I recently encountered the work of Roger Shank, a computer science researcher and educator who has lately directed his attention to the problem of education.

The videos on the main page are of keynote addresses he gave and they present a great summary of his theory of knowledge and learning.   He reasons that the task of training is to train the mind, but the mind consists of the conscious part and the unconscious part.   He describes that the unconscious mind is far more important than the conscious mind.   Compared to the conscious mind, the unconscious mind is more powerful in terms of our engaging with the world around us.   The conscious mind is good about learning facts for recitation, but these facts are fairly useless when we need to engage with real world problems in real time.   However, the conscious mind is where we process and exchanges stories.  This story telling capacity gives us a path to train the unconscious mind.

The above summary captures what I want to talk about today, but he has far more to say about it in his site.   I recommend watching his videos to get a more complete view.   I especially enjoyed the first one titled The Cognitive Processes that Underlie Learning.

He also has made progress in putting his innovative educational ideas into practice by working with educational institutions to set up online learning programs using an experiential approach.   His innovation to education is to focus on experiences rather than lectures.  He proposes that educational coursework should involve long (multiple month) periods where the students are assigned a difficult meaningful but ambiguous task to perform.   In this model, the role of the instructor is to be a mentor or adviser.   The instructors do not present lectures, nor do they give assignments nor grade exams.   Instead, the instructors simply make themselves available for when the students seek assistance or guidance in doing a particular task.   Even in these cases, the instructors do not always give the answer even if they know a good answer.  Instead, the instructor engages in conversation with a goal to suggest another approach for the student to take that will get him closer to a satisfying result.

In his speeches, Roger Shank repeatedly emphasizes the futility of the lecture and exam approach to learning.   He explains that lecturing is really just a form of entertainment.  Students sit silently to watch the lecturer perform something for 45 minutes or so.   He explains this can not lead to learning because the students can not truly listen to the lecture.   He presents the paradox of lecture learning as in order to listen, the student has to think about what was just said and contrast it with his own experiences, but if the student does that he is not hearing what the lecturer is saying next.   The idea of paying close attention to the lecture requires the same mental activity we do when we are audiences of an entertainment show.

He is very critical of the lecture approach to education and proposes that it should be abandoned entirely from the earliest childhood learning to the adult college education.   He illustrates his case by describing how PhD candidates learn from their professors is similar to how infants learn from their parents and this is completely unlike how we educate for all the years in between.    Like the parent training a child, the PhD student is allowed to pursue his investigation independently but can approach the professor for guidance when he encounters a problem, or the professor can correct him if he begins making some mistake.   For most of the time, the student independently works on his assigned task.

His education reform recommendation is to apply this same model throughout all of education.   This model of education assigns a task to a team of students to work on independently over a long period of time.   Instructors are available to supervise the activities and to provide experienced guidance when the students run into difficulties.

He promotes a concept of online learning exemplified by the XTOL site (where XTOL stands for experiential teaching online).   The online medium permits connecting multiple instructors to many teams in such a way that the instructors’ time will be used efficiently and that teams always have access to an instructor when they need guidance.   This is a radically different approach to the currently popular online education models described as MOOCs (massive open online course) that he effectively criticizes here.

As should be apparent by the my extensive writing a summary of his ideas, I am very impressed by his ideas, how long he has been working on this, and how much he has accomplished.   In some earlier posts on this blog, I have been describing many of the same concepts about learning through experiences instead of lectures.   It is nice to see confirmation of these thoughts by someone so much more distinguished.   I hope his ideas get more attention.   We need to have this discussion about how we should approach education.

I am not so sure that online learning is the right model.  Certainly, it is worth pursuing.  Given how entrenched the old lecture-test model of education is in our society, an online model may be the only opportunity to make a change.   The time may be right also as it fits with the current fantasies about disruptive technologies.   We need a disruptive technology for education and online implementation of experiential learning may be just the right tool to make this change.

However, we got here by abandoning an older educational model.   The modern educational model (as Dr. Shank discusses in his lectures) came from innovations coming out of the industrial revolution to meet the need for preparing a workforce to work in factories or other types of specialization within large organizations.   These innovations set up the concepts of grade-level progression of learning involving lectures to classrooms of students who are expected to pass tests.   This construction produced an efficient system of education that produces a large number of credentialed students needed to supply the labor needs of industry.

The modern educational system replaced an older method of education that was not as efficient.   The opportunity for modern online learning technology is to rejuvenate that older method of education with efficiency coming from online technologies.  In contrast, I ask whether we still need the efficiency in the first place.

The most beneficial form of learning is the form the trains the unconscious mind.   Training the unconscious mind involves practices that are engaged in solving a problem.   In order to engage the student, we need projects that are meaningful and relevant to the student’s lives.   In order to result in valuable learning, the projects also have to be relevant to real world needs.  As I discussed before, I learned my most valuable lessons from just in time learning.   The necessity of solving the problem motivates the learning, and the actual practice of that learning cements the lesson in the unconscious mind.

The modern educational model replaced an older model of cohort learning exemplified in rural settings with single-room classrooms.   Although students were given tasks to demonstrate though tests or papers specific age-appropriate topics, they learned in the same space as younger and older groups.  In such an environment, the teacher would not be able to simultaneously address everyone’s learning needs.   I imagine that for practical purposes, the students organized into age cohorts and did most of their learning among themselves.   They would need the teacher’s intervention when they as a group were unable to find a solution to a problem.   The teacher would mingle through the room and provide specific responses to specific requests for assistance.   Also in this environment, the proximity to the older groups is a motivation for younger groups to impress the older group, and it is an opportunity for the older group to offer tutoring for what they already learned.

