¹ This article was read, with minor changes, before the Cosmos Club of the University of California, at Berkeley, April 1936.

It is a short outline of some of the more general considerations made in the author's “Wahrnehmung und Gegenstandswelt—Grundlegung einer Psychologie vom Gegenstand her“ (Leipzig, Deuticke, 1934).

A more extended presentation of the connected experimental research will be found in a series of studies edited by the present writer under the general title “Untersuchungen über Wahrnehmungsgegenstände.“ Until now the following titles have appeared:

I. E. Brunswik, Die Zugänglichkeit von Gegenständen für die Wahrnehmung und ihre quantitative Bestimmung, Archiv für die ges. Psychol., 1933, 88, 377-418.

II. B. E. Holaday, Die Grössenkonstanz der Sehdinge bei Variation der inneren und äusseren Wahrnehmungsbedingungen, ibid., 419-486.

III. K. Eissler, Die Gestaltkonstanz der Sehdinge, ibid., 487-550.

IV. S. Klimpfinger, Ueber den Einfluss von intentionaler Einstellung und Uebung auf die Gestaltkonstanz, ibid., 551-598.

V. S. Klimpfinger, Die Entwicklung der Gestaltkonstanz vom Kind zum Erwachsenen, ibid., 599-628.

VI. T. Izzet, Gewicht und Dichte als Gegenstaende der Wahrnehmung, Archiv f. d. ges. Psychol., 1934, 91, 305-318.

VII. K. v. Fieandt, Dressurversuche an der Farbenwahrnehmung, Archiv f. d. ges. Psychol., 1936, 97, 1-30.

In English, a very brief sketch emphasizing the empirical aspects has been given under the title “Psychology in Terms of Objects” (Proceedings, Anniversary, University of Southern California, Los Angeles, 1936, 122-126). Furthermore, some points are brought out in a joint article of E. C. Tolman and the present writer, “The organism and the causal texture of the environment,” Psychol. Rev., 1935, 42, 43-77.

² It is essential to connect, in any case, the use of the term “constancy,” as it became customary in psychology, with a further clear conceptual determination of the type of abstract (measured or computed) type of physical property or “object” (Gegenstand) for which the “constant” coupling to a certain type of reaction is successfully established.

Gestalt psychology, in its successful fight against the “constancy-hypothesis“ (which has been an unrecognized premise of the old associationism), emphasized that there is no one-to-one correlation between retinal (“proximal”) stimulation-elements and perceptual reactions. The same negative statement of an absence of a “retinal-size constancy” was the topic of our first introductory paragraph. This is far from being in contradiction to the positive statement of a body-size constancy. In fact, abandonment of the “constancy-hypothesis” even cleans away for finding that other physical something—body-size—which does stay in a (fairly good) one-to-one relationship to the reaction and thus easily presenting the environmental terms which gives proper sense and meaning to the establishment of the reaction.

³ The term “independent” was used by E. C. Tolman (“Psychology versus Immediate Experience,” Philos. of Science, 1935, 2, 356-380) in order to indicate types of objects whose definition does not include a reference to a relationship of an environmental entity to the organism.

⁴ The term “proximal stimulus” was used by K. Koffka (“Principles of Gestalt Psychology,” New York, Harcourt Brace, 1935), in order to discriminate the stimulating event arriving at the sense organ from the “distant” body. For the sake of brevity we use the term “stimulus” always in the sense of proximal stimulus or even of primary physiological excitation, whereas the remote manipulable cause of the stimulation will be called “body” or “body property.” Both, stimulus as well as body property, are types of physical entities or “objects.”

⁵ Cf. K. Bühler, Die Erscheinungsweisen der Farben, Jena, Fischer, 1922, and E. Brunswik und L. Kardos, Das Duplizitätsprinzip in der Theorie der Farbenwahrnehmung, Zeitschr. F. Psychol., 1929, 111, 307-320.

⁶ Cf. E. C. Tolman, Purposive Behavior in Animals and Men, New York, Century, 1932.

⁷ According to their objective probability cues may be graded on a scale of “reliability.” No cue, of course, is perfectly reliable, i.e. inimitable and therefore univocal in its indicative value. One may think of a stereoscope counterfeiting a cue of such high reliability, as binocular disparity is.

⁸ It has been shown by Holaday (l.c.) that binocular disparity is able by itself to sustain a high degree of size-constancy, whereas its elimination (by closing one eye) becomes almost ineffective in cases in which the full normal variety of other possible distance-cues is available. A similar and very perfect “substitutability” of tactual and visual cues for the volume of a body has been found by Izzet (l.c.) in experiments on weight-constancy.

