The birth of a concept?

Since I can remember, my mind ponders on the question of concept formation within. A small introduction: A concept is a meta-physical manifestation that lets us reason about the world in a ‘computationally efficient’ way. It is a set of buckets, containers to classify, or one might say discretize, box our experience that emerges from continuous streams of sensory stimuli.

Concepts are not only computationally efficient in one’s own mind. They form the basis of sharable experiences and can be transcribed with icons in sound or shape (or touch?). The need for concepts in sharable experience seems to come from everyday relevance within our living environment. A fascinating example can be found in the language Tsimane’ as speakers only use encodings of 3 conceptual colors (words), corresponding to black, white, and red whereas an internet pedestrian clearly conceptualizes more. On the other hand, a colorblind infers a conceptual color based on the context, but experiencing the concept does not occur. These examples are interesting as they expose the plasticity of the use of concepts and adaptation/morphing in one's mind based on relevance enforced by nature and nurture. One might ask: how many buckets are there to categorize color in the mind? One might say: depends on how many buckets you need. There are obviously limits to our sensory apparatus but buckets do seem to come and go regardless.

From a computational/mathematical standpoint, my interest in formalizing the adaptive nature of concept plasticity is asking whether the sensitivity of stimuli change in combination with outcome can be translated into concept multiplication. We know that within humans outcome shapes perception in such a sense that categorization needs to be rewarded and one can become ‘blind’ without reward. Filtering invariances in stimuli that are sufficient to predict an outcome, invariances leaking into awareness, would be the first step to figure out.

To be continued…