1The two key words of this book are ‘settlement system’ and ‘transition’. The first refers to the ways in which human societies have occupied space in past and present times. It implies an interest in spatial actors and objects belonging to many levels: notably, individuals and families, settlements and transportation infrastructures, territories and networks, villages, agglomerations and urban systems. The second term refers to the passage from one ‘thing’ to another qualitatively different. Within a speech or a text, a transition (from the Latin transitio, a going across or over) makes it possible to pass from one idea to another and thus to join them together. In the social and natural sciences, ‘transition’ refers to the dynamic of change in a system which passes from a given state at a moment t₁ to another state, qualitatively different, at a moment t₂. This is the case, for example, with a regime change in a political system, with the transformation of water into gas in a physical system, with a population characterized by high mortality and birth rates that experiences low ones as part of a demographic transition. In this volume, we are explicitly concerned with transitions in settlement systems. These transitions relate to diverse processes: migrations of hunter-gatherers, settling and concentration of populations, emergence of cities and of settlement networks, fragmentation or fusion of territories. The volume consists of three parts corresponding to the three principal stages of the collective research undertaken. In the first one, the objective is to explain and discuss the different meanings that the terms ‘transition’ and ‘settlement system’ can carry, and to present the choices adopted in this work. The discussions led to the collective elaboration of a common conceptual framework. Secondly, the concepts thereby defined are put to work in a group of case studies ranging along different spatial and temporal scales: from the local level to that of the planet, from the out-of-Africa migration of Homo sapiens to the emergence of new urban forms in the present day and into the future. Finally, in the third stage, the meaning of the concepts and models elaborated and developed in the preceding chapters is re-examined in a comparative and reflexive perspective.

2This volume counts 35 co-authors representing different disciplinary traditions, who are extremely varied in their research practices and in their relations to space, time, data and models. The work is the fruit of multiple meetings, seminars, work groups and intersecting collaborations over a period of five years within the framework of the ANR project TransMonDyn.1 All of the participants had already had experience in interdisciplinary work, and the project is based on an advanced form of interdisciplinarity, nourished by a great profusion of cross-connections, whose results were not known beforehand. The co-authorship of each chapter by researchers in several disciplines, in a variety of combinations, most of them new, reflects the process of co-construction of the models developed as part of the project.

The emergence and functioning of the project TransMonDyn: ‘strength of weak ties’ and ‘choreography’ of research trajectories

3The project has benefited from a number of predecessors2 and it has forged new relationships between various laboratories and networks. The intersection of two conceptual fields derived from the social sciences makes it possible to capture the functioning of the TransMonDyn project: the theory of ‘weak ties’ of Mark Granovetter3 and the ‘time-geography’ of Torsten Hägerstrand.4

4The former distinguishes between the ‘strong ties’ and ‘weak ties’ (figure 1) that connect an individual with others. The strong ties link the individual with persons sharing the same kinds of knowledge, the same networks, the same visions of the world as himself. The ‘weak ties’ connect him with people and networks more distant. The theory of the sociologist Granovetter emphasizes the ‘strength of weak ties’, which are best able to open an individual to new things by establishing bridges between him and areas less well known to him. Thus, in the field of research, strong ties unite researchers sharing a common object of research, belonging to the same discipline or sharing the same methodological orientation, and who are used to working together. Weak ties are those that researchers form during participation in conferences and in the context of interdisciplinary research networks. They often stem from individual initiatives, even if they find a place within institutionally defined frameworks, reflecting the combination of ‘bottom up’ and ‘top down’ processes in the establishment of a research programme.

