Archivist Manifesto

By Yuk Hui, 22 May 2013
Image: Modified books borrowed from Islington libraries by Joe Orton/Kenneth Halliwell


A new archivist has been appointed. But has anyone actually appointed him? Is he not rather acting on his own instructions?... He will not concern himself with what previous archivists have treated in a thousand different ways: propositions and phrases. He will ignore both the vertical hierarchy of propositions which are stacked on top of one another, and the horizontal relationship established between phrases in which each seems to respond to another.

– New Archivist, Gilles Deleuze



§1. Archive


We are archivists, since we have to be. We don't have choice. This decision is already made, or determined by the contemporary technological condition. The ubiquity of information in digital, calculable forms has created a new situation of work and exploitation, we entered an endless process of data production, and then we also enter an endless black hole of data navigation. The internet of data is a huge archive of data and at the same time a enormous black hole that sucks all productivities. Google is the best exemplification of this double: on the one hand, we are contributing data by using the Google tools, emails, blogs, Google+, Hangout, etc; on the other hand, Google gives us searching and managing tools to survive in this milieu. Worse still, Facebook shows another face, a huge archive of data without navigability, the only navigation one can do is to search for your friend, otherwise you need to stroll down the page to find out what you have written years ago.


Under such situation, we must recognize that one of the main economical and political questions concerning the digital is the one of archives. What has happened to the concept of archive after the digital turn? What kind of power structure is present to us in this new setting? Archives for Foucault are traces of enunciations by which one can reconstruct the game of rules (le jeu des régles), that in turn reveal the power structure of its milieu. Archives in this sense are reservoirs of discourses that make possible an archeology of knowledge. In order word, archives are sources of deduction for archeologists. The will of archive further makes archives as a manifestation of power. This will expands in the modern time and set up a direct relation between institutes and archives. Each institution has its archive, has its history of discourses. In order to maintain its status quo of discourses, it needs to give its archive a proper name. As Foucault observes the expansion of museums and libraries in modern time, that attempts “to enclose in one place all times, all epochs, all forms, all tastes”1. An archive is also a symbol of authenticity and authority. This project of the moderns confronts the biggest challenge in the digital age. On the one hand, public institutions adopting digital strategies also have to develop new form of interaction between archive and audiences, so to speak they are forced to open their archives; on the other hand, within an institutional framework, they continue keeping their archive in a centralized way (even the Michel Foucault Archive) in order to consolidate the status of these institution. Though today when these institutions implement “open polices” or alas crowdsourcing, it is still a strategy of crowd sourcing under the name of humanities or digital humanities to serve its central archives.


In this respect, Google is technically moving far ahead of these organisations, Google book, Goolge Museums are efforts to re-appropriate humanities through digitisation and crowd sourcing. Google doesn't only provide services with higher qualities and higher speed, but also goes beyond the relation between archives and human subjects. Google's archives exhibit an important distinction from that of Foucault's. It is that power is posing its forces directly on archives to achieve control, that is to say archives become a mechanism of control and social actions, instead of being traces of power2. Now we have to face a new game of rules that operates on more and more automatic and algorithmic level. Metadata produced by users become materials of induction, to generate patterns for prediction, rules, protocols of control. Archive stretches from discourses to gestures. It seems to me very clear, today in order to response to this political and economical dimensions of the web, we have to politicize the question of archive. The key seems to be personal archives. It is not simply from a technical consideration that we can then mitigate controls from the service providers or to keep privacy, but rather to rethink our relation with archives and to create another technological/digital culture. In order to do so, I propose to reflect on the following questions: What are we archiving and for what are we archiving? what does it mean to be an archivist?


