The Monster Archive: from Ars Memoria to Big Data
p. 147-159
Texte intégral
1Since the ars memoria of the seventeenth century, we have known that the capacity of human memory is intrinsically tied to space and that the very possibility of developing its faculties lies in spatialization. Ars memoria is hence one of the defining factors in the shift from orality to visuality; not only is it a method for enhancing memory, but it also provides us with the first visual representations of the activity of the human brain in spatial form. Its technique, called the method of loci, rests upon two key data: “images” and “places.”1 The leap from ars memoria to big data might seem gigantic here. The purpose of this three-century-long leap across history is not to reiterate the relevance of this old technique or to show its resurgence in digital culture, but to cast light on the schematization of memory for which its representations were meant. Architectures or palaces of memory, to mention its two best-known visual paradigms, representations of memory do not emphasize that which is being recalled - the content of recollections or information - but rather the system of representation itself.
2With the technological tools available to us today, might it be possible to provide an accurate visualization of memory - of our memory? This question immediately raises another: might there exist a space capable of containing all of human knowledge? Isn’t this a problem - or rather, an enigma - already depicted by Jorge Luis Borges in his very brief 1941 short story The Library of Babel? Yet something strange stands out from the start: it takes Borges only a dozen or so pages to describe the complex and infinite structure of this library - as if, in inverse proportion, his visual description required only a very limited number of words. This literary paradox adds a third key factor of memory to those of images and places: storage.
The storage of memory
3A recent example will allow us to update these issues for the digital age. In 2013, artist Nicolas Baier exhibited two side-by-side representations of memory in which he brought together the living and the machine. The first of the two, Engrams (neurons), is an image of the flow of human brain activity, and the second, Engrams (the world of ideas), is a computer-like structure replicating the storage space of all human knowledge since the invention of writing. Their titles use a neuropsychological term referring to processes of memory inscription, the engram being the trace left in the brain by a past event. Combining “en,” meaning “in,” and “gramma,” meaning “written character and trace,” the term’s etymology underscores this material dimension - as does neurophysiology, which defines the engram as the “biological trace of memory,” a “mnemonic artifact” lodged inside the brain.2 Nicolas Baier draws from these sciences to reconstruct the activity of human memory, of which he provides a visual representation. The picture entitled Engrams (neurons) features a reconstitution of this living system; in other words, it attempts to portray human memory. The photograph shows the traces left by multiple neural connections, and is a collage of microscopic files which required months’ worth of photographs with a microscope.3 The specificity of this abstract portrait is to make visible the traces left by the inscription of recollections in memory without revealing their content.
4Whereas this picture takes us to the core of human memory function, Engrams (the world of ideas) plunges us inside a computer network by faithfully reproducing the compartments of a computer server used for recording data. The juxtaposition of the two artworks creates a tension between the human and the machine: on the one hand, the flows of the living reticular system and, on the other, the rigid structure of digital servers; on one side, the brain as the site of memory and, on the other, the computer as storage space. In spite of the differences between the world of the living and the virtual universe, we should emphasize that neuroscience and computer science define memory in a very similar way, treating it in terms of its capacity to acquire, retain, and restore information, and in general as a form of storage.4
5With its superimposed black octagonal shapes, Engrams (the world of ideas) presents an elegant structure. Built with thin volumes of polished metal and rubber wires, it sits upon a white marble surface designed to evoke, according to the artist’s intent, the origin of knowledge and philosophy in Ancient Greece.5 The space is empty: only the outline of the computer servers has been mapped out, letting visitors peer inside a complex network of galleries and tiers which, despite its extreme precision, seems undefinable. The files stored in a database this size, according to Nicolas Baier’s calculations, “could contain all the books ever produced by humanity in all languages since the invention of writing.”6
6Though it is not as immense as the Library of Babel, this architecture of memory recalls Borges’s project in several ways. There are indeed many connections between the two. Borges had also imagined a space capable of containing all the books ever written and all the books to come, but without revealing their content.7 And yet the library is not empty - quite the opposite. The first pages of the short story quite meticulously describe it as a gigantic space made up of identically arranged hexagonal rooms, galleries, or floors, a circular staircase, and vast ventilation shafts. Despite these specifications, it would be difficult or even impossible to concretely reconstruct its architecture, and this for several reasons: it is analogous to the universe, it is of considerable size, the number of books it contains is infinite, and, most important, its logic is combinatory. The memory of the world is preserved in its hexagonal galleries in the form of books, but their number is enormous and their content has been replaced by a code which, in all likelihood, cannot be deciphered:
