An archivist is traditionally concerned with gathering information and then classifying it according to largely pre-existing categories. But, in the era of “big data”, the sheer volume of information now being generated makes this process nigh on impossible. Instead of archiving, we now tend to talk of data harvesting or mining. IBM claims that back in 2012, 2.5 billion gigabytes of data was generated every day, of which eighty percent of this data is unstructured, and that there are now more than one trillion connected objects and devices generating data across the planet. We are documenting the present in ever-accelerating real‑time, and the gap between ourselves and the data is closing exponentially.
Conversation between Mariabruna Fabrizi, Fosco Lucarelli (Socks) and James Taylor-Foster.
James Taylor-Foster: Perhaps we should begin by defining the difference between archiving and data collection. This situation has evolved a great deal over the past decade, which begs the question: has archiving become data collection, or is it at the very least heading in that direction? If so, what are the implications?
Mariabruna Fabrizi: For me, data collection and archiving are two different things. Archiving requires selection, somehow. Even, for example, with a ubiquitous website like Wikipedia, there is a sort of encyclopaedic will behind it. It’s a selection – but it wants to talk about everything. When I think about contemporary archives, Asimov’s Foundation Series comes to mind: the notion that we don’t exactly know where the future is going, and perhaps our civilisation is doomed, but we somehow try to keep all of our knowledge in an ordered way. I think data collection is much more about optimising processes – making a better selection for marketing reasons, for instance.
Fosco Lucarelli: In spite of what Mariabruna argues, I still feel that the word “archive”—or the act of archiving—represents a general attempt to collect vast amounts of data. I think this is utopian in a way, and Socks is not interested in that. To me, data collection is without a curatorial (if we can call it that) element.
In the act of selecting and curating, all three of us, in one way or another, are involved in what could be described as contemporary digital archiving in relation to the field of architecture; we’re just operating at different scales and in different ways. We each deal with large volumes of information in the form of data and represent it. In the act of re-presenting, this material begins to accumulate new data in terms of who is looking at it, how long they are looking at it, where and how are they are sharing it and so on. You start getting these pathways of propagation through the tendrils of the internet; data packets extend and distend until they have departed the point of the “curated” archive. Is it more accurate to suggest that digital archiving is becoming more about data collection than conventional curation?
MB: It depends on the scale. For some large organisations like Facebook, for instance, it’s all about data. But when you work at a smaller scale, like Socks, data is a only a small part of the equation because we cannot control it; we don’t know much about what to do with it, and we try not to think too much about it. We extract some information out of it, of course, but it doesn’t really influence what we are doing. To take data out of content requires a sort of critical mass.
It’s interesting that you suggest that you have very little control over the data that you are accumulating. Large-scale operations control the small-scale but, as a collective, small-scale operations can also control the dominant forces on the internet. For instance, algorithms can be deceived—quite easily, for now at least—to present particular data through certain social media platforms, if you have the knowledge, experience and flexibility to manipulate them; if you have, as you put it, the required “critical mass”.
One of the defining parameters of an archive, in the physical sense at least, is that it can be accessed and that access is tangible. By now, we have become so accustomed to the notion that practically anybody can access the internet but, in your position as the producers of a digital archiving project, do you ever consider the concept of access? How people are—or are not—accessing the data you re-present?
MF: We imagine [Socks Studio] as a territory of images and information, and we imagine ourselves and our users navigating pathways through it. Before the internet, or before we had such a large-scale data on the internet, we had magazines with editorials that provided a set of keys to unlock your way through their content. We are looking to find other keys to give to people who want to access the data, and there are many ways to do it. We don’t give editorial guidance, but create a sort of chain between content. You begin with one item of content that is connected to many others. People follow their own paths, and these paths are a way to create access. In this sense, we are sort of at a midway point between providing something that is curated by us but also opens other curatorial possibilities for the readers themselves.
