1 A symposium on the “Spatial Turn in History” held in February 2004 at the German Historical Institute in Washington provides me with an opportunity to offer some observations on the appearance of a new approach to history. [1] The phrase “spatial turn” is indicative of the increasing propensity of historians, beginning in the 1990s, to pay attention to the spatial dimension in studying the past and, consequently, forge even closer connections with specialists in geohistory and cultural geography. [2] Yet this is an ambiguous development because it seems obvious that the culturalist approach from which it primarily derives orients it almost exclusively toward symbolic aspects, which compromises many promising presuppositions from the very beginning.

2 Consequently, in this paper, I shall attempt to define the “spatial turn” through the formulations of one of its founders, Denis Cosgrove, and through one of the central categories of his approach, namely the landscape. In particular, I shall attempt to reconstruct the way in which a reductively symbolic reading of this analytical category became predominant. At the same time, I shall attempt to show the potential of alternative methodologies, such as historical ecology, which have led to substantive and processual analyses of the landscape.

The Spatial Turn and Symbolic Landscape

3 The late Denis Cosgrove, who can rightly be considered one of the standard bearers of the new theoretical turn, [3] proposed a genealogy of the “spatial turn” and then suggested several paths that might be taken. His point of departure is quite interesting. Assembling several a number of ideas borrowed from several human sciences, he emphasized how the growing interest in space as historiographical discourse was an aspect, if not a consequence, of the culturalist turn in the last third of the twentieth century. [4] According to him, this turn should be related to a redefinition of the concept of space, i.e., the abandonment of the absolute, or Cartesian, concept of space, and the assertion of a relative concept that takes into account other processes and phenomena, and in particular interactions of scale. This new awareness makes possible, for example, the radical critique of concepts such as “region” and the launching of new – though often quite divergent – research in the relations between the global and the local. Another result is the growing interest in studying place as a process rather than ontologically.

4 To understand the success of the “spatial turn,” Cosgrove considers the concept of landscape essential since it links geography to history tightly. In his view, the concept of landscape is a “direct expression of modernization.” It is the result of the social and spatial process that modified the older, continental concept of Landschaft. [5] From a vision of local space centered on customs corresponding to (unified) social entities that are expressed in the law and in bodies and designate a spatiality acquired in practice, we move to a visual conception of local space. This implies permanent property rights and the suppression of customary rights, a change in the scale of spatial control, and a link to a specific political culture. Cosgrove then reconstructs changes in the concept of landscape in the nineteenth century and affecting its power to express and challenge,; differences in control over the concept depending on whether it is exercised by the State or a Romantic culture,; its pernicious effects on geographic science (particularly German geography),; and finally the culturalist critiques of American geographers, mainly Carl Sauer.

5 For the “new cultural geography” advocated by Cosgrove, the concept of “landscape” has a two-fold nature: it refers to natural and social processes, and it corresponds to consequences of human actions that transform the material world. In other words, the landscape is simultaneously a natural and a cultural space. In both cases, the move from Landschaft to landscape is a long journey, from its medieval roots to the contemporary capacity to develop the idea of relative space. As place, the landscape plays a key role in historical conceptualizations since it is nothing less than the move from the territorialization of space to its relativization. [6] This is a powerful statement. It is not at all obvious that we should abandon a substantialist understanding of the landscape, understood not only as territory or scenery but also as link between community, justice, nature, and environmental equity—a meaning that goes back to the end of the sixteenth century, when the term enters the English vocabulary. [7] In sum, Cosgrove defends a turn toward a symbolic interpretation of landscape, which is very important in Anglo-American culture, [8] and which privileges the study of the imagination and ways of seeing. In any case, beyond the attribution of responsibility and methodological labels, the impression of witnessing a profound and fiercely debated turn in geography is perceptible even by interested observers from outside of the field such as myself. It explains the widespread appeal of the history of geography, which has led to new and interesting research in recent years. [9]

Landscapes as Visions or as Places?

