for patients who have sports-related injuries. In the process
students produce in-depth reports on additional issues related
to sports medicine.
92 Teaching Minds
5. Designer Genes—Students consider ethical and political issues
related to genetic engineering, for example, cloning, gene
therapies, and the manufacturing of drugs. In the process
they learn about DNA and the basics of genetic engineering.
6. Plant Plague—Students working for a fictional county farming
agency are faced with an anomalous powdery mildew that
has infected wheat in their local area. They investigate how
the new strain of mildew arose and how it spread to this area
and crop in particular. They work to develop a treatment for
the current crop; then they develop a way to alter the wheat
or the way in which it is grown to prevent future fungal
outbreaks.
7. Medical Detective—Students work with the fictional county
medical examiner to conduct medical investigations. They
are asked to determine the time and cause of death for
various mysterious cases.
8. Cutting Costs Without Cutting Care—In this rotation,
students consider business aspects related to healthcare.
Students play the role of a hospital consultant whose job
it is to discern why the hospital is losing money and make
recommendations for correcting the situation. In the process
they confront ethical issues related to cutting costs in the
area of healthcare.
9. Outbreak—In this rotation, students work in the areas
of infectious disease, epidemiology, and public health
administration. They begin by diagnosing the cause of a
fictional patient’s infection that stems from bacteria in a
food item. They then learn that many people across the
country have been found with similar illnesses. Students
have to develop a plan to manage the outbreak due to the
availability of the food item to a wide population. Later they
are fictionally hired to develop a readiness plan for a possible
worldwide pandemic.
Students spend weeks in each rotation. What do they learn?
Remember, the answer cannot be that they learn about health sciences.
Why not?
Because that is really not the issue. No high school student learning
a subject really learns that subject. Students forget what they have
New Curricula for a New Way of Teaching 93
learned in a day or in a week or a year. Professors always assume that
entering high school students need to be retaught the basics. They
may pick up the general idea but in reality the content of the subject
area is, at best, vaguely understood. What they can learn is that they
like a subject or have an interest in learning more.
The only other thing they can learn involves the 12 cognitive processes.
So let’s look at how the health sciences curriculum covers these
processes.
By explicit design, all of the health science rotations emphasize all
four of the “social processes.” The students work in teams (teamwork),
sometimes even in collaboration with other teams. The students and
teams necessarily try to influence one another, negotiate with one another,
and constantly have to describe their points of view and their
results. The step-by-step instructions that are part of every rotation
explicitly discuss how best to handle this. Now let’s look at four of the
rotations in detail.
INTERNAL MEDICINE
The essence of this rotation is learning about how to do diagnosis,
in this case, of liver disease. Students watch a detailed interview with
the patient, select (after some orientation) specific tests to administer,
receive the results, and report suggested diagnoses. They must communicate
what they have discovered. So they describe the patient’s
symptoms and must analyze and discuss the causation of the patient’s
symptoms. Then they begin to plan a course of action. Planning what
to do is a major component of this rotation. To do this they must
make a judgment as a crucial part of the diagnostic work. They must
do this again as well in the ethics unit that occurs later in this rotation.
In that unit they undertake a detailed study of both medical and ethical
issues in liver transplantation—which is where evaluation comes
into play. Here the students become consciously aware of their values
as they decide how to influence medical and ethical choices.
SUPER WORM
The students plan carefully for this unit (a hypothetical redesign of the
earthworm to make it even more helpful in agriculture). Hypothetical
94 Teaching Minds
experimentation is performed repeatedly as the students predict the
effects of various modifications of the worm’s anatomy. Describing
their conclusions is important—the project’s goals; the worm’s anatomy,
physiology, and behavior; and the interaction of anatomy and
function.
SPORTS MEDIC
The core of this rotation is four athletic injuries. In each case the students
examine the patients, describe their observations in detail, and
perform a formal differential diagnosis. The students must use judgment
and modeling to predict the effect of providing a competitive
athlete with an artificial bone implant.
DESIGNER GENES
This rotation has three primary activities. In all three, evaluation plays
a role as they discuss genetically modified animals, crops, and muscles.
Describing, planning, and influence all are involved as the students
prepare a congressman for hearings. All four of the social processes
come into play in formal debates.
What should be clear here is that what the health sciences curriculum
seems to be about and what it is actually about are very different
things. It seems to be about teaching health sciences content, when
what it is really about is having students practice various cognitive
processes that occur again and again throughout life. This was the goal
of the design, pure and simple: to help students practice thinking. It
really doesn’t matter what arena they are thinking in. We get them
interested in thinking by having them think about something that
interests them and is connected to a world they may wish to explore
later on in life. They may indeed learn something about that world as
well, but the point is the cognitive process-based education, not the
subject-based education.
I thought of the idea of building SCCs instead of using the normal
set of courses that constitute most students’ school year when I took
a job with Carnegie Mellon University (CMU) as the Chief Education
Officer of its new West Coast campus, located in Silicon Valley. There
New Curricula for a New Way of Teaching 95
were no faculty located at this campus, since Carnegie Mellon is in
Pittsburgh. I took this as an opportunity, not as a problem, since faculty
want to teach the way they always have taught. Lecturing is easy
and faculty like to do it. Students have learned to cope with lecturing
and later cramming for tests, so no one complains much. No one
learns much from this, but no one seems to be too concerned about it.
I was asked to design master’s degree programs in computer science.
It seemed to me that students entered a master’s degree program
because they wanted to get a good job after graduation, so I asked
what jobs they were preparing for as I looked at each Pittsburgh campus
program. Even though the faculty in Pittsburgh had been teaching
these master’s degree programs for years, faculty members were
surprisingly unconcerned with what students did after graduation.
They just taught their courses, their specialties actually, and assumed
the students would find value in them. This seemed an odd state of
affairs to anyone who looked at it from the outside, but as professor I
know that faculty are rarely concerned with master’s degree students
at all and naturally wouldn’t have given these programs that much
thought.
There was a great deal of hostility to the SCCs that were built by
my team and me. They were seen as threatening the existing structure
of courses and lectures. Nevertheless, the students liked them a great
deal and the people who mentored in them, after some initial resistance,
began to like them and promote them. This happened in the
areas of software engineering and software development.
The faculty in e-business liked what we had built so much that
they got rid of the existing course-based, e-business master’s degree
they had offered in Pittsburgh and now offer only the SCC version on
the main CMU campus.
SCCs work and work well. Students learn actual skills and teachers
feel like they are helping students do something real. But faculty,
who are used to the old classroom-style method, often resist doing the
hard work now required of them. This is true for students as well. One
mentor in the West Coast e-business program who himself had graduated
from the Pittsburgh classroom-based program said that he felt
sorry for the kids on the West Coast campus because they had to work
so hard. He noted that they were learning a lot more than he did but
that he had liked sitting in the back of the classroom and ignoring the
teacher. It was much easier and he had done well at that.
96 Teaching Minds
A story-centered curriculum is intended to teach cognitive processes,
not subjects. Subjects are, of course, covered, but they are not
really the point. Certain things need to be done again and again in
life, but those things can be learned only in context, not as an abstraction.
Different contexts must be provided in order to motivate
students and to provide real-world skills that will be remembered, not
because they were studied and tested, but because they were practiced
again and again.
What is life like in a story-centered curriculum? The first ones we
built were built as master’s degree programs at Carnegie Mellon’s new
West Coast campus. Here is Max Soderby, a mentor in the first one of
those master’s programs, talking about his experience:
I am almost jealous in a way because I see that they are gaining
skills more rapidly than I gained them when I was a student in
Pittsburgh at CMU’s campus. They get exposure to things that
we just talked about in a lecture hall, whereas they are actually
doing it: implementing actual software and putting designs
into practice. We mostly did homework and talked about it in a
lecture hall. So I am jealous in that respect. It is also a lot more
work, but that work pays off for the students.
Subject-based education is not really supposed to be training for work.
I once proposed to the president of Yale (Bart Giamatti) that we build
a master’s degree program in an area of computer science that would
help get people jobs after graduation. He said that that was training
and that Yale does not do training. The academic subjects taught at
Yale are meant to produce scholars. But, in a way, he was very wrong.
Yale does do training. Yale and almost all other colleges are divided
into departments, and a major in a department’s subject typically
is seen by the faculty as preparation for an academic career in that
subject.
The students may well have a different point of view, however.
Unfortunately, they come away disappointed. An English major could
be hoping to become a journalist, but the education that he will receive
is more likely to be appropriate for creating an English professor.
A math major may well want to be an actuary, but will not learn
actuarial science at Yale. He will learn to be a potential mathematics
professor. And, worst of all, in my own field of computer science, the
New Curricula for a New Way of Teaching 97
very idea that a Yale degree would make you capable of getting a good
job as a programmer is frowned upon by the faculty. They are training
computer science professors. This is the logical end result of subjectbased
training.
Now, what Giamatti had in mind as the end goal of college,
training for the mind, is a noble enough ideal, and a natural outcome
of cognitive process-based education. The classic liberal arts
view of education, one that a reader might think I am not in favor
of, is actually a better model than the model that has evolved in the
nation’s top universities. The idea that you should try thinking in
a variety of fields is a better plan, and one more in line with what I
am proposing here, than the model that exists on most universities’
campuses. The latter model, the one that makes students major in a
subject and thus supposedly become prepared to work in that field,
is really just a big lie.
There is nothing unusual here. Here again, is a statement from the
Ivy League professor whom I quoted earlier:
There is an unspoken rule at places like my university that if you
are really good, you do exactly what your teacher does.
