Behold! The GO Ontology in all its glory. This only shows the relations between classes, not attributes and other fun stuff. It only took 27 hours to transform it into the appropriate format and render this graph. Partially because I’m a bad programmer ;-)
Constructivism is a specific philosophy of learning. construct knowledge after exposure and interaction with material, learn in cycles and spirals.
Constructivism acknowledges that learning is a self-directed process with experimentation, exploration and reflection. It can be done with a learning leader, groups or alone.
It can be combined with direct instruction. The instructor determines the level of each student and designs a series of questions which increase in challenge level and require the students to practice the necessary skills. Frequent feedback, ideally built into the activity itself, aids this process.
There are several types of desirable difficulty that can be incorporated into practice to increase learning and effectiveness. Two of the most well studied are spacing and interleaving. Spacing reinforces memory and aids in the construction of models, while interleaving tends to allow errors to surface and for the appropriate submodel to be selected.
Play is enjoyable, undirected activity that seems to develop skills that occurs in mammals. It is generally social and resembles both combat and sex.
Play and exploration have their own structure. It’s fractal and natural. It occurs in cycles. New information is introduced, then this is explored. Then reflection occurs and a new outing is made.
Sometimes it’s useful to use specific heuristics in exploration. For example “leap and creep” (from James Bach). Spend some time exploring the local “area”, but sometimes leap into a totally different part of the map. Repeat as boredom dictates.
“Brainstorm and analyze”, my own invention. Especially useful when troubleshooting. Brainstorm possible solutions or areas to look, then go through them, adding new ideas as areas are eliminated. Occasionally reflect on what has occurred to come up with new ideas. Also gather ideas from others.
Experimentation is important. Either guided by hunches, curiosity or hypotheses. These too require reflection. It is important to realize that all of these experiments are equally legitimate depending on the context.
Playful exploration is sometimes funny. It’s natural and non-mechanistic. It resides in the complex domain where reasons can only be discerned after the fact. If it was predictable, it wouldn’t be fun.
Sometimes play can be uncomfortable. You’ll stretch or end up a place where you haven’t been before. The important thing to realize is that these phases are transient. You’ll soon have new knowledge to integrate into your repertoire and practice. Lots of small mistakes and discomfort lead to rapid learning.
Feedback loops are one of the necessary things for a learning system. In general, imprecise but rapid is better than precise and perfect. If it isn’t rapid, the system doesn’t know which action lead to which result.
According to this article, it may also be psychological. You perform with more intensity when you know you’ll get feedback right away and hence perform better. They also claimed that those with pessimistic outlook performed better, which has interesting connotations for Stoicism, which I may investigate in a later post.
One obvious way to increase feedback is to break your large goal into subgoals. Since subgoals are smaller, they can be accurately judged much more quickly.
Errors are useful in several ways. One, they maximize surprise, which allows for maximum learning. They also tend to increase willpower and determination, as long as you don’t totally give up.
The build-measure-learn loop of Lean Startup and science in general thrives off of error. The error is the difference between the actual and the goal.
If you have written goals, this comparison between actual and desired occurs subconsciously.
Sometimes it’s useful to actively reflect upon the current state and desired state, and the changes that have occurred through timelining, journals and other techniques.
Another way to look at this is as a map. You have to know where you are and where you want to go to get there, but not necessarily the path. You can simply explore your local space to attempt and find a route to your goal. Reflecting upon where you are and where your goal is is all that’s necessary.
In day to day life, this means having goals that you write down and reflect on for short time periods seems to aid learning and change.
Now, as someone who’s a quasi-consultant, this makes me think long and hard about the work that I do. It makes clear that a system of strong, virtue-based ethics is necessary to do my job effectively: Do no harm.
Taleb’s recommendation to eliminate ethical problems such as these are for the people suggesting actions to have skin the game. On the other hand, not having skin in the game is supposed to enable suggesting greater changes.
The other issue is that for people, behaviors are changed by adding new ones, not by eliminating old ones. This is Satir’s “Law of Addition”. So clearly the trick is to eliminate bad behaviors by adding good, or at the worst, non-harmful behaviors instead.
A lot of the literature on self-improvement on the Internet and in the technology world emphasizes the struggle of improvement. The feeling of incompetence as you master new skills or bring them to a new level. A struggle between different parts of yourself that seems intense and in some ways unnatural.
Compare this to the challenging things we learn as humans. We learn to talk, to walk. To think that we have to spend our lives in a sense of turmoil to develop skills, to become better in our lives, is very disheartening to me.
Part of this comes from how we judge these feelings. The feeling of chaos is judged as “bad” or “incompetent” by our narrative self. Something which must be stamped out by tighter and stronger control
An healthier perspective is to view it as the unconscious process of increasing skill through exploration and experimentation. In this model, the narrative self only sets the direction and allows natural progress to occur. To reflect upon what has occurred and where to go next.
Another source of conflict between these two approaches is where the impetus comes from. In the “push yourself” school, this comes from your own willpower. In the “pull yourself” school this comes from your emotional energy associated with the end state. Pushing yourself is all about metrics while pulling yourself is all about desired outcomes.
Both these models need a clear way of understanding where you need to go and where you are currently. I prefer a heuristic approach to these, which includes:
and finally:
I’ve been reading the Buccaneer Scholar again and thought that it was time for me to refine my methods and techniques to learn and understand topics. I’d like a technique that’s general enough to use for skill development as well. Something that allows for create connectedness in all of my knowledge and education, letting me put my skills to the forefront of their use, whatever the situation.
So far, I have this outline:
Previously, I had talked about how learning can occur in cycles, and these cycles can be promulgated by deliberate practice.
Now, there are two parts of any skill, how to use the skill and when to apply it. I find it useful to think of this in terms of Ashby’s cybernetics model and Ashby’s law of requisite variety.
Ashby’s cybernetics model is a hierarchical learning model where individual techniques are learned at first through trial and error, the decision of what technique to use is then propagated to a “higher” unit, which initially starts at trial and error. Basically, at the end you end up with a decision tree of what actions to take.
Ashby’s law of requisite variety states that a controller for a system must have a large enough repertoire of reactions for each behavior of the system, if it is to control it effectively. Not only do these reactions need to be the same in number, but the controller has to effectively engage them at the right time. This is the part of learning which is difficult to learn in “drills” and tends to get lost by ludic activities.
There are several techniques for increasing this. First is interleaving activities to allow for some exposure of the choice of what action to take. This is one of the ways that activities can be made more effective by adding difficulty (see here). Second is simulation, where you go through the actions in in your mind or some other safe environment and the possible states that you’ll come across. Third is actual experience, which when properly combined with reflection, seems to be the most effective way of learning.
TMNT Pumpkin of the Day: Complete with carved-watermelon Krang.