6 Cosgrove convincingly relates the emergence of history and geography’s interest in landscape to a more general cultural turn. He maintains that the concept of landscape derives from the Ptolemaic distinction between absolute and relative space. This distinction is also found between geography and chorography, where the scientific and mathematical descriptive capacity of the former is opposed to the preference of the latter for visual, even artistic description. Chorography concretely connects geography and history, particularly at the level of the landscape. However, it can also grasp the specificity of each zone or locality to the extent that it was in past centuries a privileged tool of celebration at certain sites. [10]

7 Although the parallel between historiography of landscape and historiography of locality is suggestive, Cosgrove does not develop it. Yet it has the potential to allow for a comparative reading of the ambiguities and difficulties these two research approaches reveal to the attentive observer. Above all, it would make it possible to link the landscape clearly and contextually to its topographical dimension—one of the possibilities, as we will see, that the supposed “spatial turn” does not explore, whereas I consider that such a possibility could resolve some of the turn's intrinsic difficulties.

8 As with the historiography of landscape, the new prospects for the historiography of locality [11] in current research, theory, and methods are not homogeneous. Interest in the purely cultural meaning of the local, mainly in terms of identity, [12] is gradually shifting toward the study of the intellectual practices of the locality, hence specifically toward the study of local erudition. [13] This cultural meaning also makes it possible for us to note another important shift. Initially, a purely rhetorical and textual approach was privileged, such as the one Michel De Certeau advocated in 1980, [14] in which places would result from essentially written operations of boundary setting. Today, an approach has emerged that is linked in particular to the pragmatic and processual dimension – and thus to social relations [15] – even if, for an eminent representative of historical geography, it tends to be concerned above all with local consumption. [16] Yet more attention should be paid to the phenomena of production in the locality, i.e., to the continual operations of constructing the competence of subjects in a social environment defined by rituals, folklore, and oral culture. [17]

9 Similar remarks can be made about the concept of landscape. Undoubtedly, Cosgrove’s paper would have gained had he undertaken a more thorough historiographical examination. By that I mean that research on the birth of the concept of landscape in recent historiography is a fundamental problem that requires analytical efforts equal to the task. Without claiming to be exhaustive, I would like to tackle some aspects of this history to clarify the difficulties implied by use of the term “landscape” in recent historiography. In fact, Cosgrove uses and presupposes - in the paper mentioned above as well as more generally throughout his entire work – a more metaphorical, visual, and purely symbolic use of space. [18] Today, this approach is clearly successful in many historiographical traditions and might have considerable influence over the contributions of an entire field of research concerning the concept of space. [19]

10 The metaphorical and visual meanings of space presupposed by the concept of landscape goes back, it seems to me, to events and debates that occurred in historical geography beginning in the 1970s, particularly in English-speaking countries. Yet historians are not always aware of these debates, even when topics and approaches linked to space concern them directly.

At the Supermarket of Theory: British Geographers in the 1970s

11 Retrospective recollections and commentaries on the culturalist turn in the 1970s are abundant today in the world of geographers. Usually these are personal reflections, stamped with an obvious desire for self-justification. [20] Although they should not be accepted without reservation, they do offer evidence of a highly significant transformation. To summarize, I believe that there is a consensus over recognizing the formative role of the crisis experienced by geography after World War II. [21] In its French version, largely directly inspired by Vidal and thus linked to a non-determinist conception of regional formations and modes of life, the problem appears to be posed essentially in terms of the reception of Fernand Braudel’s so-called structural methodological proposition, that is, in terms of the practice of observing space on a widened temporal scale. [22] This problem essentially underlies historical geography’s search for autonomy from history proper and its temporal hierarchies, which obviously does not exclude projects and occasions for effective collaboration. [23] In any case, the strong interest in a history closely linked to the concept of space has borrowed from other approaches. Some aspects can be related to research on urban morphology and considerations of scale developed by Bernard Lepetit. [24] Other aspects can be linked to the reconstruction of the origins and intellectual matrices of the French administrative setup, linked to the rediscovery of Cassini’s map and the revolutionary debates on that subject [25] as well as to the importance of research in border changes and their sources in central archives, which can explain the structure of that setup. [26]

