Objectivism's problem of percept formation

[Grames]

Lots of organismic processes are auomatic without having developed through learning. More pertinently, the fact that many cognitive processes are automatized by learning does not prove that all are.

-- Mindy

Enough are automatized by learning to refute the generalization that "Generally, what is automatic is not learned, and vice versa."

[Mindy]

Enough are automatized by learning to refute the generalization that "Generally, what is automatic is not learned, and vice versa."

"Automatic learning" is the problem. If it isn't redundant, it is contradictory.

"Perception is an automatically learned system..." (From yFeldblum, 6-27 post.) How is that any different from saying it is a learned system?

My position remains that it isn't learned at all.

-- Mindy

[Eiuol]

My position remains that it isn't learned at all.

I don't think learning necessarily must be a process of volition. If your sense organs were repeatedly faced with the same stimuli, I suspect your brain would automatically attempt retaining such information in a manner you were describing on the last post of page 1. The sensations would be grouped differently than how they would have originally, so the process is considered learning.

[Mindy]

I don't think learning necessarily must be a process of volition. If your sense organs were repeatedly faced with the same stimuli, I suspect your brain would automatically attempt retaining such information in a manner you were describing on the last post of page 1. The sensations would be grouped differently than how they would have originally, so the process is considered learning.

Yes, that fits what I'm trying to say. The brain automatically processes what is available from the senses. It does this from the beginning, that process isn't learned. But the grouping achieved changes over time, akin to abstracting from abstractions, versus reaching a very high abstraction from concretes.

Exactly. Thanks.

-- Mindy

[Grames]

Yes, that fits what I'm trying to say. The brain automatically processes what is available from the senses. It does this from the beginning, that process isn't learned. But the grouping achieved changes over time, akin to abstracting from abstractions, versus reaching a very high abstraction from concretes.

Exactly. Thanks.

-- Mindy

There is now a technical, neuroscience based theory of how automatization works. Basically it is the migration of recognition down one layer in a hierarchically organized neural net. One way to provoke this to happen is long hours of deliberate practice such as when trying to gain a skill. Another way is simple immersion with no clear a priori idea of what the hell is going on, which is a good description of infant development.

Jeff Hawkins and Ayn Rand theory of concepts

[Mindy]

There is now a technical, neuroscience based theory of how automatization works. Basically it is the migration of recognition down one layer in a hierarchically organized neural net. One way to provoke this to happen is long hours of deliberate practice such as when trying to gain a skill. Another way is simple immersion with no clear a priori idea of what the hell is going on, which is a good description of infant development.

Jeff Hawkins and Ayn Rand theory of concepts

It's not ideally "technical neuroscience" if it relies on the term "recognition," is it? I don't believe that term has a neuroscience equivalent, does it? "Practice, practice, practice" is not a new idea, either.

The "immersion learning" is an interesting phenomenon at the adult level. It is limited to the sensory-perceptual in animals and infants. In adults, it is pretty much wide open. Have you seen the lit. on "unintentional learning?"

Of course, there are things which no degree of familiarity makes automatic, so there will have to be an additional factor added to that "neuroscience" explanation.

-- Mindy

[Grames]

"Pattern recognition" is what networks of neurons do. Neuroscience as a specialized anatomy needs no such concept but study of what the brain does and how it works requires this concept to progress.

[Mindy]

"Pattern recognition" is what networks of neurons do. Neuroscience as a specialized anatomy needs no such concept but study of what the brain does and how it works requires this concept to progress.

"Pattern recognition" in neural nets is a constrained performance, under optimized conditions. It is more akin to information than knowledge. If you are using "pattern recognition" specifically as it is termed in the performance of neural nets, you do not mean "recognition."

You seem to say it is needed, but not really. It's not material, though.

-- Mindy

[Grames]

"Pattern recognition" in neural nets is a constrained performance, under optimized conditions. It is more akin to information than knowledge. If you are using "pattern recognition" specifically as it is termed in the performance of neural nets, you do not mean "recognition."

Oh, but I do. The principle that the same phenomena is at work from the most mechanistic sensory response to the most abstract concept is an irresistibly powerful integration, at least for me. This is not a complete theory of consciousness but makes one possible.

[Mindy]

Oh, but I do. The principle that the same phenomena is at work from the most mechanistic sensory response to the most abstract concept is an irresistibly powerful integration, at least for me. This is not a complete theory of consciousness but makes one possible.

Well, sure, it would be great. But it is unfounded. Neural nets do their thing, and it is termed pattern recognition. To say that that function explains recognition in the large sense must be supported.

-- Mindy