Five shelves correspond to each one of the walls of each hexagon; each shelf contains thirty-two books of a uniform format; each book is made up of four hundred and ten pages; each page, of forty lines; each line, of some eighty black letters. There are also letters on the spine of each book; these letters do not indicate or prefigure what the pages will say. I know that such a lack of relevance, at one time, seemed mysterious.8
7Certain information is given to help decipher its content and combinatory logic, such that a “librarian of genius” is said to have discovered the fundamental law of the Library without having accessed the content of the books. He deduces from his investigation “that the Library is total and that its shelves contain all the possible combinations of the twenty-odd orthographic symbols […]; that is, everything which can be expressed, in all languages. Everything is there: the minute history of the future, the autobiographies of the archangels, the faithful catalogue of the Library […].”9
8But the mystery remains. Some theories speculate that the library’s space may be infinite and hypothesize the total number of books it contains. According to philosopher and psychoanalyst Didier Anzieu, it would be mathematically conceivable to calculate the number of books in the library based on the information disclosed by Borges. He presents the math in his book Le Corps de l’œuvre (1981) based on initial calculations carried out by the mathematician François Le Lionnais:
[…] the number of books contained in the Library of Babel comes to 25 to the power of 1,312,000. Each book contains 410 pages made up of 40 lines with 80 characters each, or 410 x 40 x 80 = 1,312,000 free spaces per book, among which the universal alphabet’s 25 letters of which all the books are composed must be divided each time in a different way. Although it is not infinite in the mathematical sense of the term, the figure 251,312,000 is so great that it exceeds any human capacity not only of actualization but also of representation.10
9The analysis of these figures confirms that representing the library is indeed inconceivable. According to the French Wikipedia, the number of books - that is, 25 (410 x 40 x 80) - would exceed the number of observable atoms in the universe and, most important, the storage space of a computer: “This number would contain 1,834,098 digits, which shows that computer technology does not have the capacity to actually create the library […]. It is also worth noting that the mere writing of this number would require more than one book from The Library of Babel.”11 The estimate is 1.4 books. Borges says so himself: the library has no limits, and even if it is mathematically possible to calculate the number of books it contains, its storage space is still undefinable, because irrepresentable. Although there have been scores of attempts at providing a picture of it, they all more or less lead to the same result: the materialization of such a library remains confined to the realm of illusion and fiction. This architecture of memory is therefore purely fictitious, and all the more so because when The Library of Babel was written, no real or imaginary space could provide even a rough schema for it.
10Didier Anzieu also draws our attention to the dizzying and irrepresentable nature of this result: “Although it is not infinite in the mathematical sense of the term, this figure is so great that it exceeds any human capacity not only of actualization but also of representation.”12 To prove his point, he takes the theme of the short story and applies it to the real world - a particularly fascinating experiment. In the same line of thought as Borges, he calculates what our library might contain, namely, all of the oral and written knowledge since the creation of the world. He starts by estimating the number of human beings since the beginning of speech and comes to the number 1013. He then measures the amount of verbal signs used on average by a human being in his or her lifetime and comes to ten billion signs, or 1010. Finally, he proceeds to one last calculation in order to estimate the number of books required by the transcription of all the “verbal or written speech of a human being in his or her lifetime,” that is, “around 10,000 books per human being”: “the sum of 10,000 billion human beings existing or having existed could have produced, at a rate of 10,000 books each, a total of up to 100 million billion books: 104 x 1013 = 1017.”13
11Even if the selected criteria are difficult to verify and may be easy to challenge, and even with a wide margin of error, as Didier Anzieu insists, this number “while enormous in absolute value […] is still insignificant in comparison to the number of books in The Library.”14 Whether it is 10 to the power of15 or 10 to the power of 2015, and even if we project this estimate into a future where the number of human beings and books is constantly growing, up to, for example, a fictitious number of 1025 or 1030, the required space, according to Didier Anzieu’s estimate, would still be insignificant with respect to the mathematical projections that Borges’s system anticipates.
