XXII.4 July - August 2015
Page: 20
Digital Citation

The ethos of context-aware computing

Sorin Pintilie

back to top 

Context is a word we use a lot in our community. But we don't really talk about it. We occasionally churn out nuggets of wisdom like "context matters" or even "context is king," but it's surprising how little we talk about what context actually is and how to design for it.

And it's not just in our community. To put it into perspective, context, a concept as old as poetry, with a history equally as rich, amounts to approximately 12,000 entries in the online catalog of the Cambridge University Library, while, say, digital, a concept no older than Walt Disney's The Three Caballeros (at least in its modern, technology-related sense), counts nearly twice as many entries.

That may be due to the fact that context is, by definition, an abstraction, and abstractions are always difficult to talk about. That makes our conceptual frameworks, in a way, a bit religious: They always break down into three dimensions, seven flavors, five this or that. We inherited this tradition from the Middle Ages, when a man named Sedulius Scottus first laid out the "rule" of seven circumstances: person, fact, cause, time, place, mode, and topic. I say religious because it sounds a bit like the seven deadly sins.

It's madness.

There is a framework, though, that helped me think about—and work with—context. As a means of understanding abstractions and talking about them, it's useful to think about the ecosystem in which these abstractions develop and how they interact.

Ideas may support, contradict, combine, influence, interact, and evolve in as many ways as we can. So, in an attempt to untangle the intricate web of relations between these ideas, I'm going to borrow a term from Gregory Bateson, who in turn borrowed it from one of the pioneers of systems thinking, Sir Geoffrey Vickers: an ecology of ideas. It's a concept that sits nicely with context in design. But to put these things together, some distinctions are in order.

Context, in the general sense, is something we create in the process of being and doing, a thing we make with our mind in our body so we can interact with another mind in another body.

It is thus more of a quasi-abstraction, one that is truly ubiquitous. Anything can be context. Deceptively simple and delightfully vague, it's one of those words that's so familiar we all use it every day, but never quite appreciate what it means. Its meaning simply won't stand still.

It has been framed from many angles—as a situation [1], as a setting [2], as a mental model (which led to an almost cult-like approach in the design community: user-centered design) [3], as an activity [4], and as a space of meaning (which lead to the creation of Siri). But if we look at context more as an ecological actor [5], in Bruno Latour's sense, then we can start to see a theoretical foundation for context as a space of material abstractions, a space between ideas and things.

The theory is that we experience the world through our interaction with it. With each moment, we bring with us our experience of being human, with a body in a physical environment.

We live in a dynamic world of patterns and rhythms, where everything hums and jiggles to the rhythms of something else. When one person talks to another person, words enter the eardrum as pressure waves and rattle a few bones and some fluid. When we touch an object, countless sensors on the surface of the skin feed information back to the brain. Our eyes have a million tiny light detectors. Our ears, just in the cochlea, have 30,000 hair-like sensors that pick up on the sounds of the world and edge them into actual, physical brain circuits.

The sheer complexity of the sequence of events involved is nothing sort of amazing. And it all happens almost entirely unconsciously. Only two percent of thinking is conscious. The rest is recruitment—circuits already learned strengthen the bond between others to link the topographies of the world around us with the two percent that's conscious. That's how the brain makes context: with a body and a mind.

Context, then, is embodied. It is an interactive property of you being in the world. A complex and vibrant ecosystem with infinite connections and relationships that relies on a fragile balance between the environment and the ideas that develop within that environment—an ecology.

Taking an ecological approach to context also makes context-aware technology—in the general, Popperian sense—a vision as ambitious as artificial general intelligence is. (Note the distinction between AI and AGI.)

However, as an umbrella term for digital assistants, wearable computers, ambient intelligence, smart objects, and sensor networks, context-aware technology is nothing more than technology that uses knowledge representation methods, such as intent-understanding algorithms, semantic processing, machine learning, and other techniques, which are all direct applications of AI, to make calculated guesses on particular, specific contexts.

