Why Educational Reform Needs a New Theory of Mind

Incogito nullo cupido. (We cannot desire

what we do not know.)
—Latin proverb

Nothing is possible, except to extend the area

of sanity little by little.
—George Orwell

The century-long failure of wave after wave of educational

reform is now such common knowledge that those of us who persist
in trying to design better ways to educate appear to many of our
colleagues as if we are engaged in idle academic amusement. Serious
reformers are elbow-deep in politics. There is no question that
education is political and that any reform that stands a chance must
have a power base. Education also seems to be so constituted that
change either from inside or outside the profession is unusually
difficult to bring about. But there is a prior question: Is the proposed
reform a good idea? Not, is it any good at all, for even the flakiest
innovations usually have some good in them. Is it good enough to
justify the price—a price that must be measured not only in money
but in all the effort and disruption and personal loss that go to make
up what Eric Hofer aptly called “the ordeal of change”? And is it
good enough to meet the need? That is a question of such stupifying
complexity that it is almost never raised.
Although the term ‘reform’ is used to cover all kinds of changes, it
will help to sharpen our focus if we distinguish reform from
transformation. Transformation is what I was talking about in the
preceding chapter: transforming education into a modern,
progressive profession. Medicine underwent such a transformation
in the the 19th century. Twentieth-century efforts to provide
universal health care and recent efforts to restrain its escalating
costs are reforms, having only indirect effects if any on the nature of
the medical profession. Reforms can be imposed from outside, and
sometimes have to be, whereas transformation can only come from
within.
Education has experienced reform but not transformation.
Univeral access to education is by far the most important reform of
recent educational history, and it is well advanced in countries where
universal health care is still far off. But teaching is still the same
traditional craft it has been for centuries. Indeed, if you look at the
Jesuits’ Ratio Studiorum, which dates from 1599, you may get the
impression that the craft has gone downhill since.
Despite the use of phrases like ‘breaking the mold,’ practically
everything going on in education today has to be counted as reform
rather than transformation. Almost all the reforms are aimed at
improving the quality of education. The stimulus often comes from
unfavorable reports of achievement test results or complaints from
employers or parents. But, with one major exception, the reformers’
approach to improving educational quality is quite indirect. It may
consist of change at an administrative level, such as firing the school
superintendent, turning the school over to a private firm, or
instituting site-based management. Alternatively or at the same
time, reform may involve try-harder strategies such as
accountability schemes or programs like the Georgia HOPE
initiative, which rewards good student performance with free
college tuition and textbook allowances. Irrespective of the wisdom
of particular reform schemes, we must acknowledge that better
management and increased motivation could improve the quality of
education, but what actually happens at the classroom level is not
itself part of the reform agenda.
The one notable exception, of course, is back-to-basics reforms,
which frequently specify not only what is to be taught but how (and
which may even go on to specify what not to teach and how not to
teach). As contrasted with the other reforms, which are
administrative, this is instructional reform. This is the kind of reform
I want to focus on here, because it is primary. The administrative
reforms can affect the quality of education only to the extent that
they somehow induce instructional reforms, and one of the criticisms
of accountability schemes, for instance, is that they induce
instructional reforms of a detrimental kind. Many instructional
reforms of limited kinds are taking place all the time in education.
Most of these originate within the profession and they range from
sober curriculum changes to sheer quackery. Back-to-basics
movements are often a massive revolt against the cumulative effects
of these piecemeal reforms.
The question I now raise is, How successful can instructional
reform be without transformation? Surely it can have some success.
We are not hopelessly in the dark. But I shall argue two points: First,
instructional reforms are not neutral with respect to transforming
education. They either nudge it in the direction of
modernization—increasing its capacity to solve problems and
address new goals—or they further solidify its traditional craft
character. ‘Back to’ movements are thus detrimental to the long-
range future of education, and this is as true of movements ‘back to’
child-centered education or constructivism as it is of ‘back-to-basics’
movements. Second, reformers are likely to fail in even their
immediate objectives if they do not become more deeply engaged
with the unsolved problems of pedagogy.
‘Best Practices’ Are No Help If You Don’t Understand
Them
The term ‘best practice’ is often heard in educational reform talk
these days. It is a conservative approach to improvement. It is about
making optimum use of the knowledge already implicit in practice
rather than flying off on the wings of some theory or radical notion. I
cannot think of any well-supported reform movement that is not
grounded in this idea. Sometimes the reform is aimed at a specific
‘best practice,’ such as cooperative learning. Other times it is aimed
at creating a process, such as a teacher network, devoted to
identifying, spreading, and refining best practices. Even the reforms
that deal only in accountability and standards must, to make sense,
presume that the knowledge of how to meet the standards already
exists. Although that is a questionable assumption, it is reasonable to
assume that knowledge of how to do better always exists
somewhere. But that does not mean that the knowledge of how to
use that knowledge exists where it is needed.
In the business world, ‘best practice’ is turning into a creative
discipline in its own right (O’Dell, Essaides, & Grayson, 1998). A
medical center imitates the registration practices of a major hotel
chain. An airline speeds up its maintenance procedures by studying
the practices of auto racing pit crews. However, such creative
copying requires, according to O’Dell (1994):
Knowledge of your own process and problems to see the
underlying characteristics and spot analogies across other
industries.
Ability to look for common themes without having them spoon-fed
to you.
 Information about outstanding companies in other industries.
This information needs to be organized by process and not only
by industry.
Language that is common to all. One of the barriers to sharing
across industries is finding a common language that describes
processes regardless of industry. For example, the medical
center had to equate its admissions process with hotel
registration in order to see the analogy and potential transfer.
These are conceptual requirements, and education fails on all of
them. Knowledge of processes and problems is so shallow that
educators have difficulty seeing useful analogies between one school
subject and another or between one kind of student and another, let
alone analogies that reach outside the classroom. That has always
been one of the frustrations for me in teaching university courses for
teachers. Teachers of different subjects and grade levels find little in
common except problems of classroom management. The common
themes they can recognize are either at a vacuously high level of
generality (learner-centered versus teacher-centered) or else they
are political. The information flow on which ‘best practice’ depends
is, as everyone in education recognizes, terribly deficient. And the
common language, once the jargon is translated, is just the language
of folk psychology and folk epistemology.
These conceptual deficiencies are all evident at this writing in a
growing movement to adopt phonics as a ‘best practice’ in teaching
reading. There is, of course, vehement opposition to the claim that
teaching phonics is a desirable practice (Goodman, 1998). Tempting
as it is to get into that argument here, it would take us away from the
point I want to pursue, which is the need for a better conceptual
structure for educational reform. We must not ignore, however, the
fanaticism that has been so prominent in debates about reading.
Fanatics have often been not at the fringes but at the center of
reading reform movements. Conceptual equipment is surely in very
bad shape when sensible people cannot tell the lunatics from the wise
counselors.1

1
A few examples of lunacy: At a meeting of the Reading Reform Foundation
there is no demur when a member proclaims that no one can become a good
reader who is not taught phonics. At a scholarly conference, after a
presentation of research on the importance of teaching letter-sound
correspondences, an official of the International Reading Association declares
“We can’t really go along with that; we believe in comprehension.” A leader
 Conceptual inadequacies become immediately apparent if we ask
what this ‘best practice’ is that is proving so controversial. Calling it
‘phonics’ isn’t enough, because when cornered every reading teacher
can claim to be teaching phonics already and every publisher can
make a similar claim for their reading program. So the legislators
and school board officials who are trying to lay down the law add
terms like “systematic” or “research-based” or “intensive.” But these
terms are equally open to everyone’s claims. When the David and
Lucille Packard Foundation undertook to back a reading
improvement effort in Sacramento, California, they went so far as
to specify the textbook series that had to be adopted.2 The outrage
that followed on the part of the unchosen was staggering. There
were strong suggestions that what the foundation had done was not
only unethical but that it marked the beginnings of a police state. But
what else was the foundation to do? They had studied the research
and had identified what they took to be ‘best practice.’3 If they tried
to mandate it in general terms there was little chance that it would
happen. Virtually anything could be claimed to fit the mandate. The
necessary concepts were missing. Packard had the choice of being too
general or being too specific. There should have been a middle range
of concepts capable of making significant distinctions among the
wide range of approaches that might be called ‘phonics.’
Those middle-level concepts are essential for any worthwhile
debate about educational methods. They are the concepts by which a
putative ‘best practice’ can be characterized with sufficient clarity
that people know what they are talking about. Without them you get
wars of words, with fanatics leading the charges. In many areas of
education the middle-level concepts are altogether absent. It so
happens, however, that beginning reading has been the object of
sufficiently intense and varied research that a rich fund of concepts

