The density of 250 persons per hectare, or sometimes 200 persons per hectare, is derived on analogy with pre-modern Muslim settlements, primarily from the Late Ottoman period in the Levant, where it is assumed that habitation patterns in the past
did not change much until the 20th century (Zorn 1994: 32). However, there is very little in common between Muslim settlements of the 19th century and settlements of Canaan in the Late Bronze Age. Not only are these settlements separated by almost 3,500 years, but the technology, culture, religion, architecture, and ethnicity are all different; a direct comparison is invalid because the adaptation is entirely different. Thus, in order to accurately assess the population of settlements in Late Bronze Age Canaan, a methodology specific to that period and region must be derived from micro studies and data specifically related to the relevant time and place.
Broad application of uniform population density coefficients have often been employed in estimating ancient population data. Baer estimated a rural population density of 184 people per square kilometer in the Nile floodplain for the entire period of Dynastic Egypt, and a slightly higher 225 per square kilometer estimated for ancient Greece of the same time period (Bintliff and Sbonias 1999: 8). This number, when multiplied by the estimated square kilometers of inhabited land in the boundaries of ancient Egypt, yields an approximate maximum rural population of 1.5 million people for Dynastic Egypt, excepting the major towns and cities—the total population including urban areas would be higher (Butzer 1976: 77). Adding estimates for the population of major cities and towns, resulting in a larger total, would vary based upon the urbanization level of the culture and the typical population density at a particular point in time. Uphill, using the same basic technique but applying it to towns and cities, generally uses a number of 250 people per acre (617.5 per hectare) or 250 per 0.4 hectares (625 per hectare) in a town for ancient Egypt (Uphill 1988: 15). This is only used for calculating urban or suburban populations, and is not applicable to calculating any possible nomadic population. According to Butzer’s calculations, with an estimated population density of the ancient city of Memphis at 550 per hectare, urban Memphis during its peak in the Old Kingdom had a total population of approximately 17,050 people (Butzer 1976: 102). Butzer’s estimate of Per Ramesses in New Kingdom times, the same period as the Late Bronze Age, puts the population of the city at 100,000 (Butzer 1999: 250). New Kingdom Per Ramesses according to this estimate reached 350 hectares, although after subsequent excavation Bietak later estimated the size at up to 600 hectares—substantially larger than any site in Late Bronze Canaan, and
indicative that no city during this period in Canaan would have matched or exceeded its estimated population (Bietak 2010: 12; Butzer 1999: 250). Even though New Kingdom Egypt coincides with Late Bronze Age Canaan in time, is a geographical neighbor, and had a similar technology level, these population density coefficients cannot be directly implemented into Canaan of the Late Bronze Age because Canaan had a culture, geography, and architectural tradition distinct from Egypt. Constructions of New Kingdom cities such as Akhetaten and Per Ramesses and the sheer size of their metropolitan areas suggests a population increase and a move to more urbanized culture during this period (Uphill 1988: 60, 62). Similar trends may have occurred in Canaan beginning in the Middle Bronze Age and initially flowed into the Late Bronze Age before settlement change occurred. The increased urbanization of New Kingdom Egypt makes accurately estimating town and city populations important for demographic estimates of Egypt in this period. The emphasis on towns and cities is also applicable to Canaan during the Late Bronze Age, since the many towns and cities discovered archaeologically, in addition to the Amarna correspondence, indicate that a major segment of the population was settled in towns and cities during this period.
Past studies of ancient Canaan have used a uniform density coefficient. The primary problematic aspect of previous Canaan population studies is the assignment of an all-encompassing density coefficient multiplied by total site area. The studies start with the flawed premise that a density coefficient of 200 or 250 people per hectare is correct, when in fact this figure is based primarily upon a study of modern villages in Iran (Broshi and Gophna 1986: 73-74; Finkelstein 1996: 244). This practice is due to convenience and availability of data, but there are obvious problems with such a simplified view of ancient population density. Interestingly, a study on old quarters of Middle Eastern cities, specifically Iraq, demonstrates a density coefficient of around 450 people per hectare (Adams 1981: 350). The building density of old quarters of cities would probably be more similar to building density of ancient cities than modern village density; the old quarters database is a closer comparison than modern village density calculation. Still, broad modern ethnographic data should not be used in place of specific ancient data.