12Whether right or wrong, these last calculations show that the concept of the Library of Babel indeed had the potential to extend beyond itself into our own reality. The threshold between the fictional universe and the real world had already been crossed before Nicolas Baier explored the possibilities of such a step and before the computer came to facilitate its mathematical calculation. Engrams (the world of ideas) imagines the storage space necessary for archiving all the books published since the invention of writing. But although the space of the Library of Babel remains, as we have shown, irrepresentable and unfigurable, Engrams shows just the opposite. On the one hand, visualizing the Library seems too complex, perhaps even impossible, because of the dizzying number of books it contains as well as the confusion of its hexagonal galleries and the “nightmarish spirit” that prevails within them. On the other hand, the computer provides a perfectly fixed, organized, and quasi-rigid representation of human knowledge.
13On a strictly formal level, the Library and the computer server can be described as an abstract framework in which nothing can ultimately be discerned. In both cases, the books are dematerialized. With Borges, they are reduced to the status of signs in a system of codes and combinations, whereas with Baier, they are reduced to the status of mere digital data in a system of algorithms. But what is then being represented? Several authors insist that Borges is representing the container rather than the content, that he reveals writing to be a series of characters that compose no precise language (hence the reference to Babel). He is interested in the structure of language, in the code - and it is no accident that the short story has been fertile ground for linguistic and structuralist studies such as Didier Anzieu’s.16 In Nicolas Baier’s work, we are dealing with an empty, translucent big data center, a stack of servers similar to how we might imagine the hexagonal structure of the Library. Once again, the emphasis is placed on the container rather than on the content; and yet we must go beyond this binary opposition. The immateriality of the data is “figured” by the empty spaces, which reveal nothing but the size of the servers, in other words, the storage space. Engrams (the world of ideas) is a material visualization of knowledge once it has been dematerialized as digital data.
14In both cases, the aim is to represent the universe, something often mentioned with regard to Borges’s reference to the Tower of Babel17 and which we also find on another scale in Nicolas Baier’s remark that Engrams offers a “self-portrait of the digital world.”18 While the two projects may seem to overlap in their desire to imagine everything the mind has produced in order to contain it within a single space, it is worth noting that the means to this end are profoundly different. Borges projects himself into imagination and fiction, whereas Baier models knowledge and its space in the material world based on real estimates. When Baier created Engrams, the computer made this type of representation possible because the storage capacity of a computer server is quantifiable and computable - even though it is ever increasing and even though this increase is, as we all know, inversely proportional to its space. It is because of these technological resources that Engrams (the world of ideas) can claim to provide an image of what a virtual library - a big data center - containing every book might be. The structure is indeed technically capable of containing “[all] the memory of the world hidden in databases.”19 In other words, Engrams is a mega- archive of all the knowledge of humankind. But there is still the issue of calculating it…
15Two mathematical operations were instrumental in creating this portrait of knowledge: how many books have been published in the world since the invention of the printing press and what amount of space would be needed in order to store all the knowledge written, produced, and saved to this day? It was therefore necessary, on the one hand, to determine the number of books and, on the other, to replicate the required storage space in line with the most recent technology. The first calculation was based on figures released by Google in 2010, at the time the company announced its project of cataloging and archiving the entirety of the world’s written knowledge. According to Google’s algorithms, the number of books would be around 130 million, or, to be exact, 129,864,880 (the equivalent of 4 billion digital pages and 2 trillion words).20 Baier used these data in order to count the number and power of the computers necessary for the largest existing archive of knowledge: around 140 million books x 3 megabytes on average = 420,000,000 megabytes.
16It should be noted that the figure is purely theoretical. It was calculated in 2010 and the mass of books has continued to grow ever since. In 2013, the year Engrams was produced, Baier estimated that, in a space of relatively reasonable size (206.5 x 388.5 x 122 cm) containing X number of computer servers, it would be technically possible to download all the books published since the beginning of humankind. Accordingly, Engrams (the world of ideas) provides a “picture” of what “will have been” the storage space of all the world’s literature based on the technology of the time.
