Compared to studies of documentation and the material of documents in geography, the disciplinary effort that has been put into reconceptualizing the archives – the locations where such documents are preserved – is far more explicit and fecund, making a notable change. Up to the 1980s, any mention of archives was usually confined to footnotes. When they were discussed, this would take place within a methodological treatise in which a rodent model of research prevailed: researchers were like squirrels who would forage for acorns of 'evidence' to be collected from the contents, taken home, and stored for later analysis; and the more experienced among them would instruct others (through a list) as to which were the more promising trees for a process of collection. In the 1990s, the signal work of Kenneth Foote and Joan Schwartz began to rethink the archives as peculiar places in themselves, places that extend the spatial and temporal range of communication, yet exercise remarkable power over what people remember and whose knowledge is marginalized. At the same time, a body of work started to percolate from social theory in which 'the archive' was a central figure. In this material, the archive remained an abstraction. For Michel Foucault, it was the level of practice that enabled statements to survive; in the hands of Edward Said and Thomas Richards, it became a whole body of literature, media, and more that constituted 'the cultural archive' of a society; with Jacques Derrida, the archive represented a figure of interpretive authority. It was not until the turn of the millennium that geographers began to follow the leads of Foote and Schwartz in appreciable numbers, to think through the specificities of particular archives (e.g., this library's collection of photographs, that record of testimony in South Africa) as a place of materiality, memory, and power, and as a peculiar cultural artifact.
If that brand of archival research has become contagious, there are reasons to be cautious before naming it 'archive fever'. First by implication, the label seems to locate its origins in 1996 with the publication of Derrida's book. That text, Archive Fever, is almost always cited in the new geography of archives. Second, as Carolyn Steedman says, many ''have assumed that Archive Fever has something to do with archives (rather than with psychoanalysis, or memory, or finding things),'' but the book is not about archives. Derrida begins with an etymology of the word 'archive' (material on the second page is often quoted in geography), but his foremost concern – which consists of a reading of Yerushalmi's reading of Freud – becomes a matter of interpretation. In the edited volume, Reconfiguring the Archive, Derrida even reinterprets Archive Fever, so as to ''leave the book and just discuss burning things here'' at the seminar in South Africa where he was speaking. In that text, he positions the book as a homage to (and failed discussion with) Yerushalmi at the Freud Museum. Third then, what is archive fever? In Derrida's rereading, it is the tension that comes from ''something in us ydriven to destroy the trace without any reminder,'' a death wish to destroy the material archives and silence the riot of relationships that murmur within them. But in Steedman's interpretation, archive fever is the impossible and insatiable desire to find the beginning of things, to recover their singular moment of inception. She sees a little irony, accordingly, in the 'great assiduity' of historical attention given ''to looking for and finding what wasn't there'' – that is, an origin – in Derrida's text.
If we were to outline the emergence of recent enthusiasms for research about archives in geography without locating its source in seminal material by a French philosopher, we could do worse than to build the sketch, again, in relation to interests in materiality. The analysis of archives in the discipline followed in the wake of critiques against discourse theory as all too textual and immaterial. Moreover, archives have proven to be a more substantial topic than documents alone, due in part to the localization of material archives. As Steedman points out, archives name ''the many places in which the past (which does not now exist, but which once did actually happen; which cannot be retrieved, but which may be represented) has deposited some traces and fragments, usually in written form.'' Because they are actual places, archives are finite. They cannot hold every trace, nor can they serve every patron at once. Some selection is necessary, which means the archives on which (and in which) geographers conduct research are inherently political. Yet they are not wholly the instruments of human agency: ''the archive is made from selected and consciously chosen documentation from the past, and also from the mad fragments that no one intended to preserve and that just ended up there y It is indexed and catalogued, and some of it is not indexed and catalogued, and some of it is lost.'' As a finite location, an actual place, ''you cannot be shocked at its exclusions'' (so Steedman soberly tells us), because the archive's ''condition of being deflects outrage: in its quiet folders and bundles is the neatest demonstration for how state power has operated, through ledgers and lists and indictments, and through what is missing from them.''
Along with materiality, state power is a second major vector with which the interest in archives emerged in geography. A historian and anthropologist, Ann Stoler effectively maps the major benchmarks in an essay titled 'Colonial archives and the arts of governance', published in the volume cited earlier, Refiguring the Archive. She sketches a move from archive as source to archive assubject across several fields: historical methods, archival science, cultural theory, the history of science, and colonial ethnography. As a student of colonialism, Stoler's own archival research has led her to look at archives not as 'inert sites of storage' for old documents, but as 'technologies of rule' and governmentality, as instruments of knowledge production that became critical to the administration of European colonies overseas over the course of the nineteenth-century. She argues that colonial states were information hungry machines that accrued power from the quantity (not the quality) of the knowledge that their administrations produced, and that their archives ''ordered (in both the imperative and taxonomic sense) the criteria of evidence, proof, testimony, and witnessing'' to construct such knowledge. Like Riles, Stoler is interested in the form this material takes, as well as its content. Among her examples are commission inquiries, which ''organized knowledge, rearranged colonial categories, and prescribed'' what colonial officials were charged to know. Anthropologists and historians need these archives for their ethnographic content as they often have, she suggests, ''but also for the content in their form'', for those forms mark out 'populations' that warrant the interest of the state, offer up historiographies of colonial research, and most of all, serve as monuments to the importance of history to officialdom.
The methodological program that Stoler recommends is rippling across the discipline, as more geographers return to archival research. A recent article in the Journal of Historical Geography is indicative. There, Ruth Cragg explores the imaginative geographies of empire produced by the Royal Empire Society Library and the practices of acquisition, classification, display, and study within its walls in London. Stoler's counsel frames Cragg's work, which turns from an extractive to a more ethnographic project in archival research so as to consider the archive itself as a cultural artifact. For instance, the layout of this library's book collection served to reproduce the regionalization and hierarchy of the empire: the form of the archive reproduced a vision of imperial heartland and peripheries. Cragg also argues that the library was imagined in ways akin to what Latour calls a center of calculation (more so or less so by era), that it was understood as a space of imperial knowledge in the heart of empire. But her final caveat about the inadequacy of the terms 'library' and 'archive' relative to the individual topography of this institution follows and completes the warrant that Stoler offers near her own conclusion: ''to understand an archive one needs to understand the institutions that it served. One needs to understand what subjects are crossreferenced, what parts are rewritten, y which [categories] become common sense and then fall out of favour ….''
In geography, recent developments in archival and document based research attend more to the form than the content of an empire of inscriptions. They read documents and archives as important cultural artifacts in themselves rather than as inert sources of other information. These new methodologies have drawn from theoretical innovations in other fields (with work by Latour, Riles, and Stoler, highlighted above), and they also collude with the influences and critiques of discourse analysis in recent years. These research projects are further energized by the dynamic interest in materiality and state power across the discipline. Yet the strongest enthusiasm that animates these fresh approaches to archival research – and one that saw little direct diagnosis in the pages above – is perhaps the open politics that they bring to issues of memory, knowledge, and power.