12 In the Anglo-American (mainly British) version, the evolution is connected more to a growing dissatisfaction with geographic research at the topographic level, marked by the canons of local history, that is, a dissatisfaction with research in the relations between the spatial characteristics of the rural settlement and the economic and social developments that characterized European history. [27] In my view, the debates engendered by such dissatisfaction are characteristic of an extreme openness, namely the search for theory at the end of an innovative experience such as the quantitative revolution of the 1960s, [28] which led some of its supporters, almost in spite of themselves, to choose theories often in obvious conflict with a long tradition of empirical research. From Karl Marx to Anthony Giddens, from Jean-Paul Sartre to Walter Benjamin or to political economy (much as was also happening in other human sciences such as anthropology), the analyst could choose everything and anything. [29] As a result, this historiographical context produced extraordinarily changeable trajectories. The career of Alan Baker (among others) appears to me to be representative of this trend. Initially interested in a geographic reading of historical sources and in processual analysis, Baker’s work acquired a strong methodological and theoretical connotation during the 1970s. In the end, he focused on the relations between ideology and landscape and systematically studied nineteenth-century rural associationism in the Loire Valley. [30]

13 However, beyond individual and collective choices, one pivotal work became vital to the participants in the debate, [31] a work in which a historian of literature reconstructed a classic subject with new presuppositions, a move that appears to me to be linked to the emergence of an intellectual climate influenced by Wittgenstein. By straightforwardly postulating that literary productions can in no way be read as reflections of social reality, Raymond Williams determined “to show simultaneously the literary conventions and the historical relations to which they were a response.” [32] However, in his brilliant research, he did not go on to explain Arcadian literary themes through the resonance of aristocratic ideals but rather as the reaction of the new English property-owning classes of the seventeenth and eighteenth centuries to their recent social pre-eminence and their economic relation to the land. [33] Such a reaction could also appear as the confirmation of a conflict between city and country. Yet this still has to be verified empirically and, thirty years later, the discussion has only begun. [34] Leaving aside the attraction exercised by an author with an extraordinarily personal and semi-autobiographical style, it should be remembered that Williams’ work was deeply involved in a debate over transformations in the English countryside, following research on British local history and its topographical approach. [35] In fact, he displaced – materially also – the center of analysis from the physical plane, from the land, so to speak (the geographer’s main concern), to the plane of intellectual production, [36] to which he attributed an autonomous (and even greater) weight and value in relation to what is called “social reality.” In particular, it should be remembered that the intellectual dimension was not conceptualized as devoid of links with social reality. Rather, it was simply asserted that the former was not derived from the latter, even though intellectual aspects were not subsequently studied in terms of this same social reality but by means of literary conventions.

14 I am emphasizing this episode because of the retrospective relevance it assumes. That is, it forms a sort of “theoretical turn,” a moment of cultural reorientation for an entire generation of British researchers, an irreversible choice, which may account, in a way, for the tendency of a large part of British geography to privilege cultural theory and, consequently, cultural geography. Furthermore, this choice appears to account for an initial formulation of the problem of landscape that ended up influencing the future choices of British geographers and the way in which they (metaphorically) approach reality.

15 Cosgrove’s ideas were born in this context, developed there, and regularly referred to it. That is, the landscape exists only in the act of seeing. It is materialized in the content of the vision, which has no origin. Moreover, the very fact of seeing permits power and control over territory. In his 1984 research, for example, Cosgrove explains in this manner why the Venetian patriciate favored Palladian architecture. [37]

16 Cultural geography, which appeared several years later, [38] tends to reduce its own relation to the land even further. Sometimes this type of geography is even confused with studies of diffusion or frequency, which are characteristic of French sociocultural history. [39] From the initial ideas, provided we have correctly understood them thanks to Williams, there remains little more than traces of the outcome and nothing of the method. It is no longer a question of a tension-response relation between literary conventions – or more broadly, cultural conventions – and historical realities but rather the distribution of cultural markers in a metaphorical space. [40] Baker’s book on the world of associations in the Loire Valley is a significant example of this. [41] Although the author undertakes precise and articulate analyses, he makes no reference to places or settlement processes. Although he studies the diffusion of the association phenomenon, he is not interested in local interrelations since his work originates in a problem of historiographical synthesis – namely modernity, individualism, and fraternity in nineteenth-century France – and not in the observation of social practices.

Landscapes and Processes

17 Undoubtedly, other disciplines have managed to give sustained attention to the strained link (strained, because it can only be a link of tension) between human society and the land. [42] One discipline in particular, environmental history, has displayed such an ambition, [43] and that from the very moment it adopted the name.