So what are these schools training students for? It could be only one
thing—to become professors. There is no attempt to teach practical
real-world applications of the ideas taught in classes, in part because
the faculty themselves don’t know those applications. Here is the Big
Ten computer science professor again:
There are roughly 60 faculty members in computer science. They
cover all the traditional areas of computer science. Ironically,
software engineering, which is what 90% of the undergraduates
do when they graduate, is not covered. It is not considered an
intellectual or academic discipline. It is considered too practical.
There is only one software engineering course and it is taught by
an adjunct because no one really cares about it.
This is a real problem because (the Big Ten professor again). . .
There are hundreds of computer science majors here. The
faculty doesn’t feel it needs to change because there are students
98 Teaching Minds
clamoring for what is now offered. Ninety-eight percent of them
want to be programmers. Almost none of them want Ph.D.s.
I cannot go to a faculty meeting any more. I get into a fight
at every faculty meeting. I argue about teaching and education
and they think they know because they are professors. I cannot
subject myself anymore to their abuse.
These problems exist precisely because of the subject-based
education system. That system is about factual knowledge, and it is
this emphasis on factual knowledge that has given rise to the testing
mania that has swept the country. These problems exist because the
real mission of the university is very different than the general public
imagines.
Here is a quote and a story that I rather like:
A university is what a college becomes when the faculty loses
interest in students.—John Ciardi
Benjamin Franklin told the story of some Massachusetts
commissioners who invited the Indians to send a dozen of their
youth to study free at Harvard. The Indians replied that they
had sent some of their young braves to study there years earlier,
but on their return “they were absolutely good for nothing,
being neither acquainted with the true methods for killing deer,
catching beaver, or surprising an enemy.” They offered instead to
educate a dozen or so white children in the ways of the Indians
“and make men of them.”(From Benjamin Franklin, An American
Life, by Walter Isaacson)
What has changed? Key life skills no longer include catching beaver.
Otherwise, things are pretty much the same. Change “catching beaver”
to any modern daily skill, and Franklin’s story is just as valid
today.
What do students really learn at a great university? Parents never
really ask this question. They just know that their kid got into a good
college and that their child is lucky to go there.
Should you avoid sending your child to top schools because they
don’t teach so well there and really don’t plan to improve the situation
any time soon?
New Curricula for a New Way of Teaching 99
No. Of course not. If your child gets into Yale, send her there. It is
a great place. But you should know what Yale actually has to offer and
what it doesn’t.
If your child wants to be a professor, Yale is the place.
If your child wants to be an intellectual, Yale is the place.
If your child wants to go to law school, Yale is the place.
If your child wants to hobnob with the best and brightest, Yale is
the place.
If your child wants to have a fun time for 4 years, Yale is the
place.
Then what is wrong?
There is a problem only if you think that there is a different reason
to go to college than the reasons I have listed above.
Oh. There is this other problem. Many of the other 3,000 colleges
are trying very hard to imitate Yale. They attempt to provide similar
experiences and they can’t pull it off. Yale is a unique place. The nation
can afford only so many unique places, however. We cannot afford
having the main university of a state thinking it is Yale, as my
state university friend suggests. If every university has as its main focus
research and not education, then the best and brightest of each
state will be trained, not necessarily willingly, to be academics rather
than practitioners. There will be a great many students who came to
school for an education that will help them in their future lives who
will be disappointed to find out that that is not the type of education
being offered.
Professors at Yale are playing the prestige game. Unfortunately,
they are hardly alone in this.
My state university professor again:
We definitely want to be part of the superstar system but we have
no superstars. If we had them, we would probably lose them to
Harvard and Yale anyway. Nevertheless, we are obsessed with
the National Rankings put out by places like U.S. News and World
Report. Faculty and deans say they are not obsessed with them,
but rankings are an important part of the evaluation process and
shape a department’s growth. We want to be in the top 4 or 5
universities. We are not but that’s what we want to be.
100 Teaching Minds
U.S. News and World Report’s annual rankings weigh heavily on the
minds of the faculty and administration of universities who are in
the prestige game. These rankings are based on numbers: average SAT
scores of admitted students, average rank in class of admitted students,
faculty publications, and many other numbers that come out in favor
of research universities with world-class faculty. But world-class faculty
means faculty who care about research and not about teaching.
While there certainly is no harm in going to Harvard or Yale, the success
of their students hardly depends on what they learned in those
places and depends a great deal more on the fact that the best and the
brightest are the ones who go there in the first place.
These places get away with teaching courses in obscure issues in
literature and history, or in economic theory or in complex mathematics,
by pretending that they are really teaching students to think.
But does knowing obscure information necessarily imply that one is
a good thinker?
A good thinker, I claim, would be good at each of the 12 cognitive
processes.
What does it mean to be good at prediction, for example? Is a
2-year-old good at prediction? Is a dog good at prediction? Is a professional
gambler good at prediction? Is a stock trader good at prediction?
Is a mother of a toddler good at prediction? Is a politician good
at prediction? Is a scientist good at prediction?
We actually are quite good at assessing the ability of others at prediction
precisely because we have data to support our conclusions. We
know how good gamblers or stock traders are at predicting. If they are
very successful, we can say they are brilliant at what they do, or we can
say they are lucky. Those are our choices.
The same is true of scientists. Most scientists make predictions,
and those that are proven right are seen as brilliant. Luck enters into
science as well and quite often scientists say that a given Nobel Prize
winner was lucky and isn’t really all that bright.
Dogs are seen as being smart (for a dog) when they can correctly
predict the arrival of their master or bad weather or threats, and are
seen as stupid when they bark at thunder. Their behavior is seen as
stupid precisely because of the erroneous prediction that barking will
scare the thunder away. A dog’s inability to predict is exactly why we
think dogs are dumb animals, and when they surprise us with an accurate
prediction, they are seen as smart. Of course, we don’t expect
dogs to predict who will win the big game. We know their limitations.
New Curricula for a New Way of Teaching 101
But it is not only the accuracy of predictions that factors into our
sense of a person’s or an animal’s ability to think. We say that a person
is intelligent even when he predicts badly, if he gives good explanations
for his predictions even though they don’t pan out.
So when a sportscaster gives his prediction about the outcome of
an upcoming game, we think he is intelligent if he has thought it out
carefully, and if his explanation is coherent, and if his reasoning is
sound, and we give extra credit if some of his ideas are surprising in
some way. So, even if he subsequently turns out to be wrong, we still
think he is good at thinking.
But prediction is actually quite complicated to judge. We respect
great predictors. We think people who can predict well, especially
those who can explain their predictions well, are very intelligent. But,
often, we see intelligence when it may not be there. The reason is
scripts. And we may fail to see it when it is there. The reason for that
is explanations.
Scripts and explanations are at the beginning and at the end of
intelligent behavior. What do I mean by that?
When a child is learning about the world, she is learning the
scripts that commonly are followed in the world that she inhabits.
I have explained this at length in two different books, 1 so I will just
summarize here.
Scripts tell us what will happen next in the aspects of the world
that repeat frequently. Anyone who goes to a restaurant knows that
when you order food, someone will bring it to you and later you will
be expected to pay for it. There are lots of variations on this standard
restaurant script, however. It doesn’t work quite like that at Burger
King. The script is different but there is a script there too and we learn
it if we frequent Burger King. The restaurant script has many variations
and we are initially confused when we encounter a new one, but
we learn through repeated practice. And, we generalize so that we can
understand that the McDonald’s script is pretty much the same as the
one at Burger King.
We can predict what will happen next in the world based on experiences
we have repeated. Following scripts is so normal that it is
not seen as a sign of intelligence to be able to do it. We don’t exclaim:
Wow, he predicted that the waitress would bring what we ordered and
she did. Amazing!
But the opposite is certainly seen as a sign of stupidity. Once one
has experienced something many times, one should know what will
102 Teaching Minds
happen next in that arena. Dogs usually know the scripts that pertain
to them, as well. They know which merchant will have the dog biscuit
for them, for example.
People seem astonishingly dumb when they can’t predict what
is obvious to everyone else. Not knowing what will happen next in
a script you don’t know, because you don’t have the relevant experience,
means nothing at all, of course. The question is: If you have
experienced something repeatedly, why haven’t you figured out that
what you have seen before will happen again?
Script following is, therefore, a sign of intelligence but a very limited
one. We can blindly follow a script, and this can make us seem
dumb indeed. Since scripts vary one from the other in many ways,
the ability to see the nuances makes all the difference. Expecting that
a fast-food restaurant will be the same as a three-star Michelin restaurant
because it is a restaurant after all is what makes people seem
stupid. Failing to make the right generalizations indicates a lack of
thought.
The real question is this: What do you do when your script fails?
This is important because scripts fail all the time. You expect something
to happen and it doesn’t. You love the cheesecake at Lindy’s and
suddenly it doesn’t serve cheesecake. Or you find that Lindy’s is now
out of business. What do you do? People recover from script failure on
a daily basis. How they do this tells us a lot about how the mind works.
When people refuse to abandon the generalizations they have
made, they immediately are perceived as being stupid. When a medical
assistant asked me the other day about the upcoming Thanksgiving
holiday, I responded that I would be eating duck instead of turkey.
She said that sounded awful and that duck was greasy and gamey and
it sounded like a terrible idea. I asked her if she had ever eaten duck
and she said no because it was game and she hated game. I told her it
wasn’t gamey. She refused to believe me. I was about to recommend
trying magret de canard but thought better of it and asked if she had
ever eaten in a French restaurant in her life. She said that she hadn’t.
