square metres occupied by its dwellings”. Naroll’s estimate of 10 m 2 per
person would make our 1445 double-count-corrected estimate for the total defended
living area of pa in 1445 of 15,607,000 correspond to a total pa population in 1445
of 1,561,000.
245
Table 14.3 Data comparing living area to total area of pa
Living area
Total area
Living area
Total area
360
445
700
800
1,667
1,780
5,932
7,860
545
750
2,500
3,900
2,135
2,318
750
1,490
383
500
1,875
4,775
1,463
1,628
550
660
915
1,400
924
1,114
2,469
3,121
1,470
2,510
2,255
2,530
800
1,120
4,756
5,819
725
2,420
5450
10,300
575
1205
2,133
2,770
3,835
4,430
2,460
3,045
815
1,300
2,200
2,325
1,090
1,800
1,582
2,032
2,516
5,670
4,720
6,500
645
1,800
675
715
677
1,810
1,195
1,900
140
150
6,150
6,900
1,274
1,365
Third, to obtain population estimates from pa data, we need to consider the proportion of living area to total area for pa. As can be seen from Table 14.3, the data
is limited.
Table 14.3 presents a list of the 38 pa, drawn from Walton’s sample of 931
pa, for which information about both defended living space (y i ) and total area
(x i ) are provided. A basic model to link these is y ~ a + bx. A regression indicated that the constant a was largely irrelevant, and that the correlation between
x and y was 0.9281. The variation inflation factor was 1, so there was no evidence
of multicollinearity. The Cameron and Trivedi decomposition of the IM test and
the Breusch–Pagan/Cook–Weisburg test both indicated heteroscedasticity. Working
with robust standard errors gave y i ~ b x i for coefficient b = 0.676, with robust
standard error 0.049 for the coefficient b. Thus a 95% confidence interval for b is
(0.576, 0.775).
The proportion of defended living space to total pa area can thus be estimated as
0.676 × total area, with 95% confidence interval (0.576, 0.775) for the multiplier.
If the total area of 90% of all pa of 2,100 m 2 or more is 23,094,000 m 2 , their total
defended living space in 1445 will be 15,607,000 m 2 .
Raoul Naroll [33] has suggested “that the population of a prehistoric settlement
can be roughly estimated by archaeologists as of the order of one-tenth of the floor
area in square metres occupied by its dwellings”. Naroll’s estimate of 10 m 2 per
person would make our 1445 double-count-corrected estimate for the total defended
living area of pa in 1445 of 15,607,000 correspond to a total pa population in 1445
of 1,561,000.