FULL ACTIVITY OF THE RETINA

A result of our previous account is that brightness, darkness, and colour, taken in the strictest sense, are states, modifications of the eye, which are simply sensed immediately. A thorough43 account of colour must proceed from this idead of it, and as a result, begin to investigate it as a physiological phenomenon. For in order to work properly and deliberately, before one undertakes to discover the cause for a given effect, one must first become fully acquainted with the effect itself, because it is only from the effect that one can collect the data for the discovery of the cause, and only it provides the direction and guiding threadf to the cause. Newton's fundamental mistake was just this: without some accurate acquaintance with the effect according to its inner relations, he went too hurriedly in search of the cause. However, from the most ancient to Goethe's most recent, all colour theories make the same mistake: they all merely talk about which modifications of its surface a body must undergo (perhaps through analysis of its components) or which modifications light must undergo (perhaps through turbidity or through some other combination of shadow) in order to show colour, i.e., in order to stimulate a specific sensation in the eye, which cannot be described, but can only be detected by the senses. Instead, obviously the correct way is first of all to turn to this sensation itself in order to discover whether one could find out what constitutes it in and of itself, that is, physiologically, according to its nature and conformity with laws. Obviously such a precise knowledge of the effect, which actually is what is at issue when one speaks of colours, will also provide data for the discovery of the cause, i.e., the external stimulus which excites such sensation. First, for each possible modification of an effect, one must at all times first be able to demonstrate the precisely corresponding capacity for modification; furthermore, where the modifications of the effect show no sharp boundaries between one another, such boundaries should not be found in the cause, but there too there must be the same gradual transitions; finally, where the effect shows contrast, i.e., allows a complete reversal of its character, then the conditions for that contrast must also lie in the nature of the cause, according to Aristotle's rule: ‘contrary causes for contrary effects’ (On Generation and Corruption II, 10).