Why I stopped calling things digital – on metaphors as obstacles by Rafael Dernbach Difference engine by Charles Babbage, 1832 at Science Museum London A short glance at today’s job market shows that we live in a digital age. Companies look for digital managers, digital designers or simply heads of digital. Certainly, this use of the term ‘digital’ has little to do with its original meaning: the representation of information in binary code, in ones and zeros. What we mostly mean when we call something digital is that it involves computers. Yet, in a society where computers affect most parts of life, is it still sensible to describe something as digital? Or has the term, in fact, become an obstacle in discussions? The following is the story why I stopped calling things digital. It is also a text on the politics of metaphors. Recently, I sat at a table with a group of hackers, technologists and researchers from the humanities and social sciences. We were discussing the name for a new research group that we wanted to form. To my surprise, all proposals including the term digital were met with outspoken skepticism from the hackers and technologists. One of them expressed his doubts very vividly: “It is not that I am against the title, I don’t care about it and I know that many people will not care about it. I am not even annoyed by it, I will not bother to read on, if I see a meaningless title”. I was surprised by this fierce resistance to a term that seemed intuitive to me. And I became curious why ‘digital’ was such a red rag for exactly those people, who are most immersed in the subtleties of what I referred to as digitalization. I could see how the term ‘digital’ and its noun ‘digitalization’ were sloppy, encompassing a broad range of phenomena, from online marketing to the storage of data, from the recording of music to computer generated images. But I had an intuition when using them, namely, the intuition to describe a transformation triggered by the use of computers. And digitalization, the totality of these transformations is real. We all have a story how an app or an online service radically simplified or complicated an aspect of our everyday life. And we have witnessed the transformation of entire industries through the introduction of computers. So, why was there such a resistance against the concept in our group? After our discussion I asked one of the technologists to explain his skepticism in detail. This technologist was Palle Dahlstedt, a professor of computer science and a musical composer who teaches at the University of Gothenburg in Sweden and at Aalborg University in Denmark. “The word digital strictly only means that something is represented in discrete steps in contrast to continuous analogue signals” Dahlstedt explained to me. He continued to describe how most phenomena perceptible to human beings can represented digitally today, which is particularly true for acoustic phenomena. What makes digital representation so attractive in contrast to analogue representation is that digital objects can be replicated without loss. A clone of a digital object does not display any difference to its original, which is not the case for analogue representations. So far the term ‘digital’ did not seem problematic to me, but Dahlstedt continued: ”A lot of things that we associate with the word digital are, however, also possible with non-digital means. And a lot of things that most people think are analog are really digital. For example all writing is digital, and has always been.” [home] Home is where the heart is. Home is where it hurts ... On the economy of machinery and manufactures, 1835 In fact, digital signals can be represented with light, with hydraulic pressure, with mechanical gears, or with voltages. As most people, I imagined something electronic when talking about digital representations. But not only are there are analogue computers that can be both mechanical or electronical. There are also digital computers that can operate non-electronically. An example for such non-electronical computers are the computing machines that the English mathematician Charles Babbage developed in already in the 19th century. Digital, it became clear to me, describes a different and a far older idea than the processes of an electronic computer. This misunderstanding seemed also to be at the core of Dahlstedt’s skepticism: “the term has become all and nothing, sometimes it means ‘electronic’ sometimes it means there is a computer involved, sometimes it means it has to do with internet.” Dahlstedt sees the reason for this in a lack of understanding by a majority of people in the humanities and the broader public for the terms of computer science. “Terms are used loosely, metaphorically, and finally lose their value.” When I asked him about an example from his own field for such a loss of explanatory value, he told me that in his work with electronic music digitality is not a crucial parameter. “Many of my tools are digital, many are analogue and many are hybrid.” He added that recording can be digital or analogue, too. And to complicate things even more, certain acoustic signals can be discreet and therefore digital in time, but not in magnitude and vice versa. It became clear to me that the categories ‘digital’ and ‘analogue’ fail to address what is driving Dahlstedt’s production process. He insisted: ”it makes no sense to talk about digital as opposed to non-digital. It is like saying that a CD with the Berliner Philharmoniker is computer music, because there were computers involved in the making of it. Of course it is digital music, in a way, but that is not the crucial parameter. It is a digital way of storing the music – it was analog until it reached a certain machine in the chain, and it is analog when it reaches our ears.” For Dahlstedt a far more relevant parameter than if something is digital or analogue is the way one interacts with one’s tools. Rather than distinguishing digital from analogue objects or practices, the particular methods used and their intentions matter. What kind of computation is involved in the composing? What forms of abstraction are applied? What data structures are used? To subsume these questions under the term digital would not only miss the point etymologically, but also would create an obstacle in practice. Analytical Engine by Charles Babbages 1834-1871 Two experimental models for Babbages Analytical Engine, 1870 And yet, Dahlstedt does not relativize the new possibilities that increasing computational power holds. “With electronic digital computers, we can do things that we could not do before, mostly because of the orders of magnitude of increase in speed and memory over the last decades. Gradually we have pushed our thinking, and our tools, and the cognitive models of the tools, to be able to imagine new processes we can do to data – but the difference