Modules and Wholes

The more design considers the emotional domain -- what makes an object compelling -- the more I, and many others, want to make objects explain human behavior. It's interesting how the old conceptual chestnut known as affordances can experience a new lease of life in the emerging era of tangible interfaces and umbitiqious computing. I'm reminded how strong, if potentially misleading, the concept is as I read a book on toys.

More from a current favorite author of mine, Brian Sutton-Smith, this time from his book Toys as Culture. Discussing the properties of toys, he quotes the elegant, opinionated and hip-shooting essayist Roland Barthes, who blasts "complicated toys" for supplying "actions without adventure, without wonder, without joy...they are supplied to him [a child] ready-made, he is never allowed to discover anything from start to finish." Barthes goes on to praise the wonder-inducing qualities of wooden blocks. Without any misindoctrination from the debates in the HCI world over the utility of affordances, Barthes has arrived at the concept on his own, and applied it with conviction.

After presenting similar quote from another French author ascribing various intrinsic properties to manufactured toys, Sutton-Smith rightly notes that the writers:

assume that one can simply look at the character of toys and make predictions as to how they will be used and what their effects will be on human creativity. It is doubtful if that is possible. One needs to know the context in which the toys is used to know much at all about its effects.

I find the discussion brilliant, because it highlights how tempting, even apparently commonsensical, it is to make assumptions about how people (even children) will behave based on how something appears. In the GUI era, the affordance concept was rehabilitated by appealing to the notion of the "perceived" affordance: what a user guesses something does based on learning or cultural conditioning. As we embark on designing new tangible interactions that don't borrow from established patterns of behavior, it will be tempting to again imagine that certain properties will elicit, automatically, certain behavioral responses. But truth will come from observation, paying attention to the highly variable contextual element. And the reality will always be muddier than we'd like. Pity.