The “Knowledge as a By-Product of Artistic Practice is Still Not Research” Objection to My “Criterial Knowledge” Post

I spoke to some colleagues about my earlier post, The Criterial Knowledge Argument for Research Through Design, who are themselves experts in research through design [EDIT: the researchers in question are Jodi Forlizzi and John Zimmerman]. While these colleagues were generally sympathetic to the claim that art and design can contribute to knowledge in general and even criterial knowledge in particular, they objected that the sorts of knowledge outcomes I describe in that post can be arrived at not intentionally but rather as a by-product of artistic practice, and therefore cannot be characterized as research.

This is a very reasonable objection. In fact, their objection could be stated even more forcefully: the knowledge outcomes I describe in that post are almost always achieved as a by-product of artistic practice and therefore almost never research as such (at least historically). We don’t view the poet Virgil as a “researcher,” even if his literary depiction of Dido has informed Western people’s cultural understanding of the character and qualities of jealousy for two millennia.

Their objection made me realize that I had left certain things tacit in that post that should have been made explicit. And upon further reflection, I developed some of my ideas about these issues further than I had before. So here are some amendments to the earlier post, addressing their objection:

  1. For a design researcher doing RtD to be able to invoke the Criterial Knowledge Argument for Research through Design in presenting/publishing her own work, she would have to justifiably claim that the processes, methods, and desired inquiry outcomes of the RtD project in question were designed to contribute to criterial knowledge of a given domain, concept, phenomenon, experience, etc., and she would likewise have to demonstrate that she did indeed achieve such knowledge.
  2. If, on the other hand, she was trying to do something entirely different (e.g., in the hope of contributing to “intermediate concepts” that characterize several successful designs) and as a by-product just so happened to contribute to criterial knowledge about X, then the latter knowledge outcome would obviously exist, but it wouldn’t be research (in the same sense that Virgil wasn’t a researcher when he wrote The Aeneid, even if the latter did result in criterial knowledge of jealousy).
  3. Let us consider another example, that of a design researcher theorizing a domain/phenomenon/matter of interest, whose methodology includes a critical examination of other individuals’ research through design projects (and sadly there is not enough of this at present, at least in HCI). If this design researcher is able critically and analytically to discover and to explicate that this corpus of designs do contribute to criterial knowledge of X in such-and-such ways, then that would be research, but it would be design criticism research, and also not research through design.

Centuries of critical interpretation and analysis of works of art as well as the theorization of that critical practice has shown that art works contribute to intellectual virtues in incredibly broad and diverse ways (e.g., criterial knowledge, improvement of our capacities of perception, rendering us more empathic and less egoistic, opening our minds to new ideas and values, and much more). In parallel ways, research through design is likely to achieve many of the same intellectual benefits.

My colleagues’ objection has helped me clarify that it is important to distinguish between those knowledge outcomes intentionally “baked into” a given RtD process as central to, even the point of, the research, and those knowledge outcomes that happen externally to such a research project. I’ve identified two categories of the latter: cognitive by-products of a design practice intended for some other purpose (e.g., to produce a good design), and subsequent critical discoveries, often made by third-parties.

I view it as a happy problem that RtD has such high potential to yield “surplus” knowledge outcomes, but less salutary that our theoretical and reporting vocabulary does not yet allow this community to do them justice.


  1. lynndombrowski

    Hey Jeff, This post is lovely. I’d like to think through this idea of intentionality and research a bit.

    I don’t know if RtD and qualitative empirical work are comparable, but if they are, its important to note that in qualitative ethnographic work, researchers often change the direction of their analysis and inquiry because they often learn that something else was interesting or important during the course of their work. As a result of learning (or deciding something else was important) likely impacts what researchers ask and to whom they might be doing the asking. This seems to happen in design work, as a designer’s starting assumptions about a design space likely (and probably should) change as they conduct their work. Other forms of inductive knowledge making (e.g., such as those working in big data and machine learning) also go through a process where they have to learn a great deal as they interpret their data and think through their analysis. Why would design be any different? Now, granted, design tends to more abductive than deductive in its reasoning, perhaps thats why you are leaning towards those claims of intentionality as being a key component of design knowledge?

    Anyway, I’m not sure how key this idea of intentionality really is to whether or not something is knowledge production or not. Given that design is about making, doing, and practice, why should we downplay the types of learning and knowledge that design seems to be well suited for?

    • jeffreybardzell

      Well, these are excellent questions.

      I think if an ethnographer engages in a study around X, and this later morphs into X(1) and even after that into Y, I’d say that X, X(1), and Y are all research outcomes (so long as they are ethnographic and relevant–topically or genealogically–to the purpose and methods of the research in the first place), even if the latter two weren’t intended as such. But some ethnographic research outcomes were intended all along, and most ethnographers are very comfortable with the idea that their inquiry is likely to have emergent qualities–all of this is, so to speak, “baked into” the research, even if not initially specified up front (because they can’t be).