§2. Care


We have been archivists from the moment we start owning things, toys, books, postcards, letters, we have our own way of organizing them, cataloging them. But it is only by now that we are confronting a situation, that we are not able to archive them, either possessing them or indexing them by ourselves. There are different levels of inabilities here: web services becomes more and more distributed, while data portability is still a problem; cloud computing is moving things from your hard-disk to someone else's server; indexation tools and personal libraries softwares are still underdeveloped, etc. These are not teddy bears, barbies, but digital objects (teddy bear and barbie dolls can be also digital). It is exactly this inability of archiving, created by the technological condition, opens up a new battlefield for search engines, social networks, cloud computing etc. But what are these objects and why do we have to re-conceptualize them seriously? In order to grasp the particular question of digital objects, we need to look at the particular evolution of the web, especially the decisive movement from Ted Nelson's vision of the web and Tim Berners-Lee's World Wide Web.


Ted Nelson's vision of the web was very much inspired by literature, it concerns indexation, by which one can jump from one link to another. For Nelson, the ultimate goal of the web was to create a micropayment system for authors. The invention of the web by Tim Berners-Lee and the recent use of ontologies to formalize data in machine-readable format, is not longer one that concerns hyper-link economy, but according to me a new economy of objects. It would be too complicated to argue the meaning of semantic of Semantic Web proposed by Berners-Lee, but it is clear that formalisation gives objects not only identity but also mobilities. It is only in such vision of Tim Berners-Lee we saw the continuation of the care of librarians, the emergence of a new type of object-hood defined by structuring metadata and a new type of care to come (HTML → XML → Web Ontologies). One of the examples that best demonstrate this is the enormous impact brought by human-machine readable web ontologies to library science. XML based Dublin Core directly confronted the conventional practice of MARC (MAchine-Readable Cataloging)3. “MARC must die” this was the slogan resonated among library technicians since early 2000s, you can find a specific website created by digital librarians dedicated to it. It is clear that the early development of metadata scheme for cataloging such as MARC hadn't taken the relationship between librarians and digital objects seriously. Technicians have to deal with abstract symbols which only machines can read, they are condemned to be assistants to machines.


It is also within this conceptualisation of object, we can go back to the question of care that centres in the relationship between librarians and books. I take the word care very much from Martin Heidegger's Sorge. Care is the temporal structure by which we can understand our existence, this theme lies at the heart of Heidegger's Sein und Zeit. Care is not only, as we say in our daily life “taking care of something”, but also a temporal structure that create a consistent milieu for ourselves. Heidegger further develop the concept Besorgen to describe the concernful mode of being with tools, and Fürsorgen to describe two positive mode of care4. The political implication of Besorgen is that today our everyday life dealing with archives, for example searching on Goolge, updating Facebook, or other activities of crowd sourcing, are naturalized as habitude, as something unquestionable.


In contrary, Fürsorgen is at the same time worries of not being able to be there and affirmation that one can have of oneself. If we are allow to say so, one is passive, like affections, emotions; the other is active, anticipations and preparation for the future. What is crucial in Fürsorgen is act to “look back”. The late Foucault's concern of care resonates with the change in the perception of power of discourse to the self- development and practice of discourses. When Foucault wrote le souci de soi (care of the self), he means the practice of care, one occupies oneself, as Heidegger's nachsicht and zurücksicht both worries and affirmations. Foucault uses the example of the figure Socrates, who keeps on asking young people on the street “are you occupying yourself” even in face of his death. This practice of care is like librarians are able to take care of books, wiping away dusts on the cover, putting them on the right places, relating books to different themes, by taking these objects they create an associate milieu for objects and themselves. Of course not all librarians are like this, but this librarian metaphor emphasize links between objects and archives5. In contrast, the deprivation of care is one that systematically destroy these care structures by alienating the relation between objects and individuals. For example, users are seen merely as producers of data and those who can contribute to different sorts of crowd sourcing. At heart of the question of archive is the question of care, and I think today in order to care, one must think of archive, the exteriorisation of our memories, gestures, speeches, movements. Search engines, social networks, what one can call info-capitalism operates in the direction to transform care into something efficient and computable. By the end, we are no longer be able to organize these traces, but leave them on the cloud to be taken care by others.


