L.Kleine-Horst: Empiristic theory of visual gestalt perception. Hierarchy and interactions of visual functions. (ETVG), Part 10, II
Comparing the ETVG with Berlin gestalt psychology
in greater detail
As I already mentioned in Part 0, the five Berlin gestaltists who received my theory seem not to have understood anything. In order to find out the probable reasons, the criticisms of B1, B2, and B3 who responded to the 1961 manuscript, will be reported and analysed here, which at the same time supplements the comparison between the ETVG and the Berlin gestalt theory already made in Part 0. (B4 and B5 received Parts 1 and 2 of this book. .
B3 wrote to me in 1963:
"The access to your problems and your procedure is probably made difficult for me because I think the arguments that take perceptual psychological facts as starting points to be already a violation of 'gestalt' psychological principles. In my opinion, the lowest unity, from which one can start an analysis of human nature, is the act itself, not the perception. I asked a colleague, who is less weighed down with my prejudices, to go through your manuscript. He shares my view that your work is without doubt an interesting attempt. We are, however, obviously ...... incorrigible empirists and would propose less forced generalizations."
I did not intend to apply "gestalt psychological principles" (with which only those of the Berlin gestalt school can be meant ); instead, I wanted to propose new principles, moreover, even at the beginning of my 1961 manuscript I did relate perception to act, in a certain sense, as I showed that
"the relationships existing between triggering stimuli and the stimuli-triggered instinctive movement of an animal are the same as those existing between the stimulus and the stimulus-caused gestalt perception of a human" (p.4).
But the last phrase of B3 has often given me pleasure: It was always through exceptionally forced generalization that I was able to solve an exceptionally difficult problem.
1. "Constancy hypothesis": none, one, or two?
One of the strongest statements of the early gestaltists was that the elementarists' "constancy hypothesis", i.e. the assumption of a set ("constant") relationship between sensory stimulus and experience ("sensation"), is wrong. B1 wrote in respect to the 1961 manuscript (probably without re-reading):
"You once say that the gestalt psychologists initially assumed the constancy hypothesis, and were then surprised each time the perceptions did not correspond to this assumption",
and stated that the gestaltists would, instead, from the beginning have raised an objection to the constancy hypothesis.
This was, however, not what I had said; I quote:
"One cannot simply reject the constancy hypothesis, as the entire gestalt psychology is f o u n d e d on it. Gestalt perception consists in the d e v i a t i o n of an a c t u a l 'sensational configuration' from that which is to be e x p e c t e d according to the constancy hypothesis. A description of a gestalt perception is based on the use of the constancy hypothesis as its r e f e r e n- c e s y s t e m. Astonishment on the occasion of the discovery of a gestalt phenomenon is possible only when one 'actually' had expected something else. The constancy hypothesis is implied with this 'actually'." (Kleine-Horst 1961, p.106., spaced as in the original).
For the gestaltists, "constancy hypothesis" is an emotive word, that easily leads to misunderstandings: B1 thought that I should have returned to an old, but flaw riddled hypothesis. B2 was subject to the same misunderstanding, when he in turn believed in a misunderstanding on my part:
"that in your opinion the constancy hypothesis must be reintroduced, is based on a misunderstanding; if one is surprised at the deviations from this hypothesis, as one must indeed be over and over again, this does not at all mean that one must hypothesize any inascertainable processes, since one has a very much more comparable object of comparison: the stimulus configurations on the retinas."
The emotive word led the two gestaltists to also overlook my contrary concept on the preceding page. I had even introduced, it's true, the constancy hypothesis into my own theory, but in quite a different form:
"The c o n s t a n c y h y p o t h e s i s is
to be established, its validity, however, restricted to the relationships between
sensory stimuli [today, stricly speaking: 'Y-matters'] and the hypothetical
functional sensory contents [today: 'Z-functions'].....
Since besides the sensory stimuli other factors cocondition perception, the constancy hypothesis is not valid for the relationships between stimulus and sensation, but only for the relationships between stimulus and functional sensory contents. It is these, to which, in the present theory, the effects also of the other perceptual factors are added. These effects of gestalt contents [today: 'gestalt factors'] and direction of attention consist in that they change the relationships between the functional sensory contents" (Kleine-Horst 1961, p.105).
As Figs 10-7 and 7-1 show, there are even two "constancy hypotheses" in the ETVG, one valid for the physical level (as just mentioned), and one for the psychic level, where constant relationships are established between the psychic functions (PF, gestalt factors) and the psychic consciousnesses (PC, gestalt qualities) which are the gestalt factors' dependent correlates that form a holistic experience.
