The Art of Deviation
From Representation to Stored-Attention
The following translation marks the beginning of a new series of Martin Burckhardt’s later lectures, essays, and chapters. The first is Do it Again!, a chapter from Philosophie des Designs1, published in 2020. While this essay was written as a description of aesthetic iterative design, it contains some of the leitmotifs from his earlier works, which matured with the publication of Philosophie der Maschine in 2018. Specifically, the changes in what an object was in Die Universale Maschine; the changes in work and rationality from the advent of the computer in In the Working Memory and Time is Money is Time; and his descriptions of the dividual that we read in Portrait of the Author, Electrified: On the Transition from Mechanical to Electromagnetic Writing and The Guillotine: The Connection Between Death Technology and Rationalization – all symptomatic of the crackling Zeitriß that have brought us forward into the current age of our discontent and unease with Modernity.
In particular, Martin looks at the SCRUM methodology as something beyond our recent theoretical discussion with Jeff Sutherland on Agile’s philosophical basis, that is: how the Artist’s Representational view from the Medieval Rädewerk Psychotope has transitioned into that of simulation in Modernity’s Digitalwerk. In examining iterative design’s relational impact on aesthetics, he brings sharply into relief how our old notions of a representational self have faded back into the more original notions of self as a simulacrum of our socioplasm – and how our psychoplasm is a reflection of our connections to it, meaning we’re an augmented dividual in a networked world. Thus, we’re no longer the sovereign who authoritatively defines who we are – after all, we killed our father, the King, long ago. In this sense, what he brings into relief by examining the aesthetic iterative design process of software is how our imagination’s power becomes stored attention in an economy where capital is no longer gold but attention in an economy of attention.
Martin Burckhardt
Do it again!
The Philosophy of Iterative Design
SCRUM, agile, test-driven, and data-driven development - for some years now, or more precisely since the 2001 publication of the Manifesto for Agile Software Development,2 the public’s been buzzing with the idea we’re dealing with a revolution in teamwork. This has resulted in everyone trying to become agile as quickly as possible. Add to this the promise of a big data revolution, and you get the impression that work will be able to do itself in the future. As with many buzzwords, the promising halo overshadows reality, with its categories and terminology becoming confused. At the same time – as perhaps the most fatal accompanying phenomenon – the reasons for this change remain largely unexplained. Historical hindsight points to a production system that Japanese engineer Taiichi Ōno pioneered at Toyota in the early 1990s. Ōno had two intentions: to diversify mass production (so that different types of cars could be assembled on the same assembly line) and improve quality. If an error occurred, the assembly line was stopped, and the entire team worked intensively on analyzing the error to find a remedy. Through this technique of incremental improvement, Toyota achieved quality improvements that seemed previously unimaginable. Independent of this development in the automotive industry, the question of contemporary project management in software development emerged – not least because programs were becoming increasingly complex and opaquely Byzantine. Here, two different aspects came together: one was the principle of test-driven design, which the American software designer Kent Beck put forward, and the other was the fixation on the most efficient group work possible, which led Jeff Sutherland and Ken Schwaber to their SCRUM logic (which could be understood as group work with an innovation-friendly role distribution focused on quality management). However, the aim of this essay isn't about tracing these different approaches - it's about what these approaches have in common, that is, the question of what iterative design is.
The Art of Forgetting
What is novel in this approach? If we apply the idea of iteration to the design process itself, we could argue that there is nothing new under the sun: the design process, which interacts with a natural material (the clay in hand, the splash of color on a canvas, the word on a piece of white paper), has always known the moment of revision: that of iteration. From this perspective, doesn't the aesthetic design process mark the non-plus ultra of iterative design? Of course, this objection fails to recognize the paradigm shift that occurred with the computer's advent. This is because, as a working memory, it can store processes far exceeding the capacity of a single person's brain but also that of a group. In 1955, Harvard psychologist George A. Miller wrote a short essay entitled The Magical Number Seven3, which dealt with the limits of human information processing. In it, Miller argued that, in actu, humans can hardly keep track of more than seven variables - which is why, with increasing complexity, we're dealing with an explicit limitation of our human ability to perceive and react. Since humans also have a deficient memory, storage media come into play, not in the sense of ars memoria, but in the sense of ars oblivionis. Seen in this light, the written text is not only an aid to memory but also a form of forgetting, a black box that reliably protects us from information overload. Already familiar to us in the form of books, libraries, and encyclopedias, our computer culture's electromagnetic writing brings into play the possibility that everything electrified can take on the character of writing.
In this sense, any work that's been analyzed and digitized can be transferred to the working memory as stored attention - and from there, it can be teleported to any place in the world, reproduced, and reactivated at will.
