I have noticed a dark pattern among papers that align themselves with critical or humanistic approaches to HCI. I myself have been guilty of contributing to that pattern (though I am trying to reform). But I still see it all the time as a peer reviewer and also as a Ph.D. supervisor.
And since I spend so much time evangelizing humanistic HCI, I thought it might also be good to point out one of its dark patterns, to encourage critical/humanist HCIers not to do it, and to encourage reviewers to call this out and use it as an argument against accepting the paper.
And of course I want to offer a positive way forward instead.
The dark pattern is:
“I love a critical theory/author; you in HCI should change your practice to use it, too.”
Characteristic features of this dark pattern include the following:
- An assertion of how naïve HCI is for not also having known this theory all along.
- No acknowledgement of nor engagement with the thousands of critical/humanistic papers in HCI.
- No articulation of a research problem or question that HCI is already asking or expressing within a given HCI domain (e.g., personal informatics, HCI4D, user experience).
- In other words, the research problem is expressed something like this: “I can’t believe you ignoramuses don’t know what Deleuze says about the Movement Image, but oh man, if you did, your HCI would be as hep as a Robbe-Grillet novel; happily, I am here to teach Deleuze to you; I’ve got a whole page devoted to his thought.” (I might be exaggerating for effect, but seriously, sometimes this is how it sounds).
- A writing style that resembles that of a drunk Derrida trying to impress someone.
- A references list full of Heidegger, Foucault, Merleau-Ponty, and Deleuze.
- A references list that literally cites no HCI papers at all.
- A response to one’s inevitable rejection by saying that HCI is too stupid or too bigoted against the humanities to get it (tip: it is not).
Now, there is a light pattern version of this that I would argue for instead.
This version completely buys into the project of importing and evangelizing on behalf of this or that critical theory. But it seeks to do so in a dialogic, rather than imperialistic, way.
The light pattern is:
“I understand current HCI practice in domain D to be Dprax. I understand that Dprob is a known problem/challenge in Dprax. I hope to contribute to Dprax by turning to critical theory T to help understand better/clarify what is actionable about/reframe Dprob.”
Characteristic features of this light pattern include the following:
- Respect for prior HCI research as rigorous, informed, and of high scholarly quality, even if it doesn’t seem aware of or make good use of your favorite theory or its ilk.
- A focus on what that prior research itself articulates as a problem, gap, opportunity, challenge, or whatnot.
- A generously cited and clearly articulated statement of the state of the art and its known challenges/problems. Your HCI readers, which might include the people you are citing, must be able to see themselves in your characterization; they must agree, within reason, that you’ve characterized their work and their challenges in a fair way.
- A positioning of your theory as potentially helping others grapple with that challenge or problem. It won’t solve it, so don’t say it will.
- A clear introduction to your theory in a way that is focused on the HCI problem domain you are trying to address. Do not offer a general “history of philosophy” overview of the theory; this is not Philosophy 101. Point readers to the Stanford Encyclopedia of Philosophy or a Cambridge Companion To and get to the point.
- Write in a style that is accessible to HCI readers. It is OK to push stylistic boundaries and be somewhat challenging to readers–you want to be true to your critical-humanist self. But remember that you are inviting the community to take up your perspective; your writing should feel like an invitation, not a Gallic Howler.
- You probably should have three different kinds of references:
- HCI references in your domain of inquiry (personal informatics, crowdsourcing, sustainability, design fictions)
- HCI references of a critical-humanistic nature (to align your approach with an established way of doing in HCI)
- Primary and secondary references pertaining to your external theory
Humanistic approaches to HCI should be generous, dialogic, critical, and engaging. They should not be imperialistic power-moves that condescend to HCI.
I had a follow up thought, which is that I want to stress that this kind of approach is not good humanities research either.
For example, a scholar might bring a new theoretical perspective to a study of Emily Dickenson’s poetry, but she won’t talk down to Dickenson scholars in doing so. She will argue instead that her new theoretical perspective helps all readers attend to something important we were overlooking before, not that everyone until now was looking in the wrong place altogether.
Thanks, Jeff! I’ve been revisiting many of the critical design and critical HCI classics with a group of graduate students at Berkeley this semester and have seen ones in either side of this, certainly, even among that illustrious group.
This is of course no excuse for it in humanistic HCI, but perfunctory literature reviews do have a long tradition across many subdisciplines of HCI – and was in fact what I initially learned as “expected” when I was an undergraduate in Berkeley CS first getting my feet wet in HCI research. (I’ve since learned better!) Kudos to holding your students, and humanistic HCI more generally, to a higher standard. 🙂
Thanks Jeff for a good posting. In design, I see similar chauvinistic tendencies coming from a variety of disciplines, not just philosophy (though philosophers are usually the worst). I routinely kill papers that tell how design has to build on theory X from discipline Y.
Two current fads are nudging and practice theory, which both are painted as the only possible foundations for the future of design. My usual response points out that there are lots of design practices that the writers do not take into account and their design implications are unclear. More importantly, I usually tell that these theories are heavily criticized in their own disciplines, and then point out some of these criticisms.
It is strange to see such paternalism, which builds on ignorance of design.
Thanks Jeff, great post! Could not agree more.
Best read today, saved to drive for PhD supervision. Keep ’em coming!
Thank you for talking about this. 🙂