of the Whole Language movement reportedly expresses in a speech the wish

that a stake be driven through the heart of the author of a book presenting
reasons why Whole Language should incorporate teaching of phonemic
awareness and phonics (Levine, 1994). More recently, this same leader has
been characterizing the move toward phonics as part of a vast right-wing plot.
2
See the Packard Foundation’s 1997 annual report.
3
I happen to know about this because I am one of the authors of the program
Packard chose—SRA/Open Court Reading. But my point here is not that they
made the right choice, only that they were pursuing a reasonable and indeed
highly ethical course.
does exist (see Adams, 1990). But they are unknown to the multitude
of practioners, reformers, and ideologues. As a result, it is fair to say
that the participants on both sides of the long-running phonics
debate don’t know what they are talking about.
Calfee and Drum, in a 1986 review, described what for the
previous decade or more had been the standard approach to
phonics—what they called the “decoding curriculum, a skill-based,
‘bottom-up,’ drill-and-practice program” (1986, p. 812). They
characterized it as follows:
Divide the letter-sound correspondences into specific
objectives (e.g., f, ff, gh, and ph make the sound /f/). Teach the
child these objectives, often through seatwork. Avoid
rules—they are untrustworthy…. (1986, p. 812)
When people talk about going “back to phonics,” this is what they are
talking about going back to. Teachers who remember teaching
phonics probably remember something like this. Opponents of
phonics object to the pointless drills that occupy hours better devoted
to actual reading and writing.
What Calfee and Drum described, however, is something that
knowledgeable advocates of phonics find equally appalling. This
anomalous practice, aptly labeled by one critic as “Mickey Mouse
phonics,” was something devised by textbook publishers in the late
60s and early 70s as a quick fix in response to public pressure. It was
quick and easy because it could simply be added on to the existing
Dick-and-Jane type programs without requiring basic revision. It
was not based on any research or theory. It was simply a marketing
ploy. Yet, because it is what people remember, it is what the
contemporary phonics debate is about. In the absence of concepts,
people naturally argue about cases, and Mickey Mouse phonics
constitutes the known case.
Phonics, to anyone who understands the point of it, is a
transitional skill that enables kids to get started reading
independently. Everything is focused on that objective. Even on the
first day, children are using phonics to derive meaningful words
from print, and in a few weeks they are reading interesting stories.
Workbook exercises can play only a peripheral role, because the
principal task is teaching children to blend letter sounds together
well enough to serve in word recognition, and such teaching requires
the children to speak and the teacher to listen and respond.4 Once
children are able to read normal texts independently, which may take
a few weeks or 18 months at the outside, phonics instruction has
served its purpose and can be ended, whereas Mickey Mouse
phonics, having no objective, typically drags on for six years.
Mainstream publishers naturally hope that the same quick fix can
get them through the current surge of public pressure for the
teaching of phonics. There is good reason to suppose they will have
their wish. The current resurgence is backed by an accumulation of
research showing the superiority of sound-it-out type phonics to the
Mickey Mouse variety, but it is unlikely to have much effect because
neither the opponents of phonics nor most of its advocates can tell
the difference. If my gloomy prediction is right, Mickey Mouse
phonics will return as an add-on. It will be boring and pointless as
before and an easy target for its temporarily defeated opponents. A
reaction against it will set in. Publishers will remove it, and we will
have seen one more swing of the pendulum, with no progress to
show for it.
Yet in this case there could be progress instead of pendulum
swings. There is no reason why Whole Language approaches have to
be thrown out in order to bring in phonics. The Whole Language
movement has made some notable contributions, especially in
finding ways to involve children seriously in creative writing and in
discussing what they have read. Letting children invent their own
spellings, the most controversial part of Whole Language, actually
builds on children’s efforts to connect sounds and letters. None of
this is the least bit incompatible with teaching phonics. And there is
no reason why the adoption of phonics has to be a backward swing
of the pendulum. A great deal has been learned about what it takes
for children to decode print, and this knowledge can lead to
improvements in method that may not be obvious to the naive
onlooker but that can shorten the time and increase the success rate
of beginning reading instruction. But you cannot simply add phonics
to Whole Language. That will give you Mickey Mouse phonics.
Whole Language teaching at its best creates a literate culture in the

4
The responsive listener could in principle be a computer, although
at present there is no reading instruction software that provides
that. The available software is thus limited to Mickey Mouse
phonics.
classroom—that is, a culture in which reading, discussing, and
producing good literature are central to the life of the community.
Teaching children how to read is no more out of place in such a
culture than teaching them how to use a word processor, but it needs
to be done in ways that keep core literary values at the center. That
requires sophisticated and creative program design.
There is nothing novel about this suggestion. A “balanced
approach” is one of the watchwords of the more enlightened reforms
(Pressley, 1998). But how do you get people to realize that this does
not mean compromise or merely adding A to B? To understand the
difference between a phonics-based program that preserves Whole
Language values and something that merely looks like phonics added
on to something that merely looks like Whole Language requires that
decision-makers and practitioners get into pedagogical analysis to a
depth that they do not even know exists. But if they do not do that,
the pursuit of ‘best practice’ is a travesty.
What’s the Problem?
Among educational problems, beginning reading instruction is
exceptional in a number of ways. Most notably, its failures are
visible (or should I say audible?) and acute, which may explain why it
has generated so much emotion on one hand and so much research
on the other. But there are many other important educational
problems that seldom appear on reform agendas. In fact they receive
little attention at all except by the minority of researchers and
practitioners who are sensitive to them. The reason is not that
people dismiss them as unimportant, it is that the educational
failures associated with them are not conspicuous enough to set off
alarm bells. As a result, they tend not to be regarded as problems at
all but simply as worthy parts of the curriculum. Yet how to teach
them is in every case deeply problematic. People serious about
teaching them ought to be looking for every piece of knowledge
available, yet the reformers’ radar seems to sweep past them with
hardly a blip on the screen.
Here, mainly by way of review, are some of the instructional
problems that remain untouched by reform efforts:
1. Number sense. Widely recognized as the foundation of
mathematical competence, it now appears regularly in
curriculum standards. But no one ever says what it is, and the
curriculum guidelines always reduce it to activities or subject
matter that misses the point.
 2. Fractions, proportions, ratios, decimals, and percents. These
poorly understood terms mark off an area of colossal and
sustained failure, unalleviated by the switch from slices of pie to
slices of pizza.
3. Scientific misconceptions. Informed educators know about these
and can recognize them, but they don’t address them as a
problem, show little concern about the failure of available
remedies, and are distracted by controversy over what to call
them.
4. Functional literacy. Study after study shows that large numbers
of school graduates cannot cope adequately with real-life
reading tasks. The educators’ response is to criticize the studies
(which is often justified, but diversionary nevertheless) or to
adopt reductionist solutions: introduce realistic reading
exercises (reading bus schedules and such) or teach
comprehension strategies as subject matter. The notion that
everybody ought to be reading better than they do cannot really
be entertained, because current pedagogy offers no suggestion
of how this might be accomplished or even what it would mean.
5. Literature. Although literature teaching has suffered from every
kind of reductionism, many teachers have enough personal
sense of what the experiencing of literature should be like that
they are continually dismayed at their limited success in bringing
it about. But there is little they can turn to except tradition and
the inspirationalism of language arts journals.
6. World knowledge. This refers to the miscellaneous knowledge
of geography, history, and current events that provides a
background for further learning and thinking (see Chapter 9).
Getting students to acquire a lot rather than a little is highly
problematic, but the prevailing response is simply to produce
lists and expect teachers to teach them.
7. Thinking skills. That the teaching of thinking should be regarded
as a straightforward matter—a collection of procedures that
can be taught through rules, modeling,and practice—is perhaps
the most extreme example of failure to recognize an
instructional problem.
Doctors freely admit there are diseases they don’t know how to
cure, and the public response is not to repudiate doctors but to
appropriate money for research. No one ever admits that they don’t
know how to teach fractions, even though generation after
generation of students fails to learn them. And so the public response
is to blame the teachers, and the response of the teachers is to blame
someone else. Ignorance never gets the blame.
Reformers can reply that, although there are of course unsolved
problems in education, as in any field, we do know enough to make
substantial improvements. Hirsch (1996) has presented this case
forcefully, attributing the failure to act on available knowledge to the
academic pretensions of education professors. “Research-based” has
started to become a watchword of reform-minded politicians and
officials (Carnine, 19xx). Some parts of the education establishment
have reacted hysterically, seeing “research-based” as merely a code
word for phonics and rote learning (Taylor, 1998). They are partly
right, but not for the reasons their inflamed imaginations suggest. It
is true that advocates of research-based methods are in favor of
teaching kids phonics and getting them to memorize number facts to
a high level of automaticity. These are as close to settled issues as
educational research is ever likely to come. The danger is in
generalizing from these instances to the full range of instructional
issues. That is a mistake that both the advocates and the opponents
of “research-based” reform make.
“Of course there are unsolved problems,” says the practical
reformer. “There always are. But we can’t wait around. We have to
act on the best knowledge we have.” No one can quarrel with that.
What I am attacking is the attitude that goes with it. In the last
chapter I wrote about the complacency that marks traditional
professions, a complacency that rests on “doing a good job.” A
similar complacency seems to characterize reformers, whether they
are fanatics who believe they have the answer or pragmatists who
make do with whatever research and common sense suggest. The
invariable accompaniments of this attitude are:
1. Absence of need to understand.
2. Reductionism—especially the reduction of everything to subject
matter or activities.
3. Absence of forward momentum.
I have remarked on all of these previously. It isn’t that people are
generally averse to understanding, although the more ideologically
zealous may be. It is that practitioners and policy-makers never find
themselves stuck for lack of understanding. That bespeaks something
deeply wrong with the system. Try fixing your automobile or your
Internet setup and unless you are a specialist you will almost surely
come to a halt where your understanding can carry you no further.
But set out to fix education and you will encounter no such
roadblock.
Reductionism permeates the education system. I have talked
about teachers reducing everything to subject matter, activities, or
self-expression. But often the reduction occurs before anything
reaches the teacher. The textbooks that come into the teacher’s
hands have been shaped by a selection process that reduces
everything to subject matter. State and local textbook selections are
dominated by checklists which specify content, with no regard to
whether the textbook affords any promise whatsoever that the
students will learn the content. Curriculum guidelines will state airy
objectives and then immediately reduce them to activities. The
activities may be presented as examples, but inasmuch as there is no
basis for generalizing from examples, the suggestions can only be
taken as requirements. Get-tough reforms introduce an even more
lethal form of reductionism. Education is reduced to training
students to perform well on tests.
By lack of forward momentum I do not mean that there is never
any improvement. A lot of what happens in education can be counted
as steps in the right direction. But there is no sense of one step
leading to another. Each step has a finality about it, even if it is
recognized as only a partial solution to the problem that motivated
it. When there is forward momentum in a discipline, profession, or
technology, one can look back at the steps that led to the present
state and infer a direction for the next step. It affords a kind of
running jump into the future, as anyone will recognize who has done
a piece of research or development in a progressive field. But in
education the best guess is that, whatever the direction of the
preceding step, the next step will be in a different direction
(frequently the reverse of it predecessor).5
Underlying all three of these symptoms is what I take to be the
fundamental malady: disengagement from the constitutive problems
of instruction. The kinds of instructional problems I reviewed above