People who have scripts often generalize them so that in their own
minds they experts on things that they have never experienced. This
is what stupid looks like.
On the other hand, we might wonder what smart looks like. Let’s
imagine the same woman with the same beliefs hearing me say that
duck was not gamey or greasy. Intelligent people respond, when they
New Curricula for a New Way of Teaching 103
are confused, or when a long-held belief is challenged, with a request
for evidence. She might have asked where she might try duck or what
duck tasted like since she had not understood it correctly. She might
have allowed for the possibility that she was wrong and asked to know
more. But she didn’t. People who aren’t intellectually curious rarely do.
This kind of dull thinking is not so much a matter of genetics as it
a matter of not having been educated properly. And, that is, of course,
the real issue here. If a child grows up in a world where questions are
expected and where long-held beliefs can be abandoned because of
new evidence, he will seek such interactions. But a child who grows up
in a world where adults set themselves up as knowing everything and
no one’s beliefs are ever questioned, you will get mindless behavior
like this.
Of course, it doesn’t matter if this woman doesn’t try duck. It
is likely, however, that this behavior pattern—learn rules and never
question them—pervades her life. This leads me to my main point.
Scripts are great things to have. They get you through the airport.
They get you through Burger King. They get you through most of the
mundane aspects of life. But scripts need to be modified. They fail all
the time. The airport starts a new check-in procedure. The restaurant
you always go to deletes your favorite item from the menu. The store
you always shop in is getting very crowded. At some point we encounter
script failure and we deal with it. The question is how.
We deal with script failure using two key procedures. It is our facility
with these procedures that differentiates intelligence from stupidity.
Thinking depends on them, and everyone must do them when
trying to think. But not everyone does them well.
The procedures are:
Generalization
Explanation
These are not new ideas in the context of this book. Explanation is
one kind of describing. Generalization is the method by which we do
prediction, make judgments, do diagnosis, and determine causation.
We generalize whenever we try to think.
This entire book is a generalization. It is an attempt to make
sense of a vast array of information. That is what generalization is all
about. The book is also an explanation of the generalizations I have
104 Teaching Minds
found to be true. At the core of thinking, you find generalization and
explanation. But it is important to remember that what starts the
process of generalization and explanation is failure. Without failure
we don’t try to generalize and explain because we have nothing to
generalize and explain. Thinking, therefore, looks like this:
Make a prediction
Prediction fails
Make a generalization
Explain your generalization
Make a new prediction
Let me explain how this works and why one cannot think well if one
cannot do this. To explain, I will tell a personal story relating to my
own thinking and learning with respect to two of the twelve cognitive
processes. I will explain afterwards why I have chosen to tell stories
and why I have used personal ones.
DESCRIBING
Let’s start with describing. There is, of course, an art to describing.
Anyone who writes and anyone who speaks publicly is learning all the
time about describing. Since I was a professor for 30 odd years, and
since I have written hundreds of papers and about 25 books, and since
I have given numerous keynote speeches around the world, I have
been thinking a great deal about describing for many years. I learn
something whenever I speak publicly because I can easily tell whether
I am being heard or not. Are the listeners on the edge of their seats or
are they half asleep (or literally asleep)? I learn when I write because I
read the reviews, and colleagues are always happy to tell me what was
wrong with what I wrote.
Once, I was given a lesson in public speaking by someone older
and wiser than me that I never forgot. I recently had been hired at
Stanford as a professor. I was pretty young (22) and full of myself. In
those days, the Computer Science Department ran a course for new
graduate students that served as an introduction to all the special research
possibilities in artificial intelligence for those who wanted to
enter that field. There were many faculty in AI at Stanford and each
New Curricula for a New Way of Teaching 105
got a week to talk about what his work was about. The goal was to try
to convince students to sign up for a special research seminar with
that faculty member the following quarter.
My champion at Stanford was a psychiatrist named Kenneth Colby.
He invited me to share his week and thus we would be a team for
which students could sign up the following quarter. I listened to his
talks to the students. He was very funny but rather light on content, in
my opinion. I wasn’t impressed. But, after I spoke, he said something
to me I never forgot. He said: If you try to say everything that you
know in an hour, either of two things is true. Either you can do it and
therefore you must not know very much, or else you can’t do it and
you will talk way too fast trying to fit it all in and you will be generally
incomprehensible.
I listened to what he had to say, but I wasn’t sure he was right. At
the first meeting of our jointly run seminar, we discovered that we
had won the student jackpot. While other faculty had gotten four or
five sign-ups, we had gotten 25. I was very proud of myself until the
students went around the table to say who they were and why they
had signed up. Not a single one of them had signed up because of anything
I had said. They had been mesmerized by Colby.
Then I reheard in my head what Colby had told me the previous
quarter. He had entertained them—not overwhelmed them. They
thought he would be interesting and fun and they wanted to work
with him. While I had looked down on his lighthearted presentation
style, it turned out he knew what he was doing.
So, what did I learn and how did my thinking change? Thinking,
as I said earlier, looks like this:
Make a prediction
Prediction fails
Make a generalization
Explain your generalization
Make a new prediction
What was my prediction? I had predicted that speaking quickly with
a great deal of brilliant content would woo the incoming students.
This was simply wrong. I needed to make a new generalization. Fortunately,
I did not have to do the work myself. Colby had helped me by
supplying a generalization that he believed to be true. (Good teachers
106 Teaching Minds
do exactly this: They supply generalizations when a student needs one
and cannot come up with one by himself after a prediction that he
made has failed.)
The generalization that Colby supplied was, more or less: Be entertaining
if you want to attract followers and be listened to.
Was he right? Is this a good generalization?
The next step in thinking is: Explain your generalization. What
this means here is that I needed to understand why this might be
true. If possible, one wants to test out the new hypothesis. Fortunately
for me, I had many opportunities to speak in public over the next 40
years. I tried many different methods of teaching and lecturing. Entertaining
always works. Colby was right.
But I said that one needs to explain why it works. This is my explanation:
People have trouble paying attention to someone who talks for
an hour. Human beings are not built for this. Our ancestors certainly
didn’t listen to lectures. People communicate best by asking questions
and interrupting. Since this isn’t possible in a lecture, any questions
they might ask, they ask themselves and try to answer. While they are
thinking about what a speaker says, the speaker keeps on talking. No
one can really hear a lecture, in my opinion. So, a good speaker, recognizing
this, does not try to make the audience do that. He makes them
laugh, he paints interesting pictures for them to ponder, he amazes
them perhaps, but he does not try to get them to, nor does he expect
them to, remember all that he has said. The less content, the more
likely they will remember. Colby knew all this. I learned it over time.
He jump-started my thinking.
By now, this new generalization of mine, together with my explanation,
is part of my core beliefs. But any belief can be challenged by
reality. And, any new belief generates new predictions. So I predicted
that if I was funnier when I spoke, people would appreciate what I
said more. I also predicted that if I didn’t speak much in a graduate
seminar and let students argue with one another, they would get more
out of it. After 40 years I still believe these things. But, and this is the
important part, there are nuances upon nuances about all these issues
in my memory. I need to explain it simply when I write or speak
about it, but when I think about it, I recall all kinds of exceptions and
caveats. I know, or at least I think I know, a great deal about speaking
and have lots of memories about specific successes and failures.
Thinking and learning require one to recall one’s experiences, analyze
those experiences, come up with new hypotheses about failures, make
New Curricula for a New Way of Teaching 107
new predictions, and be prepared for these predictions to fail. This is
what thinking looks like no matter which cognitive process is being
thought about and practiced.
DIAGNOSIS
Now let’s talk about diagnosis. Again I will start with a story. This is
one I have been telling for a long time because it informs us about
how the mind works.
I was discussing with my colleague (Bob Abelson) how it could
be that my wife could not seem to make steak rare (as I like it—she is
no longer my wife, but not because of this). Bob responded that he
couldn’t get his hair cut as short as he wanted in England 20 years
earlier.
This seems on the surface to be a rather odd response, but when
we look deeper we can see that I was saying something like, She could
do this right if she wanted to, and he was thinking, Maybe she thinks
the request is too extreme, as happened to me with a barber in England
many years ago.
Matching odd situations to other odd situations and seeing the
similarity is what creative thinking is all about. Bob was trying to diagnose
a problem that had been on his mind for a long time, and my
new story provided him with new evidence to think again about what
the proper diagnosis might be. This is what thinking looks like. It is
also what reminding looks like. People get reminded precisely because
they are trying to match a new situation to one they already know
about and thereby determine what to do next. To put this another
way, diagnosis depends on prior diagnosis. We constantly are trying
to improve our diagnostic capability because we always strive to make
better decisions no matter what arena these decisions are in. The fact
that the improvement of diagnostic capability is not explicitly part of
each and every curriculum in school is scandalous.
When we design new curricula, we need to ground them in some
realistic framework that will enable students to practice things that
they might end up doing in the real world. But that does not mean
that the real issue in designing these curricula is anything other than
teaching thinking, that is, enabling practice in the 12 cognitive processes.
However, teaching thinking in the absence of a context that
truly interests students is absurd.

CHAPTER 9
How to Teach the Twelve Cognitive
Processes That Underlie Learning
The art of teaching is the art of assisting discovery.
—Mark Van Doren
We need to completely redefine what we mean by school and what we
do in school. We need to think about education in a new way. Rather
than wanting people to be educated, which usually means being able
to quote Shakespeare or nod sagely when Freud’s name is mentioned,
we need to expect people to be able to think well. Education would be
better defined by defining an educated person as one who can make
well-reasoned arguments for what they are about to do.
