Speculative and critical design: provocation as a knowledge claim

Speculative design does not ship products. It ships arguments about what the future could be. Dunne and Raby (2013) state the ambition plainly. Design can pose what-if questions that open debate about the futures people want and do not want, and the objects are not offered for use. That is a knowledge claim, and it can be restated as a testable one. If a speculation opens debate, someone is reached who was not reached before, and something in what they think or decide changes.

This unit, the eighth of the ten, holds the field to that formulation. It argues that the claim to open debate has been treated as self-evident. The previous unit named the values a design embeds. This one takes a practice that refuses to be used. The field's own critics have pressed on who is actually reached. The recent review literature has begun to replace provocation with criteria, though those criteria remain untested and the audience unmeasured.

The claim about debate can be tested as an audience and a change

Dunne and Raby (2013) are the primary statement. Their book argues that societies, and the industries that furnish them, are losing the capacity to envision futures in order to evaluate their desirability. It argues that design should physicalise speculation beyond the prototype, but no longer deploy it for people to use. If the point is to restore a capacity to evaluate futures, two questions would measure the claim. Who is helped to evaluate, and how does their judgement change?

Tonkinwise (2014) accepts the case for speculating, and he presses on what the speculations assume. Reviewing the book, he argues that a shopping framework and a taste regime underlie its arguments and projects. The futures on offer are therefore shaped by the consumer logic the work means to contest. The objection is that its audience and register are narrower than the claim implies, since a provocation carried by an exhibition label reaches the visitor who can read the label and can travel to the gallery. This is the first charge here that evidence could settle.

The words matter because the object of the claim keeps changing

Auger (2013) examines how the fictional character of speculative work is produced. He argues that speculative design serves two purposes, thinking about the future and critiquing present practice, and these do not succeed or fail together. His concept of the perceptual bridge names how a design engages its audience. Success, he argues, depends on managing the speculation as a craft skill. His own projects show the craft. The Audio Tooth Implant, made with Jimmy Loizeau, proposed a tooth that receives radio signals and passes them to the ear through the bone. Its plausibility leaned on two borrowed facts of the day: the miniaturisation of electronics, and an urban myth that dental fillings can act as radio receivers. A resin model tooth with an embedded chip carried the idea to audiences.

The taxonomy separates the claims, and the vocabulary stays unsettled

Malpass (2013) supplies a taxonomy. He draws on in-depth interviews with designers who helped develop the practice. He identifies satire, rationality and narrative as salient, and he separates associative, speculative and critical design by their satiric devices. The categories turn on intent and on relation to the present, so a critic can say which claim an object is making before asking whether it made it. Lindley and Coulton (2015) review ten years of design fiction, since Bruce Sterling coined the term in 2005. They find it used inconsistently across disciplines, and they conclude that it is compelling but inherently ambiguous. A claim about debate therefore cannot be tested while researchers disagree about whether the object is a design fiction, a speculative design or a critical design.

The sharpest objection to the claim comes from inside the field

The sharpest objection comes from inside HCI, and it concerns where the practice came from. Wong and Khovanskaya (2018) contrast the history usually cited in the field, a critical intervention linked to radical art practice and critical theory, with speculative design's introduction into HCI through the corporate design research of the RED group at Xerox PARC. Their argument is that it works in HCI because it is compatible with conventional corporate speculation. A defence that appeals to radical origins is then arguing from a history the field does not have.

Bardzell and Bardzell (2013) reach a related conclusion by another route. They analyse critical design as a research through design methodology. They argue that its central concepts and methods are unclear and difficult to adopt. They draw on traditions of critical thought from the past 150 years, and they read two designs not billed as critical to argue a broader formulation.

The field has begun to convert provocation into evidence

Elsden and colleagues (2017) propose speculative enactments, in which participants act within speculative but consequential circumstances. They analyse three case studies: Runner Spotters, Metadating and a Quantified Wedding. The move converts a provocation into a situation with participants, and so into something observable.

The same tension runs through the evidence about what a speculative object is for. Wakkary and colleagues (2013) combine practice theory with design fiction to analyse two studies, of everyday repair and of green do-it-yourself practice. They argue that speculation can carry analytic content rather than only provoke. Pierce (2014) shows why an artefact's type determines how it can be judged. He distinguishes operational prototypes and products, conceptual and material design studies, and design proposals. A proposal is not a prototype, and prototype standards do not test one.