5
In progressive disciplines the next step is not always a running jump from
what has gone before. When there is a marked change in direction, this will
be called a ‘paradigm shift.’ Educationists have taken up this term and applied
it so liberally that paradigm shifts are a monthly occurrence. You may infer
from this either that education is the most progressive field in the world or
that it is not progressive at all.
are treated as somebody else’s business. This removes the need to
understand and it encourages reductionism, a simplistic translation
of high-level goals into familiar routines.
How Folk Theory Impedes Educational Innovation
Innovation. It is supposed to be the driving force in the new
economy. Government agencies are scrambling over one another in
their zeal to promote it. There are people who believe that
education, like religion, should be shielded from innovation; but let
us ignore them in this discussion and concentrate on the more
progressive elements, both inside and outside the profession, who
regard innovation as at least potentially a good thing.
The first and often decisive barrier to innovation is the response,
“We already do that.” Those of us involved in educational design
encounter it constantly, any time we try to move above the level of
specific procedures and offer some principled approach to an
educational problem. I used to think that this barrier could be
overcome by discovering clearer ways of explaining what was new. I
have since come to believe that the problem lies deeper and cannot be
solved on a case-by-case basis. The problem is conceptual, but it is
such a massive conceptual problem that it has produced a strong
emotional resistance to any departure from familiar categories or
habits of thought.
Educational thought, as carried out by practitioners and lay
people alike, is in its various aspects moral, subjective, and
procedural, but in no significant way theoretical.6 When a reading
pundit declared that teaching phonics is an act of violence against
children, he was speaking to the moral aspect of educational
thought, which at the elementary level runs heavily toward
protecting children from harm. For such a statement to be anything
more than sanctimonious claptrap, however, it would be necessary
to back it up with a theoretical explanation of the cognitive and
emotional effects of phonics instruction. For there is nothing overtly
injurious about phonics. Overtly, the behaviors required of children
are not unlike those that might be observed in a singing lesson, so if

6
It is also, of course, political and economic, and these aspects often become
salient in judging an innovation; but at this point I am talking about
recognizing innovations in classroom practice. These must typically be
recognized on some other basis first, after which their political and economic
implications may come under consideration.
harm is being done it must be at a deeper level. But in the public
discourse carried on about educational issues, there is no deeper
level. That is why a statement like the above, which I make out to be
sheer lunacy, can carry weight in a policy discussion equivalent to
that of statements about measured effects on test performance. In
phonics debates as elsewhere, however, the most common and
destructive response is “We already do that.” It, too, is a claim that
needs to be backed up at a theoretical level, for at a procedural level
people who make the claim can no doubt show that they are doing
things that would fall into the category of teaching phonics. The
issue is whether they are actually teaching children to recognize
unfamiliar words by means of their spellings. Although that may
sound like a fairly clear-cut issue, it calls for a level of analysis that is
simply absent from practitioner and lay discourse.
I keep coming back to phonics as a case in point because the
concepts needed to understand it and to make progress are not very
deep. And the conceptual failures are dramatic. Let us turn aside
from the lunatics and fanatics and consider what now seems to be
the majority, people who have soberly concluded that teaching
phonics is a good idea. Suppose that you come to them with an
innovation that you claim constitutes an advance in the teaching of
phonics. How will people judge whether your claim is plausible or
even whether what you propose is an innovation at all? At a
procedural level, what you propose is not going to look much
different. There are games and exercises that involve the sounds of
words, but those are already commonplace. And you have made
some changes in the teacher’s patter, but so what? Finally, all people
can do is ask for evidence that your supposed innovation produces
better results. If so, they may be prepared to adopt it, but with no
idea of why or in fact what they are adopting. To give your
innovation some identity, you might coin a name for it. Suppose you
decide to call it ‘integrative phonics.’ That would be a mistake.
‘Integrative’ is a term associated with Whole Language and project-
based learning, the sort of approach that people adopting phonics
are eager to reject. That one word, with its train of associations, is
liable to obliterate everything you try to say. The power of labels in
the marketplace of educational ideas is overwhelming. When there
are no concepts, emotive labeling fills the void, and it is a
tremendous void.
 Suppose, however, that you choose a label that catches on.
Maybe ‘coherent phonics’ would do the trick. Your window of
opportunity for advancing the ideas that this label represents will be
too small to put a foot through. As soon as the label catches on,
everyone who is trying to make money off phonics will call whatever
they are doing ‘coherent phonics.’ No one will be able to tell the
difference, because they never understood what you were talking
about in the first place. The same will be true if, instead of ‘coherent
phonics,’ you come forth with ‘coherent number sense,’ ‘coherent
proportionality,’ ‘coherent science,’ ‘coherent literacy,’ ‘coherent
history,’ ‘coherent thinking,’ or just ‘coherent knowledge.’ Unless
your innovation looks conspicuously different on the
surface—unless, for instance, it involves exotic
technology—practitioners and decision-makers will not be able to
recognize what is new about it or to distinguish it from anything else
that is being called by the same name.
If we switch from “We already do that” to “Wow! This is what
the 21st century is really about!” we view the same dismal scene from
the other side. Walk into any well-equipped classroom and you are
likely to find students engaged in an activity that has been tediously
commonplace for generations—copying information from
authoritative sources, organizing it under topical headings,
decorating it nicely, and presenting it as a finished ‘project.’ The only
difference is that now they are copying information from the Internet
rather than a book, organizing it by means of a word processor or
media authoring application, and decorating it with computer
graphics, animations, movies, or whatever the technology allows.
This is innovation of a sort, and it may even be beneficial—although
that remains to be demonstrated. What is discouraging is that only
innovations of such instant visibility are recognized. As technology
enables ever more spectacular innovations of this superficial kind, it
reduces even further the ability of educators to recognize
innovations of any deeper sort.
The Concepts We Need and Why We Don’t Have Them
Philip Agre has remarked on “the tendency of people who define
themselves against something to simply invert whatever it is they
oppose, rather than actually having a new idea.”7 That has been the

7
In “RRE]notes and recommendations,” Red Rock Eater News Service,
November 15, 1998.
character of much educational reform in the 20th century. Liberal
reforms have all been defined in contrast to the stereotype of the
teacher at the front of the room lecturing and quizzing. Conservative
reforms, reacting against what they perceive to be the abandonment
of teaching in the liberal reforms, give us back the teacher as lecturer
and quiz-master. California has provided the most dramatic
example of such unprogressive flip-flops. In the early 1990’s they
virtually outlawed direct teaching of reading and mathematics; then
they did an abrupt switch, producing guidelines that virtually
outlawed everything except direct teaching. There were people in
California who attempted to use the switch as an opportunity to
introduce some new ideas, but they were overwhelmed by the
reactionaries who, unable to make discriminations, categorized all
new ideas as belonging to the liberal pedagogy they were determined
(with fair justification) to overthrow.
Reformers should not be expected to produce new ideas, but they
should be able to recognize them and to distinguish them both from
the supposed bad ideas they are attempting to eradicate and the
supposed good ideas that are simply the reverse of the bad ideas. To
do this, they need concepts that allow them to think constructively
about issues like the following:
• Why something is worth learning—apart from its conjectured
long-term utilitity in the job market and apart from its
traditional backing; why, in other words, a student might feel
disposed to learn it.
• What different learning objectives actually mean. What is
number sense and what does acquiring it consist of? Similarly
for mathematical problem solving ability, understanding a
historical period or event, mastering algebra, learning
graphical design, understanding Dante, being able to write a
decent paragraph.
• What is teachable and what isn’t, why a certain thing is easy or
difficult to learn, in what ways the learning of it can go awry.
• What is the normal course of developing competence in
particular domains and what are optimal and suboptimal
developmental paths; what distinguishes expertise in the
domain and what learning leads to it or away from it.
Answering questions like these requires getting deeply into
subject matter and into the cognitive developmental and
instructional research in the various domains. There are not
standard answers that fit all subjects. We cannot expect reformers to
have answers or even to be in a position to evaluate them. What they
should understand is that these are the issues on which the value of
any instructional reform must rest and that knowledge does exist
that can be brought to bear on them. They should be aware that
knowledge is progressing and that the best reform will not be simply
the reverse of the preceding reform. Professional educators ought to
know more than that. In their preservice education they ought to
have had courses that immersed them in these issues as they arise,
not only in the particular subject they are preparing to teach but in all
the other major domains of education. Practitoners could then serve
as intermediaries between policy-makers and researchers.
What we have instead is instructional reform carried out by
people who not only know nothing about instruction beyond what
common sense and the mass media provide them but do not know
that there is anything to know. This ignorance of ignorance is not
limited to politicians and representatives of the public. I have run
into scientists and psychologists who study the development of
scientific understanding but who do not realize that there is anything
beyond the obvious to be understood about the teaching of science;
the same for mathematicians and psychologists of mathematics.
Obviously I have not tried in this book to set out the pedagogical
knowledge, the ignorance of which I have been lamenting. That
would require a larger book, with multiple authorship, for it would
need to delve expertly into the problems of teaching and learning
peculiar to different domains, as well as more general insights into
these problems. It would be an important book, especially for teacher
education. There have been various attempts along these lines, but
none, I believe, that has given the task the massive effort it deserves.
My chapters 8 and 9 have touched on various aspects of pedagogical
knowledge, but my main concern has been with concepts of a more
central nature—concepts of knowledge and mind. What I want to do
now is show how they are important in the practical work of
educational reform.
For perspective, I have adapted a schema used by E. O. Wilson in
his book, Consilience (1998). Wilson was representing the domains
that relate to environmental policy. I am using the same kind of
diagram to represent the domains that converge on classroom
instruction. His diagram had four sections instead of five, but what
he said about their intersection conveys the point I want to make
about education:
As we cross the circles inward to the point at which the
quadrants meet, we find ourselves in an increasingly unstable
and disorienting region. The ring closest to the intersection,
where most real-world problems exist, is the one in which
fundamental analysis is most needed. Yet virtually no maps
exist. Few concepts and words serve to guide us. (Wilson, 1998,
p. 10)
An implication of Wilson’s schema is that, although concepts may be
well developed and well understood within each separate domain,
as you move into the region of convergence the problem is not that
people have differing concepts but that they share concepts that are
poorly developed. In education, I suggest, this “unstable and
disorienting region” is occupied by folk theories of knowledge and
mind. Although the inadequacies of this folk knowledge may not be
evident in the work going on within each of the sectors, it gravely
affects the movement of knowledge between sectors. Knowledge
moving between sectors must, so to speak, pass through this
“unstable and disorienting region,” where it becomes degraded. On
the surface it typically emerges as mere folk platitudes, such as
“Every child is different” or “You can’t teach what you don’t know.”
Beneath the surface, however, the degraded knowledge propagates
as unarticulated assumptions, and in that form it is really damaging.
 Educational Policy

