The Animal and the Environment (Gibson, 1979, Chapter 1)

Gibson's first chapter introduces his notion of the environment. This is a distinct level of description from the world according to physics, even though everything in it is still made of physical stuff. The environment is the ecologically-scaled surroundings to an organism, and Gibson lays out some of the key differences between this and the physical world here. Why? Because perceiving is going to be of the environment, and not of the physical world, and as we progress this is what the word 'environment' will mean.

Note: Sabrina also blogged this chapter here.

The environment is what surrounds a perceiving, behaving animal (Gibson rules out plants etc here because they aren't animate and don't have nervous systems; the field has moved back to recognising that intentionality shows up in a lot of ways; Carello et al, 2012; Raja & Segundo-Ortín, 2021. But for the moment, we are talking about animals.)

There are two senses of the environment; that which is common to all organisms, and that which is specific to an individual organism. (He hints that he will address this with reference to the fact animals can move; see Chapter 3). So there is an ambiguity to the word already (see Baggs & Chemero, 2020, who suggest 'habitat' and 'Umwelt' for the two senses) but for now, the important thing is that an environment implies an organism (and vice versa) in a way that 'physical world' does not - there is a mutuality. Before there was life on Earth, the surface was not an environment; it was just a potential environment, made of geology. Perception and behaviour are going to work with respect to environments, and not worlds. 

One of the side effects of approaching the environment this way is that it does not have distinct, identifiable and 'correct' units of space or time. In terms of space, there are nested scales of structure, each with their own metrics (leaves (cm) are nested in trees (m) which are nested in forests (km), for example). In terms of space, there are nested scales of events (hungry/not hungry cycles (hours) nested in sleep wake cycles (days) nested in changes in the passage of sun overhead (years), for example). There is structure, and regularity, but there is nothing atomic to build perception of environments out of. Perception and behaviour are going to have to work with respect to the multiple natural units of 'space' and 'time', rather than with space and time as considered in physics. 

To replace space and time, Gibson will talk about persistence and change. An environment exhibits both - it must, or it cannot be an environment. Both are relative terms; persistence depends on the materials involved, for instance. A solid object might persist for a long time, but once melted it ceases to exist in the environment of an animal. The matter continues to exist, but that is a fact from the wrong level of analysis for perception. Objects can disappear because they have melted or otherwise implode, or because they have gone out of sight behind another object. Ecologically, these are two distinct events and are perceived as such (see Chapter 11). 

Finally, motion in environments is complex; not the abstract points and forces of Newton, but complex changes of structure. 

Summary

Gibson is laying out the ecological scale of analysis. This is not more or less true than the scales of physics analysis, but it is the correct scale for explaining perception and behaving because those thing happen at that scale. Animals inhabit environments, not physical worlds.