12 Levels of Learning



Level 1: Learn to do what works.

For an organic, carbon-based lifeform, this primarily means learning to survive.  It's controversial in the scientific community, but some scientists claim to have observed even single-celled lifeforms learning to some extent - learning to find food, and to avoid harm.  In the early 20th century, Herbert Spencer Jennings claimed that stentor roeseli responded to stimuli differently, depending on earlier stimuli. Beatrice Gelber in the 50s and 60s continued such experiments, and though her work was often sidelined and ridiculed, recent attempts have largely replicated the earlier findings.  Similar experiments have been performed on paramecia.  More widely accepted in the scientific community is the evidence of learning by large plasmodia, slime-molds, and fungi.  The fundamental principle of learning is: if it works, keep doing it.  If it doesn't work, stop doing it.

Level 1a: Learn to simplify. 

I call this "1a," because, in a way, it is simply a corollary of the first principle.  And yet, it is really the key to entire process of learning, levels 1-12.  Learning to simplify sounds, well, simple.  And yet it is the most difficult thing to learn.  In a way, all of the levels of learning are ways of learning to simplify.  Nothing is more difficult than simplicity. 

An example: a baseball player might touch his nose, and then hit a home run.  Now, every time he's up at bat, he touches his nose again, because he has learned that this "works".  Of course, he has also learned the physical, sensory-motor skill of hitting the ball with the bat.  So, presumably, he can learn to *simplify* his actions, getting rid of the unnecessary nose-touching part.  But how is a learner to know which actions are unnecessary - which can be removed in the simplification, and which must stay?  This may seem obvious, but it may be a very difficult thing to learn - or even to express, to one who has not already learned it.  The answer, of course, is that touching your nose doesn't work - and by not touching one's nose, and still hitting a homer, we may learn it.  In that sense, 1a is a corollary of 1.  But actually getting rid of things we have learned may not be that easy, and we may all have such "superstitions," in one form or another.  Indeed, these superstitions may multiply endlessly, if we do not learn to simplify.

Level 2: Learn to do that which results in a reward / positive reinforcement.

Level 2a: Learn not to do that which results in punishment / negative reinforcement.

Level 2 and 2a might look a lot like 1 and 1a, and in a sense they are identical to them- but with something added, which fundamentally changes the natures of these levels - namely, we have just entered the social dimension of learning. A perfectly solitary being can exhibit the kind of learning described in 1 and 1a.  But now we have reached the social level, which transforms everything.  

Now a learner can be trained.  The reward may be, as with a dog, a treat, or some other kind of food; at a subsequent development of learning, the reward may be a pat on the head, or may not involve touching at all - a smile, a look, some attention, clicking the "like" button....  

Of course, it's not hard to see that although 2 and 2a look similar to 1 and 1a, they are not the same and might even be opposed.  Sometimes learners might be trained to do things that don't actually work.

Level 3: Learn by imitating.

I made this one "Level 3" though I could just as easily have made it come before Level 2.  After all, there are animals that seem to "imprint" on their elders, mimicking their actions, without any direct system of punishment or reward to enforce this.  Level 2 is the beginning of the social; Level 3 is the beginning of the cultural.  But it's also important to point out that imitation is not limited to species - for instance, a human being can imitate a cat.

Level 4: Recursive imitations will evolve.

This is like a game of telephone.  B may copy A, but not copy A's actions exactly.  There are inevitably some slight differences.  Then C might copy B, but again, they will not be exactly the same.  As D copies C and E copies D and so on, eventually you will get to a set of behaviors that are markedly different from the original behaviors they were modeled on (those of A). 

At first, the variations in behaviors will be more or less random, or at least based on nothing but errors in copying.  But eventually, if the behaviors produced at the end of the chain of imitations is completely erratic and unhelpful, then the survivability of the learner will be harmed.  Contrarily, if, through a series of random copying mistakes, a learner stumbles on a set of behaviors that happens to be better adapted to the environment, then the learner will survive, thrive, and be more likely to have more learners imitate them (perhaps through reproduction, perhaps not).  Thus, a kind of natural selection of behaviors develops.  A culture will eventually learn to do that which works.  This process brings Level 3 back in line with Level 1.