§3. technics


What an archivist needs is not only his or her love of objects, but also skills or technics of care. Care like power, is not substance, but relations that is modified according to its material conditions. The technics of care and the technologies of care coincide in the context of archives. While this technological culture of archiving is not yet there, we must bring it back to the table. I want to associate this question with the French philosopher Gilbert Simondon, not only because Simondon proposed to take care of technical objects (that would be digital objects in our case), but, also, Simondon envisaged a technological culture as a possible solution to the question of alienation specifically, and the opposition between culture and technology in general. This imagination of technological culture was projected through the encyclopedianism of the Enlightenment. Simondon saw the emergence of encyclopedia as the first time in history that technics such as making glass, porcelains, weaving are presented to the public. Lets recall that one of the goals of the encyclopaedia is to ‘publish all the secrets of manufacturing’. In the 17 folio letter press volumes of the encyclopaedia, about 2,900 plates in 11 folio volumes were devoted to technology.


But the paradox concerning technological development, is that the more advance technologies become, the further we may distance ourselves from it. There are two problematics associated with it: firstly machine became abstract knowledge, what presents to the workers is an interface of control, the know-how is reduced to pressing a button, feeding machine with raw materials; secondly the mechanisation of all artisan skills, create a tendency of diskilling among workers, since automation replaces man's position as a technical individual. Industrialisation amplifies this effect. On one hand, the concretization of technical objects as technical individuals joins hand in hand with abstractions of labour and knowledges, a button or a control panel hided away large amount of practical knowledge and social relations. On the other hand, the exchange between labour and capital bring labours in the the endless circle of capital reproduction. This is what Simondon understands as proletarianisation and alienation through his reading of Marx, that somehow bypasses the question of properties. Be careful here, I am not suggesting that one should avoid abstraction and reduction; not at all, abstraction and reductions are necessary in many aspects, but the question is rather what kind of reductions and abstractions: one that favours individuation or one that produces disindividuation. For example, one can drive a car by knowing its speed, its limit, and one who dives by only knowing pressing which button, in both case we are dealing with different kind of abstractions. Today, Google and Facebook are doing the latter, we are used to press buttons, satisfy ourselves with the speed and convenience, without really understanding the problematics behind interfaces and its crowdsourcing algorithms.


Simondon argues against automation, for him automation is the lowest level of perfection of technical objects. Simondon proposes to take into account of the “margin of indetermination” in the invention of machine. It means the machine of higher level shouldn't be perfectly automatic, but needs to integrate human being in the technical ensemble as one that operates it, one that lives with it. This cannot be understood simply that human beings are users, instead they must restore their position as technical individuals. For Simondon it makes no sense to return to the craftsmanship as negation of industrialisation. According to his analysis, “human individuality finds itself more and more disengaged from the technical function by the construction of technical individual – but it creates actually a malaise, because human, always searching to be technical individual, no longer has a stable place next to machine: he becomes servant of machine or organisers of technical ensemble6”. The question is no longer one that savages machines and factories, but to invent a new conception of technical knowledge that reconstitute the culture of machine that is driven by capital and marketing. This is what Simondon calls a technological humanism, which “ aims at the most serious aspect of alienation that a civilisation behaves or produces”, so “each epoch should discover its humanism, and orienting it towards the principle danger of alienation”7. Simondon observes that work is only a phase of technicity, instead that technicity is part of work8. In our context, the problem of crowdsourcing as work is that the users don't know what they are contributing for, while they are virtually working through searching, typing status update. While this relationship between technicity and work, lays the possibility of overcoming the limitation of works by retrieving the potential of technicity.


