From Fred Nickols (2017.11.19.1434 ET)
Count me in, Rupert. I’ll probably just follow although I do know a bit about human learning, mainly what works and doesn’t work more so than any particular theory.
As a matter of fact, I’m so excited by the prospect that I dashed off a few thoughts. I’m sending them along as being representative of the kind of knowledge and thinking I might be able to offer. If they’re not a good fit with what you have in mind, feel free to say so. I won’t be offended. I know quite well that my practical orientation and the theoretical orientation that dominates this list are often miles apart. So, here are my initial reactions:
First off, memory and imagination, the ability to recall information and images and the ability to envision the way things should look or work is central to learning. I show someone how to disassemble and reassemble a globe valve and, sure enough, they can do it. Some can do it right away and others need a little practice with some commentary from an observer. In PCT terms, we are at the very least talking about establishing reference conditions, goals or standards, that the trainee will strive to meet.
Second, it is often the case in training that we provide what PCTers might call a “disturbance free” environment. We train them to do the basics, free from disruptions and distractions. Later on, the same training might entail gradually adding these factors back in until the trainee can perform a task under realistic conditions.
Our empirical knowledge of how people learn is actually quite good; we can and do successfully train lots of people to do lots of things; we know how to structure their experiences so they learn, so they become capable of doing what we’re out to equip them to do. Our theoretical grasp of how people learn is quite developed but it doesn’t take into account a view of people as living control systems so it falls short on that score, whether from the cognitivists or the behaviorists.
We know also that an individual’s learning can be inhibited by the person’s view of himself. I’ve had more than one of trainees shake his or her head and say, “I can’t do that; it’s beyond me.” I’ve also been able to turn around those negative self-images and get them to believe in themselves. So one factor to consider in all this consists of the roadblocks to learning that the learners themselves erect.
In that vein, I think there is something akin to a structured and organized knowledge base present in the human mind. It is very much affected by knowledge and examples and not everything works the same with everyone. Individuals learn and they do so individually.
When training, it is important to convey or communicate the results to be achieved and the actions that will achieve them. Enter here the practice with feedback mentioned earlier. However, when people are learning on their own – figuring out what to do so to speak – that often requires no small amount of trial and error, of experimentation to see what does and doesn’t work. That, in turn, implies an adequate reference condition for the result. In some cases, the experimenting or trial and error is done simply for the sake of finding out what happens, of identifying the result associated with this or that action. Here, the learning is focused on the relationship between actions and results, between ends and means.
Another area of learning has to do with the judgments we believe others are making about us (e.g., I’m a nice guy, a pain in the neck, competent or incompetent, rude or polite, etc., etc.). Again, we are focused on the relationships between actions and outcomes or results. Enter here some social expectations or reference conditions (e.g., I don’t want to unnecessarily offend anyone, I need to tread carefully around Dick because he goes off very easily, Shirly is easy to work with and I can relax around her, etc., etc.).
So, from a PCT perspective and as a long-time student of learning and human behavior, I’m most interested in how reference conditions for results or outcomes, are established or adopted; how we come to master the actions associated with a particular outcome or set of outcomes; how the PCT hierarchy fits in with all that; and what’s different when our learning is being facilitated from when we’re doing it on our own.
Fred Nickols
···
From: Rupert Young [mailto:rupert@perceptualrobots.com]
Sent: Sunday, November 19, 2017 1:54 PM
To: csgnet@lists.illinois.edu
Subject: LWG: Learning working group
[From Rupert Young (2017.11.19 18.55)]
Regarding collaborations I think the most useful topic to pursue is learning within PCT. Although there is a general theory of how learning, by reorganisation, happens in PCT and there are some demonstrations showing it working within simulations the theory lacks detail and the demos are quite limited. There is, therefore, a lot of scope for a more comprehensive understanding of PCT learning that can be demonstrated in working models, particularly of how perceptual functions arise. The ultimate goal would be, I think, to throw an unorganised hierarchy at a control problem and it would learn a resolution. This would be a great coup in the field of AI and machine learning, as well as in the behaviour of living systems, to show the power of PCT. An initial target application could be something like the mountain car problem.
Therefore, I would like to set up a learning working group for those interested in taking part by building models, contributing expertise or just following. If anyone is interested let me know and we’ll either set up a separate email group or tag subjects on csgnet with “LWG”.
–
Regards,
Rupert