2. Sensory and gestalt stimuli
As quoted, B2 proposed to compare the phenomenal configuration with the configuration of the primary sensory stimulus. B1, too, stressed this sort of stimulus as the only relevant stimulus concept to be used, and thus rejected the concept of "gestalt stimulus":
"Even though others have done so, I would not recommend speaking of 'gestalt stimuli'. There are indeed objective conditions on which configurations depend, these are, however, not 'stimuli' in the same sense as represented by optical and acoustical waves, i.e. establishing immediate physical effects on the respective sense organs. These stimuli impinge upon the nervous system as purely local causes. The gestalt conditions, however, are the abstract relationships one can find between such local effects, for example, that the spatial intervals between stimulus points are larger here and smaller there, the stimulations themselves stronger here and weaker there and so forth. When one or the other configuration is established in the nervous system, due to such abstract relationships, this is a matter of purely physiological interaction that depends on such abstract relationships. I would not designate such relationships as 'stimuli', because they assert themselves after local states have been established in the sense organ, between which there are then approximately the same abstract relationships."
Here, and in the literature, we find essential B-hypotheses with which we can compare the respective ETVG-hypotheses, albeit without claiming completeness.
B: The basis for forming configurated experiences are abstract relationships.
B: These abstract relationships that form the configurated experiences are
such like those between local physical stimulations of the retina.
(a) Abstract relationships between local physical stimulations of the retina can form, if they do at all, only the one experience: "here in front is something bright now" (Pmldt); and this is not an experience that could be designated a "configuration".
(b) It is particular abstract relationships between Pmldt experiences that condition further experiences: four sorts of "difference". And it is particular abstract relationships between "difference" experiences (or their underlying functions) that condition experiences of "homogeneity" and "inhomogeneity" and so forth. In this way, a hierarchy of 25 different abstract relations and relation-relationships (gestalt stimuli), founding ultimately on the sensory stimuli, are described. All of them contribute to experiences that can be designated "configur- ations". Thus, it makes sense to distinguish between "sensory stimulus" and "gestalt stimulus", and subsume both under the generic term "stimulus".
(c) The dualistic B-theory has no possibility to propose anything other than material ("sensory") stimuli, and material entities ("photoreceptors") the stimulation of which leads to phenomenal entities. In the trialistic ETVG, it is possible to propose non- material, i.e. functional, stimuli and non-material (functional) entities that when stimulated lead to phenomenal entities. Thus, the B-way leads "straight ahead" from the material sphere to the phenomenal sphere, but the ETVG-way leads from the material sphere first to the functional sphere, where memory contents are actualized which then create phenomena in the phenomenal sphere.
(d) What in the B-concept are non-identified, "purely physiological interaction" effects of the abstract relationships between the sensory stimuli, are in the ETVG-concept defined special visual functions, which, according to empirical neurobiological research, seem to be identical to functions of specific classes of neurons.
3. Explicit and implicit memory
B1 rejected any memorizing concept for the creation of configurations.
"The explanation of configuration by memory effects seems to me to be circular. Memory effects with configuration content ('Gestaltgehalt') are possible only after primary perceptual processes of approximately the respective gestalt contents have previously occurred. Memory as such seems, to me, incapable of introducing an absolutely new type of happening into processes that are unconfigurated".
B-psychologists (and many others) know of only one sort of object to be memorized: only experienced "things" can be memorized; this corresponds to an explicit learning process, whose memory contents can be remembered, i.e. imaginatively experienced a second time without actually being perceptually experienced a second time. The ETVG knows of a second sort: non-experienced "things" as, for example, abstract relationships between stimulations can be memorized, too; this corresponds to an implicit learning process. As these abstract relationships are not conscious, they cannot be recalled, after having been memorized. Only by an implicit learning process can something be experienced for the first time, i.e. something that has never been experienced before by the individual; only in this way, previously non-existing phenomenal configurations can come into existence. "Experiencing" is, according to the ETVG concept of memory, identical to "actualizing memory contents". In 1961, I needed to hypothesize an implicit memory in order to explain how all the possible configurations could occur (i.e. be experienced) for the first time, when they had never been experienced before (and thus could not have been memorized by the explicit memory). Thus I wrote, without knowing, or even using, the terms "explicit" and "implicit":
"The memory memorizes (at least) two experience contents and their relationships
(and relation-relationships) to one another.
1. It is always e x p e r i e n c e c o n t e n t s that are memorized: pure sensations, configurated perceptual contents, thought contents, imageries, feelings, vague impressions ('Anmutungen'), sensations caused by motor acts, and so forth.
2. The r e l a t i o n s h i p s between the experience contents are always comemorized" (p.150).
The memorization process due to No.1 is an explicit learning process, the process due to No. 2 an implicit learning process (although the wording lacks in final precision). Fig. 4-4 shows the concepts of memorized objects for both the Berlin Gestalt Theory (A, B) and the ETVG (C). (See also Kleine-Horst 1994b).
"Also your description of perception in early infancy, which underlies the attempt to account for the gestalt perception by memory, has not yet convinced me, apart from its remaining unverifiable and in contradiction to a series of results in observations on behavior in early infancy."
ETVG: At that time, there had not been much investigation into infant perception.
4. Toward a synthesis of Ganzheit theory and "Element theory" using the ETVG
criticized my not having used enough references; he recommended to particularly deal with the early discussions between the gestalt psychologists and their opponents.
