Interactivity, as Crawford notes, is an overused and misunderstood term.  We now use the term “interactive” to describe anything that involves user input, but the response to the user input is what actually makes something interactive.

Whether the user is interacting with another human or the computer itself, both can count as interaction as long as the user’s input (action) has some sort of impact, however cryptic or subtle, on the subsequent (responsive) action.  If two people play a game of Battleship, but they simply call out “B3” and then never confirm with the other player whether its a hit or not, the interaction does not occur.  It is when the Opponent receives the “B3,” checks his board, inserts a peg into B3, and analyzes the impact the peg has on his board and his subsequent moves that the interaction actually occurs.

An interaction facilitates a response, which facilitates another iterative response and so on until the game or experience ends.

While this might seem to limit our definition, it can actually expand it if we expand our mindframe.  For instance, when I write a note down on a piece of paper, am I “interacting” with the pen/paper?  Well technically, yeah definitely, but I think Crawford is implying that this definition would be too mushy to get us anywhere.

But what if we only define the writing as an interaction once I read my note back to myself, and then allow it to influence my other thoughts, and the next thing I write down?  Then can we say that the letters on the paper have “responded,” “interacted” with me in a way that allows me to define this as an interaction the same way I would a computer program?

I think so.  Unlike the refrigerator light example, in which the light turning on simply allows you to do what you were already planning on doing (in fact, I think a more “interactive interaction” would occur if the fridge light failed to turn on, and the user was forced to remember where he put his food and why), Writing/reading/thinking can actually veer you off your automated course, which is IMO a valuable quality of good interaction.

I like the idea that for a good interaction, both actors must “Listen, think and speak” well.  This supports his qualitative rather than boolean definition of Interaction, and allows us to start thinking about quality of interaction rather than just “there, its interactive, so it must be cool.”