Contested question: can the quality of a speculation be specified?

The first position holds that quality is specifiable. Ringfort-Felner, Dorrenbacher and Hassenzahl (2025) review 63 publications on speculative design, design fiction and critical design. They identify nine key qualities, grouped as speculative, discursive and process qualities. They offer the set as a descriptive taxonomy for creation and evaluation, including academic peer review. Their motive is that the criteria for a good speculation are scattered through the literature, so the problem is a missing standard that can be written.

The second position holds that the missing standard is not the problem, and that importing one would misread the practice. Bardzell and Bardzell (2013) argue that the field's concepts and methods are unclear and hard to adopt. They answer by bringing in critical theory, not by scoring projects against a rubric. Dunne and Raby (2013) give the strongest version. Their book insists the objects are not deployed for people to use, which makes instrumenting a speculation for effect a category error.

The third position sidesteps the question of standards and asks what speculation does. Zhu, Wang and Li (2024) review 53 speculative artefacts in HCI. They derive four categories: reflective, exploratory, interventional and heuristic. They argue that artefacts shape experiences and social norms, and that the field should recognise their capacity for participatory speculation. Kozubaev and colleagues (2020) contribute five modes of reflection, concerning formgiving, temporality, researcher positionality, real-world engagement and knowledge production. This reframes the topic from quality to function, which is more tractable but leaves the claim about debate unmeasured.

The three predictions and their falsifiers are these.

• Quality is specifiable. Ringfort-Felner and colleagues (2025): reviewers given the nine qualities should agree on which of two speculations is better. Low agreement between reviewers on the same pair would falsify this.
• Speculation is not an instrument. Dunne and Raby (2013) and Bardzell and Bardzell (2013): measuring effect misses what the practice is for. A measured audience effect that tracks critical intent would falsify this.
• Function is the tractable question. Zhu and colleagues (2024) and Kozubaev and colleagues (2020): artefacts can be classified by role. Cases that fit no role, or change role with the audience, would falsify this.

What would settle it is a study none of these sources reports. It would take one speculation and document who was reached and by what route. It would record with independent coders whether the reasoning or the decision of those reached changed, against a non-speculative comparison. Elsden and colleagues (2017) come closest, since they put participants inside the speculation. They report what participants did rather than what the speculation changed.

The three positions rest on different kinds of evidence. The specifiable position rests on a scoping review of 63 publications, which gathers the qualities the literature already names rather than testing whether they discriminate. The not-an-instrument position rests on the primary text and a methodological analysis, which are arguments about what a speculation is. The function position rests on two reviews that classify artefacts, so it describes what speculations do without measuring what they change. Across all three the reach of a single object is unmeasured. Until that study exists, the designer should state the audience and the intended change in advance, document who was reached and by what route, and record with independent coders what changed against a non-speculative comparison, so that the claim to open debate carries a record rather than an assertion. The strongest objection to the unit's thesis is that some speculations aim to prompt reflection rather than to change a decision, so a test that requires a measured change may misdescribe success and push the practice towards its most legible and least ambitious form.

Where the argument is heading

Named work points in several directions. Ringfort-Felner and colleagues (2025) turn scattered expectations into a descriptive taxonomy aimed at peer review. Kozubaev and colleagues (2020) supply interrogation questions for practice. Zhu and colleagues (2024) classify what artefacts do. Against them, Wong and Khovanskaya (2018) warn where instrumentation leads, since speculative design entered HCI through corporate speculation.

