Introduction

1Developed in the United States from the 1940’s onwards by J. Stewart and G. Willey (for a historical review, see Trigger 1967), reflections on prehistoric territories are at the heart of debates concerning the spatial structuring of collectives (sensu Descola 2005) and the notion of prehistoric cultures. We present here the polysemic concept of prehistoric “territory” from the perspective of various disciplines to propose a definition adapted to prehistoric archaeology.

What is a territory?

The geographical vision

“The map is not the territory” A. Korzybski (2007)

2Etymologically the Late Middle English term ‘territory’ we still use is derived from the Latin word territorium (Latin) that designate broadly the extent of the Earth on which a human group lives (Jaillet 2009; Kourtessi-Philippakis 2011). In the eighteenth century, this word entered administrative and legal languages with particular reference to the jurisdiction and authority of a state, taking up the Latin meaning of jus terrendi (right to terrify) established under the reign of Justinian (sixth century). It was not until the 20th century that the natural sciences, first and foremost ethology and especially ornithology (Howard 1920) took up this term to designate the exclusive living space of an animal, access to which is forbidden to other individuals, including those of its own species. In France, since the 1960’s, the term “territory” has been common in administrative language in connection with the development of national development projects and in law as a synonym for administrative subdivisions of the State (country, region, department) (Négrier 2009). However, although present in the discourse of some human geographers as early as 1940-41 (Demangeon's “territorial base” cited by Meynier 1969:67), “le territoire existe comme idée sans le mot”1 (Ozouf-Marignier 2009:32) and the term, timidly used by geographers in reference to the “geographical preoccupation” of certain indigenous peoples (Sorre 1952, 1957), was only really introduced into the social sciences in the 1980’s, laying the foundations for a new branch of geography, namely social geography (Frémont et al. 1984).

3Society has long spoken of environment, human ecology, synecology2 (Schröter 1902), terroir, landscape(s)... (Sorre 1943, 1957) and consequently social geography remained, in its beginnings, strongly impregnated by physical geography and the naturalistic paradigm with some concern for humanism and human occupation of a global earth. In this “ambient cacophony” (Séchet & Keerle 2009: 85), ‘territory’ quickly became synonymous with ‘geographical space’3 (Mérenne-Schoumaker 2002). The fundamental relationship between humans and space, a key element of the notion of territory from the 2000’s onwards, has not yet been completely assimilated across the humanities. Di Méo (1998 a) observed at the end of the 1990’s that the term ‘territory’ is balanced between two poles; an exclusively political vision of territory (nation-state) - gradually substituted for certain peoples by ‘tribal cultural unity’ (c.f. Malinowski's tribe-nation 1968) - and a naturalistic vision inherited from biology.

4At the beginning of the 21st century, the notion of territory takes shape and is established around three fundamental axes:

1. the characterization of the links between humans and space: “how does humans manage space and how is space interpreted by humans?”
2. the inscription of space in time since, according to M. Marié, “l’espace a besoin de l’épaisseur du temps, de répétitions silencieuses, de maturations lentes pour devenir territoire”4 (in Ozouf-Marignier 2009: 34); and
3. the spatial inscription of a political institution.

5In other words, a territory is the fruit of four components whose cumulative expressions allow for its establishment:

• the collective dimension: insertion of the individual into the social reference group,
• the political dimension: mode of division and control of space,
• the symbolic dimension: the set of heritage values,
• the historical dimension: temporal inscription of the territory.

6Although it necessarily stems from a societal rather than individual construction, the diversity of meaning given to the territory also stems from the multiplicity of actors and the scale at which it is being considered (Bernus 1982). In this respect, Debarbieux (2009) proposes a series of concrete examples defining not one, but several territories according to the mode of organisation of space (type of spatiality) and the scale considered (imagined territoriality) (table 12.1). In his conception, 1) the terroir is defined as an entity of proximity experienced by each person within these components, 2) the network area corresponds to contiguous surfaces whose elements are held together by networks (as in the case of nation-states5) and 3) the network of places is characterized by the arrangement of non-contiguous places attached by a system of practices and exchanges. This last category is of particular interest to prehistorians and echoes Bonnemaison's observation that a nomads territory “se rattache à un certain nombre de points forts fixes et à des itinéraires reconnus qui déterminent les territoires d’errance”6 (1981: 254).

