Operationalizing Design Fiction with Anticipatory Ethnography

Exploring Anticipatory Ethnography

The foundational properties of design fiction and design ethnography are mutually consistent. By combining these properties in symbiosis, Lindley, Sharma and Potts (2014) named a practice: anticipatory ethnography. This paper builds on that work and puts the construct to the test practically. In order to make this paper accessible as a standalone piece, here we offer an abridged account of anticipatory ethnography.

Design fiction – Speculative design practices have no direct interest in producing a finished article for production, sale or implementation. These speculative designs aim to challenge assumptions, be critical, and stimulate conversations (cf. Bleecker 2009; Dunne and Raby 2013; Tanenbaum 2014; Lindley and Coulton 2014; Lindley, Sharma, and Potts 2014). Design fiction sits within the taxonomy of speculative design, differentiated from other speculative design practices by its invocation of, and dependence on, ‘diegetic prototypes’ (Kirby, 2010). In the context of design fiction ‘diegetic’ refers to of diegesis, and diegesis refers to the world within which a story takes place. Hence, a diegetic prototype is a prototype that exists within a ‘story world’. Although other speculative design approaches, for instance critical design, may craft story worlds too, design fictions tend towards doing this with people at the core and with a character of neutrality (as opposed to criticality). In doing so they allow for a discursive space to emerge from the “diegetically situated” designs, people, and environments (Lindley, Sharma, and Potts 2014:246).

Figure 1. The nature of design fiction’s speculations. This diagram (Coulton, 2015), a descendent of PPPP (Dunne & Raby, 2013), which in turn is inspired by Voros’s Cone of Possibility, is well suited to describing design fictions. First we see that the entire spectrum of past, present and future possibilities are ‘filtered’ by the people at the center. Second it accepts that there are multiple people each with their own interpretation of the spectrum. Third it gives equal weight to perception of the past; an essential part of how one interprets futures. Fourth, the original spectrum in Voros’s cone included probable, plausible, possible and preferable futures, here it has been refined to a simpler duality of plausible and possible.

Because diegetic prototyping in design fiction naturally creates diegetic situativity, the techniques, methods and intentions of design ethnography are applicable and may be used as a way of analyzing and operationalizing the discursive spaces that design fictions aim to create. Meanwhile design fiction’s inherently speculative character, and focus on the future, offers a site for ethnographic study that is temporally free from ethnography’s usual ties to the present.

Design fiction is relatively immature and as a practice is still imbued with methodological uncertainty. Lindley’s (2015) ‘pragmatics framework’ goes some way to mediating the potential confusion inherited by this uncertainty, by proposing three categories which describe the differences between types of artifact that are referred to as design fiction:

1. Intentional design fictions – artifacts that are created as a design fiction.
2. Incidental design fictions – artifacts that can be interpreted as a design fiction.
3. Vapor fictions – usually marketing materials that resemble design fictions.

The pragmatics framework paper considers how design fiction relates to Sir Christopher Frayling’s typology for art and design research (1993). The following types of research are considered:

1. Research for design – a contextual search intended to support a specific design, research leading up toward an artistic or design endeavor.
2. Research into design – studying the theory of design, usually a paper-based activity, not necessarily involving any design or making work, per se.
3. Research through design – production of knowledge as a direct result of a design and/or making process.

Notes on Incidental Design Fictions

David Kirby, in developing the notion of diegetic prototypes, a concept central to making design fiction work, quotes the Red Hot Chilli Peppers, “Space may be the final frontier but it’s made in a Hollywood basement”. The lyric highlights Hollywood film producers’ skill at crafting fictional worlds. Bearing this in mind the incidental design fiction the authors elected to focus on for this research came from Hollywood. The authors worked with Spike Jonze’s acclaimed 2013 film, Her. Watching Her would enable the reader to gain a greater insight into this study. Before progressing into further detail about the reason for selecting Her, below, the paper will elaborate on what qualifies Her as a piece of incidental design fiction.

The most popular definition of design fiction is “the deliberate use of diegetic prototypes to suspend disbelief about change” (Sterling 2012). Lindley, Sharma and Potts comprehensively deconstruct this definition, as it relates to anticipatory ethnography (2014:240–243). Lindley and Coulton offer a paraphrased version “a design fiction is (1) something that creates a story world, (2) has something being prototyped within that story world, (3) does so in order to create a discursive space” (2015). It is easy to see how these definitions apply to Her. The movie creates a story world (Her’s critical acclaim would suggest it did this successfully). The diegetic prototypes within the story world of Her are abundant – some of these will be elaborated on later when we discuss the insights produced. Julian Bleecker, writing on design fiction, notes how these props, and the stories they tell, are so affecting (2010):

“[A] good story with its props may be more effective at materializing an idea than an engineering prototype. We might wonder why more engineers are not drawn to storytelling as a way to prototype their ideas, rather than circuit building or software prototyping. As a means to communicate and disseminate an idea, not much works as well as the circulation of a compelling story. Hollywood and the entertainment-media network has taught us this much, at least.”