Subject Matter
Knowledge

Pedagogics

Cultural Studies

The Learning
Sciences

In the preceding chapters I have tried to articulate and criticize

some of these hidden effects of folk theory. Here I will summarize
them, with suggestions as to how they impede or subvert
instructional reform.
1. Knowledge is narrowly conceived as the contents of a mental
filing cabinet. This leads on one hand to devaluing knowledge as an
educational objective and on the other hand to hapless efforts to
itemize and test it.
2. There is no idea of teachability. Tacitly, it is assumed that if you
can measure (or perhaps only name ) something then it must be
possible to teach it. The result is the imagined teaching of imaginary
skills.
3. There is excessive and unfounded belief in transfer of learning.
In the extreme, where word magic takes over, it is assumed that
what is learned will automatically transfer to everything else that is
known by the same name. Word magic and excessive belief in
transfer are essential to folk educational thought, because without
them it is difficult to justify most school teaching.
4. There is no idea of quantity. The tacit belief, applied to both
educational content and procedures, is that if some is good more is
better. Quantity as an objective is either scornfully repudiated or else
is replaced by long lists of items to be learned and tested.
5. There is a shortage of concepts for the analysis of teaching
methods. Means are lacking for distinguishing one method from
another, and ideological critique takes the place of analysis of means
and ends.
6. Reductive practices go unrecognized and are sometimes
elevated to pedagogical principles: Reduction to activities becomes
‘project-based learning,’ reduction to self-expression becomes
‘learner-centered education,’ and reduction to subject matter
becomes ‘core knowledge’ or ‘building a firm foundation.’
7. Science is deeply misunderstood. School science becomes either
purposeless activity—the mere exercise of research skills—or the
mastery of a body of truths or else a branch of social studies.
Improvability as a defining characteristic of scientific knowledge
does not even enter the educational picture.
8. Understanding, conceived of as a mental state, becomes
something that can be pursued only indirectly, whereas the direct
pursuit of understanding is characteristic of real scholarship and
science.
9. Knowledge creation is reduced to learning and belief. Terms
like ‘knowledge work,’ ‘constructivism,’ and ‘inquiry’ thus lose their
meaning. School learning comes to be seen as an impediment to the
kinds of productive work that drive a knowledge-based economy.
10. There is no way to formulate a generally acceptable idea of
the core job of schooling. With an impoverished conception of
knowledge, knowledge development cannot be accepted as the core
job. But any other definition is either too global or too narrow.
What I have tried to summarize here are the inadequate concepts
that occupy the region where the various sectors bearing on
instructional reform come together. These concepts impede the
movement of knowledge from one sector to another or degrade it
below a useful level. For instance, insightful educators recognize
that statable knowledge (the stuff in the mental filing cabinet) and
specific skills represent the two visible tips of a very large
iceberg,and they are concerned with the whole iceberg. But they have
no cogent way to talk about the invisible part. What they say comes
out as romantic froth. They spout nonsense about ‘process versus
product,’ surely one of the most ill-advised rhetorical gambits in the
perennial educational debate. They espouse catchwords like
‘constructivism’ and they solemnly aver that children’s minds are not
empty buckets waiting to be filled. It is no wonder that politicians
and business people grow impatient and launch back-to-basics
reforms, which are aimed at focusing education on the very limited
range of learnings—the two tips of the iceberg—that can be specified
in reasonably clear terms, pursued in a no-nonsense manner, and
that lend themselves to measurement of results. Other examples
could be generated from other segments of the above pie diagram.
Subject matter experts embody vast amounts of the implicit
knowledge that constitutes the invisible part of the iceberg, but what
comes across from them as educational wisdom tends to be either
cut-and-dried content or else more romantic froth. Sociocultural
researchers are learning a great deal about how knowledge is
constituted and communicated in working groups, but when it passes
through the inner circle it comes back out as learning-by-doing,
apprenticeship, and informal talk—in other words, what everyone
already knows. Learning scientists have produced penetrating
critiques of textbooks in every major area of the school curriculum,
but what reaches the policy makers cannot be translated into
standards that publishers could act on or that could be enforced; so
instead the policy makers send out content lists or standards based
on conventional tests.
Although some progress can be made by piecemeal attack on the
10 faulty conceptions listed above, stable improvement is going to be
very hard to achieve as long as the underlying folk theory remains
unchanged. Throughout I have been arguing for only two basic
changes:
1. The idea of mental content should be reduced to the status of a
metaphor, useful for some purposes and not for others, but
never to be taken literally. In order to suspend use of the
metaphor where it is proving inadequate, we must find ways to
construct mentalistic accounts that do not refer to things
residing, being searched for, or undergoing changes in the mind.
2. Abstract knowledge objects, such as theories, numbers, and
designs, should be accepted as objects in the real world, with
which people may develop relationships, much as they do with
 animate and inanimate material things. Understanding and
mastery may then be treated as characteristics of such
relationships, and the advancement of knowledge as the
creating and improvement of abstract knowledge objects.
Producing these changes remains a tall order. It is not a matter of
convincing people of propositions like the above. People have to see
learning, teaching, thought, and knowledge in ways that are
consistent with these propositions, just as, to become a Copernican,
you have to see the terrain outside your window as part of a globe.
Furthermore, they have to become proficient at thinking within this
altered framework. As I proposed in Chapter 3, it amounts to
moving out of the two-dimensional world of folk theory and into a
three-dimensional world in which there are not only physical reality
and mental states but also conceptual artifacts, rather than only the
first two. Compared to folk theory’s Flatland, this three-dimensional
(Popperian) world greatly increases the complexity of relationships
one must learn to deal with.
Conceptual change has to start where it can. It is not likely to
start among the policy makers, although some of them are business
people who have a more modern conception of knowledge than does
the typical educator. It is not likely to start among teachers, because
by reason of their daily work they are more tightly bound to a things-
in-the-head conception of knowledge than most other people. It is
not likely to start with subject-matter specialists because folk theory
of mind does not normally figure in their work except on the
occasions when they are called on to provide educational counsel.
That leaves learning scientists and students of culture, my academic
colleagues. Many of them, in different ways, are moving beyond folk
theories of knowledge and mind. I am trying to help that process
along. But ultimately, taming that “unstable and disorienting”
region of concepts where the different kinds of education-related
thought meet is going to require cooperative effort.
Crafting a Vision
One of the areas in which cooperation is most obviously called
for and where conceptualization has been most limiting is in the
formulation of goals. A goal is a conceptualized outcome. If it is
conceptualized at a high enough level, it becomes a vision or an
ideal. If the conceptual framework is strong and complete enough,
there will be ways to connect the vision with lower-level goals that
can actually be achieved. Otherwise, the vision tends to be of only
inspirational value and the actual business of reform may not
advance toward it at all.
The educational reforms that have stirred the hearts of educators
over the past century have been of a liberalizing kind. From a
distance they look like the same reform, surfacing under different
names: progressive education, child-centered education, learning by
discovery, open education, whole language, constructivism, situated
learning. In contrast, the reforms that have gained strong public
support have been of a reactionary nature—usually, in fact, reacting
against the liberalizing reforms espoused by the educators. This
contrast has led many educators to see the conflict as having deep
political and philosophical roots—as a conflict in world views, no
less. There is just enough truth in this to keep alive in ed schools a
strain of aggrieved social criticism, culminating in what is called
“critical pedagogy”8. But it serves to obscure a more mundane yet
8
I find it infortunate that the term ‘critical pedagogy’ has been adopted by
people whose critique is mainly of the school as an institution rather than of
pedagogical practice. According to Barry Kanpol (1998), “Critical Pedagogy is
the challenging of any or all forms of alienation, oppression and
subordination—no matter from what identity position one is coming from.”
Its main concern has been how schools preserve social power differences;
there has been hardly any interest in pedagogy except insofar as it pertains to
such inequalities and to ways of overcoming them. I would like to think that
what I am doing in this book constitutes an effort, however limited, at critical
pedagogy in somewhat the way that Philip Agre’s Computation. and Human
Experience (1997) constitutes a critical AI—that is, a criticism that tries to
clarify the constitutive problems of a practice, identifying where it is stuck,
and trying to help it get unstuck. Yet there is almost no overlap between the
problems I have discussed here and those that occupy the practitioners of
‘Critical Pedagogy.’ To read the literature of Critical Pedagogy, you would
think that there are no technical problems in education. That would make
education a very strange field indeed. The inspiration for much of
contemporary Critical Pedagogy is the work of Paolo Freire (1970), whose
educational efforts with peasants convinced him that the difficulties in
teaching them to read were not technical but had to do with their
powerlessness. This important idea seems to have been generalized by his
followers into a universal diagnosis applicable to all educational problems
and a corresponding disdain for any technical analysis. (In medicine, this is
known as quackery.) Along with Agre, I believe that a real critical practice
needs to get into the technical problems of a field, particularly those that are
seeming intractable, and look there for underlying and often unstated beliefs
that need to be questioned. It is Freire’s method of inquiry, but without
Freire’s answer being assumed in advance. It is what has led me to identify
pregnant truth: The public can never be expected to share the
enthusiam of educators for enlightened methods. The public cares
about outcomes. Methods stir up public passion only when they are
seen as detrimental to outcomes. The public’s position is always
going to be reactionary, until educators can offer them new and
superior outcomes.
The liberalizing reforms of the past century have not done that.
Only when carried to absurd extremes have they produced
demonstrably bad results, but evidence of good results has been
sparse and unconvincing. The enthusiasm that educators have had
for these reforms has not come from evidence that students turn out
better. It has come, on the philosophical side, from liberal beliefs
about rights, democracy, and humane treatment. On the practical
side it has come from demonstrations of process that any open-
minded teacher is bound to find impressive. Walk into a well-
functioning open classroom and you are going to see students going
about their work on their own or in small groups, seemingly
interested in what they are doing, and not being driven to it by the
teacher. The teacher, free of the need to conduct recitations and
micromanage schoolwork, has time to work individually with
students according to their needs. It’s hard to resist. You see students
doing what they are supposed to and teachers doing what they
always want to do and not doing what they dislike doing. The
amazing thing is that such an attractive reform seems never to have
caught on with more than a fourth of the teacher population
(Elmore, 1996).
Of course, it has some appeal to parents and to the public at large
as well, but not of the same intensity. It does not touch them where
they live, the way it does many teachers. And so whenever
dissatisfaction arises about results, the public is ready to scuttle
liberal reforms and go back to what they imagine to be tried and true
methods. In this they will be joined by that majority of educators who
have either been opposed to the liberal reforms all along or else
weakly committed, like the public at large. It seems to me that the
reformers and those who work at understanding and engineering
processes of change have simply failed to face up to this rather
straightforward situation. It has nothing intrinsically to do with