7This dynamic vision of territory, which concerns the division of geographical space into a patchwork of contiguous areas (q.v. “cultural mosaic” during prehistory, e.g. Otte 1992; Klaric et al. 2009; Bruxelles & Jarry 2012; Vialou 2015; Ducasse et al. 2017...), thus opposes active and fluctuating networks to the static “territory”, which is considered to be resistant to change (Dicken et al. 2001). This hypothesis ignores thus all the inter-actor and/or inter-place connections that may occur within space. We agree on this point with Dicken et al. (2001) who consider that networks are integrated into one or more territories and that these territories are themselves included in networks. Consequently, territories and co-existing networks are not mutually exclusive but on the contrary, are interdependent and “inter-active”. From the actor's point of view, G. Kourtessi-Philippakis points out that in the Anglo-Saxon world (2011: 8):

la territorialité [le fait de mettre en place un territoire] exprime la tentative par un individu ou un groupe d’affecter, d’influencer ou de contrôler d’autres personnes, phénomènes ou relations et d’imposer son contrôle sur une aire géographique, appelée territoire.7

8This idea, which stems directly from ethology and economics in the broadest sense, would thus make it possible to optimize the access of an individual or group to resources on either a temporary or permanent basis. Once again, the detachment of a social science outlook from naturalistic influence is incomplete and the concept of territory seems inseparable from that of appropriation and privatization.

9The notion of territory certainly depends on the scale considered, but also on the actor in question. One must therefore ask oneself what degree of analysis is the most relevant to respond to the problem: that of the individual, the family unit, the residential group, the ethnic group, etc. As understood today in social geography (supra), ‘territory’ corresponds to the occupation of a given space by a group that shares the same value system and whose culture is identical (Di Méo 1998b). Consequently, the search for the organization of space (i.e. territory) on the intermediate scale of the cultural group is possible on the basis of material productions8 (distribution of categories of objects, function of places, establishment of infrastructure that link them, etc. (Séchet & Keerle 2009: 90):

considérer le territoire sous la dimension spatiale de la société, c’est-à-dire comme produit de l’action collective et non le définir comme simple résultat de l’action étatique (territoire stato-centré), c’est le définir comme le produit de la dynamique des acteurs sociaux.9

10This distinction between individual and collective thinking also marks the difference between psychology (individual) and sociology (group) in the social sciences. It should be noted that for societies considered on a larger scale, i.e. civilizations, we no longer speak of a territory but sometimes of a cultural area (Retaillé 2009). This individual/group/civilization distinction is therefore found in geography, sociology and ethnology (Le Mouël 1978: 89):

Entre sila, l’univers, et idglo, la maison pérenne, existe un certain espace que la langue ne reconnaît pas comme entité, mais avec lequel pourtant, l’homme entretient des rapports spécifiques. Aire de dimensions précises dont les moindres recoins sont connus pour leurs possibilités alimentaires, espace qui se définit, dans le cas présent, par rapport au point de sédentarisation, parcouru selon des trajets propres à en tirer les ressources qu’il recèle, domaine non possédé de jure mais de facto par les habitants qui le sillonnent: c’est le territoire. Lieu de vie du groupe, d’où il tire les expériences au travers desquelles il construit sa conception du monde, ce mésocosme, tout comme l’univers qui le comprend et le microcosme de l’habitation qui s’y inclut, s’ordonne sur les grands axes qui orientent le monde.10

Territory and ethnography

11In Western sedentary and modern (Jolibert 2012) conceptions of space, the slightest fragment of land “belongs to” and “is managed by” a physical or moral person without, in theory, there being any empty zones within a group's territory (perceived as the sum of individual possessions); such structuring of space is totally different from that of current and recent nomadic peoples, regardless of their economic system. These opposing views, necessarily Manichean in light of the modes of occupation of space (table 12.1), envelop a real difference in cognitive terms describing the structuring of space, between sedentary and nomadic populations (Debarbieux 2009; Frérot 2011; Descola 201611). Within nomadic territories, named places corresponding to recognized points of interest or “hyper-places” (Lussault 2017; Banos 2009) are followed by itineraries without any status, the ‘non-places’ (Augé 1992). Voids located away from regular routes or outside identified places are parts of the landscape that are only occasionally visited and appear as if they were outside the territory (e.g. Tindale 1974; Tilley 1994; Collignon 1996; Frérot 1999). ‘Emptiness’ is therefore not only the result of an historiographical deficiency or gap imposed by geomorphology, it is one of the principles of the organization of nomadic space (e.g. Bonnemaison 1981; Bernus 1982; Debarbieux 1995; Collignon 1996; Descola 2005, 201612; Pedersen 2007; Nabokov 2008). Thus, among the Inuit of Hudson and Baffin Bays, as among many so-called nomadic peoples, the perception of the territory is not expressed by a surface, but in a linear form by routes linking points of strong interest (Rasmussen 1930; Michéa 1949, 1967) (figure 12.1).