17In so far as it is an image frozen in time, this structure curiously possesses the features of photography. Of a photograph in the post-photograph age.21 The comparison is not insignificant since Baier was first known as a photographer and because he has extensively explored the possibilities “[of] photograph [ing] the unphotographable and [of] captur [ing] the invisible,” what he views as a “quest to do the impossible, the creation of new worlds.”22 But is this “still” image crystalized in metal and marble not already an archaic image? With technology evolving at phenomenal speed, is this spatial configuration of knowledge not already outdated? Engrams points to the speed at which the computer produces planned obsolescence, anticipating, in other words, the loss of its own memory. Its empty structure may hence be viewed as an artifact or a document of an already bygone period of digital evolution. The idea seems here to be to freeze, to archive, this memory in a form that bears witness, not to human knowledge, but to the storage capacities of digital databases.
Babel versus Google: from the irrepresentable to the monster archive
18If we return to our two reference points, Babel and Google, the opposition may seem extreme. Yet it speaks to the gap separating the fable from reality: it is no longer a matter of a space capable of containing and storing all the literature of every language, but of the appropriation of the totality of knowledge; no longer the vertigo caused by the infinite quantity of knowledge, but its control by a database; no longer the incommensurable, but the measurable; no longer the irrepresentable, but the immediately graspable; no longer an architecture of memory, but the monstrosity of a digital archive able to encompass and devour everything.
19This venture of appropriating knowledge is unprecedented, and Google Books is one of Google’s most controversial development projects.23 Its original mission to “organize the world’s information and make it universally accessible and useful” as well as its accomplishment of that mission are still facing numerous legal challenges. Indeed, the function of its database is no longer that of a conventional archive, i.e., documentation and preservation, consultation and organization, but rather to provide immediate access and instant browsing to anyone in front of a computer screen or mobile device. The archive, which was once the privilege of the historian and in the service of history, is now in the hands of a multinational, with all the economic, legal, and ethical implications that that entails. The difference is crucial. In claiming to provide access to the entirety of global knowledge, Google Books has fostered the illusion that there is no longer anything that cannot be immediately accessed, that with the computer everything is potentially saved, archived, indexed, intelligible, usable, etc. Not only is knowledge, like a commodity, now subject to a monopoly, but it has also become the property of a database that handles its transmission and organization according to its own system of networks and algorithms. Treating knowledge as mere data obviously has consequences.
20Is Google in the process of spawning a monster archive? I borrow the term “monster” from historian Pierre Nora, who applied it to the “event” in an article published in the French journal Communications in 1972, and in which he laments the overproduction of events triggered by mass media.24 Has the computer in turn transformed the archive into something monstrous, just as “mass media have […] turned history into an act of aggression, and in the process […] have turned events into something monstrous”?25 What would the defining characteristics of this monster archive be? Is the monster archive unique to digital technology, capturing until all is devoured? The central features of the transformation of the event as defined by Pierre Nora could be applied nearly word for word to digital culture. The event becomes inseparable from the means of communication: the media have become, he argues, the “very condition upon which events are dependent for their existence.”26 By appropriating the event, mass media have secured a monopoly over history: “In modern-day societies,” he writes, “it is through the media and through them alone that an event strikes us, and cannot escape us.”27 Is Google gaining a monopoly over human knowledge?