18 The growing success of environmental history is fully part of the culturalist perspective examined here. Let us take a well-known example. In his work, Simon Schama emphasizes the need to define the relation between human beings and nature with these generic terms. [44] To substantiate his approach, he makes do with random and discontinuous sources, which are connected only by a continuity of meaning that is more assumed than demonstrated, still less illustrated. [45] What is more, he seems to base his hypothetical reading on a continuity of phenomena that contrasts with the irregularity of attestations. In addition, there is a surprising disregard toward the origin of sources and the conditions of production of these same attestations. That said, there are of course cases where the use of multiple sources, particularly non-textual sources, for the analysis of agrarian practices is much better handled. Yet in most of these cases, as in the work of William Cronon, for example, [46] the analyst proceeds from the context of an economic history in which the practices of land use are clearly identified from within the framework of a capitalist economy in which modern agriculture is contrasted with traditional agriculture. Native American practices of exploiting environmental resources, for example, are thus interpreted from the outside without taking into account their location or examining the origin of the accounts and sources.

19 Rather, I believe that it is precisely the systematic attention paid to the production of sources that allows an approach to the practices of power capable of including space and of exhibiting all the relevant elements for its historical analysis. In her brilliant research on Merovingian France, Barbara Rosenwein identifies a continual production of privileges of immunity she interprets as a discontinuous social and cultural phenomenon, which she calls the “power of restraint,” [47] that is, the capacity to prevent royal functionaries from accessing certain well-defined spaces. In this manner, a process is set up that gives rise through negotiation to plural spaces.

20 As a result, we obtain asymmetrical and contrasting pairs: to homogeneous temporalities (the Paul Claval view) correspond cultural continuities (the cultural geography and environmental history view), while the episodic dimension and spatio-cultural discontinuity are no longer considered relevant, which suggests an opposition that, despite everything, remains linked to Braudel.

21 The unresolved problem of the continual – though not always explicit – relation between phenomena and sources may explain the critiques leveled at one aspect of cultural geography (whether environmental or not). Of these, the anthropologist Eric Hirsch’s critique is, in my view, the most explicit. In his introduction to a collection of papers on the anthropology of landscape [48], Hirsch focuses on the purely visual definition given to the landscape by Cosgrove and Daniels, namely “… a pictorial way of representing, structuring, or symbolizing surroundings.” [49] He remarks that this definition accentuates the static aspects of the landscape and denies the processual aspects that always characterize it. He then emphasizes that this approach ignores half of the intrinsic experience of the landscape, that is, the everyday experience: “The problem with this definition is that it takes one pole of experience intrinsic to landscape (the representational) and generalizes it to the landscape itself.” [50] He then wonders what sociocultural factors justify such an elision (or self-censure), and concludes that there is an under-evaluation of those aspects of the landscape he calls “transformative,” which correspond to related but contradictory moments of representation and experience, of external and internal vision, or what he calls a “cultural and historical context.” [51]

22 With the support of these undoubtedly interesting observations, I argue that only a critique of sources can overcome the difficulties of culturalist geography [52] by means of greater attention being devoted to the genesis of documents. Recent disciplines, such as historical ecology, today offer new conceptions of context that seem destined to radically influence our notions of “phenomenon” in geohistorical research.

23 In this new approach, “resource” is the key concept. One example is the discovery by historians of ecology of practices for the activation of plant resources. [53] In this analytical perspective, the visual elements of the landscape are the active elements in a selection made by the human groups that use them. In this sense, they are the object of practices, traces of which are found in historical cartography and legal documents. Local practices selectively activate resources, and it is the articulation of these activated resources that creates the definition of the landscape, allowing us to make substantive readings of texts, including literary and narrative texts, and grasp the history of space in a primarily regressive direction, which valorizes the episodes that make up historical ecology.

24 Thanks to Oliver Rackham’s work, the birth of historical ecology [54] in the second half of the 1960s made possible a fruitful dialogue among archeologists, geographers, and historians of the local around a “science of the site” that provides protection against the imprecisions of cultural geography and allows us to imagine a topographical history that falls neither into the rigidity of positivistic determinism nor the imprecisions of environmental history. After the crisis of processual archeology and the deconstructionist critiques, [55] archeologists found in the “site” a way to contextualize an analytic discourse solidly anchored in the land, particularly for the most recent periods. Geographers tried to preserve the vital relation with the materiality of the topographical scale, while historians attempted to surmount the difficulties posed by a discourse of the local now divided into a positivistic topographical history and a historiography that makes the local a pure invention of tradition. An example of possible collaboration will allow us to illustrate these initial observations more concretely.

