Ideas

Why metaffordance isn't just Gibson again

Affordance theory, as James Gibson formulated it, describes a world thick with directly perceivable possibilities for action. A surface affords standing, a handle affords gripping, a gap affords passing through. It's a powerful correction to the idea that perception starts as meaningless sensation and gets interpreted afterward: the possibilities are there in the world, picked up directly, not inferred.

The question worth asking honestly is whether metaffordance, the idea that some things don't just afford, they invite you to interact, again and again, and in new ways, is really adding anything to that picture, or just restating it with a livelier word. Gibson's perception is already active and exploratory. You move, the array of available information changes, new affordances become visible, you act on those, and so on. Isn't that already a loop? Isn't "again and again, and in new ways" just Gibson, described with more adjectives?

No, and it's worth being precise about why, rather than just insisting.

1. It's a different logical type, not a longer list

A first-order affordance is a relation between what an agent can do and what a feature of the environment permits: a surface affords standing because of a fit between body and ground. Metaffordance is a relation where the interaction itself changes what's subsequently on offer, not just because something moved or wore out, but because the later possibility is shaped by the content of what was just done. That's not a longer inventory of affordances. It's an affordance whose object is the affording relation itself, one level up from where classical affordance talk operates.

2. Ordinary affordances update; metaffordances accumulate meaning

Climb a stair and your body is now positioned so the next stair affords climbing too. That's real change over time, but each step is independent of what the last one meant, it only depends on where you now stand. Write a word on a page, though, and what the page affords next, revise it, extend it, refer back to it, depends on the content of the mark, not merely on a changed physical layout. The first case is iteration. The second requires treating the trace of a previous act as itself something that can be afforded upon. That's the extra move classical affordance theory doesn't make on its own.

3. It names technologies Gibson's vocabulary wasn't built for

Gibson's ecological optics assumes a largely indifferent environment: structured, information-rich, but not itself watching or responding. A pencil, a phone, a video call, an AI system are different in kind. They sense what you do and answer it, in real time, and what they answer with reshapes what you can do next. That mutual responsiveness, not just the persistence of an environmental layout, is what metaffordance is naming, and it's the category that gets flattened if every case is described with the same word used for a doorknob.

4. It separates artifacts that get used up from artifacts that generate

A doorknob affords turning, and once turned, that's close to the whole story. A pencil, a language, a piece of software don't have a comparably closed inventory of what they afford. What they open up next depends on what's already been done with them, and that space doesn't run out the way a doorknob's does. This is a real distinction between kinds of things in the world, not a stylistic flourish, and affordance theory as usually stated has no vocabulary for it.

5. It survives the strongest objection

The challenge isn't a weak one: since perception is already active for Gibson, doesn't exploring a page, a conversation, a tool just count as ordinary affording, extended over time? Any environment changes as you act on it, a stepped-on stair repositions you, a walked-through room reveals what's newly within reach, a grazed field is left compacted and cropped. None of that needs metaffordance. It's ordinary ecological dynamics, and point 1 already grants it to Gibson without adding anything.

Reaching for more adjectives doesn't answer this. Noticing what kind of change is doing the work does. A compacted patch of ground doesn't mean anything, it's just compacted, and the next affordance it offers follows from its new physical state, not from what the compacting was about. A word on a page is different in kind. What the page affords next, revise it, continue it, answer it, depends on the content of the mark, not merely on an altered physical layout. Iterated affording, the stair-climbing or field-grazing kind, never requires that a later possibility take an earlier relation's meaning as its object. It only requires that circumstances changed. Getting from "the optic array looks different now" to "what's on offer next is shaped by what was just done and meant" requires adding the reflexive, type-shifting move metaffordance names, a move that only shows up once the trace left behind carries content rather than just altered matter.

Calling the loop "iterated affording" doesn't settle this either. Gibsonian event perception already explains why a walk, a climb, a reach coheres as one goal-directed event, no meaning required, just coordinated task dynamics. What it doesn't explain is why the next affordance's content depends on what an earlier act meant, rather than on the task state alone. That dependency is what "iterated affording" can't account for on its own, and it's what the reflexive, type-shifting move actually adds.

Affordances invite you to act. Metaffordances invite you to interact, and the difference isn't decoration.

References

Gibson, J. J. (1979). The Ecological Approach to Visual Perception. Psychology Press.


Metaffordance and imagination

Paul Treffner
metaffordance.com