Sean Noah

  • A family curse
    During the early 1980’s, an Italian physician was investigating a mysterious and dreadful disease that had long plagued his family. In earlier generations, this disease had killed at least thirteen […]

  • hat is the brain? Researchers conceive of neurons as information processing units, meaning that the circuits formed by neurons support logical and mathematical operations. In this view, the brain is a computer. […]

    • Hi Glen,

      Thanks for commenting and starting this lively discussion. I do not follow your argument that information in the brain is not Shannon-information. Shannon’s definition of information—reduction in uncertainty—applies here, as it does to Japanese characters as well. This is about a probabilistic reduction in uncertainty, not “physical state.” You write that, to you, “they simply are not ‘information.’” You then change the definition of information to a circular definition that requires a homunculus. According to your definition of information, you would be right. But not if we’re speaking about Shannon information.

      I’ll quote David Krakauer (president of the Santa Fe Institute) from his conversation on this topic with Sam Harris (https://www.samharris.org/blog/item/complexity-stupidity):

      “It’s like going from the billiard balls all over the table to the billiard balls in a particular configuration. Very formally speaking, you have reduced the uncertainty about the world. You’ve increased information, and it turns out you can measure that mathematically. The extent to which that’s useful is proved by neuro-prosthetics. The information theory of the brain allows us to build cochlear implants. It allows us to control robotic limbs with our brains. So it’s not a metaphor. It’s a deep mathematical principle. It’s a principle that allows us to understand how brains operate and reengineer it.”

      So this question of information processing is not just a theoretical question. It’s also an empirical question and it’s already been answered.

      -Joel

    • Hi Glen,

      Thanks for your reply, I think we have made some progress in understanding each other. I also think in some ways we’ve been talking past each other. You’re right that Shannon information does not address semantics. Depending on the application, that may or may not be an actual shortcoming of Shannon information. But semantics is not always necessary. For example, I can play (and win) a game of hangman without knowing the meaning of the word being guessed. I don’t know any Japanese and can’t understand any Japanese characters. But, if I simply know which characters are more common and which characters are less common, I can play strategically and even win.

      Regarding your point that there is some sort of dichotomy between the “physical” and “ordinary” definitions of information, Shannon’s information theory is really not a branch of physics, but probability theory. Information theory is abstract enough to deal with subjects ranging from particles to Japanese characters (the latter would never fall under the umbrella of physics).

      Sean’s article discusses very low level computations performed in the retina. Such low level computations can be handled by circuits (neural or silicon) that flexibly solve many problems using similar abstract principles. Like the hangman game, semantics are not part of the picture at this level. Since you write that you “have no doubt that Shannon-type information will probably play a role in describing the physiological mediation of behavior,” I think we may at least agree on the basics.

      I also don’t like argumentum ad populum, but since you bring it up, have a look at this Aeon essay (https://aeon.co/essays/your-brain-does-not-process-information-and-it-is-not-a-computer) that is arguing from your point of view. There are over 600 comments which are, as a whole, very intelligent and very critical (you can find the responses in the left margin).

      Best,
      Joel

    • Hi Glen,

      You write “But none of that will explain a person’s response to the note because such an explanation is bigger than a description of the marks on the page.” What are your thoughts on the process by which an embryo uses the information in DNA to multiply and divide into a whole organism? Do you also object to this description of DNA as information?

      The reason why you reject the idea that flashlights or moons can compute is that you see this as an active process, whereas I see this as a passive process. What is logically incoherent about the description of the flashlight as a simple computer? Or the description of the Earth-Moon system as a computer? This only becomes a paradox if you believe that computation is an active process, rather than a passive process. If one follows this logic, even a Macbook or an iPhone is not a computer, as each would require some magic homunculus to do the arithmetic inside its chips.

      Best,
      Joel

Skip to toolbar