The cognitive concept of the “or-collection” building up a “cue-family” has its parallel, on the action-side, in the concept of the “habit-family-hierarchy” of C. L. Hull (“The concept of the habit-family hierarchy and maze learning,” Psychol. Rev., 1934, 41, 33-54). The term “hierarchy” is—in both cases—apt to indicate the differences in “goodness” or “reliability” of cues or means respectively. W. S. Hunter (“The psychological study of behavior,” Psychol. Rev., 1932, 39, 1-24) made the fact of “vicarious functioning” a central point in the distinction between psychology and physiology.

⁹ The only question remaining open for a physical explanation is as to how natural or artificial tools (or organismic “institutions”) like collecting lenses, or the even more complex organismic systems functioning in a similar integrative way, might have developed at all. This general genetic question of living organization belongs to the field of theoretical biology and the psychologist does not need to be concerned with it, since his problems are centered more around actual achievement and functioning.

¹⁰ Multiple mediation—i.e. checking as much as possible all variations in the situational circumstances—is one way of rendering “far-reaching” couplings undisturbed by interfering conditions of causation. Another practice would be that of keeping all conditions of observation actually as constant and insulated as possible and let no uncontrolled “lateral” causal chains interfere. This latter would be the procedure usually followed by physical as well as by the traditional psychophysical experimentation, and also by man-made machinery, as e.g. electrical transmission of a message. Here the univocality remains obvious along the whole chain of mediation. This second procedure is in general the more reliable one, but also the more round-about way as far as the single case is concerned.

¹¹ Our emphasis upon the “object attained” may be considered as a kind of long-sectional figure-ground treatment of psychological research following the stimulating causal chains backwards in search for their actual “meaning” (see above), i.e. for those types of object within the environmental system upon which the reaction became focalized. The relationship of the “object attained” to “mediation” is a complement to the more cross-sectional relationship between figure and ground as emphasized in Gestalt psychology. Ground as well as mediation are both characterized as being present as stimuli but remaining lost amidst the “things” which they are “framing.” Cf. also F. Heider, “Ding und Medium,” Symposium, 1927, 1.

¹² Cf. also H. Klüver, “Behavior Mechanisms in Monkeys,” Chicago 1933. Some of the work of Lashley follows the same principle.

¹³ Even G. F. Skinner (“The Concept of Reflex in the Description of Behavior,” J. Gen. Psychol., 1931, 5, 427-458), though maintaining that “psychologists had better give up the nervous system and confine their attention to the end-terms, does apologize for doing so in pointing toward the greater immediacy of observation of these end-terms and the reduced temptation for insecure speculation. We do not find it a matter of embarrassment, but rather one of positive emphasis, to go even further in psychology and restrain—as far as at least one term of the correlations in question is concerned—completely from the organismic events as such in favor of the initial (or terminal) environmental limits or focal systems as connected with one of the organism's activities.

¹⁴ Cf. M. Schlick, Allgemeine Erkenntnislehre, Berlin, Springer, 1925.

¹⁵ As it has been pointed out above, coincidence may be used sometimes as a help to attain explicitly in perception a certain kind of object which otherwise would not be represented by a conscious dictum, e.g. the retinal stimulus-sizes.

¹⁶ The author, l.c., tried to show that even the mere naming of the apparent equality in question in conceptual terms—i.e. by indicating whether it is an equality of apparent size, or color, or weight, or density, etc.—is, in principle, omitted from “psychology in terms of objects.”

Not only the actual attainment but even the types of intentional effort or “attitude” toward attaining certain “intended” objects may be disclosed by objective methods (in fact, by analyzing the statistical distribution of judgments with regard to the number of modes).

¹⁷ I agree with H. L. Hollingworth (“Experimental studies in judgment,” Arch. Psychol., 1913, 29), who proposes, in accordance with Wells, to use variability as an objective, quantitative index of “subjectivity” of judgment. (I would not, however, as Hollingworth docs, rely upon a ratio of personal vs. group consistency, but keep the definition free from its “social” element and emphasize the more general point of non-attainment, or variable attainment, of a type of object.)

¹⁸ F. Brentano, Psychologie vom empirischen Standpunkt, Leipzig, 1874.

¹⁹ Thus the new concept of intentionality seems to be closely related to what Tolman meant in defining in objective terms the “purposive” character of behavior.

²⁰ For a concrete example of the difference between the phenomenological and the object-critical method of object-finding see page 246.

²¹ B. Russell, Introduction to Mathematical Philosophy, New York, Macmillan, 1919.

²² The intentional relation thereby might be defined as positive either in the direction from the object as initial term to the reaction as terminal limit—following the direction of the causal chains—, or in the opposite direction, from the reaction to the object as end-term. This latter would be more in accordance with the character of the perceptual reaction as a preparation to overt action towards the body in question as a manipulable means-object, and would also follow the direction of the immediately experienced “meaning” or conscious intentionality of the Brentano-type.

²³ This scale could equally well also have its ends mutually exchanged.