folk theories of knowledge and mind as the source of much educational

stuckness.
‘resistance to change,’ with ‘ownership,’ status, or power, with top-
down versus grass-roots initiative, and all of those factors that enter
the picture once a reform movement is underway. It is simply that no
reform is going to gather sufficiently widespread and ardent support
unless it can be shown to yield results that people value and value
highly.
Vision versus Test Scores
As soon as you mention results, everyone thinks of test scores.
This is not the place to discuss how it came to pass, but I think
everyone will agree that the two ideas are welded together in
people’s minds. Education can make no major advance until that
weld is broken, but the tools to break it are not in any reformer’s
toolkit.
I was once at a meeting where a host of distinguished educators
were complaining that politicians and business people, by their
insistence on objective test score gains, were making real educational
improvement impossible. Among the group, however, was Ralph
Tyler, who perhaps more than anyone then living was responsible
for the shape that educational testing had taken in America. Tyler at
last rose to say that in his experience politicians and business people
had no passionate commitment to test scores, it was just that no one
offered them anything better. If you can offer them a superior vision,
he said, you will find them receptive. Many idealistic reformers
would claim that that is exactly what they are trying to do. But what
they are offering is a vision of an educational process. I don’t think
that is what Tyler had in mind. To draw politicians and business
people away from their fixation on achievement test gains one must
offer them the vision of a superior kind of outcome. The failure to do
that is, I believe, the most profound failure of educational thought in
our epoch.
A motivating vision must open people’s eyes to educational
possibilities beyond those currently imagined. That requires
demonstrations. People are properly skeptical of high-sounding
educational rhetoric. They need visible evidence of the possible.
Moreover, we have already had enough visions of the school as an
earthly paradise. The public wants to know what young people will
carry with them when they leave that earthly paradise, and the
widespread fear is that it will be functional illiteracy, mathematical
incompetence, and scientific ignorance. An inspiring vision must not
only allay those fears but hold out the promise of higher outcomes
worthy of dedication.
Here is where the conceptual difficulties begin. Education is not
going to produce some previously unheard of ability or trait of
character. The English language contains several thousand words
for human attributes; so whatever the educational outcome, it is
almost sure to be something that has been recognized for ages and
for which there is already a familiar label. But a new vision should
give new meaning to that label, introduce a qualitatively higher
standard. Let me try to explain that statement with an example from
a different field. Art historians talk about the “Greek revolution”
(Gombrich, 1960, pp. 99-125). For thousands of years sculptors had
represented the human figure facing straight ahead, in a rigid pose,
with stylized impersonal features. There were, however, noticeable
variations in how lifelike (to modern eyes, at any rate) these stylized
figures appear. You could construct a scale of ‘lifelikeness’ and use it
to rate these figures, to examine historical trends, regional
differences, and the like. But if you then tried to apply that scale to
the sculpture of Greece during its Classical and Alexandrian periods,
you would find it didn’t work. Everything would be piled up at one
end of the scale. Yet there was tremendous change over that time as
figures became increasingly fluid and natural, eventually acquiring
distinctive human expressions. However, you could not capture
what happened simply by extending the scale of lifelikeness.
Gombrich attributed the visible changes in style to a change in
conception of what a statue should be (a change in goals) and a shift
from reliance on knowledge to reliance on observation (a shift in
method). A qualitatively higher standard of lifelikeness would be a
standard that was sensitive to the effects of these underlying
changes.
A New Standard of Understanding
I believe that something comparable to the “Greek revolution” is
beginning to appear in education. It is the emergence of a
qualitatively higher standard of understanding. As an educational
goal understanding has been around for thousands of years.
Students can be and often have been rated on it. Using commonly
observed variations, you could construct a scale of understanding for
any particular item or body of subject matter. The scale would range
from utter cluelessness at the low end to a high end represented, let’s
say, by the most convincing essay examination papers you could turn
up. What I am suggesting is that we are beginning to see among
students a sort of understanding for which that kind of scale, even if
stretched, will not work. The old kind of scale, and the conception of
understanding that goes with it, is essentially a scale of conformity to
some approved manifestation of understanding—usually a set of
propositions. Call this the conformity criterion of understanding.
The conception of understanding that I developed in Chapter 4 does
not exclude this criterion. If you understand something, this ought to
be reflected in congruence between your beliefs and those of
acknowledged experts. But a conformity scale would be insufficient.
It would be like a scale for judging sculpture that only took account of
conformity to the physical features of the model or person being
represented. Greek sculpture would score high on such a scale,
compared to the sculpture of earlier civilizations, but the scale would
be insensitive to most of what gives Greek sculpture its quality as art.
Similarly, a conformity scale of understanding would be insensitive
to most of what makes understanding a goal worthy of the fullest
commitment of people’s intellectual resources and energy.
To bring this discussion closer to earth, let me introduce a couple
of examples. They are examples drawn from childhood, for it is
easier to see what there is to understanding beyond conformity when
one looks at learners who are still a considerable distance away from
expertise. Both examples come from combined grade 5/6
classrooms, but otherwise the conditions are quite distinct: The first
comes from an inner-city classroom in Oakland, California; the
second from a middle-class school in a medium-size city in Iowa. The
first comes out of the Fostering Communities of Learners project
(Brown and Campione, 1994), the second out of the
CSILE/Knowledge Building project (Scardamalia, Bereiter,&
Lamon, 1994).
1. Students who were studying about AIDS came upon the
reassuring information that AIDS cannot be transmitted by
mosquitoes. This struck them as implausible. If hypodermic needles
are a major way that the virus is transmitted, why wouldn’t
mosquitoes, which in their view are flying needles, also transmit it?
They called a local AIDS hotline to pose the question and were
eventually put in touch with someone who explained to them that
mosquitoes transmit malaria because they become infected with it
themselves, whereas the AIDS virus does not infect mosquitoes.
Some students were satisfied with this explanation but others saw
that it did not really answer their question. The hypodermic needle
doesn’t get infected either. It transmits the AIDS virus just by the
blood that sticks to it, so why wouldn’t the same be true with
mosquitoes? They pursued the question through library research and
asking other informed adults, but as far as I know they never did get
an explanation that satisfied them.
2. Students in the class were divided into teams to study different
biomes—deserts, tropical rainforests, the Arctic tundra. The Arctic
group became interested in the question of why trees don’t grow
there. Their initial assumption was simply that trees would freeze,
but then one student questioned this explanation, pointing out that
trees survive in very cold weather. Maybe trees can’t freeze. When
they learned about the permafrost, this suggested another
explanation: If the earth a foot or two beneath the surface is
permanently frozen, roots couldn’t penetrate it and so the roots
would be too shallow to hold up a tree. But then the group studying
tropical rainforests reported that trees there had shallow roots. An
expert on the Arctic gave a talk at the school. His explanation for the
absence of trees was lack of moisture. This explanation didn’t satisfy
them either, because a check with the desert group yielded the
information that trees did grow in some very dry places. Finally they
concluded that it must be a combination of factors.
I have intentionally chosen examples in which the quest for
understanding did not culminate in anything that would score high
on conformity to established knowledge. Indeed, there was not much
culmination at all. In both cases the question motivating the inquiry
remained unanswered. For that reason, many educators would say
that the value was in the process rather than in the outcome—in
learning to pursue inquiry, to question authority, and so on. But to
treat these cases in that way is to reduce the students’ inquiry to mere
exercise. That is the bind that folk theory of mind gets you into. The
value of instructional effort lies either in the resulting mental
content—which can be judged as to its correctness, completeness, or
utility—or in skills and attitudes. That is how it was handled in
Bloom’s Taxonomy (1956; see Chapter 4) and how it continues to be
treated in curriculum guidelines and test standards.
The unique value of the educational experiences represented in
these cases does not lie either in the explicit knowledge acquired
(although that may have been considerable) or in the skills and
attitudes that were developed (although these are not to be
discounted merely because they would be difficult to assess). Their
unique value is in where they lead. The students in the first case had
embarked on a path that could lead to a far deeper understanding
than most people ever have of the nature of microbial diseases and
the complex routes by which they can be communicated. In the
second case, we see the beginnings of an understanding of how
different environmental characteristics combine to determine
whether a particular kind of flora can thrive in a particular
environment. In important respects, I would claim, the students’
understanding had already progressed beyond that of the average
educated adult.
As it happens, I had wondered about mosquitoes and AIDS
myself, but it never occurred to me to try to track down an
explanation. As for the absence of trees in the Arctic, I had never
thought about it. I guess I had implicitly accepted “It’s too cold” as an
explanation: this despite my having actually been in the Arctic and
observed that at the northern limit the conifer forests do not just thin
out and then disappear; rather, the trees get smaller until they are
the size of Christmas trees—a fact that weighs heavily against the
“It’s too cold” theory, if I had thought about it. Thus these examples
impress me because the young students were investigating questions
that are at my level, so to speak—as distinguished from the
schoolbook questions one expects elementary school students to be
investigating—and making more progress on them than I ever made
or even thought of making.
The higher standard of understanding illustrated in these
examples can be formulated like this: Understanding is to be judged
according to its ability to produce further understanding. This is a
qualitatively different standard from the familiar correspondence
standards. It represents a dynamic rather than a static conception.9
What Would It Take to Make Teaching for
Understanding a Reform Goal?

