Overview
The distinction at the heart of Agre's 1994 paper 'Surveillance and Capture.' Not a concept on its own so much as the central analytical move: contrasting the visual/spatial metaphor of surveillance (watching from above, the Panopticon) with the linguistic/structural metaphor of capture (imposing grammars on activity, making it computationally tractable).
The surveillance model asks: who is watching? The capture model asks: how has the activity been restructured so that it generates data as a byproduct?
Why the Distinction Matters
The surveillance model suggests resistance through hiding, encryption, anonymity — avoiding the gaze. The capture model suggests that resistance requires refusing the restructuring of activity itself, or at least understanding that the restructuring has occurred. You can't hide from capture by closing the curtains, because capture doesn't work through observation — it works through the grammar imposed on your activity.
Later Resonance
Shoshana Zuboff's 'surveillance capitalism' framework operates largely within Agre's surveillance model. Several commentators (including Alexander Galloway in 'Agre > Zuboff') have argued that Agre's capture model provides a more technically precise and analytically powerful framework.