It was not necessary to study the discussions in greater detail because I was discussing the topic at a higher level of abstraction: I found that both gestalt psychology and the opposing structuralism ("Element psychology") are grounded on a common axiom. If one drops this axiom, as the ETVG does, new relationships are visible that do not contradict each other. Here are the respective claims of the 1961 manuscript:
"In the preface, it was implied that the theory of optical gestalt perception seems to differ from both the element theory and the ganzheit theory by an axiom. This assumption will now be justified in greater detail.
All three theories make statements on experience as well as on the conditions of experience. Both the element and the ganzheit theories possess a common basic concept, with regards to the relationships between experience and the conditions of experience, which obviously is unknown to them. They assume that the relationships between the parts of a p h e- n o m e n a l diversity are isomorphic to the relationships between the parts of the diversity that c o n d i t i o n the phenomena. This unconscious prerequisite of the element and the ganzheit theories shall be designated 'i s o m o r p h i s m a x i o m' ('Gleichartigkeitsaxiom').
The present theory does n o t have the isomorphism axiom as its prerequisite; in this it is essentially different from the two previous theoretical basic concepts of general psychology. Hence, its entire efficiency, its great expanatory value, is based on leaving out this isomorphism axiom!
On the grounds of its experimental investigation in the function of sense organs and the function of memory, the e l e- m e n t t h e o r y came to believe that the functional entities exist independently from one another, and act as single entities, and that their effects are summed up by addition and multiplication. Since one stood on the foundation of the isomorphism axiom, one took for granted that the phenomenal entities, too, would have the character of isolatedness and additiveness: according to this view, a perception consisted of a number of sensations and the associative additions coming from memory. They did not think of the possibility that the laws for connecting the parts in the phenomenal sphere could be different from those for connecting the parts in the functional sphere.
The g a n z h e i t t h e o r i s t s, too, did not think of this possibility; their theories contain the same tacit prerequisite as the theories of the element psychologists. In the ganzheit theory, however,one does not start from the sensory and memory f u nc- t i o n, but from the p h e n o m e n o n, and one states that this is holistic, i.e. the parts of the phenomenal diversity are not summands, they are instead members of a whole.
Thus, the ganzheit theories are based on the same isomorphism axiom, as they state: the entire conditions of an experience are as holistic as the experience itself, and the single conditions are members of the whole of conditions.
Thus the Leipzig 'genetic ganzheit psychologists' introduce the concept 'structure' as a designation for the holistic psychic being that conditions the phenomena, although the structure itself is not phenomenal.
The Berlin 'gestalt theorists' even promote the isomorphism axiom with their own 'isomorphism' concept to an e x p l a- n a t o r y p r i n c i p l e, in that they reduce a perceptual configuration to an isomorphic configuration of the physiological and physico-chemical processes that condition the phenomena.
To find a common axiom in these theories does not express anything about their value or non-value; it is decisive whether one can propose a more efficient theory w i t h or w i t h o u t this axiom.
It is tempting to make the axiom discovered responsible, in a way on a trial basis, for the contemporary state of theoretical psychology, drop it and try to use the thus gained new possibility of thinking to develop a more efficient theory on certain facts of experience.
The new possibility of thinking consists in the freedom to think of the structure of the functional-dynamic sphere as being absolutely different from that of the phenomenal sphere. Because nothing, absolutely nothing, forces us to assume that the relationships between the parts of a phenomenal diversity and the relationships between the parts of a functional-dynamic diversity are isomorphic, although this assumption does have the advantage of particular simplicity.
The 'theory of optical gestalt perception' tries to make use of the new possibility of thinking and is, from the start of its development, capable of accounting for important optical gestalt phenomena with remarkable precision. It assumes the structure of both the phenomenal sphere and the sphere conditioning it to be different: the phenomenon itself is qualitative-holistic, the conditions of the phenomena, however, are quantitative-additive, to clarify the difference with a few catchwords. Thus, a s i n g l e p h e n o m e n a l w h o l e is mostly conditioned by a n u m b e r of f u n c t i o n a l - d y n a m i c s i n g l e s .
The actual-genetic structuring of holistic complexes can thus be presented as a successive addition of single functional entities. The complex-qualitative similarity and dissimilarity character of two figures can then be based on the existence, or non-existence, of a number of certain factors that condition the similarity impression due to their number and the strength of their effect.
Or, in the case of visual inversion, among others, the holistic rhythmic-dynamic happening can be accounted for by the actualization of a few certain factors and the magnitude of the attention direction toward the figures.
We see: if one leaves out the isomorphism axiom, one does not encounter any difficulties in showing the single conditions, expressed in number and measure, of the holistic, only qualitatively describable, facts of experience. The previous results encourage the thought that perhaps further areas of perception may be made theoretically accessible by limiting the element theoretical approach to the functional sphere, and the ganzheit theoretical approach to the phenomenal sphere, in order to establish in this way, and with consideration of biological-dynamic factors, a s y n t h e s i s of e l e m e n t t h e o r y a n d g a n z h e i t t h e o r y." (Kleine-Horst 1961, p. 101-104. Spaced as in the original).