This description alone should clarify that we're dealing with a first-order paradigm shift. And this is precisely why we’re no longer talking about art but about a process of iterative design - the paramount importance is the machine leaving craftsmanship behind.
Undone
Nevertheless, examining the relation of iterative design to aesthetic principles is helpful for working out what’s so unique about this new process. For this reason, the thoughts presented here will omit the iterative design’s group aspect while focusing primarily on the human-machine aspects of interaction - as if it were an individual rather than a team. The second advantage of this approach is it allows the genuinely new to emerge in comparison to what was previously impossible in the artistic process. In examining the fundamental processes of iterative design, we see there are three essential steps: a thought (designing), an implementation (prototyping), and finally: a test phase in which the change is analyzed and its effectiveness checked (evaluating). What is it that distinguishes this process from the intuitive painter making a brushstroke while quickly glancing to check if the result meets his expectations?
It’s that the artist no longer makes the decisive decision on his brushstroke success – it’s the public who evaluates the result.
In the spirit of Wittgenstein's remark, ‘The meaning of a word is its use,’ now the focus shifts from the object to its attributed meaning - from the producer to the audience. This shift is possible because the audience is now actively using the object - like the computer gamer who enters the picture directly instead of merely marveling at a futuristic vision of the future. Because of this shift in emphasis, it's not the author's intention that's decisive, but whether his addition (the brushstroke) makes the product more acceptable. But how can this be measured? By giving the object a sensorium that records the form of use. If the data shows the desired effect hasn’t been achieved, the creation process is undone and replaced by an alternative. This is the demanding part of undoing the work process, and replacing it with a better solution changes the character of the work. This means all work steps must be encapsulated and modularized to comply with such an undo request at any time. This modularization aspect has a long history in pattern design4 . While pattern designs were initially intended to facilitate the construction process, it's now a question of deconstructing a work process down into its details at will – so it can be reassembled in an improved way. We could say that doubt is built into the object by introducing the undo, while conversely, the scientific view enters the construction process. The object exists with reservations – which, of course, changes the object’s status (which always remains in a provisional state) and the author's status. Unlike the artist-engineer, whose creative act ultimately remains a mystery (a kind of instinct becoming increasingly refined over an artist's life), the undo-desideratum requires complete transparency and thorough documentation of the creative process.
To put it pointedly, techne is no longer limited to the power of creation but includes the possibility of revoking the act of creation.
Understanding engineering this way now involves the process of reverse engineering, that of construction and deconstruction, which are now married. What appears to be a coincidentia oppositorum is the programming object's prototypical counterpart. If it can be loaded into the memory with the help of a constructor, it is removed from it with the help of a destructor. The object always corresponds, if you like, to its complete, traceless disappearance.
If we take it psychologically, we have to add the human factor's fraudulence to the object's fraudulence; in other words, not only the material but also how it is processed represents a first-rate source of error.
When Marx says that work objectified in the object coagulates into dead work, he is formulating the well-known economic law of inertia, which confronts us in the form of plastic residues, mountains of garbage, and so on. Conversely, the undone logic insists on the possibility of immersing dead work in a fountain of youth. If the classical product (as dead capital and dead labor) is a lethal form, the object endowed with such a promise of reversibility promises a longer life; the labor laid down in the object can be replaced at any time by a better solution, an upgrade. Whether and to what extent the upgrade proves worthwhile depends on whether it does what it promises to do. If you must choose between modules A or B for an upgrade, part of the decisive criteria is if the module in question is easier to replace. Consequently, the interchangeability and light weight of a component are quality criteria. But how can an object consisting of dozens, hundreds, or thousands of such interchangeable modules be considered? Can we still talk about object durability here? Aren't we looking at something that regenerates completely over time, like Theseus’ ship? While the shell may remain the same for practical reasons, the inner life undergoes a constant renewal process; indeed, we could even speak of an encapsulated order of thought. Because of this liquefaction, it's questionable whether terms such as thing, product, or the like are appropriate - or whether these new things should be seen as ephemerides, transitory objects, or transitional objects. What does the artist say? Work in progress ...
The artificiality of the world
Where design patterns become software modules that are downloadable at will in the internet world, a second factor is added: a module's mass and globalized use increases the probability that any design errors (bugs) can be detected and rectified. This, in turn, increases the quality resilience of the objects using it. If the arbitrary copyability of a digital object raises doubts about the meaning of terms such as original or unique, the modularization aspect underlines the tendency toward socialization accompanying digitization. Individual programmers will stock up on open-source and freely usable program libraries and modules.
