Roughly speaking, we might think of the process of our perception as dividing our perceptions into several parts or “fractions”.
- Since the processing capacity of our brain is limited, some part of the information is overlooked or ignored because of information overload.
- Our perception applies knowledge to the perceptive stream. As a result, part of the information is recognized, some part of it is even expected or predicted. This information might then be dismissed (if it is rated as not important), so it becomes part of the background that is not consciously perceived most of the time. Part might catch our attention because it is relevant or important (which, of course, depends on our current condition and context). So there is some subdivision here. Some of the information might be used to predict or expect further information (i.e. it contains information about other information, so the whole perceptive stream contains some redundancy). In terms of information theory, the expected/predicted part of the information corresponds to redundancy, although there might be some redundancy in the stream of perception that remains unrecognized at a given time.
- Part of the information is unexpected, surprising or new. In terms of information theory, this part is information in the strict sense. What we perceive might be recognized and the new information is just the fact that an instance of this known phenomenon has occurred. Some, however, might be totally new to us. Since this part might indicate something important, e.g. a danger or a resource, it might catch our attention. Since it is new, we might not know how to process it appropriately, but we would apply at least some simple, elementary knowledge to it to gain some insight. Of course, the new information might just be one aspect of a partially recognized complex of data. For example, when we listen to language, there will be expected features (e.g. in the grammatical structure) but in many cases also some new information (or else there is not much reason to listen). Knowledge that is used to “parse” some of the stream of perception and thus recognize it forms perceptive analytical spaces.
- Some information is new, but classified as unimportant and thus filtered out. This is the “background noise” that has some known/expected (sometimes statistical) properties that allows us to dismiss it, although strictly speaking it contains something new. In terms of information theory, this corresponds to “noise”.
Too much new information leads to overload and, as a result, a lot of lost information. We become confused.
Too much recognizable (expected) information and lack of novelty leads to boredom, at least when we are not tired. Boredom will cause us to actively seek out sources of new information.
If the amount of non-noise new information (i.e. information in the sense of mathematical information theory) is not too high and not too low, we will be able to assimilate it, i.e. integrate it, at least to some extent, into our perceptive system. This changes the perceptive system, i.e. the perceptive system is not a fixed algorithm, but develops.
In the simplest case, we just memorize the new information, at least partially. This can be thought of as the creation of a new analytical space containing initially only a small amount of information. If this happens at a rate at which we can cope with the new information, we will neither feel bored nor confused. At the rate of new information at which we can just cope without getting confused, we will get into a state perceived as “flow”. We might be rewarded with a feeling of fascination.
In some cases, we will discover a regularity that we did not see before, i.e. we discover redundancy in the perceptive stream that we had not recognized before. This means that we perform a unification of analytical spaces. This leads to a compression of data in the perceptive stream, and to the possibility to predict or expect some aspects of the stream.
I think that such discoveries are rewarded with a positive emotion. In the case of visual or auditory perception, such discoveries are rewarded with a feeling of beauty. In higher-level cognitive discoveries, they are rewarded with a feeling of joy of discovery. This can be thought of as a reward for data compression. There seems to be some kind of feedback loop mechanism trying to maximize this reward.
The data compression and familiarization reduces the overload or loss of information since it frees perceptive processing capacity. The process of data compression by unification, i.e. by the creation of more comprehensive “programs” for parsing the perceptive information, is a creative process. There is no general algorithm to do this (as it has been shown that optimal data compression is not Turing-computable, i.e. every compression algorithm will have gaps in the sense that some regularities in data will not be recognized by it and thus some chances of compression not found).
These emotions of boredom, fascination, beauty and joy of discovery may be thought of as “cognitive emotions”. Their evolutionary purpose is to guide the development of our cognition so as to familiarize us with our environment, i.e. maximize the part of the information that we recognize (so we can successfully identify dangers and resources). The possibility of playing and art may be seen as side effects. In the history of mankind, art appears at a stage when economy crosses the limit from mere survival to having some excess energy.
(The picture, showing a painting of Paul Cézanne, is from http://commons.wikimedia.org/wiki/File:Paul_C%C3%A9zanne_017.jpg. One might think of such impressionist art, moving towards abstraction, as the result of a subjective filtering or perceptive processing of the artists perception.)