Figure 12.1

12Aboriginal Australia, where religion13 and totemic worldviews are in many ways quite close to Palaeolithic thought, (A. Testart 1992, 2012, 2016), provides a substantial comparative basis because “la matière ethnographique (…) est plus riche qu’on ne le croit ordinairement”14 (Testart 2006:394)15. Routes between identified and thus named places are indeed perfectly known and, with varying degrees of cognition, are shared within organized (clans, moieties etc.) and mixed (initiated/non-initiated, women/men) groups and between various totemic groups. However, these places are only one element among others of mythico/religious16 constructions of perceived worlds: points within physical space where territories belonging to myths and symbols merge in the same plane of reality. In perfect logical continuity, certain movable, rupestrian or ephemeral artistic manifestations are the expression and communication of this ‘dreamed’ but concrete territoriality in which there is a notion of possession of land-use, but where the renegotiation of land in successive generations sometimes explains the blurring of certain tribal boundaries, the latter usually depending on natural environments, but above all on factors that are largely beyond our rationality (Peterson dir 1976; Hamilton 1980; Glowczewski 1981, 2006; Testart 1992, 2016; Dussart 1993; Faulstich 1998, 2003).

13In fact in its various degrees of symbolism, the complexity of the mythico-religious system is matched by territorial complexity and we can well imagine the immense difficulty that this presents for the prehistorian in a field where the simple analysis of cause-and-effect relationships is inoperative, since the same fact has multiple interpretations depending on the context (see Sack's Unsophisticated-fused thought, 1980, or Foucault's concept of heterotrophies, 1984). The depth of Aboriginal thought revealed by the Australian ethnographic school and highlighted by A. Testart (1992) makes it possible to try to interpret archaeological facts, whereas the phenomenal accumulation of ethnological data by L. R. Binford (2001) remains unconvincing in this regard (Testart 2006). It could be claimed that a comparison with semi-nomadic peoples of the “hot” period ultimately offers little basis for understanding the territories of the semi-nomads of the “cold” period in prehistory (Leroi-Gourhan 1936), however, we consider that the conception of space goes beyond the Australian Aborigines alone and it seems that the relationship between spatial structuring, symbolization and territorialization is universal, even seen for example among the Tuareg (Bernus 1982; Frérot 2011) and the Inuit (Le Mouel 1978; Collignon 1996),

14For these reasons, and by extrapolating from the observations of B. Debarbieux (2009), it is clear that between the two extreme poles of territoriality - the bounded and continuous fragmentation of nation-states versus the open and discontinuous networks of places, there are an infinite number of territorial constructions that are closely dependent on the degree of mobility of human groups. This multi-reality and its explanations at different levels of perception and action constitute a central topic for contemporary and future prehistoric studies.

