L.Kleine-Horst: "Empiristic theory of visual gestalt perception. Hierarchy and interactions of visual functions" (ETVG). Part 1, III
The functional hierarchy of contour factors
1. The three-sphere model of contour perception.
Many scientists working on perception assume that the perception of a figure is a "primitive unity" (Hebb 1949), an elementary unity. Elementary because it is supposed to be a human ability completely functional at birth, a "unity" because it is supposed to be simple, i.e. not open to further analysis. I now intend to explain how visual gestalt perception can be understood as the phenomenal correlate of a hierarchy of perceptual functions. Part of these functions serve the perception of a "figure in its outfield", a percept of the type "moon in the sky". A "figure in its outfield" is perceived by a five-level hierarchy consisting of six gestalt factors. It is therefore not a "primitive unity", but a correlate of a relatively complex functional system, easy to analyze. Moreover, this functional system of figure perception is not innate, but has to be acquired during the individual's lifetime (Part 4).
In this Part of the book I shall not yet describe the perception of a full figure, but only the perception of a figure-contour. Fig. 1-4 depicts the hierarchy of the gestalt factors creating the experience of "figure-contour": according to the three-sphere theory of visual gestalt perception there is a set correlation between the material sensory entitiesY and the functional sensory entities Z. But there is no set correlation between these physical functions and the visual experience. The material and functional sensory entities are not correlated with sensory phenomena; one cannot see anything with them, one cannot experience anything visually, there is no visual percept; since a phenomenal sphere does not exist in the bioevolution, in that the body was developed.
Such a constant relationship between (material) sensory stimulus and (phenomenal) "sensation" was, however, assumed by the early visual scientists, as mentioned above. Unfortunately, the problem of the relationship between matter and phenomenon was still not solved in the discussion between the "Elementarists" and the "Gestalt psychologists". The latter survived since the Elementarists died, and their constancy hypothesis with them. The ETVG proposes a solution to the problem by reviving the constancy hypothesis: instead of one "constancy" (between matter and phenomenon) the ETVG states two "constancies"; it postulates both a constant relationship ("correla- tion") between matter (VM) and "its" function (VF) and a constant relationship between function (PF) and "its" phenomenon (PC). As Fig. 1-4 shows, the functions being correlated to phenomena are, however, other functions than those being correlated to matter. These physical functional entities (Z) are dependent correlates of physical matter (Y), but the psychical functions ("gestalt factors") are independent; they have their own dependent correlates: the phenomenal "gestalt qualities". A further difference between the two sorts of functions: the (psychical) gestalt functions are located at the next higher evolutionary level above the physical functions.
Thus the ETVG postulates the following relationships at the physical evolutionary level:
- The luminance (Ym) does not have a subjectively perceived, experienced, phenomenally given (degree of) brightness as a dependent correlate, but a purely functionally determined "functional brightness" (Zm).
- The wavelength (also Ym) of light does not have a subjectively perceived color as a dependent correlate, but a purely functionally perceived "functional color" (also Zm)
- The locus (Yl) of the stimulated receptor does not have a subjectively perceived location within the total visual field as a dependent correlate, but a purely functionally determined "functional location" (Zl).
- The depth (Yd) of an object, i.e. the distance from the eye, does not have a subjectively perceived depth as a dependent correlate, but a non-experienced "functional depth" (Zd), on a certain functional location (Zl).
- The time (Yt) of stimulation does not have a subjectively pereceived time as a dependent correlate, but a purely "functional time" (Zt).
That degree of brightness is actually experienced, in that color the stimulus source actually appears, at what location (in that direction in space) it will be seen, and at that point in time it will be experienced, is on one hand determined by the sensory functional values that correlate with the material condition "photoreceptor and its absorbed light energy". On the other hand, the percept is codetermined by a number of additional functions P, D, G... , the aforementioned "gestalt functions" or "gestalt factors" (as opposed to "sensory functions"). These gestalt factors are ultimately triggered, activated, "actualized" by the sensory functions, i.e. by the functional correlates of the material sensory entities "sensory stimuli on photoreceptors". The gestalt functions do not correlate (as the sensory functions do) with matter, but with phenomena. To each gestalt factor correlates a certain perceptual experience, a certain "gestalt quality". The correlations are shown in Fig.1-4 as horizontal arrows; the "gestalt stimulus" that actualizes "its" gestalt factor is shown as a vertical arrow. The term "correlation" is used to signify a one-to-one correspondence. But note: in the psyche (the gestalt perception belongs to), evolutionary development happens in the functional sphere, not in the phenomenal sphere. It follows, that, since the gestalt quality is produced by its gestalt factor, the gestalt quality is a dependent correlate to its gestalt factor and not vice versa. When the actualization of the gestalt factor has ceased, the existence of its gestalt quality has ceased as well, but the gestalt factor itself still exists, although in an unactualized, i.e. non-phenomenalized, state.