We must focus on teaching cognitive processes and abandoning
the subject-based (and test-based) education system we have now that
is clearly failing. The reason we have all those tests is simply because
we have no idea how to make people learn all the stuff that is part of
that subject-based system without threatening them. No one really
wants to learn the Pythagorean theorem or information about the Taft
Hartley Act. Let scholars know about these things; the average person
just doesn’t need to know this stuff.
But all people do know how to find and use the mathematics they
need when they have continually practiced it, and they know how to
find prior relevant experiences when they have to come up with a new
plan they want to propose. That simply does not mean we have to tell
it all to them years before they might ever make use of it.
In the age of the Internet, just-in-time learning is a serious reality.
We can change things now in part because we have information readily
available online. But the Internet has been designed by committee.
It has much in it that is nonsense, and finding what you need just in
time can be quite difficult. Still, it would not be that complicated to design
a different kind of Internet, the moral equivalent of Encyclopedia
109
110 Teaching Minds
Britannica, if you will. It would be filled with knowledge from experts
that had been vetted by other experts and delivered seamlessly without
your having to search for it. You don’t need to be a scholar in order
to make reasoned decisions. You just need to know how to find information
to help you think things out well. This means that learning
to think clearly and knowing how to assess the value of new evidence
that one has found, must be the main goal of any school system.
When does school start to falter? One way to think about school
is to ask about the significance and age-related properties of each of
the cognitive processes. Let’s see if we can rank the processes in terms
of age. Which of these processes would we expect a child who was
entering school to already be able to do? To put this another way, a
normal 5-year-old:
1. can make some accurate predictions about very simple
things, like where his mother might be and what she might
be doing and what will be on television
2. would have trouble modeling any process
3. has a limited sense of evaluation but knows what he likes
4. would be able to experiment with simple things like food and
toys
5. might be able to do limited diagnosis of what might have
gone wrong in a process, but it typically would be limited to
explanations that he had heard from someone else
6. might be able to do limited planning based on plans that had
been used before
7. might know something about causation because he would
have been told about it and remembered what he had been
told
8. can make some judgments based on his own tastes and what
he has been taught about what is good and what is bad
9. can do some describing, but typically is not at all good at it
10. should be able to influence some people, especially his
grandparents
11. should be able to work in a simple team together with kids
his own age toward a goal
12. should be able to do some simple negotiation, especially with
his parents, siblings, and some friends
How to Teach the Twelve Cognitive Processes 111
So, upon reaching school age, a child can do some of the cognitive
processes. Now the question is, How to we teach him to get better at
them?
First let’s see what the cognitive processes fundamentally have in
common. This will give us a way of thinking about how to teach them.
The first and most important thing they have in common is that
they all rely on a case base. We have all engaged in each of these
processes many times and we have a range of experiences we can call
upon to guide us the next time we find ourselves doing them. Each
process relies on a story base as well. We can tell stories about interesting
experiences we have had in doing each one of them. Stories
usually revolve around failures, or at least unexpected results, since
without these there are no good stories to tell. Engaging in any cognitive
process includes the possibility of making a mistake during the
process. We expect to get smarter each time as a result of any mistakes
we make. This is what cognitive processes are like. We learn cognitive
processes through experience and we index the failures we have so
that we can find them again and perhaps avoid making the same mistake.
When we avoid an error that we know we have made previously,
we say that we have learned.
It follows, therefore, that acquiring a case base, learning the stories
of others and learning to tell our own stories, and learning classic
mistakes and being able to analyze behavior to find a mistake are all
aspects of learning the cognitive processes.
Acquiring the case base and consciously analyzing the cases in
that case base, then, is the fundamental issue in teaching the cognitive
processes.
HOW TO TEACH PREDICTION
We live in a physical world but we also live in a social world. Children
need to understand that if they drop something heavy on their foot,
it will hurt; they also need to know that if they do something mean
to someone, the person may dislike them for it and may seek revenge.
We use predictions to figure out what will happen as a result of our
actions and then we use that knowledge to guide us in our future
112 Teaching Minds
actions. What kinds of predictions do children regularly make, then?
And, what can we teach and how can they learn?
1. Children predict the actions of the people that they interact
with, but they do not necessarily realize that they do this.
2. Children predict the reactions of objects and actions in the
physical world, but they do not necessarily realize that they
do this either.
3. Children predict their own feelings and mental states. They
do things that they think will make them happy, but they
don’t necessarily realize that they do this either.
These three worlds, the social, the physical, and the mental, are at the
center of what adults make predictions about. We predict the speed
of an oncoming car and decide whether we can cross the street safely.
Children may not do this so well. We predict what will happen when
we scream and yell at someone, but children may not predict this too
well either. We predict events that will make us happy or sad, such as
taking a nice vacation, or playing a game, or a good meal, or establishing
a relationship with another person. Children do not consciously
think about such things.
But adults do think about these things, so where do children
learn about them? At the present time, the answer is that they learn
about them as events happen randomly in their lives. If they are lucky
enough to have someone helpful to talk with about their experiences,
they may, in fact, become good at analyzing how the world works
and making their predictions conscious. Getting better at prediction
is the cornerstone of living one’s life in a satisfying way. One can, of
course, get better at prediction by simply thinking about it—this is
how most people do that today, of course. But not everyone is capable
of doing that and, clearly, most adults are not all that good at making
important predictions in their own lives. (This is one reason that
there are bad marriages, financial counselors, clinical psychologists,
and prisons.)
What helps teach one to predict is to hear about the experience of
others and to be able to reflect on one’s own experience. This means
that having experiences to reflect upon, and people who are knowledgeable
to discuss those experiences with, is at the cornerstone of
learning to predict effectively.
How to Teach the Twelve Cognitive Processes 113
This should start as early as 1st grade (or age 5—I don’t really believe
in grades).
How do we do it? We design experiences for children that are age
appropriate and talk about what will happen in those experiences before
they do them and how they can learn from and improve upon
those experiences after they happen. Then they undergo a slightly
more complex experience that builds on what they learned. The process
is simple enough. The question is what experiences to design,
how to design them. 2
I should note that prediction is used as a teaching methodology in
schools today, especially in reading. I suggested this in a book I wrote
about teaching reading in 1978, and since that time (not necessarily
because of that book) it has become more common to use prediction
to teach reading. Also, high school kids are asked to make predictions
in courses that cover current events. Kids predict sporting events or
the sex of their in utero sibling.
The idea that kids can make predictions is not a really radical
point. My point is that prediction has to be the curriculum, not be
ancillary to the curriculum. If we want children to predict well, we
need to help them do that. As it stands now, they are on their own.
As adults who have not been taught to predict well, they will make
poor life decisions, predicting wrongly about how people in their lives
(bosses, spouses, children, co-workers, etc.) will behave toward them
after they take certain actions, for example. Yes, understanding is improved
if one predicts the future actions of characters in a book one is
reading—it helps a lot. And, reading is a skill that is very important.
But it also helps to understand how to predict daily actions better and
how to find out whether you were right and how to explain why you
were wrong. Doing this consistently makes you better at predicting
something more important than what an author has a character do
in a story.
How do we get good at making predictions about the outcome of
actions? What outcomes do we need to predict?
Every action we take involves a prediction. When we put one foot
in front of the other in an effort to move our bodies forward, we are
predicting that this will work, that we won’t fall down, and that it
won’t hurt to do this. Sometimes we do fall or it does hurt. We learn
this and compensate next time. This is learning in its most basic form.
And obviously, we have been learning about walking since we were
114 Teaching Minds
very small. We keep learning about walking throughout our lives because
things change.
Babies predict what will happen when they cry. They don’t start
out making such predictions. They learn to expect results that they
have already experienced. These are scripts, and I have discussed them
at length elsewhere. 3 Scripts are acquired naturally as a result of repeated
sequences of events. We predict what will happen next because
we know the script. Scripts are not normally taught because they are
readily acquired from living. But what if you want to teach them? You
might want to teach them in a situation where someone’s new job is
a script and rather than learn that job from repeated experience, there
is a desire to jump-start the process by simply teaching the script. How
can we do this?
There is also prediction that is not script-based. In other words,
there could well be a script but the predictor doesn’t know it. How can
one learn to predict well if one does not have a script?
And lastly, there is often the need to predict when there couldn’t
possibly be a script because what needs to be predicted is novel, at
least to the predictor. How can the prediction be made? More important,
how can someone be taught to predict in that kind of situation?
These, then, are the three aspects of prediction: learning a script;
functioning without a script because it isn’t known; and predicting
when there is no script.
How do we teach these things? Scripts are learned through repetition.
No one seeks to explain to a child that since she will be doing
something again and again, she will now memorize the steps before
she tries it out for the first time. Instead, we take the child through
the steps until she has learned them. There is no need to try to teach a
child scripts (such as how restaurants function or airplane rides go or
school procedure works). We can say some words about these things,
of course, but the learning comes from repeated practice no matter
what we do.
Teachers (and course designers) often fail to understand about
teaching scripts. Talking about what a student will have to do later will
not help that student do what he is being asked to do. We don’t learn
scripts consciously and thus aren’t likely to remember what we are
consciously told about what to do. We are conscious when we execute
a script—we are thinking about what will happen next—but not so
much in words as in expectations of next events. We don’t talk about
How to Teach the Twelve Cognitive Processes 115
one foot going in front of the other foot or think about it very much.
