by Kevin Walker & John Fass
De-Computation: Programming the world through design by Kevin Walker
computational thinking has four steps: decomposition, pattern recognition, abstraction, and design. Simply put, this means breaking something down into manageable or meaningful components, then remaking it in another form or making something new from it, based on an analysis of the parts. De-computation provides a way of applying these categories specifically to design practice
(3)
De-computation is intended to shift focus from devices and systems to computing as a process undertaken by people, plants, places and other things. We define computation as broad set of cultural practices driven by a belief that algorithms can shape behaviour, opinions and actions, in opposition to 20th century notions of computers as monolithic, fixed devices. De-computation thus incorporates a view of computation different from its usual definition, one that encompasses non-digital life, microscopic and astronomical processes, and poetic data exploration.
De-construction - break down the problem, research question, dataset into smaller parts
Pattern recognition - ‘patterns can be highly abstract-the rhythms of human behaviour for example, pattterns of play in a game, circadian rhythms, or turn-taking in a conversation.’
Abstraction - Whether a pattern in the data or an innovative way to address the problem or research question, this can be then generalised in order to make some statement / encourage thinking of ‘abstract’ as in fine art-for example in terms of simplified shapes and colors; the less something resembles some specific thing in the world, the more it communicates across categories / abstraction becomes an expressive means for generalising in surprising domains
Construction - this is manifest as a nonlinear system, which accounts for unknown functions as variables, and in which the outputs can vary proportionally from the inputs;; the examples shown in the Figures illustrate a variety of outputs possible from a comparatively small set of inputs.
Our broader intention is to exploit the systematic ways of thinking found in the natural sciences, in ways that enable experimental outcomes which cannot be fully predicted due to randomness and variation introduced particularly through the use of physical materials and processes. De-computation thus strives for a balance between research-led practice which proceeds from data and a set design brief, and the kind of practice-led research more common in fine-art contexts, in which materials exploration leads to unexpected outcomes.
de- computation maintains a contextual and critical perspective.
it takes a broad view of personal, social and physical contexts, continually questioning the problem or task in relation to these. Here, De-computation overlaps with user experience or interaction design, with its focus on contexts of use (Kaptelinin and Bannon 2012). But it goes further, to interrogate how a design product might impact a user’s cognitive load (Sweller, et al 1994), or society at large. Might it add to or subtract from visual or noise pollution – or indeed the environment and climate change? Is the product really necessary, or alternatively, is it something that can be used to critique some aspect of contemporary society?
De-computation is thus about using the algorithms and systems of design process and method to question the influence of computer algorithms and digital systems, as well as the designer’s own assumptions.
De-construction - investigative journalism similarly operates around narrative conventions, but in an inverse direction to science, it often starts with ‘a story,’ then seeks to collect data (in the form of interviews or documentary research) to provide multiple viewpoints / clearly stating who, what, where, when, how, and especially why, acts as an effective design principle / journalistic stories can be regarded as de-constructed and re-constructed representations of events
Pattern recognition - We believe that spotting patterns in data is, however, a subjective process of interpretation, whether undertaken by humans or computers (which are programmed by humans) / one constant factor is time, which consistently flows linearly in one direction, and if two datasets can be separated by time, this aids in analysis
Abstraction - we can abstract rules from nature to apply to other types of systems
Construction - analogue materials
We thus advocate a reverse design process applying computational thinking to real-world phenomena in order to reveal and critique this,