I can also see something similar in the increasingly popular home-schooling approach where by necessity a single teacher (the parent) is simultaneously teaching material to multiple age levels often with all sitting at the same table.   I also noted that many of the offered materials to support a home schooling curriculum draw more upon classical education systems like the trivium.   In my biased mind, I would favor a classical education approach for my children.   Unfortunately, I think the home-schooling approach misses a key ingredient of cohort learning: each age should work with peers to get the benefit of cohort learning.   Cohort learning is impossible in homeschooling, but it also isn’t happening in public education.

In the imaginary view of how single-room classrooms might have operated, I can see some of the benefits of experiential learning in terms of having independent tasks with the opportunity to fail and yet have access to mentors (teachers and older students) to help them when they need it.   In particular, there may have been a necessity of  teamwork within an age group to get their studies completed without full time attention from the teacher.

I applaud the Dr Shank’s achievements in setting up online learning implementations of his experiential teaching concept.   I agree with him that this offers a better path to develop employable skills in students because they are working on realistic problems that require working solutions.   The online learning possibility is a modern innovation and may come at the right moment.  There is a lot of interest in disruptive technologies in all markets, and this certainly has the potential for disrupting the education market or at least the sub-market known as MOOCs.

Our ancestors did not have access to online options.   Perhaps if they did, they might have embraced the concepts to counter the productivity problem of their otherwise effective form of experiential learning.   On the contrary as I try to get my mind into the mindset of my ancestors, I can imagine that they may have rejected the notion despite the improved potential for productivity.

The productivity of an online approach includes some artificiality of the problems being solved.   As I read about the actual courses in XTOL, I can see that a lot of work and thinking went into the development of simulated problems to be as realistic as possible.  I can imagine the students experiencing something close to what it is like an actual work experience.  I can also imagine this experience falling short of a real world on-the-job scenario because the problem is not real: there are no real world consequences of the decisions they make, no real and widely felt pain for their failures.

The best place to find such relevant projects is in the work place itself.   The older education model involved this kind of education.   A young person obtains employment working as a novice or intern under supervision of a more experienced craft person.  The key part is that the person is employed performing duties that contribute to the actual operation of the business.  The novice receives challenging tasks that are within his capabilities but require some learning, but those tasks are also directly impact the business for better or for worse.

I recall reading Michael Ruhlman’s book The Making of a Chef: Mastering Heat at the Culinary Institute of America.  In that book, he describes his personal experience as a student to be trained as chef at the CIA.   The new students are given novice tasks early on such as preparing beef stock.  They would perform this one task daily multiple times during a long period of time because beef stock is essential for training more advanced students.   Contrast this with a modern education approach that may have a one-time experiment performed in a single afternoon.   The CIA required repeating this basic skill over and over again and the success or failure had real consequences for the operation of the entire school.

Although the students start learn beginner skills, they are producing valuable commodities for the actual running of the business.   I was also impressed with the real imposition of schedules.   The tasks had to be done on time in order to be usable for the next step.   An excellent meal prepared too late is thrown out as garbage because there will be no one around to serve.

I think the education model described in this book had the experiential model exactly right.   Students start immediately with meaningful and valuable tasks.  They had to learn in order to meet the deadlines so that more advanced students could do their work that will eventually serve patrons who arrive at standard meal times.   This was done in an educational setting but this educational setting fully recreates a food preparation business.

I read that book many years ago when I was struggling as a manager to find people who could come and help me with my project.   I wondered why couldn’t I have something analogous the CIA to prepare people for the broad problem-solving skills and an ability to apply those skills in time to be operationally relevant.

I operated a data warehouse instead of a kitchen.  But it seemed to me that the capabilities I most wanted from my staff were very much like the capabilities that commercial restaurants expect from their kitchen staff.   I needed to get things done on time when those things require immediate innovations to handle the surprises.   My working experiences had many parallels to the stories in that book about quick thinking required to save the day from a too-hot wood-fired oven, or the unexpectedly high volume of business on a particular day.   The skills trained by the CIA were not just being able to prepare elegant meals, but to prepare them under unanticipated stressful circumstances caused by everyday randomness.

I agree with the concept of experiential learning being valuable.   From my own personal point of view, I would have preferred an experiential approach to learning to the learning I had to endure across 18 years of formal education.   As a manager I would have preferred finding staff who had experiential training than formal education, but I had no option but to choose formal education (credentials).

When considering resumes for a position, we sometimes place an emphasis on experience.   We want someone who not only knows a particular skill but has used it continuously over a period of time.    Implicit in this requirement is the idea that someone a period of education to earn a credential followed by a period of work that required that credential is equivalent to the value of experiential learning.    I discussed that this is simply not true.   The modern work environment does not challenge the worker to tackle more advanced tasks in the same way that the CIA challenges its students.   The only thing that job experience tells a manager is that the person was never fired from a job requiring the skills described by the credentials.    That is not the kind of learning I needed for my task.

The key question is how can we recover the valuable experiential learning approach.   The online approach has merit and the current time welcomes so-called disruptive technologies.   The online approach involves a simulation of a task. The problem with simulations is that it there is no way to eliminate the understanding that it is a simulation.   No matter how realistic, a simulation will never have an outcome that has real world business-relevant outcomes.