Last Friday, I represented the Snapshots team at the ‘Filing Empire’ workshop at Columbia University. The questions the workshop was designed to address are ones we’ve often tried to approach in our own work: how was empire constructed by and through technologies, protocols and systems of paperwork? How did these bureaucratic apparatuses change over time? How did they understand, codify and produce colonial realities? And how do the entailments of these systems – structural, epistemic, ideological – persist into the postcolonial present? At the core of the day’s discussion was the big question on the relationship between documentation and empire: how systems of recording, communicating, storing and retrieving information helped shape and sustain imperial power.

The papers and book chapters contributed by participants approached this question from varying angles. Asheesh Siddique’s work focused on John Bruce (1744-1826), the Company historiographer, to illustrate how networks of patronage, manuscript circulation and influence persisted into the early nineteenth century, as the historical and administrative functions of the archive began to occupy different spaces and develop different protocols. The paper I brought from the Snapshots project studied later historiographers and antiquarians in the India Office’s archive, to ask questions about the functions of memorialization, the authorization of history and the archive as a site for divergent fantasies of bureaucratic/historiographical power. Zoe Laidlaw’s contribution looked to James Stephen’s reorganisation of the Colonial Office in the mid nineteenth century, and the ‘Blue Books’ which from 1819 onwards were assembled, supposedly, as reliable statistical digests of Britain’s colonial empire. Devyani Gupta’s work on the postal system in nineteenth century India showed how colonial infrastructures worked to efface homogeneous social and cultural realities through standardizing things like forms of address, postal routes and epistolary practice – in effect, binding South Asian territory, people and societies closer into a ‘British India’. Somewhat similarly, Brian Larkin’s paper followed the progress of a 1923 attempt to standardize the spelling of ‘Mohammed’ and ‘Mohammedan’ across the British Empire. Finally, Matthew Hull’s work on present-day Pakistani bureaucracy undertook an anthropological investigation of the entailments of the ‘Kaghazi Raj’ or paperwork empire, exploring how the practice of paperwork becomes the terrain on which politics are conducted, personal credit built, influence peddled, and careers (and occasionally fortunes) made.

The question which came up perhaps most often and forcefully, in trying to make these rather various pieces of scholarship speak to each other, was that of failure. In all of the cases studied, there’s some kind of gap between what the archive or the system claims to be able to do and what it actually does.

In the first instance, this is often about procedure. Files often fail to circulate as they’re supposed to: they get lost in offices, kept on the wrong desk, redirected by accident or by design to the wrong person. The function of any given file or protocol can be subverted, repurposed or nullified: Matt Hull’s work on the complex unofficial codes of mark-up, circulation and file construction engaged in by modern-day bureaucrats is a dizzying example of how any system of government-by-writing can be made the vehicle of very different types of business than those it was designed for. Bureaucracy leaks and spreads, generates friction and grinds its gears. The given system is always, of course, a relic of an earlier time; inasmuch as it has helped to create the world it must now engage with, that world is also escaping it, and the matter of catching up is often one of improvisation and quiet adaptation. It has other systems with which to contend and intermesh; its systems of meaning have to somehow articulate to systems of meaning with which it may have little in common. In Larkin’s study, almost all these things occurred: the file generated stacks of replies, notes and additions on a global scale, became a forum in which administrators showed off their considerable classical learning while arguing about the possibility of standardizing something that, while it reflected a global Muslim ecumene, fiercely resisted standardization. In the end, the whole affair was short-circuited by a conversation which happened outwith the bounds of the file, most likely orally, at the very top level: an emerging consensus (that ‘Mohammedan’ and ‘Mohammedanism’ would be better, both for scholarly reasons and as a “global courtesy”, rendered as ‘Muslim’ and ‘Islam’), quashed by abrupt executive fiat.

The more serious and encompassing failure, however, is that of representation – and, behind the claim to representation, the claim to power. Both the Blue Books that Laidlaw studies, and the fantastic archive of Clements Markham in our work, advert to a fantasy of statistical mastery, a fantasy that the empire can be fully known. Likewise, Gupta’s work on the Post Office shows us a system with the clear aspiration to structure reality by properly codifying it. This is the kind of enabling colonial knowledge studied by CA Bayly, Bernard Kohn, Matthew Edney and Thomas Richards; it leads into the kind of fantasy of the omnipotent archive, complete library or 1:1 scale map that haunts the work of Jorge Luis Borges and Ismail Kadare.