The realistic goals of context-aware technology—much like those of AI—were always rather more fractional than general: to produce programs that enhance and complement human thinking. To assist, rather than reproduce totally, human intelligence. But, admittedly, the more modest goal of relieving humans of repetitive tasks doesn't make good front-page material.

And this is the point where views on context divide: It's either something that can be computed or something that cannot. Looking at context as something that can be modeled and calculated follows a positivist tradition, which is objective and quantitative in nature, and comes in stark contrast to the phenomenological tradition, which is subjective and qualitative in nature and gave rise to a view of context as a form of interaction.

And for anyone not interested in reading Husserl, Heidegger, or contextualists such as Wittgenstein, Paul Dourish offers one of the most crystalline overviews of context, arguing for an interactional framing of the term and for the importance of designing systems that show their context instead of computing it [6].

Designing systems that show their context means showing implementation, rather than hiding it, so people can use that structure and make things with it so we can learn from that process. Essentially, for system designers, that means understanding what people actually do and subsequently what they experience in doing.

Systems that show their context create interactions, while context-aware systems assist interactions. These are not mutually exclusive, though; on the contrary, I think there is a common space in between these types of systems.

Bateson's idea—an ecology of ideas—fits perfectly in this space. Context is a complex, personal, and social system—a system of ideas and interactions. There's an intrinsic symbiosis playing out at the subject-concept level in these systems that makes them extremely unstable, at least at the human end. Nothing's static; everything's organic and delightfully unpredictable.

That space is a vibrant ecology. Ecologies are not new to human-computer interaction, but so far they have been focused mainly on the product [7] and the user. It's worth introducing another dimension: the ecology of ideas people live in.


Ecology of ideas in the sense that ideas matter to behavior in ways more complex than are usually considered in interaction models. They interact with technology, with culture, and with emerging social practices.

If we look at context like an ecosystem, then ideas and interactions would be like the flora and fauna of this system. That makes them subject to evolution, progress, and decay, like any other biological system. So, an understanding of interactions between ideas and the embodied forces driving and shaping them is essential in building an ecological framework of context.

With this in mind, I'm looking at an emerging theory of interaction with context-aware systems that would have to bring together notions from psychology, philosophy, linguistics, anthropology, neuroscience, and computer science, use a phenomenological perspective and an ecological approach, but ultimately would be about what all design is about: foregrounding culture and backgrounding technology.

Drawing on these foundations, this inquiry points toward an approach to design that is as much about exploration and conversation as it is about problem solving. A way of thinking about the intangible practices, meanings, personal narratives, and social and cultural implications of designed objects. It's about understanding the complexity of context.

And when it comes to technology deeply embedded in our environment, I think a new language is needed, one that considers the ecology of ideas people live in. And sometimes, that means breaking paradigm and talking about design in new ways.

back to top  References

1. Suchman, L.A. Plans and Situated Actions: The Problem of Human-Machine Communication (Learning in Doing: Social, Cognitive and Computational Perspectives). Cambridge Univ. Press, 1987.

2. Winograd, T. Architectures for context. Human-Computer Interaction 16, 2–3 (2001).

3. Norman, D. The Psychology Of Everyday Things. Basic Books, 1988.

4. Nardi, B.A. Context and Consciousness: Activity Theory and Human-Computer Interaction. MIT Press, 1995.

5. Laurel, B. Computers as Theatre. Addison-Wesley Professional, 1993.

6. Dourish, P. What we talk about when we talk about context. Personal and Ubiquitous Computing 8, 1 (2004), 19–30.

7. Forlizzi, J. The product ecology: Understanding social product use and supporting design culture. International Journal of Design 2, 1 (2008).

back to top  Author

Sorin Pintilie is a designer with a cross-disciplinary curiosity for what the world has to offer. He is interested in words and the technology behind them. He works where the social web meets the semantic web.

back to top 

Copyright held by the author

The Digital Library is published by the Association for Computing Machinery. Copyright © 2015 ACM, Inc.

Post Comment

No Comments Found