The results of studies utilizing a 200 to 250 people per hectare constant were done employing a very generalized population density coefficient derived from studies of villages and sections of cities not yet modernized in the Middle East in the 18th, 19th, and 20th centuries rather than data specifically from Canaan in the Late Bronze Age, or other ancient periods, should be considered inaccurate due to invalid correlation (Broshi and Gophna 1986: 74; Shiloh 1980: 26; Adams 1981: 349-50; Kramer 1980: 322-27;
Postgate 1994: 51). Finkelstein suggested correlating the household population trends of Muslim villagers living in British Mandate Palestine directly back onto Bronze Age Canaan by arguing that this proposed ethnographic parallel indicated a household of approximately 4 or more people for the Bronze Age as well as British Mandate Palestine villages (Finkelstein 1990: 49). This drastically affects the population density coefficient estimate by making it substantially lower than calculations using a higher per household or per family population. The primary flaw, as seen in other studies, is that all of the household population and settlement density coefficient estimates used are derived from studies of modern and primarily Islamic populations (Finkelstein 1990: 48-50). It is acknowledged, yet not utilized, that a population density coefficient for ancient settlements “based on data from some Middle Eastern towns in recent generations…cannot be applied to the study of historical demography” (Finkelstein 1990: 50). Unfortunately, this astute observation that data from later, unrelated periods and cultures should not be projected onto an ancient culture is not carried through in the majority of previously utilized methodologies for estimating settlement populations in the ancient Levant, or even other regions of the ancient world.
While useful for comparative analysis, data from the modern period is not the most precise basis for making demographic calculations in a specific region during the Late Bronze Age. Although a figure of around 200 people per hectare is used for the above studies, based on 18th to 20th century Middle Eastern villages, a population density study of an ancient Sumerian city yielded a range of between approximately 250 and 1200 people per hectare and was based on a detailed analysis of dwelling space at the ancient site to determine possible density coefficients rather than beginning with an assumed premise (Postgate 1994: 62). Postgate, the archaeologist who conducted the study, leans more towards a figure of around 450 people per hectare, perhaps
influenced by the Adams study in modern era Iraq (although allowing for the possibility of a higher density) because of the amount of dwelling space per person that this figure allows—about 10 square meters—although this 10 square meters of dwelling space per person is on the high end of the scale for dwelling space studies which demonstrate a worldwide constant between approximately 4.7 and 7.5 square meters through more recent studies (Postgate 1994: 63; Brown 1987:1-49). Thus, the population density was likely even higher than the conservatively preferred estimate of 450 people per hectare.
Uphill uses a population density coefficient of about 625 per hectare in towns of ancient Egypt, while Butzer uses a population density of 550 per hectare for an Old Kingdom city in Egypt (Uphill 1988: 15; Butzer 1976: 102). From a detailed study of house size and residential area, Zorn determined a density coefficient of between 470 and 590 people per hectare at Iron Age Nasbeh (Zorn 1994: 44). Compared to the figures referenced previously for estimates of 200 to 250 people per hectare in Middle Bronze and Late Bronze Age Canaan, a density of 450 to 600 or more people per hectare appears extremely high. However, it is important to note that the higher figures are at least partially derived from ancient data rather than modern data, and therefore are much more realistic. This indicates that the 200 to 250 people per hectare coefficients are far too low for use in Late Bronze Age Canaan, and thus would give both specific site populations and overall region populations far lower than reality. Still, a general population density for Canaan in the Late Bronze Age must not simply be assumed based upon previous studies, but based upon archaeological and textual data restricted to the period and region, and further applied on a site to site basis.
Graph 6.1 Comparison of Population Density Coefficients