So with conventional print publishing, where one or a set of keys is handed to a reader by an editor with a very clear idea of their position in the world and how they would like their readers to be positioned in the world, it is easier to control how information is disseminated and how, to a certain degree, it is interpreted. How different is your mode of editing or curation of data to the “tried and tested” print model?
MF: We collect architectural magazines from the past—like Casabella, for instance—and you can really identify which are the “Mendini” issues and which are the “Gregotti” issues: from the cover to the editorial, the selection, the way the photos are chosen, and the relationship between text and image. What happens now, as we agreed earlier, is that we are at a point at which the internet is becoming more mature as a medium and people have more instruments to navigate their way through it.
FL: All of the “free” access to data that has been the norm during the last decade is becoming more restricted today. Perhaps one of our responsibilities is to try to keep this data accessible. This might be considered a democratic dream—which is a little utopian—but we still feel that is one of the responsibilities we have today when we publish on the internet.
In the December 1997 issue of Wired magazine—back when Wired was just on paper—a theoretical physicist called Michael Goldhaber wrote an article called Attention Shoppers! He said—to paraphrase—that “the economy of attention, not information, will become the economy of cyberspace.” Information, he said, was endless, but attention – which is defined by how many “eyeballs” there are in the world able to consume information – is a finite resource; perhaps the most consequential finite resource. He argued that, eventually, the currency of money will be replaced entirely by the currency of attention because it is our scarcest and most valuable resource, collectively and as individuals.
His words were enormously prescient in terms of understanding the situation unfolding before us now. For every single major internet platform—through which people access contemporary archives and which sit somewhere between information gathering and media organisations—is the single most important resource simply because they are able to transform it most easily into financial revenue.
MF: This idea of attention as a currency really shifts the nature of editorial work that we were just talking about. Before this situation came along, a reader had fewer choices if they wanted to learn about architecture compared to today. Your attention was already captured. Now you have to make a product or a way of showing content that is engaging, and this changes everything completely. As little as ten years ago, the internet was less engaging and interesting compared to books and magazines. Now it’s completely the other way around. With magazines there was this idea of authority somehow.
Goldhaber also observed in the same article that what he found fascinating about Hollywood movies is that they developed an amazing acknowledgement of people by means of credits. When a book is made, not everyone involved in bringing it about is credited. The proofreaders or the subeditors simply don’t appear in colophons. On the internet, in the same way as Hollywood movies then and now, credits are all-important.
On ArchDaily, and particularly in relation to the Projects Archive, credits are central to the way that buildings are published. Acknowledgement operates almost as a currency. But Goldhaber was also raising a more interesting discussion: attention is only the aim of the game as long as everybody wants a slice of it, made more valuable because it is finite. Credits – metadata – are important because everyone wants their five seconds in the spotlight.
MF: It is also about authorship today, too. From the point of view of Socks, we are interested in authorship, but most of the time we are more interested in the relationships between content. More than what makes something specific, we’re interested in what makes it common. For us it is about more than the single person, it’s about a chain of knowledge – because everybody is an author of course.
FL: The fact that the importance of the individual is becoming eroded by the amount of information is very interesting. According to what you suggest, James, authors seem to be alarmed by this erosion, or disappearance, of the individual. We started Socks more than ten years ago as a way to overcome—it’s sort of a paradox—the spread of vehicles of information like Tumblr or Pinterest. They didn’t exist at the time, but they were incubating. We were far more interested in retrieving the actual information that was behind a single image or in a single text or single medium of production. When we became interested in an image, we tried to…
MF: …find the story behind it…
FL: …and, naturally, when you start investigating something you start making threads and working on nets of information; you end up creating a sort of tissue of information or relationships. The fact that these images come from an historical knowledge base, which is also related to a certain context, can also be generalised as human production. In this sense we try to create our own taxonomies. The amount of information is the amount of relationships between different works and different authors, or different groups of authors and movements. So taxonomies can be a way of grouping this information and going beyond limited historical and contextual specificities.