So if a teacher tries to teach us about it, it is doubtful that the words
will help us much. Scripts are practiced. We can prepare for them if
that makes people happy. We can tell a child what will happen when
he rides in an airplane for the first time, but it isn’t so that he can do
what he is supposed to do so much as to make him more comfortable
and less surprised by noises or procedures that might be upsetting.
The only way to really learn a script is to do it again and again.
So what does this tell us about teaching scripts? What if we don’t
have the time to allow a student to practice? This is, of course, what
happens in job-training situations. We tell someone what to do and
hope they very quickly will learn how to do it. We can’t afford weeks
of practice. So how do we teach scripts in that case?
The answer to this depends on the number of mistakes that it
is possible to make while executing a script. The real issue in script
execution is, after all, not knowing the next steps, but knowing what
to do (or what not to do) when the script begins to falter in some
way. This means that a teacher (or a course designer) must have one
question in mind when thinking about teaching a script: What are
the most common (and most important) mistakes that are made by
novices when they execute this script?
To teach people to avoid mistakes in a script, or to teach them to
get out of difficult situations properly, one must practice those situations.
This means that script teaching must focus around errors and
that it is the job of a teacher to systematically make sure that everything
that could go wrong, does go wrong in any practice situation.
Teaching scripts means helping the student form a case base of errors
and a case base of how to handle them. Here again, this cannot be
taught consciously. Script failures must be taught through practice.
The students must build up their own case bases by experiencing the
problem and then either think their way out of it, or learn standard
solutions as a way out of it by practicing them. So, we can speed up
the script-learning process by creating a simulation that has every important
mistake built into it waiting for the student to trip up and
make that mistake.
When a script isn’t known and predictions need to be made, the
usual human procedure is to adapt on old script to serve as a temporary
new one. Never been on a plane but have been on a train? Never
been in a fancy restaurant but have been to a fast-food restaurant?
116 Teaching Minds
Never been to college but have been to high school? These are situations
that happen to people. They usually assume that one is like the
other. While this assumption may not work out all that well, it is the
best a scriptless person can do in the situation.
A teacher, therefore, can take advantage of this human desire to
see unknown A as being like known B. People do this all the time.
They assume the girl they just met will behave like the last one they
went out with. They assume that their teacher will behave the way
their last teacher did. They assume that their new car works the way
the last one did. They do this with situations as well. What it means is
that people predict on the basis of experience and it is the job of the
teacher to help students understand which of their prior experiences
is most relevant when they are confused. Doing this is not so easy. But
it is possible and it is a proper area for a teacher or a course to focus on.
Here is the process. Students need to be asked what to do in a situation
that is new to them. Their natural response would be to rely on
prior experience. The teacher’s job is to make that reliance explicit. To
ask students to say which experience they will rely on for help in the
new situation and ask why they think that that particular choice will
be helpful. To ask them to analyze the differences between the current
issue and the prior script and to predict where the prior script might
not work. To ask them to think of alternative scripts that might help.
What I have described here is the basic process of indexing cases
and matching cases, which is critical to thinking—especially original
thinking. No case is really exactly the same as one that came before.
We are used to partial matches when we use an old case to help with
a new one. What we are not used to is a discussion of why one match
was correct and why another was less helpful. Making case matching
a conscious process helps us understand something that we normally
do without much conscious thought. People need to learn to rely on
partial matches and they need to know how to determine which partial
match is most likely. I have referred to this process as case-based
reasoning in many prior publications. 4 But I have not spoken about
how to teach case-based reasoning. This is where post hoc discussion
is very important. An example of such a discussion can be found in
a book written by Harvard government professors 5 on the failure of
President Ford to choose the right cases to reason from in a crisis. Usually
this kind of analysis is done as kind of an afterthought. However,
teaching people to do it actually is critical to teaching people to predict
properly.
How to Teach the Twelve Cognitive Processes 117
How do we teach prediction when there is no script and there are
no seemingly relevant prior cases? In some sense we can’t. We can teach
people how to go about trying to make predictions. This is actually what
science is about. Scientists create theories that make predictions, which
they then try to verify with evidence. This process—hypotheses verified
by evidence—can be taught in the sense that it is a way of thinking that
can be practiced in various venues. It is reasonable to start teaching children
to think in this way about the world around them. As for adults,
teaching scientific reasoning in the context of corporate training is
probably a less than optimal place to start. People in corporations need
to be able to reason from evidence and to understand what data would
confirm or deny the value of actions they have taken for the benefit of
the company. This, of course, is scientific reasoning. But, unfortunately,
the people who go into the business world tend to have never practiced
scientific reasoning in their educational careers because they weren’t
interested in science. But they must be interested in predicting well in
order to succeed in business.
HOW TO TEACH MODELING
Building a model of a process is very difficult for a child to do and not
so easy for an adult to do either. When you teach computer science,
you learn this quickly enough. Computer programs are models of processes.
People try to write computer programs by creating diagrams
that model what happens first and then what happens next, and so
on. These models are almost never right the first time.
Programmers learn to debug their programs, which means they
continue to try to get their model to be accurate. But most people
cannot do this very easily, and it is a very important skill. Knowing
how to raise money is important if one wants to start a business, for
example. The money-raising process can be understood, but one has
to examine it and go through it.
This is just as true of ordering in a restaurant. Children may seem
to know about ordering in a restaurant, but they may not understand
money or service or having a job exactly and thus may have an erroneous
model of the process. Why does the waitress bring food, is a
complex question for a 5-year-old, but also an important one. Building
models of how and why things work as they do, is significant for
children to learn to do.
118 Teaching Minds
The way to do this is to look at processes that children engage
in during the course of their daily lives and have them first explain
how and why the processes work and then try to improve upon them.
They could carefully examine the operations of the school cafeteria,
for example. Of course, this has to be done in line with the interests
that the children actually have. The idea is building the model, not
telling them what models they have to build. Every child has an interest—animals,
sports, family, cars, dinosaurs, whatever. Children need
to learn to model the processes that interest them in order to better
understand them and to make them better. Children will learn about
the modeling process from working on a car engine, for example, if
they are taught to think about what is going on in a deep way rather
than just learning a set of facts about how the car runs. This is true in
any area of interest, from medicine to government to science.
Adults have a difficult time with models at work, as citizens, at
home, and so on. They don’t always know how things really work. To
get people to be better at understanding the processes that they engage
in daily life, they need to be able to model them. This ability has
to be taught and practiced early..
Of course, kids have been building models of actual objects for a
long time. Using a kit to build a model airplane is fun but it doesn’t
teach you much about how planes fly. More detailed models of physical
objects are very helpful.
Building a medieval castle, for example, sounds like fun to me,
and there are things to be learned from doing this, of course. It is a
good activity for little kids, but modeling involves social processes as
well. Kids need to understand how the world works, so it isn’t the
castle itself that is so important but perhaps a model of the society
around the castle, and the need for the castle, that would matter in
this instance.
HOW TO TEACH EXPERIMENTATION
Everyone experiments all the time. We eat foods that we hope won’t
make us fat or might make us healthy. We take drugs that are supposed
to help us, and maybe they do maybe they don’t. We try out relationships
that may or may not work. We experiment with jobs, hobbies,
How to Teach the Twelve Cognitive Processes 119
homes, cars, toys, games, lifestyles, behaviors, hairdos. . . . The list is
seemingly infinite.
We may not see ourselves as experimenting when we try out something
new. We often experiment ineffectively. Learning to experiment
in a reasonable way is yet again something that can be done only
through experience. But, in this case, there really is an experimental
process to be learned. It can be taught early on by finding simple experiments
that small children really are interested in doing—they do
not all have to do the same ones since it doesn’t matter what they do.
They can learn to attempt to control the variables and see what happens
in a variety of circumstances. This is, again, the scientific method,
but the issue is really not teaching science so much as it is teaching
a scientific approach to gathering useful knowledge. What constitutes
evidence and how to draw conclusions are the kinds of things that a
teacher can help with. Here again, a case base is acquired and relied on
throughout this process. The discussion of findings so that they can be
mentally indexed is very important part of the process.
But how do we find out what is true? Ask any 5-year-old this
question and the answer is not very likely to be, We run an experiment.
Ask mommy, is more likely or, Ask the teacher, if the kid is in
school.
But, testing hypotheses is a critical part of learning to think. Of
course, one has to have a hypothesis first. Children are rarely asked
for their hypotheses about things. This is not exactly odd because although
children do have them, it is a weird kind of discussion to have
with a 5-year-old. Nevertheless, it is important to do. Teaching children
to form and test hypotheses is as simple as asking them to do it.
But, here again, asking them to do it must be done within the context
of something they really care about. There have been many attempts
over the course of educational history to teach kids science by having
them run experiments; sometimes they are asked simply to replicate
old experiments and sometimes they are asked to try new ideas out
and figure things out for themselves. While the latter is most certainly
preferable, these experiments tend to be about testing water quality,
or nutrients in the soil, subjects that are not exactly on the mind of a
5-year-old.
What is on the mind of a 5-year-old? He might be wondering
about how to deal with his baby sister. Or he might be wondering
120 Teaching Minds
how best to throw a rock in order to hit the cat. For many of these
things, hypotheses can be formed, discussions about what makes a
good hypothesis can be had, and ideas for testing out hypotheses can
be sought. Maybe these would just be thought experiments (as in the
case of hitting the cat), but thought experiments are important also.
Knowing what would confirm or deny a hypothesis is important.