Level 4a: Recursive imitations will evolve to adapt towards convention and salience. 

Not only will imitations evolve towards survivability, but also they will evolve toward the likelihood of being replicated.  

This is really the turning point in the whole progression of levels of learning. 

Level 5: Learn to reward and/or punish as you have been rewarded and/or punished.

This is the combination of Levels 2 and 3, obviously.  But it also makes clear the differences between 2 and 3.  "Do as I say, not as I do." 

In a certain sense, this could be understood as the beginning of morality, but in fact it far precedes morality, which is not fully developed until the development of Level 12a.  (In a certain sense, morality is a kind of technology.)  That is to say: Level 5 learning may not be particularly moral; often it is unconscious, thoughtless, mechanical, and cruel, and occasionally it is very cynical and self-serving.

Level 6: Learn to do that which will be imitated.  Learn to influence.  Learn to teach.  Learn to facilitate learning. 

This is also a combination of Levels 2 and 3, but from the other direction, so to speak.  Now power enters into the equation: the more influential you are, the more power you have. 

Level 7: Develop a theory of other minds.  Try to understand what others are thinking and feeling.  Empathize.  Put yourself "in the shoes" of another.  Personify.

This develops out of Level 6, or they develop out of each other.  In some sense, in order to teach, you must already have at least some kind of theory of other minds, however rudimentary.  The better your theory of other minds is, the better a teacher you will be.  Thus there is an enormous evolutionary incentive to develop a complex theory of other minds.

Level 8: Learn to signal to others. 

Level 8a: Learn that you are signaling to others.

Signaling is fundamental, an early development, and in many ways precedes learning.  All kinds of animals, plants, fungi, and even cells signal.  Different organs in your body send chemical signals to each other; birds and insects make movements to signal to each other that they are ready to mate; lizards signal aggression, and so on.  Flowers, in their dazzling visuals and delightful scents, can be seen as signals for bees and other pollinators.  Some part of this, or even all of it may occur on an automatic, instinctual, unlearned basis.  But with the development of Levels 1-7 of learning, signaling and displaying may become learned - that is adaptive within a life, rather than evolved over many generations, so that learners can signal more strategically.  The ability to adapt one's strategy of signalling to a specific other presupposes at least some concept of other minds, however rudimentary.

Wearing specific clothing, with a particular look and sound and other senses, can be a kind of signal.  Likewise the first artform, cosmetics, whether it be war paint, marks of status, or other kinds of marks. 

Level 9: Learn to teach others the "correct" way to signal.  Establish grammar.  

This is the combination of Levels 5 and 9 - applying a system of reward and punishment to the system of signals.  Or, more precisely, making signals into a system.  It is motivated largely by Level 6. 

Level 9a: Learn pattern.  Learn rhythm.  Learn music.  Dance.

Pattern, rhythm and repetition are the basic form of insight.  In crucial ways, pattern detection is the main technique of learning to simplify (Level 1a).  Grammar gradually grows over the system of symbols as patterns are detected and over-applied

Level 10: Learn to love stories.  

Level 10a: Learn to tell stories.  Mythologize.

I'm not sure if these Levels are really necessary for all learners, but they sure are fundamental for humans.

Level 11: Learn to abstract.  Learn to conceptualize.  Learn to essentialize.  Learn the principle of principle.  Learn to apply principle.  Establish laws, moral, customary,  and legal.

Level 11a: Learn to think critically.

These are, again, complimentary.  Level 11a is necessary for establishing Level 11; Level 11a is a fuller development of Level 1a, and in many ways equivalent to it.

Level 12: Learn science.  Learn to test theories empirically.  

Level 12a: Develop technology.

In addition to all theses, there is another, which is seemingly tautological, yet is far from simple, and in many ways presupposes all of Levels 1-12, namely:

Level 0: Notice that which you notice; attend to that which attracts your attention; remember that which is memorable.

Comments

Popular posts from this blog

Why Capitalism is Ending

Liquefactionism

Why Ayn Rand was Wrong