21If the event appears monstrous, that is because it is given immediately, without “the work of time”: “it pits the present against history, the immediate against the historical, lived experience against the legendary.”28 Live transmission and rapid access strip “the event of its historical character.”29 According to François Dosse, Pierre Nora “highlights the extraordinary capacity for amplification” of the media, which makes “the event exist in a relationship of immediacy.”30 This amplification has a double impact on the event: it transforms its nature and value, promoting the eventfulness of the event, and fosters unintelligibility. A “state of perpetual over-information” and “chronic under-information.”31
22What might be the effects of the mutation of the archive into a digital technology? If the monster archive transforms knowledge into mere data and if the library is superseded by big data, does this not mean that knowledge becomes subject to unintelligibility? Given the extraordinary storage and indexation capacities of databases, are we not faced with an ever-expanding archive? And doesn’t the speed at which the digital is capable of increasing and gathering data to the point of overload expose the archive to obsolescence?32 Not only do digital technologies produce a never-ending and infinite accumulation of information, but their storage capacity also leads to a paradoxical reversal of memory. Given the obsolescence with which the archive is threatened, should we not envision a memory of memory, an archive of the archive? This last question brings us directly back to Engrams (the world of ideas). As a representation of a database of all the world’s books, such an architecture of memory reveals the coexistence of two incompatible logics and temporalities. The beauty of its structure conceals the monstrosity of its storage capacity, the voracity of the digital, and of big data in particular, such that, at the very moment that it is displayed, it is already an archaic representation, an image of the past, an archive of the archive. At the same time, Baier’s depiction of the digital universe also immediately reinscribes this obsolescence within a new layer of time. The dematerialization of knowledge and memory is joined with the permanence of this tribute, its materiality set in stone and metal. Engram is thus a representation of the digital world beyond the digital.
Nicolas Baier, Engrams (the world of ideas), 2013. Metal, white marble, wires. Collection of the National Gallery of Canada. With the permission of the artist. Photo credits: Jimmy Limit.

Nicolas Baier, Engrams (the world of ideas), 2013. Metal, white marble, wires. Collection of the National Gallery of Canada. With the permission of the artist. Photo credits: Jimmy Limit.

Notes de bas de page
1 See Frances A. Yates, The Art of Memory (London: Routledge, 1966). Ars memoria “seeks to memorize through a technique of impressing ‘places’ and ‘images’ on memory. It has usually been classed as ‘mnemotechnics.’” It is an “art which uses contemporary architecture for its memory places and contemporary imagery for its images” (11).
2 See the definition of “Engram” on Wikipedia: https://en.wikipedia.org/wiki/Engram_(neuropsychology).
3 https://nicolasbaier.com/transmission-en/
4 These definitions of memory in neuroscience and computer science are fairly basic, but they nevertheless cast light on the similarities and on the importance both definitions attach to storage. See Laurent Petit, La Mémoire (Paris: P.U.F., 2006), 5.
5 https://nicolasbaier.com/transmission-en/
6 Ibid.
7 The library is a recurring concept in literature and is often compared to the encyclopedia projects. Borges himself drew inspiration from a short story by the German writer, philosopher, and mathematician Kurd Lasswitz, The Universal Library, originally published in 1904 – not to mention the libraries that have in turn been inspired by Borges, including Umberto Eco’s in his novel The Name of the Rose. See Paul Zumthor, Babel ou l’inachèvement (Paris: Seuil, 1997).
8 Jorge Luis Borges, “The Library of Babel,” trans. Anthony Kerrigan, in Ficciones (New York: Grove Press, 1962), 80.
9 Ibid., 83.
10 Didier Anzieu, Le Corps de l’œuvre (Paris: Gallimard, 1981), 307-308.
11 See the article “La Bibliothèque de Babel” on Wikipedia: https://fr.wikipedia.org/wiki/La_Biblioth%C3%A8que_de_Babel. For the English-language article: https://en.wikipedia.org/wiki/The_Library_of_Babel.
12 Didier Anzieu, Le Corps de l’œuvre, op. cit., 307-308.
13 Ibid., 308.
14 Ibid., 308-309.
15 These figures are provided by Sylvie Parizet, among others. The number of books in the library would “be in the range of 25 to the power of 1,312,000, according to François Le Lionnais’s calculations based on the indications given in the text.” I borrow the reference to Didier Anzieu from this article. See Sylvie Parizet, “‘Bibliothèque de Babel’: ordre ou chaos?” Vox Poetica: http://www.vox-poetica.org/sflgc/biblio/bibliafin/parizet.html.
16 Although I have left out his psychoanalytic interpretation, Didier Anzieu is a telling example of structuralist thought in that his analysis aims to identify the logical arrangement of the Library, assumed to correspond to a code reflecting the structure of language and writing.