Finding out what kids want to get better at is part of the issue here.
If they want to learn to hit a baseball, having a theory about what
makes a good swing and what makes a bad one may matter to them.
The subject matter doesn’t matter at all really, just the thinking and
the experimentation.
This suggests that the real way to teach experimentation and the
other subjects we have discussed here is to group kids not by age but
by interests. So, if a child wants to think about dogs all day, group him
with a set of other dog lovers and start coming up with hypotheses
about dogs’ behavior, needs, commands that they might learn, training,
breeding, and so on. It just doesn’t matter what the subject is at
this stage.
Suggestions for kids’ science experiments are everywhere. Here is
a typical one:
Gravity
The Earth tries to pull everything down toward its center. This
pull is called the force of gravity (the invisible force). When you
lift things up, you have to pull against gravity. If you drop a
pencil, gravity pulls it to Earth. If you rest its midpoint on your
finger, gravity will pull down equally on both sides of the pencil
and it will balance in the air.
Here is another:
Sound and Noise
Have one person fill each of the plastic eggs with a different item.
Put some rice in one, some dried beans in another, and so on.
Keep track of what you put in each egg by writing numbers on
the eggs. Have a different person try to see if he can figure out
How to Teach the Twelve Cognitive Processes 121
what is inside each egg by shaking and listening to the sound
generated. After he takes a first guess, show him the list of what
items are in each egg and have him guess again. See if he changes
his mind about some of the previous guesses. Now open the eggs
and see how close the guesses were to what was actually inside
each egg.
Now, let’s make a musical instrument called a kazoo. Cut a
small square of wax paper about 1 inch larger than the end of the
cardboard tube. After doing that, wrap the wax paper over one
of the ends of the tube and put a rubber band over the paper to
hold it in place. Now, put the open end of your kazoo up to your
mouth and hum a tune into it. Notice how the kazoo buzzes and
vibrates to amplify (make louder) the sound of your voice.
Of course, this is not exactly what I have in mind when I suggest focusing
school on experimentation. Say the word experiment and the word
science usually follows right away. This is unfortunate because most
kids won’t become scientists any time soon. Also, most kids already
know what they are being asked to “experiment” about. They know
the pencil falls and they know you can make noise by blowing through
things. They may not understand how all this works, but they won’t
understand much more after doing these so-called “experiments.”
Experiments in the larger sense are about attempts to find out
what is true about things you are uncertain about. The issue is how
to do that when you are wondering about something. To do this we
need to constantly deal with what kids are worrying about and ask
them to determine how they can find out what is true. This is experimentation
that can be helped by teachers. It needs to be individually
focused, however. You can’t have a class worry collectively about any
one thing. Each kid has his own concerns.
HOW TO TEACH EVALUATION
For an issue that is so important to so many people, it is astonishing
how difficult this seems to be. Just say no to drugs campaigns don’t
work and people wonder why. Abstinence campaigns don’t work
and people wonder why. It is not all that mysterious. You can’t teach
122 Teaching Minds
evaluation verbally. Since children naturally copy their parents, it is
not that complicated to figure out where children acquire their initial
values. And, since values are not typically stated (my father never said
liquor at 5 p.m. is a good thing), it certainly isn’t through lectures that
we learn values. I learned to drink at 5 p.m. from my father. He wasn’t
trying to teach me that. I learned to gamble from my father. He wasn’t
trying to teach me that either. I learned to be argumentative from my
father. He wasn’t trying to teach that. I did not learn algebra from my
father. That, he was trying to teach me.
My son wanted to grow up and sip Pepsi. He told me this constantly
as he was growing up. It did not matter that I hadn’t drunk Pepsi for
10 years at the point when he was saying that. He was impressed that I
had done this when he was 3 and he was frustrated that I didn’t allow
him to drink it. Presto, another family value is learned.
Children learn the family values that their family actually has.
Teenage mothers who warn against getting pregnant at a young age
may say that in words, but their actions say that their kid is alive and
well and it all worked out.
So, it follows that we don’t have to consciously teach values because
we teach them without saying a word. Values are held subconsciously
and learned subconsciously. We can only hope that we have
set a good example. That having been said, there will still be those
who ask how we can teach values. You can’t expect that “you can’t,”
will work as an answer.
I mentioned in Chapter 3 that there are lots of things that you
can’t teach. I mentioned honesty as an example. Honesty is, of course,
a value. Now let’s ask whether you can teach people to be more honest
than they are naturally inclined to be. The answer is that to do
this, you have to turn a subconscious process into a conscious one.
You would need to provide case after case and experience after experience
to a student that all led to the conclusion that honesty simply
works out better in the long run. This is, of course, what abstinence
and say no to drugs campaigns endeavor to do. They want to argue
kids into believing that things they think are fun, are bad. But how
can we make that argument? The argument we can make is that these
things aren’t as much fun as you think, so try them and see for yourself—but
that kind of ruins the basic premise about not doing them
in the first place. So, in principle there is no way to argue kids into
not doing what looks like fun to them and what doesn’t seem to have
How to Teach the Twelve Cognitive Processes 123
hurt anyone very much. The idea of showing pain is big in values
campaigns. Campaigns against drunk driving like to show students
dead drivers and awful car crashes. But they miss the real point. Have
the students seen their parents drink and drive or friends drink and
drive? Are they dead? If not, these campaigns will have little impact
since values are subconsciously held.
There is a way to teach these things but it isn’t easy. Imagine that
you wanted to teach teenagers not to drive drunk. You could create a
simulation of drunkenness that asked students to drive drunk while
not being able to hold their heads steady, blacking out from time to
time, and seeing very badly. In other words, instilling a new emotion
into the mix can alter values. Make people afraid of something they
want to do and that fear will manifest itself when it is time to do it.
Emotions can be induced into subconscious processes and decisionmaking
through experience. Emotions can change values.
There is a sense in which appreciation cannot be taught. You like
it or you don’t. I have two grandsons. The 5-year-old (my daughter’s
child) has announced that he doesn’t want to have anything to do
with a ball. The 3-year-old (my son’s child) goes wild with excitement
when he sees ballgames being played and responds excitedly when
balls are given to him. What is the difference and how did this happen?
The difference is obvious: One appreciates the art of it and one
doesn’t. How this happened is less clear, but the parenting is very different
with respect to balls in each house. There are other, biological
differences as well. The 3-year-old has much taller parents and is
already the same size as the 5-year-old. He is much more physical as
well.
So, the question is, Could we teach the 5-year-old to love balls and
ballgames, and could we teach the 3-year-old to hate balls and ballgames?
The answer to this is obvious. We could do this. It might be
hard but people are, unfortunately, quite good at negative reinforcement,
so it is possible to make people change their attitudes by using
it as a method.
Is there a way to teach the positive? Can we get someone to appreciate
a work of art who does not respond immediately to works of
art? Can we get someone to appreciate classical music who does not
have any interest in it?
This is, of course, what many art history or music appreciation
courses endeavor to do. Their methodology is always the same:
124 Teaching Minds
repeated exposure and analysis. It always involves a lot of talking. And
therein lies the problem. Since appreciation and enjoyment are subconscious
processes, it stands to reason that these processes are best
addressed by a methodology that is less conscious. Telling me why I
should like something is not likely to do much more than teach me
how to talk about something as if I liked it. I can point out the finer
points of baseball to you, but if you don’t care, I won’t be able to get
you to go to the next baseball game.
On the other hand, of course, you certainly won’t like baseball if
you have never seen it played. Exposure is the key to teaching subconscious
processes. Add to that an enjoyable atmosphere surrounding
the experience, especially if it is early on in one’s life, associated with
the artistry you want to teach, and it is likely that the learner will learn
to like whatever he is being exposed to. So, evaluation, which starts
out as a subconscious process, must be taught by enabling copying
and repeated practice, but cannot easily be taught verbally.
HOW TO TEACH DIAGNOSIS
Diagnosis is the same and it is different. Many different people do diagnosis
under many different circumstances. But, the process is always
the same. So it seems normal to ask why an expert in doing diagnosis
in one area cannot do diagnosis nearly as well in another area of
knowledge. Why can’t a doctor fix his own car? Why can’t a detective
figure out why a business is failing? It is all diagnosis after all.
Diagnosis is best taught early on but it can be taught at any age.
In the end, it is just about knowing what counts as evidence and how
to create and rule out hypotheses. The general process of gathering
evidence and testing hypotheses is the same no matter what you are
doing. This is true in principle, of course, not in actual fact. In reality,
doctors cannot rule out hypotheses by tests that might harm the patient.
Businesspeople cannot rule out hypotheses by running control
businesses that may lose money while others make money. Detectives
cannot spoil a crime scene by altering the evidence. Mechanics cannot
try something that might make things worse. Investors cannot control
world events that might make a seemingly correct hypothesis still not
work out all that well.
To teach diagnosis, simple problems can be worked on that lead to
more complex problems. What is being taught, apart from the process
How to Teach the Twelve Cognitive Processes 125
itself, is the knowledge that underlies the hypothesis and understanding
what constitutes evidence and the consequences of evidence. This
knowledge is very domain specific and is the reason that doctors don’t
necessarily make good car mechanics, and vice versa. This knowledge
can be acquired only through practice and experience and can be acquired
only consciously. Diagnosis is thus a conscious process that is
very knowledge dependent. There is no seat-of-the-pants diagnosis,
namely, diagnosis that is seemingly subconscious, although it may
well seem that way. Having an intuition is usually just the result of
having a great deal of experience, so much so that hypotheses just
jump out at you because similar cases are so easily recognizable to an
expert. Someone who is good at diagnosis would be good at diagnosis
in any domain of knowledge if they knew how to gather and interpret
evidence in that domain.