The gestalt factors are hierarchically ordered. The hierarchy of the gestalt factors used by a perceiver to detect a figure-contour is depicted in Fig. 1-4. Contour perception is, in accordance with the concept shown above, a step by step process. The sensory functions actualize the lowest gestalt factor Pml, that in turn actualizes Dm and Dl, that in turn actualize gestalt factor Gml. Certain functional relations at level Gml actualize factor Ll, that in turn actualizes Fl. In the process of actualizing a gestalt factor the correlated gestalt quality is introduced into the percept. Thus, the percept already in existence through the actualization of lower-level gestalt factors is enriched by the introduction of the next-higher gestalt quality.
Perceptual experience is holistic. The early Gestalt psychologists have pointed out the truth in this basic assumption time and time again. No exceptions have been found. Therefore, the Pml experience is already a holistic experience. Each additional gestalt quality can only enrich the previous percept; it cannot eliminate the gestalt qualities that have been previously introduced. Since Pml experience is already holistic, all experiences built on the Pml factor are holistic as well. An experience at Ll-level is also always a Gml experience and therefore a Dm, Dl, and Pml experience. These relationships are depicted in Fig.1-4 by nested ellipses. Pml creates Pml'. When the subsequent gestalt qualities Dm' and Dl' are added, a percept results that includes the quality Pml'. The percept at Gml-level includes the Pml'-Dm'-Dl' percepts etc. Y matters and Z functions, too, are hierarchically ordered, for the domain of visual perception in at least three levels.
A horizontal borderline was drawn in Fig.1-4 between levels Z and P for good reason: the part of the total system of perception above the line is completely different to the part below it. Sensory functions are of a completely different nature than gestalt functions; sensory functions correlate exclusively with matter, gestalt functions exclusively with phenomena. Two apparently different systems take part in visual perception. I shall describe, in more detail, the hierarchy of the visual psychic system depicted in Fig.1-4, that determines figure contour perception.
2. "Here is something bright" (Pml)
I understand brightness to be the quality common to all visual experience. The factor that mediates the basic visual experience of a "bright world", shall be denoted by the symbol "Pml". In the following text, "Pml" shall always apply to perception in the visual domain, since I only concern myself here with visual perception. I now assume that this factor exists within us, whether we see something at the moment or whether we are asleep. If I perceive something visually at this moment then I can explain this by saying that "the factor Pml has been actualized". If factor Pml is actualized, I experience Pml, i.e. I perceive something of the gestalt quality "Pml", I experience "something bright".
Therefore "Pml" has a double meaning: Pml is the symbol for the gestalt quality "something bright" as well as the symbol for the gestalt factor whose actualization results in the gestalt quality Pml. (As a gestalt quality I represented Pml in Fig. 1-4 as "Pml' ".) In addition, Pml also has a third meaning; it is the specific gestalt stimulus for the actualization of the gestalt factor Pml. The reader might think that this triple meaning of one symbolic term will result in confusion; as it turns out, however, the opposite is the case. Much confusion is thus avoided, as this is a language of correlations which, among other things, serves the purpose of describing the connections of functional and phenomenal entities. Each perceptual phenomenon (gestalt quality) correlates strictly with a certain functional entity (gestalt function, gestalt factor). Each of these gestalt functions corresponds to a certain gestalt stimulus. It often (but not always) does not matter whether one speaks of the gestalt quality X, of the actualization of the gestalt factor X, or of the specific phenomenal effect of the gestalt stimulus X.
Once again: You aRE only ABLE TO experience through THE ACTUALIZATION OF the GESTALT FACTOR Pml. In the domain of vision, this means that you are able to see something at all, regardless of what you see. You cannot see anything without the actualization of this gestalt factor. You would not even see darkness, for darkness, too, is a visual phenomenon. You do not see anything without the Pml factor; your perceptual system is not "switched on" without the actualization of Pml; Pml is the "subjective light-switch". Without actualization of Pml one can only functionally perceive with the physical Z factors. It is indeed possible, however, that a gestalt factor is activated to a lesser degree than necessary to exceed the factor's actualization threshold. But this subthreshold activation, too, does not produce a gestalt quality.