9
Dewey clearly had such a dynamic conception. He defined education as “that
reconstruction or reorganization of experience which adds to the meaning of
experience, and which increases ability to direct the course of subsequent
experience” (1916, p. 76). For him, this idea was part of the larger idea that
education is growth. Unfortunately, it was a simplistic version of the larger
idea that survived, while the idea that the main value of school learning lies
in its enablement of further learning (1916, p. 53) suffered neglect.
 There is a problem in getting any objective taken seriously in
education. If you argue, as I have done, for the importance of
knowledge and depth of understanding, the most common response
of critics will not be to disagree but to point out that other things are
important too. Then they will give voice to whatever is uppermost in
their minds, be it numerical skills, media literacy, or spirituality. Of
course many other things are important and of course everything is
related, but the effect of this kind of criticism is to make everything
matter less and to avoid any serious criticism of the ideas that were
actually set forth. There is a time for talking about education in its
fullness but the rest of the time it should be possible to focus on
particular problems or objectives while tacitly conceiving of them
within a larger picture, which is for the time being not brought into
question. Why this should be so much more difficult in education than
in most other domains I do not understand.10 It may have partly to do
with the complexity of the enterprise but it also, I believe, bespeaks
the inadequacy of the folk theories of knowledge and mind that
frame the discourse.
At this writing, teaching for understanding is not high on reform
agendas in North America. Skills are on top. But understanding was
up there not long ago and it will be high again. Education focused on
skills is inherently mediocre, and its ascendancy usually means that
public confidence in education has sunk so low that mediocre looks
like a step up. Teaching for understanding cannot amount to much as
a reform goal, however, if it is just another item on a list of
standards. It has to be a goal that alters all the other instructional
goals. In order for that to happen, three things are required:
1. People have to understand what it means.
2. They have to realize why it is important.
3. They have to realize that teaching for understanding embraces
a host of unsolved problems.
The last point might seem dispensible or perhaps even
countervailing, but I shall argue that, unless the unsolved problems
are recognized, teaching for understanding will remain just another

10
Educational discourse is pervaded by a mistrust of other people’s motives,
which makes for a paranoia that every once in a while bursts out, as it has
among the anti-phonics people. The rest of the time it is kept in abeyance by
continual manifestations of good intentions. This, however, can lead to
hypocricy, which breeds further suspicion, and so on. Beyond that, however,
it further defeats efforts to get on with solving an educational problem.
slogan to be tossed about during ideological debates, while the
pedagogical pendulum continues to swing.
In the previous chapter I discussed why teaching for
understanding is an elusive goal for educators. The ordinary
reductive moves—reducing goals to activities, subject matter, or
self-expression—eliminate the understanding from teaching for
understanding. More generally, teaching for understanding is
undermined by folk conceptions of knowledge. It is thought of as
something in the mind, which right away makes it intangible and
inaccessible. So we see the goal of understanding replaced by the
goal of producing demonstrations of understanding (Bloom, 1965;
Perkins, 1992). The criterion becomes correspondence between what
the learner demonstrates and what an expert demonstrates. This is
reasonable as far as it goes, but it is not a very elevating conception.
But how to elevate it? The answer, to liberal-minded educators, is to
replace product with process—to espouse ‘constructivism,’ ‘social
construction of knowledge,’ ‘inquiry,’ ‘critical thinking,’ and other
noble processes. That is but another road to reductionism, however,
to activities that have lost contact with their goals. And it is also the
road to loss of support from a public that wants results, not faddish-
sounding processes.
As I argued in Chapter 4, the way to make sense of understanding
as an educational objective is to conceive of it dynamically and
relationally. To understand something is to have an intelligent
relationship to it. The main value of scientific and scholarly
understanding is that understanding one thing increases your
capacity to understand other things. That is its practical value, not
only for pursuing further education but also for achieving expertise
in any knowledge-based occupation.
If this dynamic conception of understanding is accepted,
however, it demands that instructional reform be approached at a
deeper level than in the past. As I discussed in Chapter 8, evaluation
needs to be concerned with the trajectory students are on—where
they are going in their efforts to understand the world—rather than
only with what they have managed to understand so far. We have to
stop thinking of ‘constructivism’ as a pedagogical method of
questionable virtue and start thinking of it as something that
students should come to adopt as a way of life. We should be
concerned about the students who do not adopt it, no matter how
good their present performance may be, because they are
disqualifying themselves for participation in knowledge society.
The Myth that Defeats Reform
To many skeptics, sympathetic as well as unsympathetic, what I
have been saying about a higher standard of understanding will
sound like just more inflated educational rhetoric. They will want to
know what basis there is for believing that the instances I have
narrated can be replicated on any considerable scale and that they
will have a cumulative effect. Some, less astute, will want to know
how “this approach” ensures that graduates will be literate and
mathematically competent—not recognizing that at this point I
haven’t advocated any approach but have only indicated something
to work toward. And then there are those, perhaps not great in
numbers but enormously influential within educational
organizations, who will reject what I have presented because of their
gut feeling that it is overly intellectual, scientistic, Eurocentric and in
some undefined way racist and sexist.
Whether these objections have merit and how they might be
answered are, of course, important questions. But taken together
they suggest a matter of more profound importance. It is the great
difficulty—perhaps the impossibility—of having any worthwhile
discussion about educational goals. Across ideological boundaries
discussion about goals is quickly overwhelmed by animosity. Yet it is
only at the political extremes that we find groups advocating goals
that others reject. Across a very wide ideological range, everyone
wants students to be both proficient in basic skills and able to think.
They want students to have factual knowledge as well as
understanding. They want economically useful learning and they
also want moral and cultural values (and there is not much
disagreement about what those values should be). The animosity
arises because of mistrust of motives and because goals cannot be
separated from methods. It does no good for A and B to agree that
children need to become fluent readers if A wants to teach phonics
and B considers phonics to be a form of violence against children.11
But discussion of goals is difficult even among people who agree
enough that they should be able to disagree constructively. Such an
occasion was a conference on the future of liberal education (Smith,