Under these theoretical interpretation, the archivist manifesto is a proposal to regain the knowledge and skills of organising data and digital objects; only with the technology, can we take about the technics of care, that is also a humanism called by its technical reality. In summary, what I tried to do above is to re-conceptualize archives and archivists after the web, and propose to take this possibility to re-appropriate archives, by reflecting on the question of industrialisation of archives vis-à-vis the question of care, and the development of archival tools embeds a large extent of diskilling. The current discussions on search, open access, archives, preservation of information and digital objects, often hide away the politics of individuals under the disguise of “users”. Users to technological capitalism, are consumers to consumerist capitalism. Archivist manifesto is a call for the reinsertion of knowledge and skills for developing personal archives, that on one hand, reinstall the culture of care and a technological culture; on the other hand, develop an infrastructure that allows sharing of information on individual level and bypasses marketing tools such as search engines and commercial social networks. The following points briefly summarise three practices toward the archive culture:


  1. on the technological side, one should become archivist instead of users, and manage one's own digital objects and data, in order to create personal archives; software developers should pay attention to the development of softwares for personal archives of digital objects. This includes indexation, annotation of digital objects and the portability of data and metadata from one individual archive to another archive or another system; these metadata and annotation can be used for search use.


  2. opening up institutional archives and allows self-archiving, meaning archivists can download these digital objects. Institutional archives can still keep their own objects in a single place, but if users can download, share and annotate their own collections, and then contribute their metadata to institutional archives, this will significantly vitalize archives and move further from the ambition of the moderns described by Foucault (e.g. decentralized archive).


  3. individual archives can share with each other. This may recall us of the early idea of Napster, an idea based on P2P sharing. But it is not exactly the same, since in the framework I proposed, at centre is not the question of exchange of good, but rather of care, of preserving and giving, to get away from the crowdsourcing logic and most importantly to imagine a technological humanism, as was once proposed by Gilbert Simondon.




1 The idea of accumulating everything, of establishing a sort of general archive, the will to enclose in one place all times, all epochs, all forms, all tastes, the idea of constituting a place of all times that is itself outside of time and inaccessible to its ravages, the project of organizing in this way a sort of perpetual and indefinite accumulation of time in an immobile place, this whole idea belongs to our modernity. Michel Foucault (1967) “Of Other Spaces,” Diacritics 16 (Spring 1986), 22-27.

2 M. Foucault, Sur l'archéologie des sciences. Réponse au Cercle d'épistémologie », Cahiers pour l'analyse, n° 9, été 1968 «l'appellerai archive non pas la totalité des textes qui ont été conservés par une civilisation, ni J'ensemble des traces qu'on a pu sauver de son désastre, mais le jeu des règles qui déterminent dans une culture l'apparition et la disparition des énoncés, leur rémanence et leur effacement, leur existence paradoxale d'événements et de choses. Analyser les faits de discours dans l'élément général de l'archive, c'est les considérer non point comme documents (d'une signification cachée, ou d'une règle de construction), mais comme monuments; c'est - en dehors de toute métaphore géologique, sans aucune assignation d'origine, sans le moindre geste vers le commencement d'une archè - faire ce que l'on pourrait appeler, selon les droits ludiques de l'étymologie, quelque chose comme une archéologie».

3 MARC includes : MARC standard, MARC dialects, MARC issues, ISO2709 MARCXML, AACR2 etc.

4 See §26 Das Mitdasein der Anderen und das alltägliche Mitsein, in Sein und Zeit, Max Niemeyer Verlag Tübingen, 2006

5 These books are what Bernard Stiegler rediscovered in the work of Donald Winnicott transitional objects. The first transitional object is the thumb, babies suck their thumbs in order to compensate the disappearance of their mothers. Please see, Bernard Stiegler (2011), Ce qui fait que la vie vaut la peine d'être vécue : De la pharmacologie, Flammarion

6 Gilbert Simondon(1958, 2012), Du Mode d'Existence des objects techniques, Aubier, p.101

7 Gilbert Simondon, MEOT (1989) p.101-102, quoted by Xavier Guchet(2011), Pour un Humanisme Technologique – Culture, Technique et Société dans la philosophie de Gilbert Simondon, PUF, p.110

Jean Marie Vaysse (2006), Heidegger et Simondon : Technique et Individuation, in Technique, Monde, Individuation : Heidegger, Simondon, Deleuze, ed. J.M. Vaysse, Georg OLms Verlag