Figure 1 : Schematic representation of ‘strong ties’ (solid lines) and ‘weak ties’ (dashes) according to the approach of Granovetter

5Thus the conception of our project had its origin during the French-American ANR-NSF5 conference entitled ‘Dynamics in the Human Sciences: Cognitive, Behavioural and Social Complexity’, which took place in April 2009 at Reims. Present on that occasion were Arnaud Banos, Denise Pumain and myself from Géographie-cités (Paris), Laure Nuninger from the Maison des Sciences Humaines et de l’Environnement (Besançon), Timothy Kohler from Washington State University (Pullman), Jean-Marie Hombert from the UMR Dynamique du Langage (Lyon) and many others. The exchanges were intense during this conference, and especially between those researchers who later played a key role in the establishment and functioning of the TransMonDyn ANR project. Each of them had ‘strong ties’ with the researchers with whom they worked regularly and ‘weak ties’ with researchers coming from other disciplines, encountered in different research networks6 and sharing similar methodological and epistemological approaches. A double linkage underpins the project, involving both the groups having ‘strong ties’ (archaeologists and historians on one side, geographers on the other) and a number of researchers engaged in the project as individuals. The project was therefore fully able to make use of the ‘strength of weak ties’. Figure 1 illustrates the principle of this construction.

6At this stage, it is interesting to evoke the concepts of ‘time-geography’ introduced by Torsten Hägerstrand in 1970 and often applied today in the sciences concerned with space. They describe the choreography (a term often used by the author) of individuals in time-space, a choreography whose motive force is the ‘project’ and whose organization notably depends on constraints of capacity and co-presence. Varied and intertwined individual trajectories (of life and of activity) move within the time-space, intersect, travel in the same direction, separate, intersect again… To realize a ‘project’, a group of these will cluster within a spatio-temporal ‘bundle’, a shared time-space, sheltered from interference by the daily life of research, where they will create a “local pocket of order” (according to Hägerstrand’s terminology). In the case of TransMonDyn, the ‘project’ concerns research on the transitions in the settlement systems and the ‘local pocket of order’ refers to a series of seven cross-disciplinary seminars, inscribed within discontinuous time but forming a whole whose existence has allowed for creativity and the emergence of new collaborations. As an example, Figure 2 represents the space-time ‘bundles’ corresponding to two of these seminars (A and B in Figure 2): all the trajectories of individual researchers and sub-groups with strong ties intersect and share the same space during the same period. These space-time bundles made it possible, by the interplay of the multiple face-to-face exchanges that took place, to set out fully and confront varied points of view on the transitions, the settlement systems, their conceptualisation and modeling. These moments, thoroughly shared, fostered reciprocal understanding and created a “local pocket of order” which issued in an ‘organized’ diversity of points of view rather than a one-track thinking. It was in this context that a common conceptual framework could be developed (cf. chapter 2).

Figure 2: Organization of the trajectories of the participants in the TransMonDyn project

7In parallel with the collective work developed within the space-time bundles uniting all the participants (A and B in Figure 2), the exchanges which took place led to the emergence of a number of sub-groups, varied in composition, certain of which functioned over a short span, while others were more long-lasting. Thus, sub-groups (indicated from 1 to 10 in Figure 2) formed around case studies and gave rise to the co-construction of models (cf. chapters of Part II of the volume). Other sub-groups subsequently formed around transversal work dealing with the comparison and the confrontation of the concepts and models elaborated during the case studies (cf. chapters of Part III).

8Finally, further sub-groups correspond to a “pocket of local order” distant in space but concentrated in time. Such is the case with a seminar co-organized with Laure Nuninger, Timothy Kohler and Stefani Crabtree, which took place in Pullman at Washington State University. Ten or so members of the project TransMonDyn participated, and this encounter made it possible to confront the American and French points of view on the modeling of settlement systems over the long term. A number of ‘weak ties’ already existed with the American team – on the one hand, with the archaeologists of Besançon, to which it was closely related in terms of discipline and objects of research, and, on the other hand, with the geographers of Géographie-cités, with which it shared research methods concerning the dynamic of complex systems, along with the practice of multi-agent system modelling. The meeting contributed to transform these “weak ties” into “strong ties”.