Figure 2. Diegetic prototypes in Her. Left we see the portable device that Theodore uses to interact with his operating system, Samantha. Top right we see the same device placed with the camera facing outward so Samantha can see the world, Theodore uses a safety pin to position the device in his clothing, also highlighted is the ear bud that allows Theodore to hear Samantha speaking in private. To the bottom right we see Theodore ‘installing’ the operating system on a familiar-looking desktop computer.

The third and final property of a design fiction is the creation of a discursive space. This is where Sterling’s definition differs slightly from Lindley and Coulton’s. Sterling simply says that these diegetic prototypes are invoked to suspend disbelief about change (2012). Lindley and Coulton (2015) modify this such that the reason for that suspension of disbelief is accounted for: to create a space for discussion (and thus to facilitate the emergence of insights). Any good fiction should, if it is done well, engender some kind of suspension of disbelief: that is what good storytelling does. However when we use that storytelling not solely for its originally intended cultural purpose but apply it in design fiction terms, then that suspension of disbelief helps create a discursive space. Putting this point very simply, a good story should act as a stimulus for further conversation.

Looking more closely at the quality and content of the movie itself, Her exhibits an affecting diegesis, and, although this is an intrinsically subjective arena, in the opinion of the authors the movie certainly invites a suspension of disbelief about change (cf. Farago 2013; Vanhemet 2014). The film is laden with a range of diegetic prototypes, in fact one of the lead characters – the artificially intelligent software called Samantha – is a diegetic prototype herself. Although the artificial intelligence side of the film transcends the usually mundane and everyday quality of design fictions (as well as pushing the limits of plausibility), the other diegetic prototypes are clear extrapolations from today’s technology. Examples include a smartphone-like device, a wireless ear bud/microphone for interacting with the mobile device and electronic services such as online dating, letter writing, and video gaming. The setting of the film too played into our hands, although the interior architecture, external environments and sets are clearly ‘different’ from today, they are not so dissimilar as to prevent the suspension of disbelief. Exterior filming in both Shanghai and Los Angeles helps to create a comfortable but ‘other’ feeling.

While Her is one good academic test bed for anticipatory ethnography, there is a range of other possibilities. If a design fiction relevant to the ‘thing that is being designed’ exists, it would be apt to work with that resource. For example, Her, which features a diegetic prototypes of voice interfaces, would be a logical site for research about that technology, using anticipatory ethnography. Similarly, the movie Robot & Frank or the television series Humans may be fertile ground for research about robotic appliances, using anticipatory ethnography, as they both feature diegetic prototypes of domestic robots. Where such a relevant design fiction resource does not exist, however, the method described here is not viable. The ‘Reflecting on the Method’ section will return to this question and discuss ways in which anticipatory ethnography could be used when a relevant incidental design fiction is not available.

Details of the Method and Deriving Actionable Insights from Her

Explaining the concepts of anticipatory ethnography and the nuances of design fiction requires significant wordage, in part because of the nascence of each idea and partly because of their novelty. In contrast, describing the method is, thankfully, less complicated. The method described here is not the de facto anticipatory ethnography method but rather one intended to test the feasibility of anticipatory ethnography in the simplest way possible – in effect a rapid prototyping of the anticipatory ethnographic method.

A group of four ‘anticipatory ethnographers’ watched Her in its entirety. Among the group of four, two were trained design ethnographers, whilst the other two had previously participated in design research. While watching the film verbal dialogue between the researchers was minimal, and post-it notes were used to capture observations of the characters and action. Immediately afterwards the group went through an affinity mapping process (Kawakita 1982) developing themes from the raw observations, and moved toward producing actionable insights. Since this was an academic endeavor, the authors did not have a specific client brief to work towards, and the range of insights produced was incredibly broad. In order to make the account of the produced insights more coherent, the authors have limited what is included in the paper to those insights with the potential for further research exploration. Each category begins with a general note on the space, including some links back to the content in Her, and concludes with a list of insights.

Sound, voice and audio interfaces – In design fiction terms, the system that facilitates the character Theodore’s (human) interactions with Samantha (artificial intelligence) is a diegetic prototype of an advanced voice interface. Similar technologies have existed for some time in the form of speech recognition with appropriately programmed responses. More recently applications such as Siri, Google Now, and Microsoft’s Cortana offer pseudo-intelligence voice recognition systems underpinned by big data and ubiquitous connectivity. Despite offering elements that appear intelligent, these systems are far from adaptable. Similarly, Bluetooth headsets are a staple for anyone wishing to use their mobile telephone while driving or using their hands. The design provocations contained within the diegetic prototypes in Her, however, demonstrate that these interfaces are far from satisfactory (in terms of both function and aesthetics).

Despite the video revolution that we have seen on the web, facilitated by increasing bandwidth and the ease with which video content can be created, the web remains primarily a text-based medium. Other contemporary services like ‘Chat Roulette’ utilize easily available video cameras and bandwidth to connect users on the fly. Meanwhile, internet-telephony is so commonplace that ‘to Skype’ has become a verb. However, a purely voice-based communication has significantly different properties from text-based methods, this looks like a space ripe for development, particularly in terms of interacting with machines.

