New Metaphors

Can creating new metaphors help us design new kinds of interfaces?
Could they even help us understand the world differently?

Workshops: Universidad del Desarrollo, Santiago, June 2017; Google SPAN Pittsburgh, September 2017; Carnegie Mellon University, September 2017 to date, various; Interaction 18, Lyon, February 2018. Forthcoming: UX Lisbon, May 2018.

Download the New Metaphors cards, v.0.3 (February 2018) — 129 MB PDF, 300 dpi

Download a poster/leaflet from Interaction 18 including thumbnails of all the cards, and a shortened version of this article — 2 MB PDF

Download templates / worksheets — 400 kB PDF

Through a series of workshops in 2017–18, we’ve been exploring a process for generating new kinds of metaphors, and then using those metaphors to inspire concepts for new kinds of interface design which could potentially help people understand things in different ways.

The intention of the workshops is that the process might be something designers can use or adapt for idea generation, or to provoke new kinds of thinking about interface design. The extent to which the metaphors merely provide initial ‘seed’ inspiration, or actually form the basis of the resulting design, varies.

A selection of the metaphors generated so far:

Plant growth as a metaphor for overwhelmedness Different coloured windows as a metaphor for people’s accents
A flock of birds as a metaphor for Wi-Fi signal strength The hum of a fridge as a metaphor for half-remembered dreams
The hum of a fridge as a metaphor for air quality Shadows as a metaphor for power relations between people
Sweetness as a metaphor for trust in a brand or product Taste as a metaphor for your emotions

 

Qualitative interfaces

We’ve mostly been thinking about these new types of interface design as being qualitative. All this really means is that they focus on the qualities of things rather than solely being about quantities and numbers. There is some thinking behind this: while big data, the quantified self, and other trends offer a lot of promise for helping people make decisions, there are many things in our lives which are not easily reduceable to numbers—or where trying to do this can actually make our understanding worse by oversimplifying concepts and relationships between them. We might ask whether defaulting to quantification in our interface designs in some way limits the boundaries of how we can think, and what we can think about. How might we design to go beyond those boundaries? It’s not necessary, of course, for new metaphors to be used only for qualitative interfaces, but they are a good area of application for the method, offering the potential of some radically different kinds of design.

Why new metaphors?

A metaphor is just a way of expressing one idea in terms of another. This project is a nightmare. The city is a playground. You’re a gem. Interaction design often uses metaphors to introduce people to new ways of doing things, by relating them to familiar ideas, from desktops, files and windows, to the net, web, websites and browsers, cloud storage, even blockchain. Many of these are so familiar now that we perhaps no longer even think of them of as metaphors. But they are not inherently ‘right’; they can be and are being challenged by designers and researchers exploring new approaches—including creating novel metaphors, which can persuade us to think differently and accetp new ideas, or help us reframe the ways we think at present.

Poets may be the expert practitioners in creating new metaphors, but as well as within design, many people in other fields from anthropology (e.g. Margaret Mead) to politics (e.g. George Lakoff) have suggested the potential of novel metaphors as a way of helping people understand big concepts differently—in science, society and politics, global issues, economics and environmental matters. There is potential for application both in interaction design, and in more systemic approaches to larger-scale design for change, such as transition design.

Carnegie Mellon design undergraduates came up with ideas including plant growth as a metaphor for how to tell yourself you’re doing OK, erosion as a metaphor for social or peer pressure, and the hum of a fridge as a metaphor for half-remembered dreams.

A simple generation process

In the anthropologist and cyberneticist Gregory Bateson’s words, metaphor is a “pattern that connects” two concepts. This could be the basis of a method for creating new metaphors: find a pattern that connects two otherwise ‘unrelated’ ideas. There is more to it than that: why do some metaphors ‘work’ and others not? There are competing theories in academic cognitive science on why metaphors work and how they may even form the basis of large amounts of how we think. Michael Erard has some discussion of this here, in the context of his work.

