Very simple question about color perception

@dfx the simple answer is that you (ie your neurons) don’t know what the “meaning” of an input is. If you’re getting input right now from a particular axon, you just know that “whatever that axon represents” is happening right now. Each neuron has access to a relatively small number of possible inputs, and all it does is learn to decide to fire or not, based on how many and what combination of inputs it is getting right now. If it fires, its output represents something to other neurons which again is just “whatever that axon represents”.

So, at the level of individual neurons, there is no sense in which any individual neuron is aware that “this signal means orange” or “this signal means green”. Yet somehow we (as minds) do have that awareness. How this works in detail is called The Hard Problem in neuroscience (and psychology and philosophy).

One very poor answer is that our reptilian brains have evolved deep structures for representing or labelling things in the physical world, such as “mother”, “tribe member”, “friend”, “threat”, along with solid objects, liquids, animals, etc. (see Japanese Counter Word for this surfacing in the Japanese language). The evidence for this is found in birds and reptiles, who instinctively “know” what such things are just by seeing or hearing them. In the mind of a reptile, the “meaning” of a percept is simply the presence of activation in one or more of these innate “label structures”, and the “label structures” only have meaning in that they have evolved to trigger some kind of relevant behaviour. For an extreme version of this, see Julodimorpha bakewelli which caused the outlawing of brown beer bottles in Australia).

As we experience the world as babies, these structures are (hand-wave) activated by the old brain as it receives the appropriate sensory stimuli, and our cortex (probably in or near the inferior temporal lobe, where object recognition happens) learns to identify these associations too. The reptilian mechanisms develop more quickly than the cortical ones, but feedback loops involving ever-improving cortical functions allow the reptilian components to do their job even better, and before long (maybe well before 12 months of age) the two parts of the brain are so interdependent that perception itself is a distributed phenomenon and becomes this conscious thing the philosophers call “qualia” and others call “feeling”. Many people believe that the qualia is not directly perceived by the neocortex, but is inferred from previous associations between similar perceptions and the identifying signals coming from non-cortical centres.

4 Likes

On the “meaning” of individual neurons vs large-scale patterns, try Hofstadter’s Careenium

1 Like

Dear @fergalbyrne, thanks so much for your clear and detailed answer, now I’m no longer confused any more. :slight_smile: So Great!