Localize Landscapes: A Ligurian Example

25 The edited volume by Ross Balzaretti, Mark Pearce, and Charles Watkins, entitled Ligurian Landscapes, originates in a convergence of multiple experiences and interests. [56] It collects pioneer research reports, primarily from archaeology, and results from an intense and productive collaboration between British and Italian researchers belonging to the three disciplines listed in the subtitle: Studies in Archaeology, Geography, and History.

26 This work is also the fruit of didactic experiences that remain atypical of European universities, namely the “Landscape History of Liguria Field Course” offered by the University of Nottingham [57] and the “Seminario Permanente di Storia Locale” offered at the University of Genoa and founded by Edoardo Grendi. [58] These initiatives from the 1990s, which were intended for students, stressed the link between teaching and research and resulted in new and interesting objects of study, as the large range of problems taken up effectively demonstrates. As the editors argue in their introduction, [59] these new objects arose from the encounter between British experiences of environmental history and Italian micro-history. Further, the change of scene for academics from both countries, combined with the field experience and the documentary evidence, led to the formulation of fascinating questions.

27 Interest in spatial relations at the topographic level has thus been revived, thereby giving support to a theoretical concern clearly affirmed by Italian micro-history, namely the view that personal relations constitute society and are at the same time the context in which to understand social practices. Grendi consistently embodied this cultural path when he attempted to situate practices within relational systems the spatial perspective reveals as fragmentary and discontinuous. [60] Sadly, his demonstration is posthumous since the Genoese historian died in 1999 during a seminar in Nottingham. However, his paper on “Meat Provisioning in Ancien Régime Genoa” impeccably illustrates the provisioning system of an Ancien Régime city. Instead of interpreting the provisioning in light of the economic categories of liberalism, as is usually the case, he interprets it in terms of spatial relations. [61] Thus he emphasizes the analysis of the organization of relations among butchers, merchants, and cattle breeders, even when only temporary, as well as the dependence of this organization on the economic context (type of meat, time of the year, or the urban space itself). For his part, Osvaldo Raggio returns to Grendi’s cultural path and emphasizes his rediscovery of the concrete spatial aspect of interpersonal relations. Based on a rejection of historical synthesis and a mental approach to the study of culture, Grendi advocates a historical ethnography grounded in an analysis of the social practices of past societies undertaken at the topographical level.

28 Following these suggestions, researchers have applied techniques from each of their disciplines to illustrate the active contribution of human communities to the history of the Ligurian landscape. The first series of studies could be defined as “contextualized environmental history.” These include a study of changes in vegetation at four sites over several thousand years, [62] models of fragmented settlement and mixed exploitation of resources between late Neolithic and Roman occupation, [63] and a study of prehistoric mining activities using geo-referencing techniques. [64]

29 The active contribution of human communities also emerges in another series of studies in historical ecology conducted by geographers Diego Moreno and Roberta Cevasco, who following Rackham, [65] study the “activation of plant resources.” For his part, Moreno begins with a critique of the purely geographical or visual concept of the agrarian landscape [66] (so dear to Emilio Sereni) and the pseudo-historical generalizations it represents. Against this concept, he proposes a historical ecology based on new documentation and composed of sources derived from observation and sedimentary sources, and in his case, devoted to exploring the historical connection between site and population in the vegetation of the northwestern Appenines. In this approach, landscape takes on a new dimension since it is the result of plant resource activation practices. The central concept present in the rest of the volume is thus identified, namely wood pasture and its management practices. Moreno’s aim is to propose a research path. He begins with investigations from the end of the eighteenth century that draw up and then condemn an inventory of an agrarian practice, namely the use of a potentially toxic plant, the hop hornbeam (Ostrya carpinifolia), as animal feed. He continues by reconstructing a local system of animal production based on a site study, finally ending with the contemporary experience of possible internal connections within this system. His work also results in a critique of the domestic-wild dichotomy that has held such a large place in the cultural history of the environment in recent years.