Let’s focus on images. The image has become the primary means of communication through the internet. Images and moving images have taken pole position at the top of the consumption chain, perhaps stemming from the image-centric dissemination of culture, media and entertainment that developed through film and television. While an image on the internet is comprised of the same bits and bytes that construct text, the retrieval systems developed over the last couple of decades have been basically designed to give priority to the image. And when Socks, for example, takes a single image in order to investigate it and add a whole tissue of information around it, it’s still that single image that first gets people and invites them to enter the rabbit hole that you have dug for them. The image is still paramount even after the research is conducted and presented.
MF: For an image to be able to do this, means that it’s very dense. When an image is able to attract people somehow, it always means there is something more behind it. So we cannot say that it’s “just” an image.
FL: In its singularity, an image can convey all these tissues of information. Take for example a recent post that we published on Socks about the work of the photographer Jeff Wall. It’s not just photography that he produces; it’s an entire world of information – a constructed landscape; a relationship with other pictures in the past; a relationship with culture and so on. I don’t mean that every image that we publish is so dense and so structured, but in a way every image is a text; every image is a body of information.
The fact that images are easy to read, and very quickly, means that they hold a lot of weight and value in terms of their role in the economy of attention. So who owns these images, and who holds control of their propagation, becomes very interesting. In the world of print, ownership rights are usually addressed within the realm of national laws. But on the internet things are a little more complex in terms of the lack of control over how they spread. Do you think that ownership of information like images and text, or even film and music, is less important now than it has been?
MF: I think it is just as important as it was before. There are people and work behind this information – it’s just the incredible mass of it…
FL: It is culturally important, much more than in terms of economic profit. It is culturally important in the sense that we need to know who produced an image, under what conditions and in what kind of relationship to other things.
Perhaps authorship is becoming more important simply because there is an ambiguity about ownership?
MF: Sometimes it’s very difficult to define the limits too. For example, we are seeing a lot of collages at the moment. Who owns a collage? The person who modified it? The one who put the parts together? Just as in the case of remixing music, who is the ultimate author? On the other hand, there is also much more collaboration taking place. Wikipedia is a good case in point – who can we say writes the articles?
FL: We are in an ambiguous stage. The importance of crediting is also in an intermediate state. At the beginning of the internet nobody really cared if their content was on there or not because you couldn’t profit from it, neither economically nor in terms of recognition. Today, if you are on specific platforms you can be recognised and your authorship recognised and, of course, you can profit from that. You need your name to be there but if you affirm your authorship for profit, it clashes with the work being shared, being used, being remixed, and so on.
The notion of archiving, or the intelligent archive, is no longer simply a passive entity or simple repository then. It has become, and is becoming, something in which, as you say, information can be written, can be mixed, can be reused, and notions of authorship are now pursued in favour of collaboration. In terms of context, where does this leave representation in architecture? Does this mean that architects should be taught differently about how to handle and deal with context and precedence – two important things which are, more often than not, lost once an image has entered cyberspace?
MF: Yes, I completely agree. We cannot be moralistic about being fascinated with an image and just taking it at face value. It will always depend on individual students, but teachers of architectural studios do have a responsibility there now.
FL: We have a responsibility to teach what’s behind a human product…
MF: …and to give a critical approach. The fact that we don’t have access to certain information about images and projects doesn’t mean we can just pick some parts of them and glue them together…
FL: …and decontextualise them.
MB: We need to teach students what’s behind an image, to say an image is not just something you use because it looks nice. They are artefacts in themselves; there is a culture behind them that you cannot take lightly.
FL: At the same time, as internet producers, we have been accused of fostering this very attitude that you’re talking about. We need to acknowledge the fact that we have huge power in this respect; this power must be used according to the potential it has intrinsically, but in a critical way. We cannot assume that it’s just interesting to put stuff on the internet and for other people to use it. We need to invite people to interrogate content themselves, investigate what’s behind a singular human product and the relationships that it has: in the past, the present and also in the future.