Why does the world appear stable as we move through it? "This is because our higher-level agents don't 'see' the outputs of the sensors in our eyes at all. Instead, they 'watch' the states of middle-level agencies that don't change state so frequently.
Why then are "inner agencies" stable? Becausy they are implemented by frame arrays. This is accomplished in part by using direction-nemes to select frames from frame arrays (e.g. to predict world-appearance as we move through it).
NB: The idea of internal stable reps (at least viewed in terms of time) is neat. Why might the visual system be sluggish, as in the visible persistence effect? Perhaps to build internal stablity. Are early effects like visible persistence magnified by interior processes that construct intermediate representations?