Von Helmholtz’s theory has an inferential and constructivist nature, but it does not overlook the function of the phenomenal component of perception. As Hatfield (2002) remarks, the minor premises are the sensory aggregates, and the inductive conclusions result in the appearances of things. However, the appearances are determined more by the epistemic interpretation of sensations than by the phenomenal characteristics of the latter. Von Helmholtz claims that the sensations are usually noticed only insofar as they are useful to recognize external things. The general rule of vision is “we always represent such objects as present in the field of vision as would have to be there in order to produce the same impressions on the nervous apparatus, the eyes being used under ordinary, normal conditions” (, 1925, iii: 4). He emphasizes the role of the “normal conditions,” that is, the knowledge of the ordinary circumstances to which the sense organs are usually adjusted to bring about the sensations for an effective inference of the external things. This knowledge enables the subjects to interpret the sensory signs as well as to discount the alteration of sensations that may be contingent on the aberrations of sensory responses or external circumstances. Yet it may be insufficient. For example, the light reflected by an object on the retina and a mechanical pressure on the outer corner of the eyeball bring about a sensation that can be represented as something appearing in the direction of the bridge of the nose. In normal conditions that retinal point is mapped onto stimuli along that direction entering the nasal side of the eye. At any rate, von Helmholtz claims that in the ordinary use of perception subjects develop the habit of disregarding any phenomenal property that is not needed by the inferential process or that risks averting the proper reference to objects.
In order to be functional to science, the theoretical decomposition of perception has to be realized in experimental controlled conditions. The explanation of perception requires the analysis of the sensory and psychological, that is, the inferential and knowledge-based, components of perception to specify to what extent each contributes to the perceptual representation of external objects for a given stimulation. It is yet clear that the nature and role of sensations pose essential questions for the method of experimental research. As often as not, compound aggregates of sensation are in reality the sign of a simple object, and it is impossible by direct observation and without external help, due to training or artificial support, to resolve them and discover the corresponding relevant impulses and stimulations. Moreover, the more a sensation is used as a sign, the more difficult the analysis will be. For example, in ordinary experience colors are used to make correct inferences about the constant properties of things, up to the point that one is unaware of neglecting all phenomenal properties contrary to this aim. The subjects have learned that green surfaces change their hue beyond a certain distance. Accordingly, they discount this variation; hence the green meadows and trees in the distance look like the green things nearby. In general the colors of faraway objects vary a lot because of the interaction between light and the particles of the medium. Far mountains appear to have a vague blue-grey color, because at increasing distances the particles in the air interfere mainly with short wavelengths. As distances decrease, the mountains appear in homogeneously whitish, grey or off-white shades of air, due to the particles that interfere equally with all wavelengths in the low air layers. Besides, the interaction of colors in the visual field gives rise to the illusory sensory effects of simultaneous contrast. For instance, if a piece of grey cardboard is put on a red surface, it will acquire a green tint. Likewise, an achromatic or chromatic surface always displays a complementary tint of the color of the surrounding field. Von Helmholtz points out that the clear blue sky, the red-yellow sunset, the vivid green meadows, which meet at the borders of visible regions in the outer world, tend to induce this kind of contrast. However, as subjects discount color variations for constancy, so they tolerate this phenomenon and do not ascribe the contrast colors to any definite object. Yet von Helmholtz claims that, if normal conditions are modified, color sensations can be seen as they really appear and no longer as sensory signs. If subjects assume an unusual position, for instance by putting their head under their arms or upside down between their legs, the inferential interpretation will degrade so that the color sensations and their variable phenomenal properties are perceived.
The possibility of departing from normal conditions and thus observing the perceptual contents that do not usually appear is the key to the methodology of research into perception. Indeed, von Helmholtz ( 1925, iii: 30-33) claims that both ordinary experience and science are informed by empiricist epistemology. To discover the cause of the expansion of liquid mercury, the changes in volume of the substance have to be tested by varying at will the humidity at constant temperatures or the heat at constant rates of humidity in conditions where the air is saturated with substances at all temperatures. The observation of the changes in volume selects the best fitting inference on what is the sufficient cause among the conditions arbitrarily produced in the experiment. Likewise in ordinary experience, adjusting the sense organs to diverse voluntary motor impulses in normal conditions provides a test for the correct inferences that make perception an effective interpretation of objects. If normal conditions are constant, the observation of the changes of sensations, which are coordinated with movements, may select the best inference of the object. Hence if normal conditions are altered – as can be the case even in ordinary experience – in artificially designed circumstances that make it possible to change the conditions at will, the unobserved components of perception may emerge or be inferred. The association between sensory and learned factors is broken down, and both can be manipulated in cross-conditions in which they vary or are kept constant. Von Helmholtz ( 1925, iii: 13) proposes the following rule to separate the factors: “nothing in our sense-perception can be recognized as sensation which can be overcome in the perceptual representation and converted into its opposite by factors that are demonstrably due to experience” (here “experience” means learned knowledge). In abnormal conditions, if the practice, the concentration or a sort of external aid, for example adequate experimental training, cannot make unusual phenomenal effects disappear, then it is legitimate to hypothesize that the latter are the sensations that are not usually observed. On these grounds, it is possible to look for the inferences and the knowledge base that are requested for the sensations to yield the expected appearance that in normal conditions provides useful information about the external things.
Source: Calì Carmelo (2017), Phenomenology of Perception: Theories and Experimental Evidence, Brill.