17 Sylvie Parizet, “‘Bibliothèque de Babel’: ordre ou chaos?” op. cit.
18 https://nicolasbaier.com/transmission-en/
19 Ibid.
20 https://en.wikipedia.org/wiki/Google_Books
21 Post-photography is a much more complex notion than the one I am mentioning here: as a photographic view of the world beyond the image and the use of a camera. It is also a concept that refers to the overproduction of images and the dematerialization of their material media. See Joan Fontcuberta, ed., The Post-Photographic Condition (Montreal: Le Mois de la Photo à Montréal, 2015).
22 https://nicolasbaier.com/where-are-the-definite-poetic-sites/
23 Among the legal issues facing Google Books, the question of royalties and copyrights is central. Accusations of imperialism are also looming over Google, which has been looking to appropriate the knowledge of the world without making its own objectives clear, a situation that opens the possibility of its exploitation of knowledge for purely economic ends. The legal meanderings of Google Books are reported on Wikipedia: https://en.wikipedia.org/wiki/Google_Books, accessed January 18, 2017. It is also worth noting that one of Google’s objectives is to control time 24/7 by increasing our screen addiction. As Jonathan Crary argues, we are currently facing a “vast disproportion between those human […] limits and the quasi-infinite amount of ‘content’ being marketed” by Google and Facebook. See Jonathan Crary, 24/7: Late Capitalism and the Ends of Sleep (New York: Verso, 2013), 75.
24 Pierre Nora, “L’événement monstre,” Communications 18 (1972): 162-172 [in English translation, quoted here: “Monster Events,” trans. Lee Hildreth, Discourse 5 (Spring 1983): 5-20], modified and published under the title “Le retour de l’événement” in Faire de l’histoire, vol. 1, eds. Pierre Nora and Jacques Le Goff (Paris: Gallimard, 1974), 210-227. I am basing the present study on the original version published in 1972. Also see François Dosse, Renaissance de l’événement. Un défi pour l’historien: entre sphinx et phénix (Paris: P.U.F., 2010), especially the chapter borrowing the 1972 article’s title, “L’événement-monstre,” 241-259.
25 Pierre Nora, “Monster Events,” op. cit., 8.
26 Ibid., 5.
27 Ibid.
28 Pierre Nora, Faire de l’histoire, op. cit., 212.
29 Pierre Nora, “Monster Events,” op. cit., 10.
30 François Dosse, Renaissance de l’événement, op. cit., 241.
31 Pierre Nora, “Monster Events,” op. cit., 13.
32 See for instance Jussi Parikka, What is Media Archeology? (Cambridge, UK: Polity, 2012).
Auteur
-
Marie Fraser
Professor of Art History and Museology at the Université du Québec à Montréal and a member of Figura, Research Center on Text and the Imaginary. Her research explores the transformation of narrative and temporal regimes in contemporary art, as well as forms of updating and anachronism in artworks, exhibitions, and museum collections. She is part of the research and inquiry group CIÉCO (Collections et impératif événementiel/The Convulsive Collections), endowed by the Social Sciences and Humanities Research Council of Canada (2014-2018). She has recently published articles in such journals as Muséologies, Culture & Musées, and Stedelijk Studies. As a curator, she has organized over thirty exhibitions in Canada and Europe, working as the chief curator for the Musée d’art contemporain de Montréal from 2010 to 2013 and curating the BGL exhibition for the Canadian pavilion during the 56th Venice Biennale in 2015.
Le texte seul est utilisable sous licence Licence OpenEdition Books. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
Le Sujet Digital
Claire Larsonneur, Arnauld Regnauld, Pierre Cassou-Noguès et al. (dir.) Stéphane Vanderhaeghe, Géraldine Bertres, Hélène Soldano et al. (trad.)
2015
Le Comportement des choses
Emanuele Quinz (dir.) Lise Thiollier, Gabriele Stera et Armelle Chrétien (trad.)
2021
Artistes-chercheur·es, chercheur·es-artistes
Performer les savoirs
Boudier Marion et Déchery Chloé (dir.)
2022
Architectures of memory
Jean-Marie Dallet et Bertrand Gervais (dir.) Armelle Chrétien et Joshua David Jordan (trad.)
2022
Angles morts du numérique ubiquitaire
Glossaire critique et amoureux
Yves Citton, Marie Lechner et Anthony Masure (dir.)
2023