Geoscience Reference
In-Depth Information
- a rational dimension, leading to choose according to the respective profitability
of different educational levels considered, these being perceived differently
depending on the social environment of the individual;
- an interaction dimension that will make the choice of the individual dependent
on its exchanges with other agents within the scope of its contact network, by
imitation;
- an effect of the interaction between the individual and the aggregate level of
the school. Indeed, the social profile of the school is determined by its composition
which has emerged from the individual choices. In turn this profile influences the
choice of the students.
Thus, the first two dimensions are related to the individual level, on the one
hand, its own rationality and, on the other hand, the imitation of persons from his
network. The third dimension refers to the fact that the social aggregates (here the
social profile of the school's pupils) have a retroactive effect on the action of the
agents.
The simulated data are then analyzed by means of statistical processing.
12
This
step allows to put forward the type of educational stratification that emerges from
the rules formalized by the MAS. The effect, notably, of the importance of the
interactions between agents of the same social group is explored. The confrontation
with the empirical data observed in France and Italy shows that in these two cases
the configuration of these interactions played a key role in the educational
stratification that is observed there.
4.3.3.3.
Simulating the abandonment of a territory: examples in archeology
In the field of archeology, different research teams have used MAS in order to
study the population dynamics of environmental change in the long run. The
objective here also consists of creating an artificial world in which households
evolve when confronted with a certain environment, where objective is to extract the
resources necessary to live.
An emblematic case concerns the Anazasi in the “Long House Valley” in
Northeast Arizona between the years 800 and 1350, the date at which the valley was
abandoned [AXT 02, GUM 03]. The challenge is to simulate the population
evolution during this period, as it has been reconstituted by the archaeologists from
different vestiges. This evolution shows two peaks, one peak toward 1100, the other
a little before 1300, and then a collapse toward 1300, followed by an abandonment
of the valley. Environmental data have been reconstructed for the whole of this
12 It should be emphasized here, that the same statistical tools actually make sense, whether
the data is derived from the empirical observation or whether it results from simulations.