But the approach we’ve taken in these workshops is to focus on the idea of features (or verbs) in common between two things—Thing 1, and Thing 2. As you can see below, groups choose from an array of (currently) 75 ‘Thing 1’ cards, which each names a concept or idea that we might want an interface (or visualisation) for, but which is difficult to quantify or visualise in obvious ways, and 75 ‘Thing 2’ cards, which each shows something (many from the natural world) which could, in some ways, work as a metaphor for something else. The selection—download the PDF—which has been evolved through the series of workshops with students and design/UX practitioners, is designed to enable plenty of combinations that make sense, but there should also be many surprises among the 5,625 possible juxtapositions:

75 Thing 1 cards, which each names a concept or idea that we might want an interface (or visualisation) for, but which is difficult to quantify or visualise in obvious ways

75 Thing 2 cards, each showing something which could, in some ways, work as a metaphor for something else

From a subset of these cards (or the whole set), participants pick some Thing 1s and some Thing 2s that interest or intrigue them—or create their own. For example, if you were working on a specific brief to design a new interface or display for a particular phenomenon which is already decided, this might become your Thing 1. Or, if you have a particular metaphor you already know you want to use for an interface (or perhaps an existing technology, and you’re trying to find applications for it), you might start with this as your Thing 2, and find Thing 1s which ‘fit’ it.

Birdboxes as a metaphor for personal data securityBirdboxes as a metaphor for personal data security? A combination chosen by Interaction 18 workshop participants.

The selection process might be random—a forced juxtaposition as suggested by Edward de Bono in Lateral Thinking in the 1970s, or similarly to Arthur Koestler’s process of bisociation (from The Act of Creation), or some of the generative creativity approaches being explored by people such as Kate Compton. Or, it might be that there are certain permutations that can be seen to ‘work’ by mere inspection. We then do a simple ‘structure-mapping’ between the Thing 1s and Thing 2s we have chosen to identify which features might actually make sense for the metaphor to work, assuming we want to arrive at “Thing 2 is [a metaphor for] Thing 1” (e.g. windowsills are [a metaphor for] a Friday night feeling — they share characteristics such as anticipation, perhaps looking outwards at possibilities, a way of escape, a place where one is crossing (or looking over) a threshold from one place or part of the week to another — but do not share (probably) lots of other characteristics such as being a place to put objects).

(It’s worth noting here that assuming “Thing 2 is a metaphor for Thing 1” is the assumed structure, but there are also situations where the reverse might also be interesting: tree bark [Thing 2] might work as a metaphor for insurance [Thing 1], but could insurance work as a metaphor for tree bark? Probably, yes, in this case—but in others the metaphor might break down.)

A simple template (download) facilitates this mapping, by encouraging participants to list / elaborate features or characteristics (often adjectives, or nouns) that Thing 1 has or is associated with, and then those for Thing 2, and then making connections between them:

Mapping characteristics of Thing 1 and Thing 2 to each other: below, Thing 1 was TripAdvisor comments, and Thing 2 was facial expressions (by a group at Google SPAN).

An alternative method here would be to use a process derived from Gregory Bateson’s syllogism in grass. A syllogism is a form of logical reasoning where a conclusion is drawn from two premises. Bateson contrasted the conventional form—where a deduction is made—with a different kind of syllogism which takes a more metaphorical angle. The format can be a way to generate or explore new metaphors (one of the clearest explanations is Andrew Roffman’s here):

Conventional syllogism

Syllogism in grass

Men die

Men die

Socrates is a man

Grass dies

Socrates will die

Men are grass

This approach could entail focusing on a verb that Thing 1 and Thing 2 have in common. For example, fame ebbs and flows; waves ebb and flow; fame is waves. It seems like this could be a useful structure for producing more ‘active’ concepts where one thing “doing something” (as opposed to just having a particular characteristic) acts as the metaphor for something else.

(A Bateson’s Syllogism Twitter bot produced using George Buckenham’s Cheap Bots Done Quick is still work in progress: it does the Thing 1 and Thing 2, but the verbs don’t agree at present!)

Either process—the characteristic mapping or the syllogism in grass—hopefully leads to each group arriving at one or more new metaphors which are not just Thing 2 is a metaphor for Thing 1, but also have associated suggestions as to which features or characteristics might enable them to be built on or developed as the basis of a piece of design, or at least enable a reframing of a situation around Thing 1.