But there’s always a difference between the map and the territory. At Columbia, perhaps unkindly, we settled on describing these kinds of claims to mastery ‘Trumpian’: they betray a conviction that if you imagine something to be the case, it becomes the case. Markham’s all-encompassing registry barely got off the ground; the data on which the Blue Books were based was often faulty, returned by colonial officers who filled in their questionnaires carelessly and as a matter of bureaucratic box-ticking, and was barely consulted within the Colonial Office. Later statistical digests for public consumption, like RM Martin’s Statistics of the Colonies of the British Empire, bore little relation to the colonial realities they were supposed to represent, but became increasingly central to the work of representing and imagining the empire; whatever their failures in terms of accuracy, they became the materials of the imperial imaginary and its politics. The bells and whistles, the impressive visual rhetoric of statistics – tables, columns, sheer weight of numbers – conveyed the impression of authoritative knowledge where there was none. The fantasy here became a kind of reality, and there is a strange momentum in the process by which actuarial happenstance, laziness, prejudice or misprision becomes reified as cultural force.

And this brings us, inevitably, to the question of ritual and performance, not just as things which occur within bureaucracy but constitutive elements of it. In some ways, maintaining a sense of bureaucratic administration’s performativity can help us to trouble some of the more simplistic narratives of its technological and cultural histories. It can be a bit too easy, perhaps, to identify a ‘bureaucratic turn’ (say, for our purposes, about the beginning of the nineteenth century) before which all administration was carried out through discursive, personalised and situated means, and after which everything became Weberianly mobile and routinized, all subjectivity cleared out by process, and all content submitted to form. Almost all the work we shared showed how ‘modern’ bureaucracy is stuffed with all kinds of performance, affect, desire, sociability and animosity; if bureaucracy, as Matt Hull pointed out, is ‘the terrain on which we do politics’, it’s also often the terrain on which we do various kinds of selfhood. Brian Larkin’s colonial officers, asked to agree on the best transliteration of ‘Mohammed’ and its cognates, took the opportunity to show off their classical and Orientalist education with some fairly dizzying displays of linguistic knowledge – none of it, however, of much use to the question at hand. The Pakistani bureaucrats of Matt Hull’s work make a performance, in some senses, of the bureaucratic form itself, in which mastery of the all the system’s official, semi-official and distinctly non-official protocols becomes a form of social peacockery, a platform for a kind of sprezzatura. The speaking human subject always returns to fill the spaces of the forms from which it has supposedly been banished.

But, too often, only a certain type of subject gets to speak. If there’s a danger to the focus on a kind of archival bathos – the distance between Trumpian claims and, well, Trumpian failure – it is that it is  perhaps too easily parlayed into a denial of the power of the systems we’re studying here. If every case study we read for this conference involved a kind of failure, that failure always occurred within the context of a resounding success: the job got done, more or less, and the power of the British empire (or the modern postcolonial state and its apparatchiks) endured. (Again, ‘Trumpian’ is more instructive than we may’ve at first anticipated: Donald Trump’s uniquely vexed relationship with reality hasn’t yet seen him ejected from the Oval Office.) Our last discussions of the day played around the edges of archival absences: we discussed ways and means of recovering subaltern subjects from an archive with little apparent room for them, and the various ways in which archives can be dispersed, destroyed, and misread. In an age of nationalist reaction, cultural nostalgia and anxious retreat to the national, racial or linguistic community, the matter of what’s in the archive of empire and what’s isn’t, and how we read and manage and interpret that archive, is uncomfortably urgent. The files of empire, their contents and dispositions, are still the terrain on which we do politics.

Many thanks to Professor Brian Larkin and Asheesh Siddique for organising the workshop, and to the Heyman Centre for the Humanities at Columbia University for hosting it; especial thanks to Brian, and to Gustav Kalm, for showing me the hospitality of New York.

— Peter Mitchell