‘Conversations’ directly with services, digital corporations or platforms, are incredibly rare. Although, as a customer of Amazon, for example, one can easily speak with a person who is working on behalf of Amazon, it is rare (or impossible) actually to speak with Amazon itself. Whether this possibility would be predicated on artificial intelligence, or on some kind of ‘wizard of Oz’ style deception, the authors feel that there is space for innovation here. If you could have a conversation with Facebook, Twitter, Vine, or LinkedIn, what would you say?

1. Contemporary voice interfaces are unsatisfactory; they are unintuitive, slow down information exchange, and don’t substantively alter the way we interact with computers. Although they attempt to make our interactions more natural, they feel unnatural.
2. Despite inherent issues voice-controlled computers are ubiquitous (Siri, automated answering services, etc).
3. Contemporary smartphone ‘Bluetooth headsets’ (or ear-buds) are unsatisfactory; their functions are limited, and their use is somewhat stigmatized (i.e., they’re ‘not cool’).
4. The web primarily operates around text-based media, whilst voice is a powerful and evocative medium that remains underrepresented.
5. Systems, machines, servers, and networks do not engage in ‘conversation’ with their users. Is this an oversight, or due to technological challenges?
6. The gender, voice, and nature of a personified computer or information system, will impact on how it acts in the world and alter the way interactions occur.

Reflecting on the Method

This simple method – a group of researchers watching an incidental design fiction, recording diegetically situated observations on post-it notes, and developing themes and insights via an affinity mapping process – was productive. The synthesis and analysis method was accessible to people unfamiliar with anticipatory ethnography. However, without a specific brief to pursue, the process did meander into many, perhaps too many, different domains. It is noteworthy that majority of the observations made, from all participating researchers, were actually made within the first 15 minutes of the movie. This made us consider creating abridged versions of incidental design fictions.

The authors cannot account for situations where a relevant design fiction does not exist, but where the temporally unbounded character of anticipatory ethnography might be desirable. This paper does, however, offer some (as yet untested) alternative ways of invoking anticipatory ethnography.

The Sans Duty project (Duggan and Lindley 2015), which explores the future of taxation, used an intentional design fiction as a stimulus to provoke reactions from members of a community. These reactions were captured on video, and then edited into a ‘fictional documentary’, which was in turn used to stimulate further discussion amongst the community and generate insight. Iterating the production of design fiction, consultation with stakeholders, incorporation of stakeholders’ insights into new design fictions, provides an opportunity to do anticipatory ethnography where a relevant incidental design fiction is not available by using the other modes of anticipatory ethnography. This is a different type of proposition to the study described in this paper which, deserves further consideration in its own right.

Does Anticipatory Ethnography Produce Actionable Insights?

Anticipatory ethnography should nurture and produce actionable insights applicable to plausible futures. The previous section detailed some such insights that emerged from a very quick application of a prototypical approach to anticipatory ethnography. Although the insights produced were undoubtedly interesting, and in some cases quite clearly ‘actionable’, some of these insights had an entirely different feeling or flavor to what we might expect. It is not surprising that some of the insights produced feel strange and unsettling, when one considers that they are derived from unreal characters, existing only in a contrived world, which contains technologies and phenomena, with which we are unfamiliar. In this anticipatory ethnography of Her it was, in particular, the broader societal insights, and those pertaining to artificial intelligence that had this peculiar flavor. One can assume that the odd taste is a product of the futurity of the source material and our inability, from our present perspective, properly to comprehend cohabiting with some of the diegetic prototypes depicted.

The authors tentatively refer to these strange inklings of insight as ‘plausible outsights’, as opposed to actionable insights. The word ‘outsight’ is related to ‘insight’, but incorporates an externality, which is relevant because of the ‘otherness’ associated with these findings. The plausible element simply refers to the believability, yet contingency, which is inherited through the suspension of disbelief. These plausible outsights would be, if considered solely within the diegesis of the design fiction, ‘insights’. However when viewed from our reality, they are external. This does not necessarily mean irrelevant, but they certainly have a different character from the actionable insights originally sought by the authors. Here are two illustrative examples: the findings that pertain to email management software are straightforward actionable insights. On the other hand, the findings related to loving technology and notions of ownership vis-à-vis artificially intelligent machines fall into the plausible outsight category. The plausible outsights generated by the anticipatory ethnography method, seem to reflect the ‘complete’ nature of design fiction prototypes. Design fictions provide visualized use case scenarios, personae, and user journeys all at once, and the fruits of these factors are interrelated; situated.

Strengths, Weaknesses and Next Steps

Our account of operationalizing design fiction with anticipatory ethnography provides some grounds for optimism: the method worked well, it generated interesting insights, and, as was suggested in the peer reviews for this work, there is some significant potential in future-orientated research-for-design approaches. It is, however, clear that despite the generally positive outlook many questions remain. The most significant challenge is how reliably to transform the novelty and excitement of the approach into real-world tangible results. To reiterate: a question that arose at EPIC 2014: can you bill clients for this?