The notion of territory in prehistory

Brief critical review

15In Europe, in the naturalistic and postmodernist tradition that characterizes the Old Continent, the approach to prehistoric territories was essentially based on the origin and processing of lithic raw materials at the local level (e.g. Masson 1979, 1981; Demars 1982; Torti 1980; Geneste 1985, 1988; Mauger 1985) and more recently on antler processing and/or bone industry (Averbouh 2005; Pétillon 2013), on adornment for long-distance relations (e.g. Taborin 1993, 2004; Vanhaeren & d’Errico 2006; Rigaud 2011; Peschaux 2017) or on art styles (see below). References to the work of Anglo-Saxon archaeology or cultural and social geography have only recently enriched the debate (e.g. Féblot-Augustin 1993; Bracco 2004, 2005; Aubry 2005; Audouze 2007; Jarry et al. 2008; Rasse 2010; Aubry et al. 2012; Goval 2012; Sauvet 2019; Fuentes et al. 2019) with some exceptions generated by the “prehistoric ecology” movement since the 1970’s, e.g. H. Delporte (1978) pertinently citing the relationships between particular places (Elkin 1967). This observation was made at the beginning of the 20th century and, except for a few precursors (Deffontaines 1924; Champier 1955; Nougier 1959), paleogeographic discourse was established on the fringes of these disciplines. Until now in prehistoric studies it is the strictly economic vision of space (wrongly confused with territory) that has prevailed. As if any space can accommodate a society, as it constitutes the setting and background for action: space is simply exploited and travelled through and not perceived and socialized. Despite art specialists having integrated the extra-economic dimension of territories into their analyses early on, since the sharing of symbols between individuals is a stronger link than the “simple” sharing of technical behaviors sensu lato (e.g. Leroi-Gourhan 1981; Gamble 1982; Fritz & Tosello 2005; Cazals et al. 2007; Pinçon 2007; Djindjian 2009; Bourdier 2010; Lorblanchet 2010; Lenssen-Erz 2012; Rivero 2015; Honoré et al. 2019), their works nevertheless often remain focused on the sole distribution of one or more symbols (or objects of symbolic value) in space17 and only rarely do they consider the extremely variable underlying structures and/or organization, which are dependent on the register considered (see e.g. the spatial subdivisions of the Middle Magdalenian in Angevin & Delvigne in press). In this respect, the territorial approach in prehistory remains too often disconnected from other disciplines and requires a change of perspective, a clear definition, and tools for reconstruction, both contemporary and future.

16Territories only exist because of their occupation by a society, since the territory of a group emerges from the intimate intertwining of space and society18 (see above). In other words: beyond the perception and knowledge of a space, it is practices and planning that lead to the structuring of space and they transmute space into territory. The foundations of the territory therefore lie in the structuring of spaces (and places) and the geometry of their inter-relationships. So far, in Palaeolithic archaeology, the territory is at best understood as 1) the geographical space travelled over and exploited from a site or 2) as the area in which specific types of objects or technical traditions are distributed, (Féblot-Augustin 1997, 2009; Djindjian et al. 1999; Kozlowski 2005; Jaubert & Delagne 2007); such variety of actors and scales do not lend themselves well to large scale reconstitutions (Gamble 1998). If this semiological drift of “territory” is partly due to the recent acceptance of the link between space and societies in geography (see § 1.1.), it also stems from the usual methods of study resorted to by prehistorians. Archaeological research has been, and continues to be, built up in the field following the excavation of sites (often over several years). Generally, this approach systematically places the site at the heart of reasoning about the related territory. Thus, territorial models of prehistory were therefore biased and established from a centralist perspective, structured around major centers (the sites) from which various forms of knowledge were disseminated and within which resources were collected. This analysis seems difficult to apply to Paleolithic hunter-gatherer peoples reputed to be nomadic, who occupied various sites in turn during cyclical travels, travelled through diverse areas and frequented dissimilar places of varying economic and social value, including various habitats, water points, hunting areas, places where plant or lithic resources were obtained, sacred spaces, fords, etc. Such a “site-centric” analysis of prehistoric territories provides at best a truncated vision of the space exploited and managed by nomadic groups and does not allow the identification of territories in the past. Although such a constrained approach was justified in the process of building the discipline, the accumulation of data and new methodologies available allow us now to go beyond it.

17Identifying territories in prehistory is therefore a challenge, as it requires taking into account a great deal of data relating to the spatial distribution of material and immaterial objects and places, the analysis of which is inherently heterogeneous, carried out by various disciplines and different teams coming from varied schools of thought. Furthermore, it is necessary to link sites that are not strictly contemporary and therefore represent discontinuous spatio-temporal records. On this point, however, social geography and ethnography make it possible to avoid the pitfalls, insofar as places are the guardians of a society's history and memory. Their organization (i.e. territory) remains similar if the cultural systems that contribute values to it are analogous (e.g. Bernus 1974, 1999; Bonnemaison 1981, 1986; Di Méo 1998a; Collignon 1996). In this respect, if territoriality is the terrestrial dimension of the human condition (Debarbieux 2009), then territory corresponds to the spatial imprint of culture (Sack 1980), or at least reflects an area in which a society’s culture is reproduced, and the Australian example is enlightening in this respect. It therefore seems highly probable that recurrences in the function of prehistoric sites and in the use of certain geomaterials are reflections of a phenomenon that gives structure to past behaviour and pro parte allows relationships between spaces to be deciphered.