      And surely something like that also happens in design. The conception about the design space evolves in ways not fully anticipated. But the fact that the conception of the design space evolves is part of the plan in the first place; it is baked into design as a way of knowing and doing.

      The intentionality argument offered up by my colleagues, and which I have accepted here, is mainly meant to preclude someone from undertaking a design project and post hoc claiming that they actually did research (on the basis of having learned something in the act of doing design). Imagine if I observed some people, and I really did learn something, and only later wanted to say I had been doing ethnography all along; ethnographers would not be pleased (or they might be inclined to just call that “HCI”–I kid, I kid). Anyway, the knowledge achievement is not in doubt as such (Virgil really did inform our sense of the qualities of jealousy; my hypothetical designer really did learn something; and my hypothetical observer also learned something); only the ability to call that outcome “research” is in doubt. This is different from the intended yet emergent outcomes that you describe. Your hypothetical ethnographer’s emergent learnings are all effects of her research methodology, while what I am trying to exclude are effects of a non-research methodology later on asserted to have been research all along.

      As for your last question, what is at stake is not any desire to exclude bona fide intellectual outcomes regardless of origin; it is rather to help clarify the scope of research through design as a research methodology, to help us better teach it, recognize it when it is successful, and report on it. People in the community are asking how we can better tie RtD research claims (or contributions) to RtD methods, and this post is intended to clarify that relationship as well as a few other related but not identical relations. There is a white hot debate about whether research through design can do research at all, and a lot of claims are being made on behalf of RtD. Some defenses are stronger than others, and the weak ones hurt the whole endeavor. The claim, “I did design, and I really learned something, and therefore I did research through design” is one such weak claim. I believe that concern motivated the objection from my colleagues.

  2. buridan

    I wonder what happens when you let go of the myth of intention and intentionality… and replace it with either ‘claim toward future outcomes’ or desires. Because I think the work that intention is doing above is really misleading and is constructing some faux rationality or planning which may or may not be present, and is providing, i think, a place for people to feel good that something is ‘research’ but I’ve done research, as have you. I will say that it doesn’t necessarily have anything like intentions to find what it finds, and it is generally messy in a good way if it is done honestly and openly.

    • jeffreybardzell

      Thanks for this comment. It’s interesting to me the pushback I am getting around the use of the term “intentionality.” I recognize that that is a thorny issue, and clearly I will have to invoke it with more care in my published writings.

      Now, by way of response:

      My Ph.D. training in Comparative Literature made me sensitive to issues of intentionality, and for much of my early career I also accepted the modernist (New Criticism) and postmodernist (Barthes-Foucault-Derrida) stance (I won’t use the word “myth here, but I might) of anti-intentionalism: the call to focus on “the text itself,” or the political suspicion that a focus on authorial intention amounts to a celebration of the status quo (socioeconomic, gender, colonial, etc.). I am not a fan of intentionalism in the E.D. Hirsch sense or anything like it (I know you know this Jeremy, but for other readers: Hirsch developed a theory of “validity in interpretation” based around the idea that the reader’s job is to try to understand what the author intended). Indeed, in HCI I have been accused of being too cavalier in my design criticism by not taking sufficient consideration of designers’ intentions for their work. I have even blogged on the “intentional fallacy” applied to interaction design here:

      My point is that I do not use the term “intention” lightly nor in the naive sense that we learn in high school literature classes.

      All of that said, my more recent readings in the Anglo-American philosophical tradition have made me see the intentionality question in a more nuanced light, which does not simply reject and replace the modernist-pomo anti-intentionalism, but it does complicate it. The Anglo-American tradition argues that a certain amount of intentionality is legible in any act of communication and that communication breaks down without it. We can be anti-intentionalists about Dickens if we bracket aside his politics or worldview or whatever, if we play down moral, political, and aesthetic styles and sensibilities we can see across his works (though I wouldn’t recommend doing so), but if we reject his intention to present us with novelistic fiction, with characters who represent different forms of life, or even his intention that someone reads any of his writing at all, we end up in a very weird place.

      Anyway, as I replied to Lynn who raised a similar issue just above, the problem I am trying to address is a claim like this one:

      I did a design project A. Along the way, and by means of it, I learned something B. Therefore, B is an outcome of research through design.

      To me, that is a very different sort of claim than the following:

      I am an ethnographer. Originally, my inquiry was focused on A. Once I began to engage in my ethnographic practice, and through reflection, I came to realize that I needed to shift my focus to B. B is an outcome of my ethnography.


      I did a research through design project. Originally, my inquiry was focused on A. Once I began to engage in problem framing and materials exploration, I came to realize that I needed to shift my focus to B. B is an outcome of research through design.