The crossing of ‘thematicians’ and ‘modellers’ skills for an interdisciplinary co-construction

9As appears from the Table of Contents, the researchers grouped themselves according to variable geometry. This procedure attests to the interdisciplinarity practised both at the level of the project as a whole and at the level of each case study. The terms ‘thematician’ and ‘modeller’ are often used in the volume. These two terms are understood as follows:

• ‘thematicians’: this applies to those researchers possessing expertise and research experience with respect to settlements. In this group one finds historians (François Favory, Samuel Leturcq and Marie-Jeanne Ouriachi), archaeologists (Frédérique Bertoncello, Jean-Luc Fiches, Pierre Garmy, Elizabeth Lorans, Laure Nuninger, Elisabeth Zadora-Rio), geographers (Anne Bretagnolle, César Ducruet, Denise Pumain, Lahouari Kaddouri, Céline Vacchiani) and linguists (Christophe Coupé and Jean-Marie Hombert), who are concerned with settlement systems from various points of view and in different time periods. The work of these thematicians consists in: reconstructing and understanding the organization, particularly in spatial terms, of the settlement system in present or past periods; relating this to the political, economic and cultural organization of societies; analysing the parallels between its evolution and such events as climate change or conflicts, as well as such processes as, for instance, the transmission of languages. Each thematician anchors his reflection in precise areas and epochs, while taking account of similar spatio-temporal phenomena situated in other contexts.

• ‘modellers’: here, too, we are dealing with researchers of different orientations, coming from the formal sciences, mathematics (Alain Franc), computer science (Jacques Ferber and Thérèse Libourel) and philosophy (Pierre Livet and Franck Varenne). These researchers are interested in the generic characteristics of systems. The objective is to formalize their functioning, to identify the conditions of their appearance, transformation and disappearance, and to study the properties of the models of these systems. The models are varied, certain aiming to analyse data and ‘make them speak’, others to explore some thought experiment by means of simulation.

10Far from constituting two homogeneous categories with fixed boundaries, these terms make it possible instead to distinguish two fundamental and complementary ‘roles’ within the project, as much in the elaboration of the common conceptual framework as in the co-construction of the models of transition. Rather than a separation between these two categories of researchers, there is a near-continuum (Hélène Mathian and Xavier Rodier, for example, are at the interface between their profession of geomatician and modeller and their particular thematic domains, geography in one case, archaeology in the other). Several ‘thematicians’, moreover, have also acted as modellers in the project (Arnaud Banos, Lena Sanders, Cécile Tannier and Denis Phan, for example), with some even combining the two roles (Timothy Kohler, Florent Le Néchet). Finally, we should note that even within a sub-group of researchers a priori connected by ‘strong ties’, there are differences of orientation. Such is the case, for instance, with regard to the historians, divided between antiquarians and medievalists. It is also the case with the specialists in spatial analysis, comprising researchers who are inspired by methodological individualism and give priority to individual choices and others who formulate their hypotheses at the level of spatial entities. Likewise, with regard to the modellers, the points of view on modeling differ between mathematicians and computer specialists in multi-agent systems. The expounding of these different points of view and the resulting recompositions of the strong and weak ties enabled the development of a dynamic interdisciplinarity concerning the transitions of settlement systems.

11A further interest of the collective structure lies in the continuum of generations present in the project: researchers with long experience who had participated in the Archaeomedes programmes (see note 2); advanced researchers in the middle of their careers; doctoral students completing their theses (Sébastien Rey and Clara Schmitt); and even younger ones, whose thesis subjects emerged in the course of the project (Stefani Crabtree, Robin Cura, Julie Gravier, Ryma Hachi and Lucie Nahassia).7 This association of researchers of different generations, with all of them contributing to the collective reflection and to the composition of the book, made it possible to closely connect the transmission of acquired knowledge and expertise with the emergence of new directions to pursue.

A volume structured in three parts

12The book is organized according to three principal parts, which record the productions and results associated with three major stages in the evolution of the project.

