For our first homework assignment in Introduction to Physical Computing, Tom Igoe asked each of us to write a blog post about the definition of physical interaction. Chapters 1 and 2 of The Art of Interactive Design by Chris Crawford along with “A Brief Rant on the Future of Interaction Design” by Bret Victor were assigned as background reading.

Lately, when I want to consider the definition of a word, I look it up using a somewhat unusual resource: Webster’s 1913 dictionary. If you’d like to know why I use this particular dictionary, I suggest you read this excellent blog post by James Somers. Anyway, here’s what Noah Webster had to say about the word “interaction” over 100 years ago:

In`ter*ac”tion, n.

1. Intermediate action.

2. Mutual or reciprocal action or influence; as, the interaction of the heart and lungs on each other.

In my opinion, the key words in that definition—”mutual” and “reciprocal”—are as relevant to any definition of interaction today as they were in 1913, especially w/r/t modern interactive technology. After a piece of interactive technology processes an input, it must reciprocate. That means it must do more than simply provide an output (a reaction); it must provide an output that responds to the user’s input and provides the opportunity for further input.

I believe Chris Crawford’s definition of interaction as “a cyclic process in which two actors alternately listen, think, and speak” is accurate and appropriate. However, his definition does not specify that the nature of each actor’s “speaking” must provide the opportunity for reciprocal speaking. In his book, Crawford successfully circumnavigates this issue by providing numerous examples and counterexamples to define what is and is not interactive. But I think he could have benefitted from Webster’s 1913 definition.

The framework of interaction as “mutual or reciprocal action or influence” can also guide us to a better understanding of the conditions that comprise good or meaningful physical interaction. In alignment with his listen-think-speak cyclical model of interaction, Crawford states that good interaction requires good listening, good thinking, and good speaking. Unfortunately, that’s a rather vague assertion.

As stated above, my definition requires the output (the “speaking” part) in an interactive system to provide the opportunity for further input. I believe that the quality of interaction depends heavily on the quantity of possible inputs that can be used to respond to a given output. As a student of economics, I learned that having more choices makes us happier consumers. I believe this concept applies to interaction as well.

For example, imagine two friends are asking you what you would like to do tonight. One asks, “Would you like to go to the Whiskey Bar on 8th Street?” You can respond with yes, no, or maybe. The other asks, “What would you like to do?” You can respond with any number of answers, to which she can in turn respond, and you can have a meaningful conversation about the particular options amenable to both of you. Obviously, friend #2 is providing a better interaction experience.

The methods we use to interact with machines also dictate the quality of such interaction, and Bret Victor’s article was concerned primarily with that aspect of the experience. Victor asserted that many modern interactive devices, such as iPhones and iPads, merely allow us to manipulate “Pictures Under Glass” and emphasized that this is “obviously a transitional technology.” He argued that “Pictures Under Glass sacrifice all the tactile richness of working with our hands, offering instead a hokey visual facade.”

Victor was concerned that Pictures Under Glass currently dominate our society’s vision of the future. However, in his critique of the paradigm, he failed to address its primary utility: that despite its lack of tactile richness, it allows for the creation of a virtually unlimited variety of interaction experiences. An iPad can provide any array of button, knobs, switches, text fields, and other interaction methods alongside visual content, which in turn can be manipulated by the user through multitouch swipes and taps.

A device that can fulfill a variety of interactive functions should be able to produce a variety of interaction methods because good interactivity requires methods suited for the task at hand. Accordingly, any technology that seeks to replace Pictures Under Glass must provide the same limitless (or nearly limitless) possibilities. After considering this requirement, two ideas came to mind:

  • A board of tiny pins, each displaying one or more pixels on top, that can raise and lower themselves into different arrays of buttons, switches, and other input elements. (Like this toy, but with a screen distributed among the pinheads.)
  • A screen filled with fluid that can morph into various input elements.

In discussing interactive technology, it’s easy to forget that not all digital technology is interactive. The internet, in its original incarnation, was arguably not interactive, and parts of the internet remain devoid of interactivity. Digital cameras are not interactive, and neither are other digital imaging devices such as MRI machines and sonograms. Digital television and radio, like their analog counterparts, are not interactive. Clocks, scales, thermometers, and other measurement devices that rely on digital technology are not interactive. GPS and other location detecting devices are not interactive.

Aside from these non-interactive examples, experiences shared between humans and machines will continue to become richer and more meaningful as technology advances. Only by understanding interaction and the qualities that improve it can we hope to make meaningful contributions to the realization of such a future.