CHI 3 Towards a theory of Interaction design a critique of the workshop at CHI2011

Towards a theory of Interaction design a critique of the workshop at CHI2011




I had the pleasure of attending the special interest group at CHI2011 entitled UX Research: What Theoretical Roots Do We Build On – If Any? The objective of the session was to attempt to investigate to see if there might be a common definition of and understanding of UX theory.


This is not the first time someone has attempted to do this. As just one example Paul Dourish in his book Where the Action is: The Foundations of Embodied Interaction suggested that to find an area of common ground it is necessary to abandon disembodied rationality and focus on skilled engaged practice using phenomenological practices.
The move to find some common theoretical ground is a positive one but during the workshop I witnessed a number of conflicts which I believe where unnecessary and counter productive.
Many of these conflicts arise from confusion over the meaning of theory from the differing component disciplines. By understanding these misinterpretations over the role of theory I believe it is possible to eliminate a lot of the problems and move more rapidly toward some kind of fundamental interaction theory.


What is theory ?
Theory is a term from ancient greek derived from theoria meaning a looking at view or beholding. It refers to contemplation as opposed to praxis (action or practice). This description still assumes that all academic disciplines use the notion of theory in the same way, which I contend is not true. Being on the boarder between the arts and the sciences HCI has come across a collision in paradigms between the two areas of practice. The domains of the arts and sciences have different notions over the role of theory and as such have different theories.


Science theory


Science theory is a well studied area but basically has a number of components. Firstly and to refute one misconception commonly used during the workshop there is no ONE theory or grand unified theory of science. Each domain has different theories depending upon context, so chemistry has the periodic table of elements while physics has theories such as electromagnetism, biology has evolution theory.
Confusingly physics has experienced intertheoretic reduction – were two theories such as the theory of electricity and the theory of magnetism have been combined to form one theory of eletro-magnetism. These intertheoretic identities are how ever rare and important steps in the progress of physics as an endevor. intertheoretic identities might give the impression to the outsider that there is one theory in physics but infact there are still a multiplicity of theories.
What makes science theory different is that if there are two theories which explain the same phenomena then it is generally possible to perform an experiment which appear to support one theory over another. This happened for example when the medical theory of the four humors (blood,yellow bile,phlegem and black bile) succumbed to more complex theories including bacterial infection theory. Scientific theories can be conceived of as compact descriptions of the phenomena they describe. As such well-collected empirical evidence is an important material along with experiments which appear to support or refute one theory over another. Finally generally scientific theories are also predictive. The limits are known to a strong extend and the significant parameters are known well enough that predictions can be made and ultimately tested. Scientific theories also tend to be reductive, they describe one phenomena really well but don’t seek to explain everything. Thus Galileo explained the presence of gravity but did so assuming the absence of air resistance. From an outsiders perspective he didn’t explain everything that falls ( the theory of gravity cannot explain the fall of a feather) but he did explain one aspect of movement which he considered widely important.


In terms of interaction and design scientific theories don’t tell you what to do just permit the inferring of consequences of doing so. An architectural analogy might help. Structural theories used by engineers permit them to predict if a design might stand up. Equally it is possible to say if you want a platform to stick out this far, from this material then it must be this thick. If you went to a structural engineer and said ‘I want a museum please calculate what should I build’ they wouldn’t have any idea. Scientific theories are descriptive but not generative. This can be quite liberating, for architects this means they are free to design anything they feel expresses the idea they want. With out this structural theory they would be forced to rely on precedence – previous successful buildings – something common in pre- structural theoretic cultures such as the medieval cathedrals of Europe. From this we can conclude that science theories are flexible and liberating for design.


Art theories


The second group of theories that are relevant are art theories. Art theories or design theories don’t have the property of being exclusive. Consider Cubism vs Dadaism it doesn’t make sense to consider one to be better than another,. they are not mutually exclusive and so intertheoretic reduction doesn’t apply. From this perspective art theories tend to be more prescriptive – they give a domain of consideration and a background of rhetoric to launch from. Thus when you begin a design there is already a condition to begin from, this is importantwhen a designer could be faced with creating something from nothing. Art theories gives them the property of being broad and cohesive, you can have a modernist painting, a modernist building, a modernist chair. Consider the process of creating a website and mobile version of a product ( not to mention physical documentation) scientific theories ( narrow and deep) will not help create a framework for action at this point but art theories do. Art theories often exist as a response to current conditions and can prefigure the works that compose them. As such art theories are very powerful generative tools in the hands of a skilled designer.


Summary
I have described two types of theory –art and science, but there are others, Mathematics, History, Politics and Philosophy all of whom use the notion of theory differently. It is important not to use scientific views of theory as bench marks for artistic theory, as it is not to use art theories to understand scientific ones. This said the diversity of theories creates for a rich structure and description of evaluation of user experience. If user experience is to thrive it needs to embrace the multicultural inheritance by admitting it’s presence and stop false comparison. This is not to suggest that there should and could not be a group of theories which lay purely in the domain of interaction design, just that they will have to be constructed between the two primary regions.