A Clockwork Orange? (A brief musing on the concept of a neural “code”)

I was asked this question on Quora:

Are there many layers of neural codes from the human retina to the optic nerve and the optic nerve to the brain, or are they essentially same signals relayed?

Here’s how I responded:

Here’s a question: in a system composed of clockwork, is there a “code”?

I ask this because I find that the “code” metaphor is often misleading when thinking about biology. Codes are composed of symbols. But it is not clear that neurons communicate using symbols.

The way a neuron affects other neurons is more like how a gear affects other gears. There is no code — there is causality. An active neuron releases some neurotransmitter, and this in turn makes other neurons more active. It’s like a complex network of dominoes.

Does the idea of a “code” help us understand how one domino affects the next one in the chain?

I admit that by the time a human is thinking in terms of words and symbols, “code” is probably a useful metaphor for what is going on. But the origin of coding schemes remains a great mystery in neuroscience, cognitive science, and artificial intelligence. So I recommend starting with a much less loaded metaphor, such as clockwork or dominoes. Thinking in mechanical terms helps us realize what exactly neuroscience and AI research are trying to achieve.

For now, there is a fascinating gap in our understanding of what exactly codes are in the first place.


Anyway, if you are interested in the causal “domino effect” that starts at the retina, have a look at this answer:

Yohan John’s answer to In which format is information stored in the brain?