      In the former case, the mere fact that someone learned something along the way is post-hoc asserted to have been research. In the latter cases, a researcher through her research practice emergently adjusted her inquiry objectives, methods, whatever–but she was still doing inquiry, still acting as an ethnographer or research through designer, and still operating within the language games of her research community (insofar as she understands them).

      But one reason I blog this before I put it in papers is to receive this kind of pushback (so I thank you and Lynn). So let me turn this back to you:

      I assume you agree with my claim that just because Virgil made a huge knowledge contribution to western culture through his depiction of jealousy embodied in the character of Dido, you wouldn’t call him a researcher. And I assume you would broadly agree with the reasons why I claim that: surely Virgil intended to write a poem, to create a national epic, etc. And just as surely, Virgil did not intend to make a durable contribution to western culture’s criterial understanding of jealousy for millennia to come. That is the extent to which I am using the concept of “intentionality” in so far as it applies to research.

      (I guess the interesting question here is whether one can do research by accident, which is not the same as asking whether, in doing research, one can learn something unexpected.)

      If you (or any other reader) wants to propose an alternative way of distinguishing between my two cases above that doesn’t appeal to intentionality, or if you want to argue that the distinction I am making above is wrongheaded to begin with, I’m all ears.

      • buridan

        Pasted from Facebook.
        Jeremy Hunsinger: Yes I saw. I was responding and then my response was eaten. Mostly you seem to be using a term for a mental state to distinguish between what is and is not ‘science/research’ which there is a term for the debate around that which I’m not remembering right now.
        Like · 1 · 12 mins

        Jeffrey Bardzell: Ah, there is a useful touchpoint. I don’t think of “intention” as merely a mental state, but something manifested and interpretable. That is what I need to state more clearly going forward. Otherwise we get caught up in the private minds debate.
        Like · 9 mins · Edited

        Jeremy Hunsinger Well mental states are manifested and interpretable, i have no problem with that, i just think that from my point of view that ‘intention’ is a particularly problematic concept given its genealogy. But really i think we are talking about what we use to distinguish two things which is the accidental discovery of knowledge and the purposive discovery of knowledge and the relationship between those two and ‘design research’ or ‘design process’. I strongly believe in accidental discovery of knowledge as a product or more usually a byproduct of research, but I’m wondering if you see for instance the design knowledge described as ‘tacit knowledge’ in the book leviathan and the airpump as purposive? it was knowledge that you needed, but could only be gained in relation to the objects, but even when you’ve built the objects you didn’t necessarily have the knowledge, though some did.
        Like · 2 mins

        Jeffrey Bardzell: Could you copy and paste this over to the blog? I want to keep this conversation and FB will swallow it up…

  3. buridan

    Jeffrey Bardzell Short version: I love accidentally discovered knowledge and do not want to diminish its legitimacy in any way. The point here has to do with (I like your term) “purposeful” inquiry. You can have a great insight about a way to frame a design space accidentally, but you cannot do ethnography or research through design accidentally.

    Jeremy Hunsinger I don’t actually think you can do ethnography through design, but you can do design through ethnography. it is sort of like you can do phenomenological ethnography but you cannot do ethnographic phenomenology. you can get knowledge through research through design accidentally. however, i think you are after what makes the design process require a plan that thus would require an interpretable intention, and could that plan be divorced from the outcome ‘knowledge’ in a way that would also divorce it from the term ‘research’

    • jeffreybardzell

      Quick clarification: I wasn’t talking about ethnography through design or vice-versa. All I meant was that you can’t do ethnography accidentally, and you cannot do research through design accidentally. It is possible to discover by accident the same sorts of knowledge outcomes one might expect to get from ethnography or research through design, but even so, that fact doesn’t mean that one has done ethnography or research through design.

      • buridan

        is there other knowledge that you get from ethnography which is outside of the scope of the ethnography? i think there is. here i think in terms of all the curricula that we have for learning, hidden curricula, etc. even if there is a primary and states goal, there are also many other things going on from which one learns.

      • jeffreybardzell

        Yes, of course there is knowledge you can get from ethnography that is out of the scope of ethnography. And there is knowledge that is akin to what one might find through ethnography that one might achieve by non-ethnographic means.

        I am not proposing the ridiculous position that knowledge is only ever acquired through academically sanctioned research methodologies, and that any such knowledge is scoped and bounded by whatever that academic sanctioning says is appropriate to that methodology. (Or if I am offering such a ridiculous position, it is an accidental unforeseen consequence of something I’ve said, and I don’t yet perceive that.)

        Again, all that I’m trying to rule out is a post hoc claim that one did something one didn’t: I learned by doing design, therefore I did research through design. I learned something by observing people, therefore I did ethnography.

