The notion of habiter (inhabiting in both a transitive and intransitive sense), developed in the course of the second half of the 20th century, made it possible to place the human individual in the center of spatial theories. Of phenomenological inspiration, this notion has nevertheless contributed to widen the gap between on side, a quantitative geography oriented towards at spatial statistics and at mathematical models of the territory and, on the other side, a geography of spatial existence, seeking to understand the geographical space as a world lived by individuals. The aim of this thesis is to ill this gap by articulating both approaches in a comprehensive understanding of the inhabited space in general and of the Swiss territory in particular. The latter serves as the empirical context of the author's investigations. The thesis consists in three parts: a purely theoretical part, founded in an interdisciplinary literature and in concrete examples serving to illustrate the central notions, a first methodological and empirical part, dedicated to the synthesis and the analysis of Swiss statistical data, a second methodological and empirical part, dedicated to the construction of dynamic models of the inhabited space. The aim of the theoretical part is to clarify three central notions: the inhabiting individual, the formal model and the inhabited space. The concurrent aim is to assess the possibility to quantify cohabitation and to study it by the means of formal dynamic models. A particular attention is attached to the role of these models as pragmatic and existential operators of the habiter of individuals and of the cohabitation of communities. In both methodological/empirical parts, the theoretical conclusions of the first part are applied in the construction of renewed formal models of the contemporary Swiss inhabited space: Firstly, the author adapts the method of population measurement in order to reflect characteristics of inhabiting individuals which have been identified in the theoretical part but that are traditionally not accounted for in such measurements: namely individual mobility and "polytopicity". The result of this approach is a new synthetic image of the population distribution in the inhabited space. In other words, the author proposes a method that allows bringing the diachronic dimension of the Swiss territory on a sheet of paper. While the time-scales of habiter taken into account in the first methodological part are of the order of a day to a month, the second part is dedicated to larger time-scales of residential mobility. This stance allows the author to propose a way in which to model space both as objet and as a constraint of individual intentionality. In this approach the author leans on existing multi-agent models, while providing these with a renewed epistemological foundation, conceived in the perspective of an articulation with phenomenological approaches of the inhabited space. Three models have been constructed in that scope, each examining a different degree of individual acting freedom. The first model studies the structuring of a urban network in a highly constraining environment particular to first Neolithic urban communities. The second model examines retroaction phenomena induced by the deployment of individual actions in the environment and shows the potential divergence between the rationality behind these actions and their actual results. The third model insists on the role of individual actors by showing the impact of their interiorized models of cohabitation on the structure of the urban space; furthermore, it studies the leeway of public policies to inflect thus emerging tendencies. Taken as a whole, this thesis leads to the construction of a plurality of articulated models of the territory. This articulate plurality allows for a commensuration of individual practices and for a comparison of places and spaces, without renouncing to the phenomenological project of an understanding of the modeled realities as emergent and subject to an ongoing process of ontological transformation. Besides its general theoretical outreach, the statistical model reveals a highly more hierarchized Swiss urban network then perceived in preceding models, as well as the demographic importance of some still underestimated places (notably touristic destinations). As to the multi-inhabitant models, these reveal the great impact of individual expectations and expectation-lead actions on the structure of a territory, thus calling for new strategies in public policies towards a sustainable urban planning.
EPFL2010