Diagnosis is clearly very difficult to learn. Most people are rather
bad at it outside of their own areas of interest. Even inside their knowledge
base they can be sloppy in the reasoning and leap to wrong conclusions.
This is true of all analytic skills. It is possible to never learn
to do them well.
HOW TO TEACH PLANNING
Planning is extremely important and typical of an analytic cognitive
process; it is something that some people simply never learn to do
well. Teaching planning must be focused around the assembly of a
case base. Planning is taught in many domains of knowledge and is
almost always taught wrong. The classic error is to teach the theory of
planning, means–ends analysis, a theory of urban planning, spatial
planning, military planning, and logic-based artificial intelligence
planning. Such courses all make the same mistake. Course designers
think people use theories when, in fact, when people plan, they simply
try to adapt old plans that have worked before to new situations.
Often people don’t plan at all. They simply assume that a set of actions
they intend to perform will work to achieve an expected goal.
Teaching planning is therefore a tedious process that is best begun
in childhood. It involves making plans, seeing how they play out,
and performing an analysis of what went wrong. Often people are
not even aware that they have made a plan, and are simply frustrated
when things go wrong. They almost never perform an after-action
126 Teaching Minds
review (as the military calls the analysis process that takes place after
a plan has been implemented).
Teaching planning means teaching about goals, how they typically
are achieved, what obstacles might be encountered, and how to
deal with them. The principles of planning are the same no matter
what domain you are in, so children can learn to plan birthday parties,
hikes, class trips, how to deal with their problems, how to treat
their little sister, how to get along with their parents, and so on. This
process can be learned by copying, seeing how others deal with these
things, adapting a plan that has worked before, and so on. A teacher
who tries to teach planning from first principles is teaching something
that is easily forgotten as well as teaching a process that doesn’t occur
that much in real life. Chefs adapt old recipes or parts of old recipes.
Generals adapt old battle plans or parts of old battle plans. Computer
programmers use code that has worked before. Planning without a prior
plan in mind really is quite unusual and generally not a good idea.
Planning, like diagnosis, should be a basic part of all curricula in
school at all ages. People make plans all the time. They plan their
lunch, they plan their day, they plan their trips, they plan their errands,
and, of course, they plan their lives. It is astonishing that we
don’t teach planning all the time in every aspect of life. But we don’t
because this doesn’t seem very academic. Since it is not explicitly
taught to children, it is reasonable to ask how we might best teach
planning to adults. Corporations want to teach financial planning,
resource planning, supply chain planning, creating business plans,
creating marketing plans, and so on. Planning is, in fact, one of the
major preoccupations of business, as well it should be. So, how do we
teach it?
The problem here is that planning really works in only one way.
It is relies on a case base. We plan by adapting old plans. That’s how
we do it. We store old plans and we retrieve them when we need them
again; we change them so they apply to the new situation or change
them so that this time they will work out when they failed before. But
we always start with an old plan. New planners, those we are trying to
teach how to plan, cannot help but do this, even if they do not have
a relevant old plan to work from. They simply will choose the best
plan they have, even though it might not be all that germane to the
current situation. Proverbs—for example, to a man with a hammer,
everything is a nail—don’t come from nowhere.
How to Teach the Twelve Cognitive Processes 127
Thus, when we teach planning, there is either a lot to undo, or we
must start from the beginning. We can try to explain why each and
every old plan is not really helpful in a new situation, or we can teach
a series of plans that are relevant. In other words, if you are trying to
teach people to write a business plan, you need to start with a lemonade
stand and work up. If you are trying to teach financial planning,
you need to start with a child’s allowance and work up. If you want to
teach battle planning, try a tug of war first. This is what should have
occurred in childhood. If it didn’t, it needs to be restarted that way for
adults. We need to use, again and again, plans in different situations
that are simple and begin to analyze why they fail. (And these plans
must fail, at least in simulation, or no real learning will occur.)
Planning is very difficult. It must start simple and be practiced
simply for a while or it never becomes second nature. Plans must fail,
at least in simulation, because analysis of what went wrong is a critical
part of planning. If you aren’t analyzing what went wrong, you aren’t
learning to plan. Your case base will not end up having been indexed
well enough to enable you to pick and choose appropriate plans in
the future.
HOW TO TEACH CAUSATION
At the root of diagnosis and planning is causation. Detecting cause
is an essential part of diagnosis, and anticipating cause is an important
aspect of planning. Causation must be understood in order to do
many things in this world. One needs to know what causes what. Science
courses in school attempt to teach causation by having students
memorize F = ma, or having them imitate chemistry experiments, or
having them dissect a frog. While there is nothing wrong with any
of that in principle, it really doesn’t teach causation in a way that is
particularly useful to a functioning adult.
While diagnosis and planning may not be recognized as critical
skills by schools, causation is, although not under that word. Causation
is understood as being what science is all about, and when
schools endeavor to teach science, they are in fact trying to teach
causation. This is true for social science as well. History is about causation,
as is psychology. The fact that these subjects are not talked
about in this way indicates something important about them.
128 Teaching Minds
Subject- based education makes the academic disciplines the center
of what needs to be learned, when there is really something else at
the base of learning. All human learning and all scientific inquiry is
about causation; attempts to determine what causes what, and why,
are what it means to be a scientist or an academic. Theories of causation,
and tests to see whether those theories are true, are what it
means to be a scientist.
The problem is that telling students that causation is part of science
makes them think about physics formulas and fact memorization
when the real issue in teaching causation is how to determine what
causes what, rather than how to memorize what causes what. There
are, of course, facts about causation that are worth knowing. It is nice
to know that if you let go of something, it may fall, but it is not necessary
to know that gravity is the issue in this. The world went on quite
well for millions of years before Newton. People certainly understood
before Newton that things would fall when you let go of them and
nothing else was supporting them. Scientific explanations of causation
are nice for scientists but not necessary for everyday humans.
What everyday people need to know is how to determine what causes
what in areas of their own interest. They can hear you tell them about
causation—the stock market always goes down when a Democrat is
elected president—but they need to be able to decide whether what
you said is true and whether it is the election that causes the decline
or something else.
Understanding about causation is much more a function of being
able to figure out what caused what in any given instance than it is the
memorization of facts about science. Of course, with known cases, as
we have seen, being able to extrapolate from one case to another is a
good way of determining what is likely to happen. There is no harm in
knowing prior cases and great value in being able to use them. But, as
always, cases are better learned from one’s own experience than from
being told about them.
Teaching causation, therefore, means teaching the process of determining
what happened in any given situation. Since there is a
great deal to learn about any domain of inquiry in order to determine
causation, the main issue is how to know what the facts are
and how to reason from them. This means that, yet again, it is the
domain of knowledge that needs to be learned, and this entails constant
practice in that domain. And, the methodology of determining
How to Teach the Twelve Cognitive Processes 129
causation needs to be learned. This, too, can depend on the particular
domain of inquiry. Reasoning from all this takes practice as well.
Determining cause is a critical cognitive process that underlies nearly
all thinking.
HOW TO TEACH JUDGMENT
How do we get good at making a judgment? Judgments are a kind of
prediction, of course. When a judge sentences a criminal, he is, in a
sense, making a prediction about what will happen in the rest of this
person’s life. But he also is making decision that is no way a prediction,
but simply serves as punishment. Similarly, when we decide that
a certain restaurant is our favorite, we are predicting something about
how much we will enjoy future experiences, but we also are making
a decision that may or may not matter to others, that is, a recommendation.
Recommendations are also predictions, but they have a
different feel. When a boss decides whom to promote, he is predicting
something about future behavior but, again, the prediction isn’t the
key point. A judgment is a decision that has some import.
Nevertheless, as different as judgments and predictions may or
may not be, the process of teaching them is identical. Good judgment
is learned by making judgments and analyzing the results or
truth of those judgments as more information becomes available. After
a judgment is made it too becomes one of one’s cases and stories.
Cases about judgment can be learned only by making simple judgments
and getting smarter about the process over time on the basis
of experience.
Judgments can be made in two ways that matter here. Either someone
can decide to do one thing versus another thing based on ethical,
moral, avaricious, or emotional grounds, or for many other reasons.
Judgments aren’t so different from decisions in this aspect. Teaching
someone to make a judgment of this sort, between A and B, can be
done by putting students in situations in which such judgments need
to be made and then going over with them how they decided and
what they may have left out of their thinking.
A different kind of judgment is made when one judges the behavior
of others. Judges do this, of course, as do compliance officers in
companies, and teachers with respect to student mistakes.
130 Teaching Minds
Children normally make judgments about the behavior of their
parents and friends as well. In all these cases, judgment is best taught
by having a student watch the behavior of others, keeping himself out
of the issue and seeing what factors motivated and determined the
behavior of the actors. It is a lot easier to teach proper behavior when
it is not one’s own behavior that is being judged. One can learn to act
by judging how others act.
So, children make judgments all the time. Typically those judgments
reflect the values that they have been taught at home. Children
decide what is good and what is bad mostly based on what they have
been told. No child discovers for himself that George Washington was
an admirable man. No child decides on his own that the United States
is the best country in the world. These things are taught by parents
and by schools. School, to the extent that it serves as a place of indoctrination,
has always succeeded at producing citizens who believe
what they are taught to believe at a young age. There obviously is a
great deal of sentiment for keeping indoctrination as a key part of
education, but teaching judgment means allowing children to come
to conclusions based on their own experience and not merely what
they were told.