        The problem isn’t tacit knowledge, emergent knowledge, or academically unsanctioned knowledge. The problem is the post-hoc claim that one did something one didn’t as a way of claiming legitimacy for that knowledge by claiming it was an outcome of a process when it was not. (And I am not randomly picking this target: it is an issue in current RtD debates.)

    • jeffreybardzell

      Anyway, I think the language of “purposeful inquiry” and “manifest intention” might be what I need to get the nuances (about) right. Thank you!!

  4. John Zimmerman

    My re-reading of the post leads me to two interrelated questions.

    First, does having the intention to produce knowledge as an RtD researcher change the type of knowledge that is produced when compared to knowledge that comes from artistic or design practice? Does the researcher’s intention actually have any influence? I would like to think it does, but I have not read anyone talking about this or seen any analysis of similar artifacts that emerged from practice and from RtD.

    Second, I think it is important to note that in many of the early cases proceed RtD, the intention was not so clearly linked to generating knowledge through the constructive activity. Here I am thinking about the development of Participatory Design. The creators of this method wanted to address two problems. 1) They wanted to reduce the risk of making technology that people failed to adopt. 2) They wanted to create new technology that protected the rights of workers and preserved their craft skills. The made things in order to develop an improved way of making things, and overtime, they began to view the things themselves as a research contribution. So maybe RtD as a research approach is in fact an unanticipated outcome of constructive research.

    • jeffreybardzell

      These are really interesting questions!

      Response to first question: As prior comments have already made clear, my use of “intentionality” needs clarification. What I was trying to get at was a sense of research purposefulness. What I was NOT trying to get at was any exclusion of or devaluation of emergent or serendipitous knowledge outcomes, whatever their provenance. Again, the case I want to exclude is: “Virgil’s depiction of Dido informed the West’s criterial understanding of jealousy; therefore, Virgil was a psychology researcher.” And it is relevant for us in HCI and design because of the prevalence of what I see as a parallel claim, “I am a design practitioner who learned something in my practice; therefore, I did research through design.” That claim undermines the research community’s ability to establish RtD as a research practice. Or at least I think it does, and so apparently do you and Jodi, or you wouldn’t have raised your objection/clarification request in the first place.

      You ask: “does having the intention to produce knowledge as an RtD researcher change the type of knowledge that is produced when compared to knowledge that comes from artistic or design practice?

      There’s two cases here. In the first, RtD is distinguished from artistic practice. But it is hard to take on this question without making ontological commitments about the line between art and design. On some level, most of us intuit a difference, but we could also name designs that would give us fits in terms of that difference. So my non-response here will simply be that this question cannot be answered at this level of abstraction. But if we choose a design or artwork, and a reason for asking questions like these, then at that level we could probably get some traction.

      In the second, RtD is distinguished from design practice. I would argue that the latter is not in itself research, whereas the former is. That said, a design produced by design practice can be taken up by a separate act of inquiry (e.g., a design critique or an empirical study of the design’s use). In the latter case, the design would yield research knowledge, but not because it was made as a result of RtD, but rather because it was subjected to a different research activity (e.g., criticism or user study).

      You then ask: “Does the researcher’s intention actually have any influence?” I do think that in obvious ways (perhaps so obvious that we forget them) the researcher’s intentions do have influence. If one believes she is a researcher, and believes she is doing research, then it is quite clear those facts will (or at least should) have influence: I can’t think of any research in any discipline that does not posit an important relationship between process/methods and product/outcomes (though what that relationship is varies greatly by discipline). That sensibility is “baked into” any act that someone pursues as research, and any reporting of the outcomes (to researchers) inevitably includes an account of the process/methods as well. So how could one intend to pursue research while not having that fact influence their process (and therefore their product, and therefore their research)?

      Response to second question: The PD example is an interesting one. I’ll think on it more, but since this is a blog comment I’ll feel free to stick my neck out and speculate a bit. Your example seems to suggest that during the genesis of a new methodology, when the links between processes and products are still fluid or not well understood, it is possible to have research outcomes that are more emergent/serendipitous than the emergent insights a researcher using a more mature/stable methodology (such as ethnography) has. One of these is the construction of the methodology itself. I can buy that.

      In all of this, two of my own tacit assumptions have been revealed for me:

      • It is possible to distinguish between design as a practice and research through design; that is, one can do the former without doing the latter (and vice-versa?)
      • It is not possible to do research (properly understood) by accident, unintentionally, and/or lacking research purpose, because doing research entails a commitment to linking inquiry process/methods and products/outcomes (although of course it is possible to have unanticipated, emergent, and/or serendipitous knowledge outcomes as a result of either research or non-research processes).

      I’m not sure, now stated so plainly, that I am ready to commit to either of these (other than the fact that I already have…).


Leave a Reply to jeffreybardzell Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s