June 1979
A Note on Substances, Surfaces, Places, Objects, Events
J. J. Gibson, Cornell University
The World Wide Web distribution of James Gibson’s “Purple Perils” is for scholarly use with the understanding that Gibson did not intend them for publication. References to these essays must cite them explicitly as unpublished manuscripts. Copies may be circulated if this statement is included on each copy.
In the Ecological Approach to Visual Perception I propose what animals perceive are the substances, surfaces, places, objects, and the events of the environment instead of objects in space. What animals discriminate are the meaningful properties of substances, surfaces, etc. instead of the primary and secondary qualities of physical objects. What they move around in is the medium instead of space. What they see is layout, the dihedrals and curves that surfaces make to one another, instead of depth in space.
Substances vary in substantiality, and a substance is not to be confused with matter. Surfaces are the interfaces between substances and the medium, not geometrical planes. Places are semi-enclosures in a cluttered environment of larger and larger places, not loci relative to a coordinate system. Objects are substances with a topologically closed surface detached from the substratum, not the vague entities of physics or the latter half of the subject-object dichotomy in philosophy. (A substance with a nearly closed surface attached to the ground is also an object. This means that trees, houses, rocks, tools, dogs, and humans are objects, the first two being attached, the last four being mobile without breaking the surface, and the last two being animate and mobile objects.) Events are changes in the environment of any sort, some reversible and some irreversible, not just the rigid motions and rotations of bodies along and around the axes of space.
These realities constitute the environment of animals. They comprise the niche for each species of animal, and the habitat of each individual animal. They are appropriate for the study of what animals perceive, how they behave, how they learn, and what they know, that is, for psychology. The realities of physics (particles, atoms, elements, compounds, the motions of particles and the radiation of energy) are not appropriate for the study of perception and behavior.
Ever since Descartes, human psychology has been held back by the doctrine that what we have to perceive is the “physical” world that is described by physics. I am suggesting that what we have to perceive and cope with is the world considered as the environment.
Note that substances, surface planes, and events are nested. They are not denumerable, which is to say they cannot be counted like discrete objects. There is no definite number of them. A surface, for example, can always be incorporated in a superordinate surface. A place is not separated from an adjacent place by a sharp boundary. This seems to imply that they cannot be classified, categorized, or grouped (only discrete objects can be grouped) and that hence the mathematical theory of sets does not apply to them. If true, this is alarming to the psychologist who assumes that, in order to be scientific, he must compute or measure as the physicist is able to do (cf. memo on surface perception, May 1979). But perhaps an ecological psychology would gain in relevance what it lost in mathematical simplicity. Moreover it might be able to solve the problem of meaning and value in a simple way, which metric psychology has been unable to do.
Substances, surfaces, places, objects, and events provide utilities and dangers. They have positive and negative affordances. If there is information in the normally obtainable flux of stimulation, as I have tried to show, the invariant meanings can be learned by extracting them instead of constructing them. Thus explanation is much more parsimonious than traditional explanations based on receptors, physical stimuli, sensations, psychophysics, and the epistemological paradox of assuming that one gets knowledge of the world from these inputs because one already has knowledge of the world.
The traditional psychology of stimuli, exteroceptors, proprioceptors, interoceptors, their sensations, and the “scientific nerve energies” of these sensations has also not been able to solve the problem of self-awareness. The complexity of the problem is forbidding. But the hypothesis that the body of the observer is specified in the flux of stimulation obtained by a perceptual system (although not by the inputs of a sensory nerve) is actually quite simple, and so is the theory of the facing surfaces viewed not from here. The assumption that both locomotion and status of the self relative to the surface layout are specified by change and non-change of the ambient array is elegantly simple.
Substance and surface are taken relative to an observer; they are of no interest to physics. Place is part of the habitat where an observer can stand; the standpoint is something that physics tries to eliminate although, in the last analysis, it fails to do so. An object has affordances for an observer; these supposedly subjective values do not enter into physical science. An event at the terrestrial level as distinguished from an atomic or astronomic event shows its meaning for the observer in the transients and sequential structure; orbits, spins, states, and purely relative velocities do not show their meanings.
Behavior is motivated by substances, surfaces, planes, objects, and events. They have valences in the terminology of Lewin. (It is also motivated by hunger, sex, pain, and the need for shelter, of course, but these two facts are complementary, not discrepant.) A substances that is nutritive invites eating, water invites drinking, pouring, or washing (but not walking on), clay invites molding, and dry wood affords fire-making.
A surface support invites sitting, standing, walking, or running; a surface that is a barrier to locomotion demands a halt; a double surface that is flexible affords wearing; a warm, soft, suitably shaped, animate surface invites caressing.
A place that is enclosed affords getting out of the rain, a place that is hidden and safe affords sleeping, a place where prey is found allows food-getting but a place where predators lurk affords danger; a grocery store also affords food-getting but a six-lane highway with trucks is as bad as a place with saber-toothed tigers.
An object that has a handle can be grasped and used as a tool, an extension of the hand, and the forms of manipulatory behavior are endless. Every tool has a corresponding action sequence, the perception and the movement running concurrently: hammers, pliers, awls, knives, and so on.
An event like fire affords being warmed and also being burned, so that locomotion and manipulation are adjusted to it. The fire invites approach up to a certain limit and then repels; the limit can be both seen and felt. Behavior of this sort is controlled, but it does not seem to be reducible to definite responses to discrete stimuli. The popular phrase for it is that one perceives the “margin of safety”.
According to this formula, behavior consists primarily of acts that take advantage of the existing substances, surfaces, places, objects, and events of the environment while avoiding painful encounters with them. Another kind of behavior, most fully developed in our own species, consists of acts that change the substances, surfaces, places, objects, and events of the environment by manipulation and manufacture. Fundamental to both is the kind of behavior that obtains stimulation, orienting and adjusting the perceptual organs so as to extract information about the substances, surfaces, planes, objects, and events of the environment.