Recently interaction-design.org sent me a preview of an upcoming encyclopedia chapter on aesthetic interaction, entitled “Visual Aesthetics in HCI and Interaction Design,” by Noam Tractinsky. I read through his piece and the existing commentaries that were online at the time. I couldn’t help but notice that Tractinsky himself and all of the commenters all represented a cognitive perspective on aesthetic interaction, that they described and championed this perspective, and that they collectively failed to acknowledge the existence of any other perspective on aesthetic interaction, even though some of that work has been extremely influential in the field (e.g., the work of Bill Gaver, Phoebe Sengers, Anthony Dunne, Fiona Raby, Mark Blythe, Jonas Löwgren, John McCarthy, Peter Wright, Jay David Bolter, and many others). I wrote to Mads Soegaard to point this out, and he invited me to write a commentary myself.

I took him up on his offer.

Because HCI is an interdisciplinary field, and because I am committed to the view that discipline-based spats are counterproductive, I have sought in my work thus far to represent my own position honestly and faithfully but also to do so in a way that respectfully situates it within broader traditions, some of which I am not a part of but whose legitimacy I accept and acknowledge. In other words, I generally try to avoid taking a confrontational stance.

I somewhat departed from such decorum in my commentary, and here’s why. Tractinsky (and his fellow commenters) not only failed to acknowledge the contributions of HCI researchers outside their perspective, but worse in my view, Tractinsky seemed to eliminate the possibility that such work can be valuable, by staking out (what I read as) an unabashedly positivist stance with the following core characteristics: a categorical rejection of academic theory; an exclusive embrace of empirical scholarship, methods, and findings; and an insistence that such work is “descriptive” and not “normative.” Work not embracing these values is positioned as un-scientific, “confusing,” and unnecessary. He also contrasted his own views with rival aesthetic positions by claiming for himself everyday legitimacy, asserting that ordinary people hold more or less the same views that he does, that these views are adequately reflected in standard dictionary definitions, and therefore we can simply dispense with the whole lot of aesthetic philosophy and theory and replace it with 7-word dictionary definitions. In short, in my reading, Tractinsky had taken a disengaged and dismissive, and yet aggressively territorializing position by claiming that his position represented the whole of visual aesthetics in HCI. I felt compelled to respond.

My commentary therefore seeks to achieve the following: to demonstrate that Tractinsky himself embraces a philosophical position with normative commitments, that this position is neither reflected in dictionaries nor intuitively available to ordinary people, and that these facts render his dismissal of other such positions invalid and disingenuous; I then take a step back and propose 3 normative positions about why HCI and interaction design would want a more rigorous understanding of aesthetic interaction in the first place; I then critique Tractinsky’s position vis-a-vis those three normative positions; I then introduce what I pointedly refer to as “aesthetics, according to the rest of the world,” to stress both that Tractinsky’s position is a marginal, not mainstream one, and also to open up alternative resources that our field can tap into (are more, new ideas “confusing” or “empowering”?); finally, I introduce a range of works in HCI that do tap those resources in one way or another and have made their mark on HCI, hopefully reflecting these achievements in the way that (in my view) Tractinsky should have done in the first place.

Some of my argumentation parallels Boehner et al’s critique of affective computing for relying on the information processing model, a critique that Kia Höök explores alongside rival theories in her own interaction-design.org chapter on affective computing, and one that is itself counter-critiqued in Picard’s commentary on the same. However, my critical tactics are different. Rather than surveying a range of literature and drawing forth a critical synthesis of it (as Boehner et al. do), I instead offer a close reading of Tractinsky’s chapter itself, quoting him at length in places and demonstrating through more quotations some of the diverse and important ways that he contradicts himself. I also explore the consequences of such contradictions both for his own theory in itself and more broadly for aesthetics in HCI as a whole.

In this, my approach is also closer to that of analytic aesthetics, which carefully investigates the specific claims made by those propounding aesthetic dogmas, to explore the extent to which such claims make logical sense and to assess how well they apply to diverse examples that they assert their own ability to explain (for example, a theory of painting should be able to explain not just the paintings used to express the theory but other paintings to which readers might want to apply the theory). I raise this methodological point to remind HCI researchers and practitioners that if we are to import aesthetic concepts (as Tractinsky does, e.g., his extensive use of “expressive” vs. “classical” aesthetic modes), we should also import the reasoning and evaluative strategies that accompany them. Aesthetic philosophers know full well what the strengths and limitations are of aesthetic vocabulary such as “expressive” and “classical,” and it diminishes their value if we import them without this knowledge and related intellectual approaches.