It is not easy to imagine a perception one does not have under normal stimulus conditions. In "normal" everyday perception all gestalt factors are "switched on", are actualized, not just Pml. It is important, however, for your understanding of the hierarchical organization of visual perception to first imagine what you would experience if you perceived solely with the aid of the gestalt factor Pml, and next to imagine what you would perceive if only the gestalt factors Pml, Dm, and Dl were actualized, and then, what the effect on the visual experience would be if the gestalt factor Gml were added to those previously mentioned, and so on, proceding upwards through the functional hierarchy. In this manner, a contour of a figure develops with increasing complexity before your inner eye in a process of a few steps Such a progression of percepts resulting in a full-faceted percept is called "actual genesis" or "microgenesis". Since this is a thought experiment and not an actual perceptual process, I call this experiment "virtual microgenesis".
Now look at Fig. 1-5 and disregard that the perceived object is a meaning carrier, specifically the meaning "moon in the sky". Try not to look at the form of the object, its curvedness or roundness, how "circlelike" it is. Also ignore that it is a "figure", a field bordered off by a contour against a square outfield. You cannot perceive all this with only Pml. Try to perceive it simply as "something bright". I now have to point out a second feature of the object you are able to perceive with Pml alone; with Pml you not only see something bright (m), you also see it "at a location" (l). You experience: "here is something bright".
You cannot immediately perceive the whole moon as "here is something bright", because each receptor is stimulated individually at first. So assume that any given miniscule area of the moon transmits the impression "here is something bright" to you. The areas in the dark sky, too, send light rays into your eye. Therefore you see "here is something bright" as well when looking at the night sky.
Figure 1-5. "Moon in the sky" Figure 1-6. Receptive retinal Pml areas
You do not yet say, however, "here is something dark", since something dark, too, possesses the quality "something bright" (Pml), whether there be little or much of it. Also please do not say, "here is something bright and there is something bright", because that would sound as if you could distinguish between "here" and "there". You cannot distinguish anything with the Pml factor, you can perceive only "here is something bright"; you can neither distinguish the brightness of the moon from the darkness of the sky nor the "here" from the "there". Of course, it is a fact that we, as humans, can distinguish degrees of brightness from each other, but we cannot do it with the gestalt factor Pml! Pml can detect only "brightness" and that this brightness is "here", and this as a whole and undivided percept: "here is something bright".
So how do we perceive the "moon in the sky", if not as a bright blob on a dark background? How bright is the moon? And where is the brightness? Since one cannot perceive two different degrees of brightness simultaneously, one may experience only a single brightness. Since one cannot experience two different locations simultaneously, one may perceive only a single one. I assume that, at the P-level, the "moon in the sky" is experienced with an "average" brightness of the moon and sky together, and this brightness is always "here", i.e. at the place that I am looking at. Where else should it be? This must also hold true in the case of the moonlight impinging on the retina peripherally. A prerequisite for this interpre- tation, however, is the postulation of a computational system.
Fig. 1-6 depicts the three possible cases for seeing the moon in the sky with factor Pml. Since we humans are able to perceive even extremely small light stimuli, the smallest retinal areas that carry out the reception of these light stimuli must be involved. This could be a single receptor or a small group of receptors; here it does not matter since we do not have to concern ourselves with the material basis when using a "functionological" way of looking at things. For the everyday-perception of a minimal light stimulus as a "point of light", however, we require the actualization of our whole visual system. When perceiving exclusively with the Pml factor, even the smallest light stimulus appears (if at all) as "Ganzfeld" brightness, i.e. as brightness of the entire visual field.
Let us assume that we have a restricted visual field such as in the small circular area in Fig. 1-6. When looking at area (c) we would have an intense experience of brightness. Looking at area (a), however, we would experience only a very weak "here is something bright". If we exclusively perceived the light rays impinging on area (b) we would see an "intermediate" brightness because the high luminance of the light rays from the moon and the low luminance of the light rays from the sky would form an average value. This is because it is not possible to experience both degrees of brightness solely utilizing the factor Pml. And, to reiterate that it does not matter whether the eye is looking at (a), (b) or (c); it is constantly transmitting solely the percept "here is something bright" by means of Pml. It cannot perceive that: (a) is darker than (c), and (c) is brighter than (a) and (b), although the brightness perceived by looking at (a) is not as intense as that perceived looking at (b) or even at (c).
"Brightness difference (Dm) and location difference (Dl)"
back to contents