In the workshops we have run so far, the process only goes as far as groups’ rapidly generating possible design concepts, usually for an interface or service, which they then present to the rest of the participants, either through sketching, very rapid paper prototyping, or ‘acting’ out a scenario in which the metaphor is employed. As the process becomes more refined, and participants gain more confidence—and as we work out what the most appropriate tools might be—the aim is to get to the stage where the workshops result in some actual new, functional interfaces or displays.

Four presentations of concepts: waves as metaphor for people’s accents; plant growth as metaphor for overwhelmedness; a flock of birds as metaphor for Wi-Fi signal strength; plant growth as metaphor for other people’s emotions and thoughts.

Wider applicability beyond interface design

It has been argued that metaphors and analogies are central to much human reasoning, understanding, and creativity (e.g. by Douglas Hofstadter), as well as the language we use (George Lakoff and Mark Johnson). Here we use the term ‘metaphor’ in a broad, intentionally imprecise way, to refer to a class containing a variety of ways in which one thing can be understood in terms of another. One simple reason for metaphors’ prevalence is that by mapping features of an existing or familiar situation onto a new or unknown one, we are enabled to grasp and (be more confident that we can) understand it more quickly. As such, metaphors are often used strategically in design (e.g. as discussed by Dan Saffer, Nazlı Cila, and others). Nevertheless, metaphors are not the thing itself—they are always an abstraction, a model of the situation rather than the situation modelled. They can be a map to a territory, but should not be mistaken for the territory. As with models, all metaphors are wrong, but some are useful. The constraints, affordances, and assumptions that a metaphor suggests or imposes can themselves condition or structure our interaction with, or approach to, a new situation, as we understand, or come to understand it in terms of the old. Metaphors become “enabling constraints” in Katherine Hayles’ words. The hunt for “defensible metaphors”, to use cybernetician Gordon Pask’s term, is not trivial.

When thinking about some of the major issues facing the world, including transition to more sustainable ways of living, we might consider that attempting to understand the existing metaphors in use in a situation—and then but actively generating, proposing, and following through the implications of new metaphors—could be part of a visioning or futuring project. This is beyond interface design, more like working at a system level, something closer to Klaus Krippendorff’s notion that designers could “create and start using new metaphors, new vocabularies, and new ways of languaging, like poets and science fiction writers do, thus bringing forth new ways of conceptualizing the world and encouraging new practices.” Mary Catherine Bateson, in her own work, and in discussing the work of her parents Margaret Mead and Gregory Bateson, has also frequently employed the idea of reframing societal issues through using new metaphors, for example “the idea of ‘home’ as a place to give and receive nurture” becoming “a new metaphor for the workplace”.

One significant area where new metaphors might offer opportunities for transition is the economy and political discourse in public life (an excellent BBC Radio documentary by Zia Haider Rahman explores some of these questions with contributors including Dina Nayeri and Molly Scott Cato). A number of economists have noted the ways in which the metaphor of ‘the national economy is a household budget’, commonly employed by media and politicians alike, is not just an oversimplification but a structural error in terms of many key features of the systems under discussion, such as fixation on ‘balancing the books’. This leads to specific decisions being made (austerity policies for example) that arguably cause harm or restrict the ability of the system to adapt to changes in circumstances. How would public political discourse on the economy be different if a different metaphor were used? (We can imagine ideas such as ‘the economy is a garden’ or ‘the economy is a loaf of bread being baked.’) Would it be better used to explain, or to persuade? Or both?

The art of designing new metaphors and framings is well advanced in political contexts and increasingly in corporate settings, but has been underexplored in design and futures, and offers potential for transition designers to enable communities to think about, envision, and understand their current situation and possible futures, both locally and at global scale, in new ways. A participatory process in which communities co-design the new metaphors, involving people in understanding their own and each other’s understanding as the metaphors are constructed and explored seems preferable from a transition point of view to one where new metaphors are imposed by an authority seeking to persuade. This is another angle we will attempt to explore as this project develops.


If you’d like to learn more about the project, or contribute to its development, please get in touch! danlockton@cmu.edu

Download the New Metaphors cards, v.0.3 (February 2018) — 129 MB PDF, 300 dpi

Download a poster/leaflet from Interaction 18 including thumbnails of all the cards, and a shortened version of this article — 2 MB PDF

Download templates / worksheets — 400 kB PDF

Thank you to Sarah Foley and Alejandro Pantoja Sanchez for their help with developing and running the workshops.