11
I’m not being fanciful. This very charge was made by one of the leading
spokesmen for Whole Language in an e-mail discussion group.
in press). We (I was one of the organizers of the conference)
restricted the participants to those who could be expected to agree
that a liberal education is a good thing. Within this large category,
however, we invited participants with interestingly different
viewpoints on the matter—philosophers, economists, and
educationists of different persuasions. The result was indeed an
interesting conference, but the discussion never got as far as what
you could call constructive disagreement. Instead, all the effort went
into clearing up misunderstandings, questioning one another’s use of
terms, denying supposed disagreements and sometimes denying
supposed agreements. Some discussions continued after the
conference and a year later were just beginning to bear fruit.
Why is it so difficult? Why is it so much harder for a school system
to discuss and arrive at a set of realistic and motivating goals than it
is for an information technology company, for instance? Well,
interestingly, IT companies sometimes also lose hold on reality when
they tangle with educational goals. In the middle ‘90s, a major IT
company announced that it was going to shift from marketing
educational software to marketing educational ‘solutions.’ Having
some acquaintance with the educational arm of that company, I
wondered where these ‘solutions’ were supposed to come from.
Where was the educational problem solving capacity? The answer it
seems, is that there wasn’t any. The bold venture quickly faded away.
No company would announce that they were getting into the genetic
engineering business or the machine translation business unless they
were sure they had the capacity to do it or knew how to acquire it.
The company’s marketing people, I suspect, had simply adopted a
myth that pervades the whole education system. It is the myth of
unlimited problem solving capacity. Adopt any educational goal and
means can be found to achieve it. All it takes is marshalling the
resources already available in the system.
To John Dewey means and ends were inseparable. An aim was
the ‘foreseen end’ of an activity (1916, p. 106). He spoke of “having a
mind” to do something—a quaint expression that he used to convey
a quite modern notion:
To have a mind to do a thing is to foresee a future possibility; it
is to have a plan for its accomplishment; it is to note the means
which make the plan capable of execution and the obstructions
in the way,—or, if it is really a mind to do the thing and not a
 vague aspiration—it is to have a plan which takes account of
resources and difficulties. ((1916, p. 103).
Yet Dewey seems not to have reckoned with education’s incapacity
for this kind of foresight and planning. A clue to why he overlooked
it comes from his comparison of education to farming:
The educator, like the farmer, has certain things to do, certain
resources with which to do, and certain obstacles with which to
contend. The conditions with which the farmer deals, whether
as obstacles or resources, have their own structure and
operation independently of any purpose of his. Seeds sprout,
rain falls, the sun shines, insects devour, blight comes, the
seasons change. His aim is simply to utilize these various
conditions; to make his activities and energies work together,
instead of against one another. It would be absurd if the farmer
set up a purpose of farming, without any reference to these
conditions of soil, climate, characteristic of plant growth, etc.
(p. 106-107).
Farming, at the time Dewey wrote, was a traditional craft, much
as teaching still is. The goals Dewey imagined for the farmer were
formed within the scope of that traditional craft and attaining them
required no advances in it. They were to plant certain crops that the
farmer already knew how to plant, and to cultivate and harvest
them, taking into account the variables that the farmer already knew
how to take into account. That seems to be how he saw teaching as
well, and it is how teachers quite reasonably see it on a day-to-day
basis. They are to teach certain things that they know how to teach,
taking into account the variables that they already know how to take
into account.
Although Dewey (1929) did acknowledge a role for research in
shaping educational practice, he seemed to have little sense of large
unsolved problems, of worthy educational goals that we don’t know
how to achieve and that call for advances in knowledge and
invention. Instead, he too seems to have accepted the myth of
unlimited problem solving capacity, so that all that was required for
education to advance was clear thinking, resolve, and the abundant
resources of experience. In the decades of dispute that Dewey’s own
ideas helped to ignite, all sides have tacitly accepted the myth, as do
most reformers of the present day. In the opening chapter of this
book I asserted that education is stuck. This myth is what it is stuck
on.
 Here is the worst of it: The belief that the necessary knowledge is
already in hand not only inhibits the search for means, it undermines
efforts to construct a new vision. For, as the Latin proverb says, we
cannot desire what we do not know. It takes advances in knowledge
to create new visions of what is educationally possible.
Beyond ‘Excellence’
When U. S. President George Bush, in 1989, vowed that his
nation should be first in the world in science and mathematics
achievement by the year 2000, serious educators across the land
groaned. It was obvious that it couldn’t be done, but that was not the
cause for dismay. The dismay was over what would happen to
education as a result of pursuing a goal that was not only impossible
but retrograde. Instructional researchers in science and mathematics
felt at the time that breakthroughs were imminent. The previous
decade of research had probably yielded more insight into the
difficulties of learning school subject matter and academic skills than
all preceding decades together. Instructional experiments were
beginning to show promise of student learning of a quality not
previously imagined. Now all of that stood to be forgotten in the
rush to raise scores on tests—tests that had been created long before
any of the knowledge advances that were raising new hopes.
I happened to be in Singapore, meeting with educationists, when
word came that Singapore had scored first in the world on the
computational part of the Third International Mathematics Study.
People were naturally elated, but they scarcely missed a step in their
pursuit of what they saw as the next objective, which was to improve
mathematical problem solving and invention. Their vision was not
borne aloft by romantic fluff but by the economic drive for which that
small nation is famous. A front-page newspaper headline appearing
in the same week gave the flavor of it:
Can Singapore Compete with Bill Gates?
The article’s reasoning was straightforward. Singapore has a
population of a few million. Microsoft can select talent from a pool a
hundred times larger, plus being able to recruit talent worldwide.
How, then, can Singapore hope to compete in an industry based on
technically sophisticated inventive and design talent? Answer:
Liberalize immigation and do a much better job of developing talent.
There was no suggestion that being first in the world in arithmetic
skills was even relevant to the problem.
 If the current back-to-basics reforms in the United States work, it
is just possible that mathematics test scores will in a decade or so be
up to where Singapore’s were in 1996. But where will Singapore be
then—or any of the other high-scoring countries, almost all of which
are seriously pursuing higher-level goals? I’m not sure. The road to
thinking skills and creativity is lined with quacks. Some countries are
liable to fall victim to quackery and may even lose ground as a result,
but I expect others will not. The Dutch, for instance, not only score
well now in mathematics but they seem to have better quack
detectors than North Americans do.
International economic competitiveness is not an exalted goal, of
course, but it is a notch up from winning a test score race. A still
higher notch is that of producing the kinds of citizens who can thrive
in and contribute variously to a knowledge society. For strategic
purposes it is well to think of these as hierarchically ordered rather
than as competing goals. That is, the goal of producing economically
competitive talent ought to include the goal of raising conventional
achievement test scores. The goal of full citizenship in a knowledge
society ought to include development of economically competitive
talent. And if a still higher goal is to be proposed, it should include all
of those. This is strategically if not logically necessary, because
support quickly evaporates when people think they have to give up
their sensible goals in order to adopt your riskier ones. Educational
innovators would prefer not to be bothered with goals of a lower
order than those they intend to reach, and some of the effort that
may have to go into raising conventional tests scores is effort that, in
a more ideal world, could better be spent elsewhere. But innovators
who want to be taken seriously realize that the admission ticket for
being taken seriously is evidence that their innovation satisfies other
people’s goals as well as the goals they are trying to get people to
adopt. Easier said than done? Of course, but no one has suggested
that instructional reform is easy.
What are these higher level goals, and how can the sights of
educational reform be raised to target them? The metaphor I have
just used suggests a major conceptual problem to be overcome. As
soon as you move to goals above the level of test scores or specifiable
performances, the target-shooting metaphor no longer works. You
must conjure up a state of things, a network of conditions that must
be met altogether rather than as separate targets. (This does not
mean, of course, that they cannot be reached by stages, only that the
stages themselves are complex.) All sorts of organizations must find
ways to pursue complex goals. The top-level goal of a for-profit
company might take the form of a clear target, such as achieving a
certain level of return on investment; but if the company determines
that achieving this goal entails providing a superior quality of
service, then they face a problem not much different from that of a
school system that determines its goal is to provide a superior quality
of education. Each must craft an image of something beyond what
presently exists but that looks achievable. Creativity is required, not
only to achieve the goal but to imagine it in a sufficiently realistic
way that it can be pursued.
An educational vision, I have been insisting, must offer an image
of an outcome, not of a process. It is time now to look more closely at
this stricture. When innovators move beyond test scores and other
immediate indicators of achievement, they find themselves issuing
rather long-term promissory notes for future benefits to the learner
and to society. Unless they are among the few who manage to
achieve celebrity status, they are likely to find those notes rejected:
no credit rating, no collateral. As observed in Chapter 7, creditable
ideas about what should be done now to prepare students for the
uncertain future are in short supply. Back-to-basics, liberal
education, and developmentalism offer at most slightly renovated
versions of their traditional answers. Futurists generally offer
nothing of substance. The people attached to these schools of
thought are not fools, they are just faced with a problem that is very
difficult.
To keep matters in perspective, it is important to realize that test
scores are promissory notes, too, when they are used to judge the
worth of an educational program. They require us to believe that if
School A’s students score higher than those of School B, then School
A’s graduates will do better in the future than those of School B on
worthy criteria of membership in society.12 These promissory notes
have been in circulation long enough that they have come to be
accepted as legal tender, but that does not necessarily mean there is
anything behind them. There is ample evidence, as one might expect,
that reading test scores have wide-ranging predictive value, much
12
That prediction may be self-fulfilling, of course, if universities and
employers believe it and award opportunities on the basis of it. All that does
in the present context, however, is add further obstacles to the selling of a
new educational vision.
like the IQ, with which they are closely associated. But if we move
from there to more specialized kinds of achievement, predictive
validity trails off. At the extreme, we have no evidence at all that a
program that increases scores on thinking skills or creativity tests
has any effect on performance in later life.13
Futurists stir our imaginations by telling us that the future is
going to be qualitatively different from the past. Therefore, the
futuristic reasoning goes, education must produce a new kind of
educated person. This is an exciting notion, but one that quickly trails
off into verbalism and technohype. I would like to see educational
reform take on some of this excitement, without the verbalism and
hype. We must dismiss suggestions that surfing the Internet or
processing hypertext requires a differently functioning brain.
Previous discussion has suggested two important constraints that a
realistic futurism must honor:
1. New human traits are unlikely to emerge. Whatever may be
posited as desirable characteristics for tomorrow’s citizen, they
are likely to have been recognized as desirable for centuries and
already to figure in lists of educational objectives. If that is true,
reformers can forget about inventing new objectives and must
craft a new vision out of familiar objectives.
2. Whatever new or elevated vision may be pursued, it must be
translatable into objectives for the here-and-now. Achievement
test scores will continue to hold sway unless they can be
replaced by other immediate indicators of effectiveness.
These constraints may at first seem to rule out everything new
and exciting, but as I tried to show earlier, they do allow one
possibility. It is the possibility of qualitatively higher standards—that
is, standards that are not merely an extension of current metrics but
that introduce a new and higher realization of things educators have
been trying to accomplish right along. The prototype for introduction
of a qualitatively higher standard was the so-called ‘Greek
revolution’ in sculpture. The educational examples I gave of a
qualitatively higher standard of understanding may seem trivial by
comparison, but I would not reject the possibility that they mark the
13
The same, actually, can be said about score increases on any mental test. The
fact that reading test scores correlate with future academic performance does
not mean that doing something to raise reading test scores will have a
corresponding effect on future academic performance. That is a separate
