Geoscience Reference
In-Depth Information
between the entities considered as elementary (the agents) are the driving force in
the dynamics of the system. It follows that these interactions lead to the emergence
of a phenomenon of interest that can be observed at a higher level of organization. It
involves clearly a
bottom-up
procedure (just as for the CAs). The emerging
configuration at the higher level can in turn influence the behavior of agents
following a
top-down
logic. The properties and usefulness of these models in
geography and archeology are discussed in Chapter 4.
In conclusion, we have presented four issues (or “challenges”) (section 2.1) in a
linear manner. We will see later, when presenting case studies that each issue
“challenge” can have a more or less significant weight depending on the questioning
and data. In addition, the sequences are not necessarily linear, and loops can be seen
to occur between certain issues (for example, exploration-simulation or conception-
analysis...). The whole procedure contributes to highlighting new knowledge about
the phenomenon being studied. From an epistemological point of view, the approach
is thus abductive, with a multiplication of points of view to better observe the
information and test hypotheses on the relationships between different phenomena.
2.2. Challenges and models: the possible misunderstandings
The previous section makes account of the diversity of the types of issues and
models that a research about the dynamics of a spatial phenomenon can mobilize.
Conceptual models, data models, exploratory models, statistical models, spatial
analysis models, explanatory models and simulation models can sequentially follow
one another in order to answer to a given problematic but in practice such a
succession is rarely the case. According to the question, its field of competence and
disciplinary practices, a researcher will in general mobilize a subset of this chain of
models. As a matter of fact, depending on the question asked, several sequences of
this chain can be drawn. Some questions will give rise to a long construction stage
leading essentially to a representation and exploration step of the created database.
For others, the stage of the construction of the objects will be short and the analysis
or simulation step will be the main challenge. It should also be stressed that this
sequence is not always linear. For example, the analysis may lead to the construction
of new objects, and simulations generate data that should be explored.
In practice, the term model is often associated with a thematic specification
(growth model, hydrographical model or movement model) and less frequently to a
specification related to the type of model (descriptive
vs.
explanatory model, or,
following another register: data model, statistical model or simulation model). This
is without any doubt the reason why misunderstandings can occur in the
interdisciplinary collaboration between domain experts, computer scientists and
geomaticians. Figure 2.3 provides an example of this case: it illustrates a meeting