30 In the second contribution to historical ecology, Roberta Cevasco makes a detailed reconstruction of systems combining pasturage and forgotten arboriculture, from the alder to the chestnut and the oak, by combining historical cartography and a floral analysis of vegetation at the site. [67] She thus demonstrates the role of multiple resource use in species enrichment. [68] In this way, she identifies and reconstructs cartographic sections of the exploitation of land that used to be subject to multiple uses (wood pasture) and was abandoned during the last 150 years, with a few toponyms and traces in the vegetation and the soil being all that remain.

31 The topic of resource activation in an agricultural system recurs with equal force, yet from different angles, in the work of Ross Balzaretti, [69] Charles Watkins, [70] and Sandro Lagomarsini. [71] Balzaretti also attempts a topographical interpretation of an exemplary document, the Relatio dell’origine et successi della terra di Varese (c. 1557) by Antonio Cesena. He applies the method of historical ecology, [72] which leads him to localize the information in a narrative text and then bring it back to a dataset still observable today. Moreover, this method makes it possible to verify the validity of the observations made by those who formulated the hypothesis of the mobility and cumulative effect of land use practices. Meanwhile, Watkins’ paper provides later confirmation of this interpretative element, since he attempts a geographic analysis of terracing systems using the techniques of regressive history. Clearly, the realistic deciphering of terracing is possible even if it lacks textual attestations in the cadastral, notarial, and even cartographic documentation. By contrast, field observation reveals a mobile world made up of uncommon conservation practices and a greater tendency to abandonment. Such observation makes it possible to give general confirmation to oral testimonies on the management of this landscape. The paper concludes with an observation on the pernicious effects of conservation policies, such as UNESCO’s inclusion of the nearby Cinque Terre as a “World Heritage Site,” and emphasizes the need to define models for the political management of terracing.

32 Finally, Lagomarsini interprets the abandonment of the highest elevation localities in the Valle di Vara in terms of opposition between cultures. One of these consists of an urban and idealistic culture characterized by hostility toward common and civic uses of wooded areas in contrast with the free circulation of humans and animals throughout the territory. The other consists of a local culture for which the use of plant resources such as sheep rearing and the cultivation of timber resources protect botanical diversity and even encourage it and make it prosper (this is a reference to the work in historical ecology the late Giuseppina Poggi). The rest of the paper, which contains an interesting and surprisingly personal story, gives an account of the astonishingly rapid effects of abandonment in terms of impoverishment, even disappearance, of species. Lagomarsini then attempts to reconstruct the local history of gentian, raspberry, blueberry, and asphodel. Paradoxically, a major part of the impoverishment comes from neglect followed by belated recognition by local authorities of collective rights (such as the rights of hamlets, for example) [73] over plant resources. The true leitmotif of this methodologically rich collection of papers clearly appears in the extraordinary analytical possibilities offered by the topographical interpretation and in the discontinuous character of landscape history (or historical ecology).

33 In summary, it is clear that the choice of the local level has implications in terms of the relation between general mutations and specific cases. In Ligurian Landscapes, the heuristic value of the case seems capable of deconstructing the idea of the permanence of time—which up until now has played a major part in establishing the foundations of environmental history, as Catherine Delano-Smith aptly points out. [74] In contrast to the image of the history of landscape based on continuity, she mentions the positions of researchers such as the archaeologist John Bradford, who emphasized irregularity in landscape changes and the dynamic contributions of the humans who occupy it. This is a history made of events of varying significance and types: not only disasters but also, and undoubtedly above all, the imperceptible occurrence of daily local changes. Cutting down a tree or burning a plot of land is an episode just like an innovation applied to cultivated land. The revaluation of detail and discontinuity in local space mark the boundary these papers invite us to cross but without leading us to think that trees and plots of land are not the result of intentional human efforts.

34 We are now able to formulate more successfully our dissatisfaction with the proposition, which is both visual and relativistic but simultaneously linked to exogenous categories such as power structures or the Loire community, identified in the “spatial turn” as Cosgrove understood it. Instead of using a purely symbolic concept of landscape, it becomes necessary to conceptualize via case studies – perhaps we should say “site studies” – the study of the environment and its specific historical dynamics with the help of the concept of “activation of plant resources.” This concept makes it possible to avoid separating elements of the botanical and organic universe from social and cultural practices, with their legal and relational values, which presages a return to a terrain closer to historical analysis.