²⁴ Cf. E. Brunswik, Zur Entwicklung der Albedowahrnehmung, Zeitschr.f. Psychol., 1928, 109, 40-115, and l.c. The degree of failure with respect to body-size is indicated by deviations either below or above 100.—Cf. also R. H. Thouless, Phenomenal Regression to the “Real” Object, I, Brit. J. Psychol., 1931, 21, 339-359.

²⁵ In a quite analogous way, in experiments with figures, interferences have been found between area and length of the edges, etc. (see below).

Some of these interferences, especially those occurring in the field of the original type of constancy research (as size-constancy, etc.), may be satisfactorily accounted for in terms of the causal texture of the environment in its relationships to the organism. Perceptual cues for remote objects always remain ambiguous, and therefore perception is forced to co-include in its basis for reaction a large number of more or less indirect and unreliable cues which in some cases may stay in but a very low correlation to the type of object which they are admitted to indicate. (This would hold, e.g., for retinal size per se, when taken as indicating body size. Indeed, the characteristic odd admixtures round in the results of size-constancy experiments may be represented as functions of the product retinal size times indicated distance [i.e. body-size], on the one hand, and of retinal size, per se, on the other.) As has been pointed out in the author's article “Psychology in Terms of Objects” (l.c.), the effect of such a general way of functioning is a decrease in the probability of exceedingly large perceptual errors.

For some other types of perceptual compromises an explanation in terms of the technicalities of the sense-organs at the command of the perceptual system might be found, by studying physiologically their structure and the way of their functioning. This seems to be especially true for pitch and loudness which both turned out to be joint functions frequency and energy of the sound. Cf. S. S. Stevens, “The relation of pitch to intensity,” J. Acoust. Soc. Amer., 1935, 6, 150-154, and E. G. Boring, “The Relation of the Attributes of Sensation to the Dimensions of the Stimulus,” Philos, of Science, 1935, 2, 236-245.

²⁶ Some relationships could also be found with the manivalued logics, in which the absolute alternative of true and false is given up in favor of a continuous scale between these two cases as mere extremes.

²⁷ It could be objected that a presentation of psychological results in terms of functional dependence of the reaction on various stimulus-factors would be as short and less confusing than to talk in terms of “in-between objects.” The reason why we believe it more profitable, however, to make this terminological distinction is the following. Strict functional dependence would have to be expressed in terms of proximal stimulation. We would prefer, instead, to express the (functional) relationship in terms of the remote significate, instead of using the signifying stimulus, in all cases in which the operational criterion of meaning given in section I is fulfilled. This would necessarily introduce a certain ambiguity. But it may be still considered more favorable to do so from the standpoint of illuminating the essential cue-rôle of the stimuli in establishing the organism's ability to master its environment.

Let us consider an example. For a most direct functional analysis apparent size would be simply one or another function of retinal size and distance-cues, and the case of a perfect constancy would not be especially emphasized as against cases of incomplete achievement. In terms of “objects attained,” however, the response would appear to be a “function” of body-size alone (i.e., in strict functional terms, of retinal size times indicated distance) in the case of a perfect achievement, and of body size and projective size, per se, in the case of an incomplete achievement. In the latter case retinal size would enter the function twice in different rôles, once as a constituent of body-size, then “per se” (see the note above).

A further complication for a functional analysis in terms of retinal stimulation seems to be that all equi-potential members of a cue-family (see above), e.g. all stimuli and stimulus-configurations indicating distance, would have to be enumerated explicitly in terms of intrinsic properties of their own instead of simply being comprehensible in terms of their common “significate,” as “distance-cues.” By using the “object”-terminology the particular kind of environmental-functional direction of interest in stating the dependences found could be indicated at once.

²⁸ Since the ways of mediation will always determine the achievement, the highly abstracted type of object-critical analysis as outlined above would lead, ultimately, to a statement of all psychologically relevant types of “how”-problems and -findings in terms of “what,” i.e. of objects attained.

²⁹ D. Katz, The World of Color, London, P. Kegan, 1935.

³⁰ As our examples show, we are dealing here with the “psycho-physical,” not with the “psycho-physiological” relationship.

³¹ It has been found that even ambiguous cues (i.e. those of a reliability of their actual indicative value lower than 1) will be conditioned, though the response seems to retain a higher degree of tentativeness (reduced strength). In a recent unpublished study of the present author, conducted in order to throw further light upon this problem, rats were confronted with objectively ambiguous situations, possessing various degrees of probability of success or of punishment (“danger”). Even relatively small differences in probability were discriminated by these animals (whose functioning is, in fact, not very dissimilar to the relatively primitive cognitive system, called perception). As the writer hopes to be able to show later in detail, an evaluation of even some of the slightest correlations between possible cues (symptoms) and more remote traits seems to be true for perception, especially in making accessible the character-traits of other persons in social intercourse. Reactions to mere probability have also been found by Thorndike in some of his recent learning experiments.

³² See the discussion of the problem by V. F. Lenzen, “The interaction of subject and object in observation,” Second internat, congress for the unity of science, Kopenhagen 1936.