Events and the Information for Perceiving Events (Gibson, 1979, Chapter 6)

In the previous chapter, Gibson introduced the notion of the ambient optic array and discussed the kinds of information it can have (flow, and invariants). The focus there was what is revealed by a moving observer. This chapter is about what is revealed as the environment changes; when events occur. 

Sabrina blogged this chapter here and here. 

Ecological scale events come in three main varieties; change in the layout of surfaces, change in the colour/texture of surfaces, and change in the existence of surfaces.

Changes in layout (due to complex forces): surfaces and objects can translate, rotate, and these can be rigid (distances between points on the surface are preserved over the transformation) or non-rigid (they are not preserved). Surfaces can deform (e.g. water and waves) or crack and explode. These things can happen by themselves or in combinations, and they can happen in sequence or concurrently or both. 

Changes in colour/texture: surfaces can alter without moving, in terms of colour and texture. These are often chemical reactions and can have meaning to an organism (e.g. the ripening of an apple). Again, perceptually, the change of the surface is the important thing, not the underlying chemistry. 

Changes in existence: surfaces can come in and out of existence (e.g. as water freezes and then melts). The matter continues to exist regardless, but the surface doesn't, and again, perceptually, that is what matters. 

Events have some key properties. First, they have a reality; they are things in themselves (at the ecological scale) and cannot simply be reduced to the underlying physical changes. Second, perceptually, events are primary with respect to time. In the same way that surfaces do not fill space, they constitute the environment, events do not fill time, but constitute our experience of time' "Time and space are not empty receptacles to be filled; instead, they are simply the ghosts of events and surfaces." (pg 93). 

Third, events show repetition without repetition; sunsets are all basically the same, but never quite identical. Fourth, some events are reversible in time and some are not; breaking and mending are opposites but one is not the reverse of the other, for example (see this post on some great empirical work by Bingham). Fifth, events are nested, and have a structure that can be perceived (this is the basis for prospective instead of predictive control; "If we can understand these nested sequences, it may be possible to understand how it could be that in some cases the outcome of an event sequence is implicit at the outset...so that it is possible to forsee the end when an observer sees the beginning." (pg 94). Bootsma has referred to this as 'the current future'. Finally, events can have affordances. 

Gibson then discusses the optical information for events. Things change in the world, and this leads to changes in the array; but the latter are not copies of the former. This introduces a crucial distinction: 

Once again, let us remind ourselves that events in the world should not be confused with the information in light corresponding to them. Just as there are no material objects in an array but only the invariants to specify objects, so there are no material events in an array but only the information to specify events. No object in the world is literally replicated in ambient light by a copy or simulacrum. And as for what happens in the world, it could not possibly be replicated or copied in the light. (pg 95)

In modern terms, the world is dynamic, and information is only kinematic. Information therefore cannot be identical to the world, but it can specify the world and so serve as information (see this post). 

Gibson discusses some examples. First, mechanical events (e.g. a rigid surface translation) lead to changes in the array like occlusion (as it moves), looming (as it approaches), shearing of optical texture (as it spins) and so on. Next are chemical events; Gibson has less to say here for lack of data, but notes that whatever the information changing colour is, it will be present. Next are surfaces coming in and out of existence; optical texture comes in and out of the array as surfaces do in ways specific to the transition. He then lists some types of event related optical changes, notes that formally describing these will be hard but not impossible, and that again, they are specific to but not identical to the event in the world causing them. 

Finally he notes a case where one event causes another (e.g. a collision causing a ball to roll) and points to developing (at the time) work by Michotte and Runeson that we can perceive these causal relations as such.