I recently (May 2021) gave a talk at a symposium celebrating the work of Charles Eastman, one of the pioneers of the field of design cognition, as well as a pioneer in the field of computer aided architectural design. Eastman’s early career is notable for the people he was taught by and worked with. He studied at UC Berkeley where he encountered two giants of the design methods movement: Horst Rittel and Christopher Alexander. He worked at Carnegie Mellon University alongside Nobel Laureate Herbert Simon and Alan Newell who shaped the modern field of cognitive psychology. Finally, he worked at UCLA alongside two well-known design and planning theorists George Stiny and Lionel March. With these influences he was able to establish a unique identity in design research, one with deep roots in science and design and a broad canopy in research and practice.
Whilst doing my PhD in cognitive psychology in the early 1990’s, one of Eastman’s papers formed a touchstone for my own research. It is a study of designers using what was then a new research methodology called protocol analysis. For my symposium talk I wanted to return to that paper to consider again one of the foundations of a new discipline and assess its contemporary relevance. What follows is an edited text of my talk.
What is it that designers do when they solve seemingly ill-structured problems like design problems? More importantly how can we develop methodological, computational, and educational tools to support designers? Charles Eastman’s paper On the Analysis of Intuitive Design Processes [1], published in 1970 from research funded by DARPA [2], addressed the first question; arguably his subsequent career addressed the second. This talk discusses the foundational contribution of Eastman’s paper, looks at some of the problems he uncovered, and assesses the paper’s relevance for contemporary studies of design activity.
One clear thing that designers do when faced with a problem brief is to embark on a process of understanding, piecing together the fragments of information they have or can discover and forming a way to overcome the original problem. This process involves focussed acts of cognition – perceiving, retrieving, and conceiving – supported by the generation of representational artefacts – “words, numbers, flow diagrams, plans, sections, and perspectives”[3]. The resultant activity transforms the initial problem state into a goal (solution) state.
That was the theory in 1970 anyhow, when Charles Eastman published the first study of the design process to use protocol analysis as a research methodology. The new science of cognitive psychology, with Allen Newell and Herbert Simon at the helm, drew on a computational metaphor for the human mind. Information came in and transformed information flowed out.
In the terms of this metaphor some aspects of cognition had begun to be understood quite well, the features of short-term memory for example [4], though much still remained a mystery. The metaphor just about held if you looked for rationality in the right places.
On the Analysis of Intuitive Design Processes analyses the 25-minute protocol of an industrial designer with seven years of professional experience. His task was to reconfigure a bathroom layout after negative comments of an existing layout by potential house purchasers.
Each minute of the protocol is itemised and codified in the paper’s appendix. Mostly the depiction of the activity is routine enough to be displayed in a ‘Problem-Behaviour Graph’ with constraints being identified, design elements manipulated, and a number of alternative proposals produced.
There are four key points where the attribution of activity is glossed as ‘an association with deep structure memory’. These are value-based decisions and what we might consider as the crux of intuitive decision making. But if this is where ‘intuition’ takes place, then the ‘intuitive’ of intuitive design might seem beyond the grasp of protocol analysis. In this sense On the Analysis of Intuitive Design Processes can be read as a paradox and a challenge to the metaphor of information processing applied to human behaviour and the method of protocol analysis that derives from it. One of the paper’s two stated findings is that drawing on past experience is a superior designing strategy than relying on external cues. But what is ‘past experience’ in this context?
Protocol analysis – the articulation of the contents of short-term memory as well as other processing information – had been successfully used on highly constrained problems, where the state of the problem is objective and finite and it is relatively clear where the focus of attention is. For problems where the constraints are much more loosely connected, subject to interpretation and sometimes even rejection, and where the thoughts of the designer relate to past experiences and feelings, short-term memory only provides part of the story. Eastman puts it down in numbers at the end of his paper:
“currently the density of concept utilization which these studies represent is about one every twelve seconds. Protocol studies in other areas show that processing of familiar information may take place at a rate greater than one concept per second. Studies of design have a long way to go to fill in what’s happening in that eleven-seconds gap.” [5]
Has the eleven second gap been filled in the intervening years?
To some degree. The more naturalistic and social elements of designing have characterised the activity more effectively. Less a feat of mind and memory, more of a creative dialogue with others. Donald Schön in particular showed how past experience and intuition might be analysed [6], the Design Thinking Research Symposium (DTRS) series [7] has provided many insights into the nature of design thinking, and neuroimaging and machine learning are finally beginning to correlate brain activity with design activity [8].
To the present day, protocol analysis is still used as a key research methodology in many studies of design, though very few achieve the depth, cogency, elegance and insight of Eastman’s original study. The language of design cognition that Eastman proposed is still prevalent. The idea of an evolving ‘problem and solution space’, for example [9], or simply representing the links between information states in something like a problem-behaviour graph [10].
We might not have filled the 11 second gap, but we now have a rich technical language to describe design activity that can be traced all the way back to Eastman’s original research.
References
[1] Eastman, C.E. (1970) On the analysis of intuitive design processes, in G.T. Moore (ed.) Emerging Methods in Environmental Design and Planning, pp21-37.
[2] DARPA the US Defense Advanced Research Projects Agency. It is interesting, though obvious when considered, that defence would have an interest in cognitive psychology. Working out the capabilities and limits of human cognition has obvious relevance to theatres of war where both design and split-second decision-making are necessary. Two other well-known psychologists, Abel Tversky and Daniel Kahneman, were closely involved with the Israeli military both for research and in service. They drew on Herbert Simon’s theory of bounded rationality in building prospect theory, the foundation of behavioural economics and ‘nudging’. Their relationship is explored in Michael Lewis’ excellent book The Undoing Project.
[3] Eastman, C.E. (1970) op cit, p.21
[4] Miller, G. (1956) The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological review, 63(2), pp 81-97.
[5] Eastman, C.E. (1970) op cit, p.21
[6] Schön, D. (1988) Designing: Rules, types and worlds, Design Studies, 9, pp 181-190.
[7] Cross, N. (2018) A brief history of the Design Thinking Research Symposium series, Design Studies, 57, pp. 160-164.
[8] Goucher-Lambert, K., Moss, J., Cagan, J. (2019) A neuroimaging investigation of design ideation with and without inspirational stimuli—understanding the meaning of near and far stimuli, Design Studies, 60, pp 1-38.
[9] Dorst, K., Cross, N. (2001) Creativity in the design process: co-evolution of problem–solution, Design Studies, 22, pp 425-437.
[10] Goldschmidt, G. (2014) Linkography: unfolding the design process. MIT Press.