is only in terms of orders of magnitude in (storage) space and (processing) time.” Dahlstedt emphasized that the modes of computation are nothing new. What has changed is mainly the speed in which information can be computed and the storage space and, thus, the amount of information that can be computed. “All methods, for example, are well-defined and could have been done one hundred years ago, but they were unimaginable. Any computation, search, data manipulation, calculation, was in theory possible before the age of computers. But since it would have taken unthinkable amounts of time and storage space to compute it was unimaginable.” For Dahlstedt this means that our thinking develops with the tools. But he also insists that our thinking does not become more complicated. Rather than that we as humans deal with higher levels of abstraction thanks to the increase in speed and size of computers. After our conversation the term digital did not longer feel so intuitive. I started to wonder about my impulse to call things digital and realized the term often hadn’t added much to my conversations. A digital manager is just a manager, a digital designer just a designer. So why was I bothering to add the term at all? Because calling something ‘digital’ implies a narrative of a disruption. It implies a future with shiningly new things and ways of living. More than an explanatory I used it as an aesthetic category. By calling things digital I was taking part in this meta-narrative. From Dahlstedt’s perspective the term had just become meaningless and annoying. But it was dawning on me that ‘digital’ was more of hype than explanation, a metaphor that had detached itself from the fields of its origin so much that it now was meaningless. I wondered if ‘digital’ had even become an obstacle for understanding how computers interact with our lives. Could the term even foreclose insights about our relationships with information technologies? A provocative essay by the psychologist Robert Epstein came to my mind. Its title is also its premise: “your brain does not process information and it is not a computer”. In the essay Epstein criticizes the dominant framework of cognitive neuroscience, namely, that human brains are information processing networks. Epstein points to a range of misconceptions that come with the metaphor of the human brain as a computer. In the essay he elegantly debunks for instance the idea that consciousness might become downloadable at some distant time in the future. Epstein reminds his colleagues and readers that they operate with a model and, thus, with a metaphor. And he argues that this metaphor has become an obstacle framing many problems in neuroscience as unexplainable. What I really liked about Dahlstedt’s explanations was that they pointed to the historical continuities of computational methods. At the same time they were conscious of the transformative potential that their development holds. All these nuances are veiled by a narrative of disruption that comes with the term ‘digitalization’. When calling something digital I was buying into this hype, rather than understanding a particular technological context, its histories and implications. I was focusing on the new, while bracketing out where the new had its beginnings and in which directions it might drift. Just in the case of the brain as a computer, digitalization as a disruption is a faulty metaphor. As Dahlstedt described, even our ancestors relied heavily on digital media, if we count writing one of them. Describing something as digital frames the computer as a technology without history. It desensitizes us for the historical conditions and human agencies that have contributed to the development of a particular technology. But how could we now imagine a term, that takes into account the continuities of computational methods and still describes the transformations we are facing today? It might seem to you that this is a call for more precise terms. After all, careless imprecision seems to have turned ‘digital’ first into a useless metaphor and later into an obstacle for understanding. But rather than calling for more precision, I would propose a more careful imprecision. The theory of epistemic objects by the German philosopher of science Hans Jörg Rheinberger shows that certain types of imprecision are as necessary as the conceptual clarity. His theory also might help us to find terms that embrace both continuity and transformation. Rheinberger’s research examines how new ideas come into existence in science. In contrast to the idea of a divine inspiration or divine disruption of individual researchers or a technology, Rheinberger argues that science generates new knowledge by iteration. A single experiment without its context remains meaningless. But if experiments are repeated, adjusted and contextualized to a body of knowledge new ideas come into existence. Rheinberger calls this not-yet-formulated body of knowledge of a discipline its epistemic object. Epistemic objects are produced through an Experimentalsystem, the totality of all research technologies, experiments, instruments and infrastructures of a discipline. They are per definition imprecise at first. That means that they are not entirely knowable as they are the imagined knowledge that an Experimentalsystem strives for. Accordingly, an epistemic object has to be precise enough to generate knowledge and imprecise enough to incorporate unexpected results of experiments. Rheinberger’s theory shows that we should be carefully imprecise with our terms and that we should value iteration more than disruption. Rafael Dernbach is a writer, researcher and currently PhD candidate at University of Cambridge. As a Gates scholar he researches the social construction and aesthetics of futures, from science fiction to scenario planning. His essays, videos and interviews have been published by several media outlets including German Life and Letters, Teknokultura, ZDFinfo and Zeit Online. If the ‘digital’ is an epistemic object, we were not only blind for its continuities , but also carelessly imprecise with it. What could be better the epistemic objects when trying to make sense of the transformations that computers have brought to so many domains of life? I don’t have an answer to this question. However, I have an intuition that these terms will as hybrid as our production processes have become today. After all computing is not a phenomenon limited to computers. For our research group we ended up with the title Cybernetic Symbioses: shaping human-technology futures. It involved enough imprecision embrace the social impact and imaginaries of technologies. And it was sufficiently discreet to communicate clearly the merging of ideas from two systems, namely informational studies (cybernetics) and life sciences (symbiosis). This is when I stopped calling things digital.