Networks and Ethnogenesis 103whereas on smaller islands, the need for networked relationships was increased. At this
time, treating islands as individual entities is meaningless, and “the establishment of
populations is better assessed in terms of networks of mutually sustaining communities”
(Broodbank 2000: 187) which transcended insular boundaries.
Knappett, Evans, and Rivers developed their analysis of Bronze Age sites of the Aegean
from Broodbank’s PPA, incorporating into their model variables to account for asym-
metry, directionality, and costs of maintaining interactions, so simulating different social
conditions and measuring the effects that these factors had on site size and influence.
Recognizing that “humans create space through social practices, while also acknowledg-
ing that physical parameters are not entirely redundant in this process” (Evans, Knappett,
and Rivers 2009: 2), they used a network model to take account of both physical and
relational space. The computer program they built allowed them to change parameters
and simulate expansion and collapse in the Aegean basin. They found that the site interac-
tions affect site size and status, inverting localized, “site-centrist” explanations that focus
on surplus and exchange, by treating sites as secondary and theinteractions between them
as primary.
Social Network Analysis
Use of social network analysis in studies of the ancient world has also increased in the
last two decades. An important point is the difference between “whole-network studies,”
which examine the properties of a network at the global level, and “egocentric” network
studies, which look at the network from the position of an actor located within it. The
methods explored in the preceding text have mostly examined material relationships in
antiquity and often have “whole-network” aims. Although they can be complementary,
in practice it is difficult to move from the structure of the network to the individuals
who form it (Marsden 2005: 8). Moreover, the egocentric network of individuals in
archaeology can often be difficult to find, but when it is possible, social network analysis
offers unprecedented opportunity to humanize networks in the past.
Graham’s work explicitly does this. He views past space as an itinerary located within
a perceptual sphere, and this extends to his work with networks: “networks do not exist
independently of the people within them, and it is not enough that mere interconnections
should exist. Individuals matter. Individuals must make something of these interconnec-
tions, for the networks to work” (Graham 2009: 675). Graham used stamps on bricks
to reveal a small-world network configuration in the imperial Roman brick industry: and
discovered that Domitia Lucilla, mother of Marcus Aurelius, occupied a position that
allowed her control over the flow of information in the brick trade. As a function of her
network position, Lucilla knew about clay sources, building contracts, profit margins,
and distance of trade, and was able to exploit the discrepancy between the local and the
global to achieve her own ends (Graham 2009). A word of caution, however, about mis-
using social network analysis: “only when the interactions between people are analyzed
directly, can social network analysis with its specific quantitative tools and interpretations
be applied” (Brughmans 2010: 7).