6

If single cells or populations highly dependent of sensory input can often be modeled with so called receptive fields. For neurons and populations close to some motor output, they can be modeled with limb positions, trajectories or muscle activities. But how can we model what other neurons/populations do if their response is dependent on a highly dimensional input-space which we cannot visualize or grasp?

My question is therefore: what does it mean to understand the brain? Do we understand the brain when we have 10000 formulas that describe the mayor functions of populations in the brain but nobody gets their meaning? What if we can reproduce a working brain by copying the main principles without the possibility to illustrate or explain what exactly it does - do we understand it?

  • A (linked) definition of understanding may be useful. Making this question "good" material for philosophical confusion (likely soon to follow) is the apparently open (to interpretation) problem of whether a concept (which features dominantly in the main definition) is itself a brain pattern. :) – user3164 Jun 20 '13 at 16:51
  • I consider a definition of understanding already a huge part of the answer to my question. – Brandli Jun 20 '13 at 16:58
  • Searching a bit for "understanding understanding", I found Second-order cybernetics and New cybernetics. I have no idea whether this is an established/mainstream discipline, but at least it looks related. [Note that I'm just commenting away, not in any way suggesting an answer.] – user3164 Jun 20 '13 at 17:25
  • Even though Second-order cybernetics is an interesting aspect of understanding the brain, it's not the core of my question. I am more asking what we consider "understanding the brain" from the point of view that in most physics a reductionist approach leads to a simple formula that explains the phenomenon. But intelligence and consciousness seem to be emergent phenomena of a highly complex system that doesn't seem to be describable in a reductionist approach. – Brandli Jun 20 '13 at 18:39
  • References for emergence and reductionism for interested readers: Emergent Properties, Emergence, Reductionism in Biology, Reductionism – user3164 Jun 20 '13 at 19:11
2

We do not yet know whether the brain is "understandable" in the sense that we accept now. If the brain operates mostly on the basis of several dozen key principles that can be expressed mathematically or with some other formalism, then it will be understandable in the conventional sense. On the other hand, if it is in fact tens of thousands of formulas with no apparent organizing principle, then we probably won't say we understand it in the sense we accept now--but we may be increasingly willing to hand over the details to computers.

There's every reason to believe that the brain is understandable in the sense that we can predict what it will do given adequate knowledge of its present state and inputs over time. Some aspects may be chaotic or random, but we know how to characterize such things. But what's not clear is whether "mind" will be like "life"; we generally have a sense that we understand what "life" is since we can break it down into a few basic principles--replication, metabolism, etc.--which in turn can be broken down into a few basic principles and so on. Even though maintaining a living organism is incredibly complicated and we are missing very many of the details, we're mostly comfortable saying that we understand what life is and how something is alive vs. not alive. There is no particular reason why the brain must be understandable in the same kind of hierarchical fashion, however. We can even build things that we can't understand. For instance, rather frighteningly, nobody really understands any more how the electrical grid in the United States behaves in the face of a large failure; this is because of exactly the feared brain-problem: thousands of local decisions about how to handle failure and thousands of local structures contribute to the stability or lack thereof of the whole system.

But in the longer run--assuming we maintain an advanced civilization--we don't really need to worry about whether we understand correctly or whether we plus computers together understand (in the high-fidelity-of-prediction sense).

1

There is nothing in modern science to suggest that we cannot one day understand everything general about our own minds.

"Intelligence" is the label we use for the processes we do not understand yet.

Everything else is a "Cognitive Algorithm" or similar nomenclature.

We understand the brain when "Intelligence" is superfluous: when you can just refer to a specific cognitive function by name and write a formula for.

More generally "understanding" is when you can predict the behaviour of a system better than random chance, or throwing your hands up and saying "I don't know".

  • 1. Nobody claimed we cannot understand the brain at one point (I am a modern scientist) 2. I strongly disagree with your idea of intelligence. Intelligence is an attribute to systems capable of finding solutions to problems. 3. According to your definition we already understood the brain thousands of years ago since we are capable of predicting many behaviors very well. 4. From what you wrote I have to infer that your answer to my question is: Understanding the brain is having a formula and we will find one day a formula that fits all behaviors so that we can predict them. Is this correct? – Brandli Jun 21 '13 at 9:19
  • That is correct. – Karl Damgaard Asmussen Jun 23 '13 at 14:32
-2

let's see - God made human and able to improve it, people made computers and can improve them too. once we will be able to improve God's construction, we can agree that we unserstand it (to more or less degree). although, being programmer, i should say that it's easier to make your own equivalent program rather than understand someone else's existing one ;)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.