Excerpt from
L.Kleine-Horst: Empiristic theory of visual gestalt perception. Hierarchy and interactions of visual functions. (ETVG), Part 3 I

Correspondence of Functions Hypothesis

1. Increase in theory deficit by neurobiological facts

In the first half of the 20th century, one did not know much about the structure and function of brain areas involved in perception. Thus the early structuralists and gestaltists were not able to take our contemporary knowledge into consideration, when developing their theories. They knew, however, that the brain is full of neurons, which are somehow connected to one another, and that it is involved in perception. This makes it all the more astonishing to see the theories they developed regarding brain processes:

Well, every opinion, even the strikingly strange is quite acceptable, if it leads to plausible explanations of basic facts. This is here, however, not the case.

Today, we know that visual information is processed hierarchically (for example: Livingstone and Hubel 1987, 1988; Zeki and Shipp 1988). Moreover, we know that there is an immense number of neuronal connections between the cortical areas involved: more than 200, as reviewed by Felleman and Van Essen (1991), so that it has been claimed, that one, under these conditions, can perhaps no longer speak of a hierarchical process. Despite a yearly increasing number of papers on this topic, a neurobiological theory of sight is not to be expected. The neurobiologists Hubel and Wiesel (1962,1965, 1968) received the Nobel Prize for their pioneering examination of the humanlike visual system of cats and monkeys. Hubel, however, summarized the results of neurobiologists' research into sight, up to 1987, as follows:

Eight years later the understanding had still not improved, so that Hubel only needed to replace "1987" with "1995".  

Many experimenters in perceptual psychology laboratories try to relate their results to, or even explain with, neurobiological facts. Some accordance can be achieved, however, a comprehensive theory of sight, capable of describing the crucial relationships discovered in neurobiology, which then extends beyond neurobiologically oriented science, has yet to be produced. Thus, neurobiology has indeed made the already high mountain of facts yet higher, and, at the same time, the gaping lack of a theory yet more apparent.

2. General functionology of both the neuron and the gestalt factor

Let us think about what is called for to develop a theory of sight, which can reflect the neurobiological facts in such a way that it will be accepted in neuroscience.

  1. As the neuron is the crucial function carrier in the neurophysiological perceptual process, the theory must present a function carrier that possesses the same general properties as the neuron.
  2. As neuronal information processing is extended over several hierarchy levels with different functions of classes of single neurons located there, the theory, too, must present a hierarchy of classes of function carriers, that carry out different functions at different levels.
  3. As the particular perceptual functions of particular single cell classes at particular hierarchy levels are known, a theory should provide the same perceptual functions for the function carriers located at the respective hierarchy levels.

No explanatory theory of visual phenomena, known to date, fulfills these demands, not even allusively, except the ETVG. The hypothesis, required for No.1, is called "correspondence of functions hypothesis". The crucial function carrier of the ETVG, which possesses the same general properties as the neuron, is the "gestalt factor", as shown in Tab. 3-1. The fulfillment of No.2 and 3 is reported on in Chapters II and III.

In the case of the neuron, the general term "stimulus" does not mean the sensory stimulus, because this sort of stimulus exclusively activates the receptors. Instead, with "stimulus" of a neuron, a particular input is meant, identical to a particular output of "its" lower level neurons. Only this sort of "stimulus" corresponds to the "gestalt stimulus" of the ETVG.              


gestalt factor

1. A neuron is excited

A gestalt factor is actualized

2. by its stimulus

by a gestalt stimulus

3. to a certain magnitude

to a certain magnitude

4. provided that the stimulus possesses a minimum magnitude (stimulus threshold)

provided that the gestalt stimulus possesses a minimum magnitude (actualization threshold)

5. When the stimulus does not reach this minimum magnitude, the neuron is capable of being excited by additional stimuli.

When the gestalt stimulus does not reach this minimum magnitude, the gestalt factor is capable of being actualized by additional stimuli.

6. The stronger the stimulus, the stronger the excitation

The stronger the gestalt stimulus, the stronger the actualization.

7. There are two antagonistic classes of exciation of a neuron: facilitation and inhibition.

There aree two antagonistic classes of actualization of a gestalt factor: a positive and a negative.

8. A visual neuron is excited via a retinal area of a certain shape ("receptive field").

A visual gestalt factor is actualized via a retinal area of a certain shape ("receptive area").

Table 3-1.  Correspondences between the properties of both the function carriers "neuron" and "gestalt factor"  

In the following section, the specific properties of the lowest gestalt factor, Pml, which shows several peculiarities, are described.  At the same time, the structure of its receptive field will be derived from a general principle for forming receptive fields. Due to this principle, the receptive field of the brightness and darkness factor Pml possesses the same structure as the neurobiologists found for the receptive fields of the brightness and darkness neurons in the retina.

"Correspondence of Levels Hypothesis"

back to contents