Unlike the original artistic genius, who creates his own world, the programmer is a dividual per se: a social being dependent on the logic of sharing a common ownership.
Here, the frame of reference is no longer extended art but an open-source world where we no longer draw from anything but a pool of freely floating usable building blocks. This form of incorporation needn't be seen as an argument against the process of artistic creation, as it can just as easily be celebrated as its liberation. The possibility of integrating the stored work of others into your own work considerably increases the options for creativity. Freed from having to ‘reinvent the wheel,’ the creator can now precisely concentrate on the developmental step, which releases something genuinely new into the world. This addition (like Duchamp's famous ready-made mark) is a specific iteration that transforms a universally usable commodity (the urinal in Duchamp's case) into something else. In a sense, the medieval notion that every contemporary is a dwarf standing on the shoulders of a giant becomes more relevant. Because what appears as an individual gesture in the iterative design process is based on a logic of socialization. Thus, an individual gesture is an attempt to add a new object or use to the encyclopedia of usability. If successful and a social custom is modified, then it always goes hand in hand with increasing intelligence. Just as in the case of computer chips, where a given quantity of silicon has an ever greater information density, the intellectuality of objects increases, and hardware is increasingly replaced by software – leading to a strange paradigm shift. Although the classic craftsman-artist belongs to a declining era, cultural artifacts remain exempt from this loss of significance. On the contrary, a massive increase in artificiality and spirituality is discernible here. Thus, iterative design could lead to an art without artists, to a form of society that is transformed into a work of art.
From posterity
You could say every work that's been digitized has entered a museum of work. Because such a work is retrievable at any time (without having acquired a patina or undergone any other aging processes), it has something timeless about it. In this sense, a world-tested work module is always also a thing for posterity. Conversely, because time is suspended, a creator who uses an object from the thesaurus isn't just communicating with contemporaries but also with the spirits of the long dead. This prospect may seem strange, but as a social process, it's unavoidable. If you want to experience a glimpse of this future, all you need to do is look around YouTube, where the talks and lectures of the deceased are juxtaposed alongside those of their contemporaries. This kind of do-it-again means entering a state of historicity inaccurately described as post-histoire. As stored work, the thesaurus certainly extends into the present - not just as a dead symbol (or as a symbol believed to have been overcome by history), but as an active present that can be reanimated. In a sense, history and the present are married in a way that's difficult to grasp. Processes that are digitally monitored in the present are inscribed in a chronicle of current events, while conversely, a historical process is transferred into an omnipresence as a successful solution - or as the beautiful film title announces: Our corpses are still alive5. Because every present, like a radiant reactor, points far beyond itself, we must deal with the past very carefully in the context of iterative design. If reality is what works, keeping an eye on the forces at work is essential - making the archaeology of knowledge a vital necessity.
Intersubject - Work in progress
But what does all this mean for the design process itself? To summarize, first, we can note the designer's de-centering, then the reversibility of the design process, and finally, that an object can be a composite of different time layers. The designer who enters such an environment obviously assumes a new structural position. Although many things may reveal a close affinity to previously practiced artistic techniques, what distinguishes iterative design is that the work isn't generally finished but is a work to be understood as a) a work in progress and b) to be continued by others.
For this reason, the designer isn't necessarily compelled to behave as a self-sufficient, autonomous subject but to look at himself with an outsider's eye.
This skeptical view assumes that the designer's actions are reversible and need improvement, transferring the artistic process into a post-narcissistic stage. While this dictum may be irritating, it's a direct experience of the strangeness of dealing with working memory itself. For insofar as the programmer has already reached the limits of concentrating on a manageable number of variables (Miller's Magic Number of Seven), it doesn't take much imagining that a program, ultimately based on hundreds, even thousands of variables, will grow into an impenetrable strangeness – indeed, at some point, it will give the impression of a hopeless mental confusion. If the programmer succeeds in moving effortlessly through the immediate work's labyrinth, they will be struck with the irritating realization when looking at the code months later that large parts seem alien and puzzling – commonly called ‘spaghetti code.’ Here, Günter Anders' formula of Promethean shame (‘I did it, but it wasn't me’) hits home with full force. Because in the absence of an external view (the common thread of the documentation, the clarity of the modular design), intuition runs the risk of getting lost in its self-created labyrinth - an experience that can also be understood as a form of Alzheimer’s disease during a person's lifetime.