The sources this unit rests on

• Dunne, A. and Raby, F. (2013). Speculative Everything: Design, Fiction, and Social Dreaming. MIT Press. https://mitpress.mit.edu/9780262019842/speculative-everything/ Dunne and Raby argue societies are losing the capacity to envision futures in order to evaluate their desirability, and that design should physicalise speculation beyond the prototype, not deploy it for use. A manifesto, not a study.
• Tonkinwise, C. (2014). How we intend to future: review of Anthony Dunne and Fiona Raby, Speculative Everything. Design Philosophy Papers. doi:10.2752/144871314X14159818597676 Tonkinwise accepts the case for speculating and argues a shopping framework and a taste regime underlie the book's projects, and that class difference is absent. It is a review essay without measurement. It measures nothing.
• Auger, J. (2013). Speculative design: crafting the speculation. Digital Creativity. doi:10.1080/14626268.2013.767276 Auger examines how the fictional character of speculative work is produced, argues it serves two purposes that do not succeed or fail together, and introduces the perceptual bridge. It is a reflective account of his projects. Self-reported, not measured.
• Bardzell, J. and Bardzell, S. (2013). What is critical about critical design? Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. doi:10.1145/2470654.2466451 Bardzell and Bardzell analyse critical design as a research through design methodology and find its central concepts and methods unclear and hard to adopt. It supplies the unit's objection about importing a standard.
• Malpass, M. (2013). Between wit and reason: defining associative, speculative, and critical design in practice. Design and Culture. doi:10.2752/175470813X13705953612200 Malpass draws on interviews with designers who helped develop the practice to identify satire, rationality and narrative as salient, separating associative, speculative and critical design by their satiric devices. The taxonomy rests on practitioners' accounts.
• Wakkary, R., Desjardins, A., Hauser, S. and Maestri, L. (2013). A sustainable design fiction: green practices. ACM Transactions on Computer-Human Interaction. doi:10.1145/2494265 Wakkary and colleagues combine practice theory with design fiction to analyse two studies, of everyday repair and of green do-it-yourself practice, reporting a sustainable design fiction, where speculation carries analytic content. Only two studies.
• Elsden, C., Chatting, D., Durrant, A. C., Garbett, A., Nissen, B., Vines, J. and Kirk, D. (2017). On speculative enactments. Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. doi:10.1145/3025453.3025503 Elsden and colleagues propose speculative enactments, in which participants act within speculative but consequential circumstances, and analyse three case studies. It converts a provocation into a situation that becomes observable.
• Pierce, J. (2014). On the presentation and production of design research artefacts in HCI. Proceedings of the 2014 Conference on Designing Interactive Systems. doi:10.1145/2598510.2598525 Pierce reviews examples of research through design in HCI and distinguishes operational prototypes and products, conceptual and material design studies, and design proposals. An artefact's type determines how it can be judged.
• Lindley, J. and Coulton, P. (2015). Back to the future: 10 years of design fiction. Proceedings of the 2015 British HCI Conference. doi:10.1145/2783446.2783592 Lindley and Coulton review ten years of design fiction and find the term used inconsistently, concluding it is compelling but inherently ambiguous. A claim about debate cannot be tested while the object changes name.
• Wong, R. Y. and Khovanskaya, V. (2018). Speculative design in HCI: from corporate imaginations to critical orientations. In New Directions in Third Wave Human-Computer Interaction. Springer. doi:10.1007/978-3-319-73374-6_10 Wong and Khovanskaya contrast the history usually cited in HCI, a critical intervention linked to radical art practice, with speculative design's introduction through corporate design research at Xerox PARC. The defence appeals to a history the field lacks.
• Kozubaev, S., Elsden, C., Howell, N., Sondergaard, M. L. J., Merrill, N., Schulte, B. and Wong, R. Y. (2020). Expanding modes of reflection in design futuring. Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. doi:10.1145/3313831.3376526 Kozubaev and colleagues examine design futuring and contribute five modes of reflection, concerning formgiving, temporality, researcher positionality, real-world engagement and knowledge production. It reframes the topic from quality to function. It measures no change.
• Ringfort-Felner, R., Dorrenbacher, J. and Hassenzahl, M. (2025). The quality of speculation: a scoping review. Proceedings of the CHI Conference on Human Factors in Computing Systems. doi:10.1145/3715336.3735794 Ringfort-Felner, Dorrenbacher and Hassenzahl scoped 63 publications on speculative design, design fiction and critical design and synthesised nine key qualities, speculative, discursive and process. It gathers criteria without testing whether they discriminate. No test of agreement.
• Zhu, L., Wang, J. and Li, J. (2024). Exploring the roles of artefacts in speculative futures: perspectives in HCI. Systems. doi:10.3390/systems12060194 Zhu, Wang and Li analyse 53 speculative artefacts in HCI and derive four categories, reflective, exploratory, interventional and heuristic, arguing artefacts shape experiences and social norms. It measures no audience.