Towards an approach to networks of prehistoric places

18Since they constitute the category of vestiges which, by their very nature are spatial tracers, geomaterials (shells, mineral pigments, collected or worked rocks etc.) are privileged witnesses of inter-spatial relationships. Biomaterials (animal parts, fossil woods, bones), and isotopic compositions that can be analyzed also provide information in terms of spatial analysis, but for the moment at a lower resolution. The same applies to the spatial distribution of knowledge, storytelling and ceremonies which may show the existence of spatial and/or temporal links sometimes over exceedingly long distances. While technical systems coded by conscious and unconscious norms, (Lemonnier 1983) play a role in the spatial distribution of techno-systems (in the sense of Roux 2000), they do not respond to the same constraints as the circulation of materials. The analysis of their coherence as technical entities and their distribution comes as a second step in their analysis in order to ascertain any correspondences between material circulation networks and the networks connected with technical manipulation of the resource (Delvigne et al. in press). During analytical studies, it is necessary to consider all variables in an exhaustive manner, using methods adapted to archaeo-materials which consider their evolution and transformation, characteristics which may modify their spatial signal19. If exhaustiveness is a lure in archaeology, it is nevertheless true that the observation of a large quantity of remains makes it possible to isolate some materials which show up as single specimen. This is also true for complex objects (ornaments, dyes) just as it is for so-called “everyday” objects forming the lithic industry. These symbolic versus utilitarian subdivisions are, however, often the result of prehistorians’ imagination than of prehistoric reality and the symbolic value of the lithic industry - if the symbolic is perceptible at all - may be expressed unexpectedly in the form of an object that appears to be non-utilitarian but is of a very specific origin and is present as a single exemple (Gould 1980) and thus meets the definition of “precious” (Godelier 1996).

19In addition to these studies, research into prehistoric territories is based on a combination of three complementary approaches (figure 12.2). Firstly, the identification of litho-spaces20; secondly the search for operating patterns indicated by raw material type and then by geotope21; and thirdly the implementation of reticular analyses.

Figure 12.2

20Determining the diversity of materials found in an archaeological level and their geographical spread allows the delimitation of a litho-space. The establishment of chaînes opératoire and reduction diagrams for each type of material are also powerful tools for interpreting the exploitation of resources and material acquisition methods, either direct or indirect22. By cross-referencing data on the diversity of geomaterials from the same geotopes and their quantities and modes of introduction into the sites, one can recontextualize the objects. The significance of the materials in terms of human behaviour is dissimilar depending on whether a single type from a space several tens of kilometres away is introduced into the site and treated in a particular way, or whether different genetic23 types from the same space are brought to the site and treated in a manner similar to the materials that outcrop nearby. Usually however, geomaterials have been treated in large groups (e.g. “local”, “semi-local” or “distant”) which place the site at the centre of the analysis and generate numerous artificial subdivisions with little real value in regard to the spatial behaviour of human groups.

21This is paradoxical from the point of view of spatial management and the study of territories. Social and cultural geography have clearly shown that only the set of interconnections between places is meaningful for each space and allows territorialisation (see above). Therefore, only the network of litho-spaces obtained for (sub)contemporary sites and their integration into the group of all geographical values, allows a glimmer of understanding of the management of space by a given paleolithic group. In order to mobilize these concepts, however, it is necessary to integrate into and adapt for prehistory the techniques of network analysis which have developed in spatial archaeology since the early 2010's (cf. network analysis; e.g. Wasserman & Faust 1994; Classen 2008; Brughmans 2010; Knappett 2013; Blake 2014; Collar et al. 2015; Mills 2017; Amati et al. 2019; Peeples 2019). These techniques allow the establishment of models that are no longer focused on the site or network node, but on the relationships or network links maintained between several sites by taking into account various parameters like types of material found, distances to sources, distribution of object types, etc. Adapted to our problem, the so-called “small world models” from proximal point analysis considering all significant locations, including those of raw material deposits, are particularly revealing (figure 12.3). This process frees the site from being the central object of study and makes possible the broader structuring of spaces and geotopes (white circles in figure 12.4) and highlights the modalities of material circulation within the networks (i.e. the territories).