To prevent such a disaster, the creator must design code so that anyone else can understand it, reproduce it, and, if necessary, modify or thoroughly revise it. This means that the code must be modular and self-descriptive. Every line and every thought step must be documented, reproducible, and interchangeable. Once again, there is a simultaneity of construction and deconstruction - only this time, it relates to the object and the creator's perspective. Paradoxically formulated, artistic genius now consists of making the artist superfluous. How little this is an exaggeration becomes apparent when we look at the self-disciplining techniques that have spread in the field of programming: exemplary is when the programmer subjects their syntax to strict notation rules, quality controls, and documentation requirements.
In this sense, the program monitoring the input of the code becomes the actual sovereign, the maitre absolu, who ensures the individual's contribution doesn't get in the way of socialization logic.
Describing the designer's position here, we'd use the concept of the intersubject: a subjectivity that experiences itself from the outset as individual but at the same time as interchangeable, as an indispensable yet arbitrary link in a web of communication. I, insofar as I am the-one-in-the-other.
According to the authority
This begs the question: is such a radically changed concept of the subject compatible with the old notions of society? Looking at the resonance between iterative design processes and business life, we can see evidence of hierarchical processes and planning sovereignty rupturing across the board, resulting in many voices fearing anarchy, chaos, and a lack of planning. There's no doubt that a team subordinating questions of hierarchy to teamwork operates differently than our customary business-as-usual approach - making conflicts of authority inevitable. On the other hand, whenever the iterative design approach is successful, unimaginable proportions of productivity gains become apparent. That an exceptionally well-made product scales immeasurably doesn't have to be a result of the contribution itself - it can also be ascribed to every digitized object being global and scalable at will (something which is putting an end to regimenting the factory and mass production). In this respect, the paradigm shift is attacking the very foundations of society.
What is work, how is the subject constituted, what authority ultimately consists of - all of this becomes questionable.
If we want to resolve this paradox and trace the concept of authority, we see that authority isn't derived from authorship (auctoritas) but goes back to the augmentare, which is the increase and promotion that we add: for the common good.
Seen in this way, iterative design marks a new form of authority that's both mindless and intersubjective. The position of power isn't derived from a claim to power any more than it can be associated with a specific person or group. Authority is based on the thing itself, which finds general acclamation as res publica or social sculpture. In this sense, another Wittgenstein quote could also be applied to authority: ‘The meaning of authority is its use’ [namely, it offers solutions that are socially acclaimed and valorized]. How does this fit in with the impression of anarchy, chaos, and the lack of a plan that iterative design processes often appear to have? Answering this question requires us to examine how an iteration cycle works closely. You develop a hypothesis with modular boundaries and add it to the program. Then, you evaluate the community’s reaction to the new feature. If the response is encouraging, the approach is pursued further; if it doesn't meet expectations, the solution is discarded, and a new approach is tried. Although this is an act of authorship, the decision as to whether to pursue the given solution is delegated to the users - a fact that's also reflected in how the ‘user story’ takes center stage in the design process (and not what the author wanted to say with it). This moment of decentralization marks a disarming of planning sovereignty. Because as the designer proceeds like a scientist whose hypothesis is either validated or falsified, they're acting in a mode of constant self-criticism and self-questioning. Now, there isn't a final authority on questions of taste or social customs. If the natural scientist relies on nature as an arbiter, the designer operates in the realm of social sculpturing. However, this in no way detracts from the results. Indeed, insights into social processes emerge - both psychological and sociological. Reality, if you like, is transformed into a laboratory. In the context of this open-air social experiment, insights that weren't foreseen in the original planning may emerge. Seen in this light, each iteration not only validates or falsifies a logic but can also open up entirely new spaces and possibilities for thought. This openness connects social sculpture with art and detaches it from the world of closed worldviews while allowing for an open future. Anything goes? Perhaps. But only what you like will stay. And because good design is only the appearance of something better: Do it again!
Translated by Hopkins Stanley and Martin Burckhardt
Burckhardt, M. – 1.12 Do it again! Zur Philosophie des iterativen Designs, in Philosophie des Designs, eds. Feige, D.M., Arnold, F. & Rautzenberg, M., De Gruyter, 2020. Volume 1 in the series Schriftenreihe des Weißenhof-Instituts zur Architektur- und Designtheorie. [Translator’s note.]
This is often and misleadingly called the ‘Agile Manifesto’ – see https://agilemanifesto.org.
Miller, G.A. – The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information, Psychological Review, 1956. [Translator’s note]
See Christopher Alexander, A Pattern Language: Towns, Buildings, Constructions, Oxford University Press, 1977. One could, however, trace the first tendencies in this direction back to the history of the Bauhaus.
Unsere Leichen leben noch [Our Corpses Are Still Alive] is a 1981 film by Rosa von Praunheim. [Translator’s note.]
