September 1968
Wave-Train Information and Wave-Front Information in Sound and Light,
with a Note on Ecological Optics
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 References to these essays must cite them explicitly as unpublished manuscripts.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.
The train of the sound waves at a point in the air is specific to the mechanical disturbance at the source and thus carries information about the vibratory event; the concentric wave fronts are specific to the direction of the source. These two kinds of information make it possible for as animal in the vibration field to identify the kind of event at the source and to localize it (Senses Considered, p. 81 ff.).
In optics, as in acoustics, it has been assumed that a train of light waves at a point of observation is specific to the radiating light source, the spectrum corresponding to the atoms of the substance and the transients to the temporal fluctuations (or the onset and cessation) of the electromagnetic events; likewise the wave fronts specify the direction of the source. And indeed this is the way we see the stars at night. Auditory perception is said to be the of sound-sources by the transmission of acoustic energy, and visual perception is thought to be the seeing of light sources by the transmission of radiant energy. The ambient light converging to a point of observation has to come from luminous things; otherwise there is only darkness. This formula works very well for an environment of stars, lamps, fires, lightning, and fireflies. But visual perception is usually perception of an illuminated environment of reflecting surfaces, mostly opaque, of varying composition, and facing in various directions, with the earth below. What is the information for the perception of this kind of environment? If one looks to physical optics for an answer the assumption is that each particle of a diffusing illuminated surface is excited to radiate energy, each having concentric wave fronts and a continuous wave train (e.g., Ch. 1 in Graham et al., Vision and Visual Perception or September issue of Scientific American). The approach is necessarily elementaristic and the implication is that, in the last analysis, what one sees is luminous particles. Some physicists have made this inference. We can, however, appeal to ecological optics for an answer to the question. Its basic assumption is that in a terrestrial medium light reverberates and almost instantly reaches a steady state. Illumination “fills” the space, and one can then neglect the transmission of energy, the wave trains, and the wave fronts. The properties of the reflecting surfaces, together with the faces and facets of the these surfaces, structure the ambient optic array at each possible point of observation in the medium. The implication is that we see surfaces and objects, and that we do so by putting our eyes at different points of observation. But note that ecological optics does not apply to the seeing of the stars in the night sky. The light from the stars must be treated by relativistic physics. The light from the earth need not be.
Thus, in both physical acoustics and physical optics we are concerned with the wave-train information that specifies the cause of the waves and the wave-front information that specifies the direction from which the waves come. An observer can thus detect the nature and the direction of what we loosely call a “sound” and a “light,” that is, of a mechanical vibratory event and of an atomic vibratory event. But if we are concerned with the detecting of illuminated objects in an illuminated environment as distinguished from luminous objects in darkness we must go beyond the information carried by waves and consider the information contained in an arrangement, that is, in the ambient array. The invariants in the structure of this array correspond to the reflecting surfaces of the environment, not to the excited atoms of these surfaces.
In a medium of air, as distinguished from outer space, the endless echoing of electromagnetic energy produces a stationary network of perspective projections. The rules of information transmission by energy no longer apply to this equilibrium state of illumination. The flow of environmental events is simultaneous at all station points in the medium. The speed of light is irrelevant. The space-filling network does not consist of light rays but of sight-lines. Illumination creates perspective, which is the same when the amount of illumination changes. Opaque surfaces occlude other surfaces that do not face the station point, but information does not have to be carried by light from place to place; it exists in all places at once. The observer need only move from place to place in order to perceive all the surfaces of the world.
Sound and electromagnetic radiation can be used by man to carry signals, or coded information, and the laws of this communication are becoming well understood. It is tempting to think of auditory and visual “stimuli” from the environment as of they were signals. Despite the germ of truth in this metaphor it is a mistake to do so, for the environment does not send out messages, and the activity of perception is not one of learning a code or breaking a code. The information about vibratory events, either mechanical or atomic, in the flowing sea of energy surrounding an organism is not transmitted in code. Moreover, and this is what needs to be emphasized, the information that everywhere exists simultaneously in a closed system of reverberating light is still more radically different. The perspectives of reflecting surfaces do not “come” to an eye; they are available at all points where an eye might be stationed, and the perspective transformations are available for all possible paths of locomotion an observer may take. The eye therefore explores; it does not merely receive.
Note that sound in this respect is different from sight, in that it does not result in a state analogous to perspective projection.
Ecological optics provides from the outset for what has been called the “subjective component” in perception, whereas physical optics does not. The points of which ecological space consists are abstract mathematical fictions. Ecological space is bounded by reflecting surfaces, whereas physical space is unbounded, empty, and transparent. Ecological optics does not imply “consciousness” at the outset but it does begin with the concept of organism, whereas physical optics does not.
Ecological optics is in no sense an alternative to physical optics, or a competitor with it. What it does is provide a better foundation for the psychophysics of perception than physical optics does. Physical optics is still the foundation for the psychophysics of the “light sense” (physiological optics) and it is necessary for the prediction of the stars visible to an ideal observer, or the distance at which the signal from a lighthouse can be detected, or the level of energy at which the eye is dazzled and that at which the rods take over from the cones. But ecological optics promises to yield an explanation of perception, of the direct contact observers seem to have with the world, and of the immediate awareness they have of existing in the world.