Learning to make judgments is a process of deciding for oneself
what is true, which is, of course, not so easy. This should be the role of
school but it usually isn’t. School wants to teach us the truth when, in
fact, truth is best discovered, again, from experience.
How would one discover the “best country in the world,” if that
is a meaningful idea, or whether George Washington was all he was
cracked up to be? Obviously, travel helps teach one about countries.
Kids can learn about countries by simulated travel in the modern era.
But the point wouldn’t be so much to teach them that they make good
cheese in France, which is the kind of thing school does today, but to
think about what makes France different from the United States. Similarly,
we can read and learn some facts about George Washington, and
these are indeed taught to children in primary school. I do not believe
that children are equipped at a young age to determine for themselves
whether Washington was a good man. Perhaps that would be a worthwhile
assignment in high school, as long as students were interested
in the question and were allowed to come to any conclusion that they
could reasonably defend. But children of 5 or 6 can understand what
How to Teach the Twelve Cognitive Processes 131
a hero is and what leadership is, and they can determine for themselves
who they know or have seen who is good at both. Again, real
experiences and discussions are how one learns to think about this,
but it must be done in an environment of possible truths, not predetermined
truths. One doesn’t create a nation of people who can think
by telling them what they should think. Kids know who is the leader
among them. They should learn to discuss what it is about their leader
that makes them want to follow. This is difficult discussion to have
with a 6-year-old, but it gets easier with age.
Once again, asking kids to make judgments isn’t that unusual.
Here is a remark from a parenting book that I happened upon:
When adults praise their kids for smart judgments, the kids glow.
But here’s the clincher: kids earn more and more freedom and
independence when parents trust their judgments more and
more. 6
The issue here is, yet again, not whether kids make judgments, but
whether they are taught, as a central part of what they study, how to
make judgments.
The cognitive processes depend on reasoning from evidence in a
way that makes sense. This is not something people are naturally good
at. They often exhibit faulty reasoning. Practicing reasoning means
practicing within particular domains of knowledge. Reasoning is the
same process no matter what you are reasoning about, but we don’t
reason about nothing. Learning the actual facts is important, but it
is the idea that this is important that has sent the school systems on
the wrong path. Academics study the facts, as well they should, but
they also teach the facts, which is a grievous error. How to determine
the facts and how to determine their effects on a situation is what the
processes of diagnosis, planning, and causation are all about.
HOW TO TEACH INFLUENCE
This is yet another childhood skill. Children learn how to influence
their parents and their siblings and their friends very early on. Of
course, they may not learn these things in a good way. They might
learn that temper tantrums or bullying works very well.
132 Teaching Minds
Teaching people how to influence people involves putting them
in situations where it is possible to influence people and seeing how it
goes. There is nothing to learn exactly. We try behaviors and see what
works. What works for one person may not work for another. One way
or another, we learn how to get what we want, or we learn to hang
out with people who will respond to our needs. This is basically a subconscious
process. We are so busy working on this at a very young age
that we may not have any idea what it is we know or how to improve
what we do. Of course, there is a conscious part as well. Someone can
tell us that we will catch more flies with honey than with vinegar,
and we can, if we think about it, adjust our behavior. But aphorisms
about what works and what doesn’t work aren’t always correct and
are highly idiosyncratic. Although there are books about how to win
friends and influence people, the reality is that, apart from adopting a
phony personality, people are who they are, more or less.
But this does not mean they can’t be taught what works. It usually
does mean that they can be taught what works as a result of their own
experience. And, they can be taught what doesn’t work as a result of
their own experience. But this isn’t at all easy. If it were, psychiatrists
would not be able to make a living. You can tell a person to change
his behavior, you can even tell him exactly what to do when, but he is
likely not going to be able to do what you say.
The way influence is taught currently is probably the way it has to
be taught, then, by use of mentors who look at your behavior and talk
you through why you do what you do. This same mentoring method
can be used in corporate settings. Simulations may not be so effective
because while we may know and be able to say the right answer, this
doesn’t mean that we can execute the desired behavior in reality.
HOW TO TEACH TEAMWORK
Teamwork is learned by working in teams. It is a mixed process because,
here again, we behave in ways that are not so conscious but
we can make conscious changes. Leaders learn to manage teams by
thinking about what works and what doesn’t. Quarterbacks must
learn to manage their teammates. If they don’t, it really doesn’t matter
how well they can throw the ball. Team members have to want
to work hard for the leader, and the leader has to know how to motivate
each team member. People are different so what works for one
How to Teach the Twelve Cognitive Processes 133
may not work for another. A leader learns to figure out who is who
and what works for each member of the team. The conscious part of
this is about learning who needs what from the leader in order for
the leader to get the most out of each individual so that the team’s
goals are achieved. The subconscious part is about interacting with
others, which is rarely conscious behavior. We get smarter through
experience. If our team wins because we functioned well as a team,
we learn to repeat the behaviors that worked. If we win because our
team was simply much bigger than the other team, then we probably
won’t learn much about teamwork.
Teamwork can be taught only by examining how a team functions
and attempting to make conscious the subconscious behavior that is
not working. Thinking about what we have done that may not have
been helpful to the team, and making sure that team members’ goals
are aligned, is pretty much the only way we can learn to improve our
behavior.
HOW TO TEACH NEGOTIATION
My daughter was a little over 2 when we moved back to the United
States from Switzerland. The enormity of U.S. toy stores overwhelmed
her and it seemed that she wound up crying every time we entered
one. She wanted everything. So I had what I thought was a clever idea.
I told her that she could have two toys of her choice but that if she
cried she couldn’t get any at all. We talked about it and it was clear that
she had understood what I said. She ran around the toy store and ended
up selecting three toys. I told her one would have to go back—that
our agreement was two. She started crying hysterically. I then said she
had to put them all back as she had violated our no crying agreement.
All of sudden, she sucked up all her tears and said in a breathless voice:
I’m not crying now. I said that we would compromise on one toy.
That was possibly her first lesson in negotiating. I say possibly
because kids and parents negotiate all the time. She and I are still negotiating.
Now it is about when she will come to visit or when she will
send her son down to visit or a range of other family issues.
Negotiation is so important that it is nearly absurd to ask how we
teach negotiation. We can learn it by copying, of course, which I did
when I watched my father get a good price on a used car I was buying
that I was ready to pay much more for. But really we negotiate with
134 Teaching Minds
our wives and children and friends and co-workers all the time.
It is possible to teach negotiation, of course. My team once built
a course on negotiation, working with a Harvard Law professor 7 who
taught negotiation. The course worked by having people negotiate.
The situations were artificial so there is some question as to how valuable
lessons can be learned from negotiating when nothing important
(except ego) depends on it. What I found most interesting about that
course were the stories that the expert told from his life as a professional
negotiator. I can’t say that I was ever able to personally make
use of the lessons that those stories taught, but other people’s experiences
are interesting to think about. In the end, what we really know
about negotiation is what has worked well for us in the course of our
lives when we were negotiating. Coaching can help, of course, which
implies that the best way to teach negotiation would be with a mentor
watching you do it for real and offering tips. Psychologists perform
this service in couples counseling, and presumably real estate agents
perform this service for homebuyers and sellers. Just-in-time advice is
always helpful.
HOW TO TEACH DESCRIBING
There is a famous quote: “I apologize that this letter is so long—I
lacked the time to make it short.”8 As long as people have been talking,
they have had to learn to talk well. When they learned to write,
they had to learn to write well. Communication is a very big part of
living in a society and those who communicate well gain all kinds of
advantages. It is difficult to attain public office without speaking well,
or to become an important academic without writing well, or to make
sales or convince anyone of anything without making your case well.
This takes practice and coaching, and there is no substitute for either.
One also has to have something to say, so this means one has to
have had experiences to talk and write about. Further, it helps if one
is writing about something that one is passionate about. Asking kids
to write about their summer vacations doesn’t necessarily make them
into good writers. Asking kids to give speeches about George Washington
fails for reasons of lack of passion. People need to learn to describe
well what is most important to them. And, they must be doing this
How to Teach the Twelve Cognitive Processes 135
in every task they undertake. They must talk about and write about
what they do until the description process becomes second nature to
them. So, describing cannot be taught in and of itself. It must be part
and parcel of other events students undertake. Writing classes make
no sense, therefore. They exist because of the subject-based divisions
in school. Writing and speaking must be part of everything that is going
on.
SUMMARY
Proficiency at all the cognitive processes depends on discovery and
being able to extrapolate from one’s experience about what has been
discovered. These processes depend strongly on prior cases, and prior
cases are best learned slowly in childhood. They also depend on an
analysis of those cases, which is best done with help from a teacher.
Discussion, reflection, and analysis of prior cases make one better able
to deal with new cases. New cases must be compared with old ones
in a way that helps one reason better from them. This comparison is
the basis of teaching analytic cognitive processes. Learning cognitive
processes means having prior experiences with events that are similar
to current events and being able to extrapolate from them. When we
go to a doctor, we want one who has seen our problem, and described
it to others, many times before. Only then can we detect the nuances
of difference that will determine an effective course of action.
Teaching cognitive processes means providing students with experiences,
hopefully each one more complex than the one before,
and helping students discuss those experiences and compare one with
another. Knowledge is experience, but it is experience that has been
analyzed so that it can be retrieved again just in time as needed. This
will happen only if we have thought about what we have experienced.