April 1965
Two Different Usages of the Term Information in the Study of Perception and Discrimination
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 References to these essays must cite them explicitly as unpublished manuscripts.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.
When one speaks of the “information” in an array of stimulation or in flow of stimulation (or in a flowing array of stimulation), one can mean either of two things. One can refer to information about the sources of stimulation in the environment (distal objects or events), or one can refer to information considered only as differential stimulation (proximal stimuli). The former is relevant to ordinary perception; the latter to discriminal sensitivity. When light does not make a differentiated array (limitless white fog) and when sound does not make a differentiated flow (endless white noise) there is, of course, no information in either sense of the term, and consequently information considered as discriminable stimulation is prerequisite to information considered as ecological stimulation. The study of the dimensions of sensitivity of an observer can be carried out without any concern for what the stimuli might mean to him, or (on the other hand) the study of “useful dimensions of sensitivity” (Amer. Psychol., 1963) can be pursued, in which case the stimuli chosen must be typical of those in the observer’s environment. Contrast these two meanings Of “information.” The structure of stimulation (either with respect to adjacent order or with respect to sequential order) can be conceived as pure structure (structure as such) or as representative structure (specifying something beyond). In the former case there is specification as mere distinctiveness or difference (the stimulus being different from what it is not, but might have been) while in the latter case, there is specification of an object or event.
Information as mere otherness is quantifiable in the way Shannon discovered, but it is not concerned with the meaning of the input. Information that specifies its source carries external objective meaning, but it does not seem to be quantifiable, at least by “information theory”.
Information as controlled laboratory stimulation permits the study of perception in an operational manner, as discrimination. Stimulus information with ecological validity enables us to study perception as knowledge, as true or false, as being in touch with reality, as being adaptive or maladaptive.
Information in the first sense is the reciprocal of uncertainty (Garner) while information in the second sense is the reciprocal of ignorance. What else can be said about it?
Hence there are different kinds of meaning (or meanings of meaning):
1. Meaning of natural stimuli and their replications: The sound of a waterfall (splashing) “means” a waterfall (also the reproduction).
- – The sight of a waterfall “means” a waterfall (also the representation)
- – The odor of an individual (or species) “means” the individual (or species)
- – The feel of wetness on the skin “means” water.
2. Meaning of conventional (coded) signals:
- – The sound of the word for waterfall (in any language) “means” a waterfall
- – The written word “waterfall” (in any language) “means” a waterfall -A combination of auditory or visual signals of Morse code “means” a set of graphemes that mean a waterfall.
- – The gesture in a sign language for waterfall “means” waterfall
3. Aesthetic meanings (qualities and values):
- – The different sounds of music have different “meanings”
- – The different structures of painting have different “meanings” (Hence, “nonsense” forms have “meaning.” )
- – Different perfumes have different “meanings”
4. Subjective meaning: (still a different kind, i.e., different from, any of the above)
- – The “meaning” of a painting lies in the beholder
- – The “meaning” of a stimulus depends on the past experience of the O
- – The “meaning” of the object consists of the responses one makes to it or the associations aroused by it
On the above analysis, the first and third kinds of meaning are learned in the sense of being differentiated, or detected. Only the second kind is learned in the sense of learning an association or a connection as being a fact of the social world, as external fact (the code), the association is something that is discovered in the course of repeated experience, not something that is formed within the individual. The theory that all learning consists of the formation of associations within an organism(as distinguished from discovery of associations outside the organism) has led many psychologists to take it for granted that all meaning is subjective. But on the above analysis stimuli have objective meanings of several sorts and experiences can have subjective meanings that do or do not correspond to the objective ones.