13The first part (three chapters) is devoted to the conceptualization of a transition in the settlement system. Chapter 2 represents the backbone of the work. It describes the collective project, the mode of proceeding, the key concepts, and the conceptual framework developed collectively, besides introducing the case studies. It was composed by the 35 researchers, who each, at a certain point, contributed a brick to the collective reflection. The conceptual framework constituted collectively provided the solid basis for the reflection and the work effected in the different case studies (a dozen). This chapter is framed by two others: one presents a ‘state of the art’ survey of the concept of transition and its modelling in the fields of the human and social sciences, the socio-environmental sciences and computer science (chapter 1), while the other proposes an introduction to the practice of modeling and simulation on the basis of a comparison of different models relating to human migrations in contexts of competition for resources (chapter 3).

14The second part comprises 10 chapters, each focusing on a transition within a settlement system. These chapters couple a narrative with a precise description of the facts that led the thematicians to designate the change in question as a transition. The narrative follows the interpretive plan developed collectively. Each of these chapters then applies the collectively constructed model to one aspect of the transition identified. Seven case studies were produced in the context of the project and are the fruit of new collaborations. The work groups, comprising modellers and thematicians, constituted themselves according to a process of self-organization which followed the collective reflection on the concept of transition and the presentation by the thematicians of the cases which seemed suitable to them. The composition of these groups varies, as is clear from the Table of Contents. A few examples suffice to illustrate the diversity of disciplinary combinations. Geographers played the role of modellers for linguists, their common interest being to simulate the spatial diffusion of settlement. The same mathematician modeller collaborated, on the one hand, with archaeologists to study the power games played out between Romans and Gauls by means of game-theory and, on the other hand, with a geographer to analyse the co-evolution of networks of cities and networks of roads by means of graph theory. Other chapters involve more substantial collectives featuring, at the same time, multiple archaeologists, geographers and computer scientists. In two chapters, the authors have combined the roles of thematician and modeller, one on the formation of Pueblo villages during the first millennium, the other on the emergence of polycentric urban regions, a transition currently in progress. In both cases, the models benefit from previous studies, which have been adapted to take account of the conceptual framework defined collectively. A final case, dealing with the emergence of cities on the basis of the diffusion of innovations, also takes account of work previously developed, this time in a close collaboration between geographers and computer scientists.

15The four chapters of the third part are each founded on all the case studies taken together, in order to propose a transversal and comparative reflection on the transitions whose models figure in part 2. Chapter 14 concentrates on the relations between narratives and models, raising the question of the differences between these two approaches in producing scientific knowledge. Its authors take up the notion of ‘narrative as model’ and discuss the place of the model between data and theory. Chapter 15 considers how to capitalize on and transmit the knowledge acquired in the programme, posing the question of the reproducibility of models in the human and social sciences. Following a theoretical debate on this question, two applications are proposed: one concerns the transfer of the conceptual framework to case studies dealing with intraurban space; the other deals with the reproducibility of one model in particular. Chapter 16 proposes an ontological procedure for comparing the totality of the transition models set out in part 2. This procedure is based on setting out in detail the objects, properties, relations and processes at issue in each transition and considers the choices made during the conceptualization of the models. Finally, chapter 17 proposes feedback on the processes leading from the elaboration of a transition to the co-construction of its modeling. The approach is reflexive and compares the ways of co-constructing models as a function of the profiles of ‘thematicians’ and ‘modellers’. It also considers different individual perspectives on the collective experience.

16This introduction has placed a great deal of emphasis on interdisciplinarity. That approach has indeed played a key role in the project and in the composition of the book. It has necessitated having at our disposal enough time – a ‘time-space pocket’ sufficiently capacious – to make possible the accumulated exchanges needed for the co-construction of the models of transition and the collaborative writing of the chapters relating these experiences. The objective has been to communicate both the variety of points of view and the convergences effected among them in order to arrive at valid models.