Figure 12.3

Figure 12.4

22While network analyses allow the establishment of models, they should be considered as analytical tools and not as images of the territory of human groups. As such, their structure is variable and eminently dependent on the accuracy of the data (Isaksen 2013); a fact which further supports the need to record data in a comparable manner with equivalent degrees of quality for the widest possible sample of sites.

Conclusion

23Definitions of territory vary from the simple to the most complex. This is evidenced by the comparison between the definition given by an Australian Aboriginal man, “My country (i.e. territory) is the place where I can cut a spear or make a spear-thrower without asking anyone” (Tindale 1974: 18) and that of A. Moine, a geographer from the Université de Franche-Comté, France (Moine 2007: 45):

24[le territoire] est un système complexe dont la dynamique résulte de boucles de rétroaction qui lient un ensemble d’acteurs et l’espace géographique qu’ils utilisent, aménagent et gèrent24

25From these two definitions from radically different contexts, one empirical and pragmatic and the other scholarly, two aspects stand out: that of the spatial domain (geographical place and space) and that of its anthropization (the pronoun “I” and a set of actors). The territory is therefore a plural object, built on the perception of its geographical essence by society (actors). Space is experienced, apprehended and managed in a permanent process which puts it into perspective by, with and within a culture. A society interprets and constructs its geographical environment by giving meaning to perceived natural objects which themselves increase in importance as they become recognised and valued. “C’est par l’existence d’une culture que se crée un territoire, et c’est par le territoire que se conforte et s’exprime la relation symbolique existant entre la culture et l’espace”25 (Bonnemaison 1981: 254).

26The conception of nomadic space is structured in the form of networks of places (significant ecosystem elements and therefore the geographical space in the sense of Di Méo (1998a), whose symbolic value varies according to society's value systems) and merge pro parte into the networks of materials (raw and manufactured objects) circulation. The territory is made up of named and well-identified full places, the hierarchy of which varies according to their cultural value, of “non-places” which are itineraries linking the well-identified places, and of voids, empty spaces that are rarely frequented or opportunistically used for specific events (e.g. Bonnemaison 1981; Bernus 1982; Debarbieux 1995, 2009; Collignon 1996; Nabokov 2008; Frérot 2011). Even though in prehistoric studies we are dealing with peoples reputedly nomadic, territorial thinking was based on models of nation-states developed by researchers from the sedentary, closed, and continuous Western world. A different perspective is necessary when considering Palaeolithic territories. The opposition between space apprehended from emic/etic points of view must be overcome because it opposes and dissociates in a sterile way sensitive experiences and representations of the world belonging to different knowledge regimes. On the other hand, the complexity and dynamic dimension of the experiences and representations of the world of societies, whatever they may be, must be taken into account. Consequently, descriptions of the social and spatial logic of human societies must consider their movements, supply of resources, etc. —and must be formulated with great methodological caution and, above all, epistemological clarification of their scope and meaning. This epistemological break, some would say paradigm shift (Kuhn 1962), although not a sharp break with the traditional approaches on which it is based, consists not only of a (re)contextualization of all the elements of reflection at our disposal which can no longer be dealt with in a global manner, but with attention to their uniqueness and interdependence. This must accompany a change in the status of the archaeological site, which must be integrated more globally into the socialized space, its pertinence brought down to a place level only and considered as it among other significant elements of the landscape.

Acknowledgements

27We would like to thank the editors of this book, Aline Averbouh, Nejma Goutas and Sophie Méry, for inviting us to present our ideas and offered a very stimulating criticism. We are indebted to an anonymous reviewer made very relevant comments which forced us to clarify some sensitive points of our argument. Collaborations with Raphaël Angevin, Pierre Allard, Jean-Pierre Bracco, Peter Bindon, Paul Fernandes, Audrey Lafarge, Mathieu Langlais, Damien Pesesse and Christophe Tufféry fueled our reflections about the notion of “territory” and usefully questioned the relevance of its use in prehistoric studies.