empirical issue and one that has hardly been researched at all.
beginnings of another Greek revolution. Early Greek sculpture does
not look dramatically different from its predecessors, either, except
that we can now see where it was starting to head.
Reading reform has of late been almost exclusively concerned
with bringing poor readers up to an average level.14 This is reform
aimed at a quantitatively higher standard, albeit an extremely
important one. But for education as a whole, we should be
considering the possibility of a qualitatively higher standard of
reading. Assuming that the world’s knowledge keeps increasing not
only in quantity but in level of complexity, competent citizens of the
future will need to be able to read and understand harder stuff. A
competitive edge will be gained by those who can get it right the first
time and not require cycles of e-mail messages and phone calls before
their initial misreadings are straightened out. (When researchers talk
about reading comprehension, they mean comprehending the kinds
of discourse that ordinarily come in written form; that these may be
received aurally rather than visually is, except for dyslexics, largely
irrelevant.)
By a complementary argument, people in a knowledge-based
society will have more complex things to communicate and therefore
will need a qualitatively higher level of writing or text composing
ability. More communication will be at a distance and so will require
more of the literary skill of putting necessary context into the text
(Olson, xx). People whose communication skills depend overly much
on pointing will be at a disadvantage.
Qualitatively higher standards of understanding, reading, and
writing are so closely tied together that it is hard to imagine getting
very far with one without also attending to the others. Reading is the
primary means of access to the conceptual tools, facts, and ideas
necessary for sophisticated understanding; and writing, when
carried out in a knowledge-transforming way, is a powerful means
of developing understanding (Scardamalia & Bereiter, 19xx, 1991).
I am speaking very broadly here. If you look into the cognitive
research in any subject area you will see qualitatively higher
standards taking shape. Some of the ideas come from studying
experts in the disciplines. Others come from studying students who
14
Whose average is an issue in this context. One highly-touted approach,
Reading Recovery, is aimed at bringing the poorest readers up to the average
of their class. But this typically means bringing them up to the average of a
class that is itself markedly below the national average.
show qualitatively advanced abilities and knowledge. Still others
come from educational experiments that yield results beyond the
kind anticipated. The last are especially important, because they
show that qualitatively higher outcomes can be produced, that they
are not just lucky accidents of birth and experience. Taken together,
these ideas represent a new kind of educated person. The
characteristics of such a person are the same ones that have been
honored throughout modern times and many of them go farther back
than that. What is new is a quality of understanding, a quality of
intellectual skills, a quality of interest, sensitivity, and appreciation
that have heretofore been found only in exceptional people but that
may now become expected results of regular education.
This, however, is a promissory note backed by only scattered
experiments of limited scope. How could one ever hope to get
reformers to honor it? There have to be convincing demonstrations.
There have to be evaluations which assure people that conventional
standards of achievement are not being sacrificed in pursuit of
qualitatively different goals. But demonstrations can only be
convincing if people understand what you are demonstrating, and
that is where the conceptual limitations I have been discussing
become crucial. People have to understand knowledge as more than
stuff in people’s heads. They need a conception of what it is to create
and work with knowledge. They need to believe that the hidden
iceberg of personal knowledge is as important and much larger than
the two peaks of it that show. They need to judge educational
reforms according to what they do for the whole iceberg.
Strategy: Reforming the Reformers
Are we not, then, in an impossible bind—where reformers cannot
adopt a qualitatively higher goal until they have seen it being
achieved and they cannot see it unless they understand it and they
cannot understand it unless they see it? Broadly speaking, yes. But
there are some breaks in the vicious circle. The breaks are of recent
origin. They have come about because of things going on outside
education, especially the rise of knowledge-based industries and of
occupations devoted to knowledge work, and the spread of ideas
such as intellectual property, knowledge management, and
knowledge evolution. More and more people are acquiring, through
their work, concepts that should help them grasp a new educational
vision, if someone can help them make the connection.
 At one time our research funding required that we attend a yearly
technology fair in Silicon Valley and show off our accomplishments
to a flock of engineers. The software we had to demonstrate at that
time—CSILE (Scardamalia et al, 1989)—was not of a kind to
impress techies. It didn’t do much. All that made it interesting was
what kids did with it. So instead of showing off the technology, we
showed videos and transcripts of kids working on problems such as
why humans can talk but chimpanzees can’t or what happens in an
asthma attack. Within minutes the engineers quit acting like
engineers and started acting like parents. Their overwhelming
reaction was that they wanted this for their own children. They
recognized that they had never done anything comparable to that in
fifth or sixth grade, and that kids who could do it were at a
tremendous advantage over those who were memorizing facts or
piddling away the hours on multimedia ‘projects.’
These were not average parents, of course. Two things set them
apart. First, they appeared to be little concerned about achievement
test scores. If there was a competitive slant to their interest, I would
guess it was toward how their children might gain an edge over
other high-achieving kids—and we were showing them what that
edge could be. Second, they were themselves employed in the
creation of knowledge, and so were able to recognize it when they
saw it and to appreciate its significance. Yet there is no reason to
suppose they were especially sophisticated in matters of education.
They merely had a better idea of what counts in the modern world.
The lesson I draw from this experience is that if a higher level of
reform is to have a chance, it must build a consituency among people
like those Silicon Valley engineers—people who are already part of a
knowledge building culture and who have a sense of what it takes to
belong. This does not automatically make them favorable to
knowledge building in schools. Back-to-basics reformers are often
recruited from the same population, and many others are romantic
believers in the magic of the Internet and multimedia. All I am saying
is that they have many of the conceptual resources needed to
understand. They are not stuck in the vicious circle and so they may
be educable.
To build a power base among such an elite, however, requires a
shifting of priorities. Educational reform, especially in the United
States, is overwhelmingly concerned with improving education for
the most disadvantaged. This has a great deal of political will behind
it. It appeals both to the liberal interest in welfare and equality and to
the conservative interest in productivity and self-sufficiency.
Although the reasons for this emphasis may be compelling, they
introduce a bias into educational policy-making that ultimately
works against even the people it is most intended to help.
Educational policy making typically consists of people with
university degrees and secure economic positions deciding what is
best for children whose prospects are not of the brightest. In this
situation, proposals such as teaching for understanding are unlikely
to get far. They will be seen as failing to address the main problems
and likely not to work anyway. There will instead be a bias toward
quantitative improvement on low-order objectives. This does not
mean that the policies these people might devise for their own
children would be much different or that a more representative
group of policy-makers would arrive at more enlightened policies.
The crack in the vicious circle is very small indeed. All it amounts to
this: People are more educable when they are thinking about
educating their own children than when they are thinking about
educating someone else’s. Not because of their emotional
involvement; that may work against reason. They are more educable
because they can connect their own experience to that of the students
who are the targets of their policies. If you can show knowledge
workers a much closer connection than they had imagined between
things going on in school and what goes on in their professional
lives, they may become receptive to a whole new range of
possibilities for education. The place to start, I am suggesting, is by
getting reformers to want something different for their own
children. After that you can try to show them that less advantaged
children can have it too.
Sometimes it is possible to merge these two steps into one. In the
Jasper Project at Vanderbilt University, they have an ingenious way
of presenting their educational approach to business people.
“Jaspers” are complex realistic problems conveyed through
professionally produced video dramas. Representative Jaspers are
one having to do with the logistics of setting up a dunking tank for a
school money-raising event and another having to do with the
problem of flying a wounded eagle out of a distant field. The central
problem may require formulating and solving a score of
mathematical subproblems. The Jaspers are aimed at middle school
students, who work on them in small groups. The idea is to get
students to use mathematics the way it is used in real life, where the
needed information is embedded in the environment and in the flow
of events rather than being neatly set out as it is in traditional word
problems. To give adults a feel for this approach to mathematics
learning, the presenters show them a Jasper video and then divide
them into small groups and set them to work trying to solve the
problem. After the adults have worked at the problem long enough to
have run into difficulties, the presenters announce that they are
bringing in consultants to help out. On signal, a line of sixth-graders,
with visible minorities well represented, file into the room, causing
an outburst of laughter. The young students fan out to the working
groups and proceed to help them move ahead on the problem. It is
easy to see why such a demonstration has impact. Here are the kinds
of adults who are busy across the land agitating for higher standards
in mathematics and a return to the basics, encountering a
mathematics problem that is too hard for them now, that would
have been unthinkable for them when they were in grade school, and
they may well suspect is also beyond the capabilities of their own
children. Then suddenly they find themselves being coached by
students for whom this kind of problem is familiar fare—students,
moreover, who look like the ones they have assumed to be in need of
emergency repairs. How can they not believe they have been missing
something?
The crucial first step, as illustrated in both the Jasper example
and the Silicon Valley technology fair example, is to get influential
adults to see themselves, their own work, their own aspirations and
those they have for their children, as part of the educational reform
picture. This is just a first step, but in order for it to amount to more
than a momentary glimmer, education and knowledge work have to
be brought into the same conceptual space, and that is a big order. As
it is now, education talk has almost no relation to real-world
knowledge talk. Some of the same words are used, but the contexts
are perceived so differently that there is little common meaning:
Word What it Means in What it Means in
Knowledge Based Schools
Enterprises
Knowledge Producing intellectual Learning by doing
Building property
Achievement Recognized contributions to Test scores; grades
the success of the
enterprise
Understandi Knowing the non-obvious Having beliefs that
ng stuff correspond to those of
experts; being able to
explain
Creativity Innovative design and Uninhibited self-expression
strategizing
Problem Figuring out ways to Solving word problems in
solving achieve goals when off- which all the necessary
the-shelf ways won’t do information is provided
Cooperative Collaborative knowledge Collaboration on seatwork
work building or in producing reports,
displays, presentations, etc.
Science What you need to know in A body of knowledge about
order to understand the physical and biological
technology or the causes of things; what scientists
many problems believe about these things;
research methods
Mathematics Conceptual tools needed in Procedures for solving
many kinds of technical numerical problems
work
Research Finding out things you Exercising research
need to know in order to methods; collecting
solve a problem or produce material for a report;
a result pursuing curiosity
The discrepancies obviously are not just semantic. What normally
goes on in schools is very different from what goes on in knowledge
based enterprises and no fiddling with the vocabulary is going to
bring them any closer. But what I have been arguing in recent
chapters is that they could be much closer. Schools could be places
where knowledge is a public product and not just something in
students’ minds; where its production is a collective, collaborative
effort and the focus of students’ individual efforts is on the success of
this joint enterprise; where what is learned is put to use in the further
creation of knowledge; where the problems to be solved are
problems in the advancement of knowledge. Much of the more
innovative educational research is heading in that direction. But how
you talk about things, the concepts you bring to bear, do make a
difference. Talk about education tends to sound like the right-hand
column no matter who does it. This not only makes it difficult for
people outside education to connect it with their own work, it
impedes the movement of education toward real knowledge
building.
 Note that I am not saying business people have the answers and
educators should learn from them. The business literature on
knowledge is just as primitive as the educational literature and
equally bound to folk theory. It is just that the nature of their work
has led people in knowledge-based enterprises to adopt a more
pragmatic approach to knowledge, less encumbered by ritual and
fetishism.15 Conceptual advancement is needed on both fronts. It
ought to go on in concert. This book, although tilted heavily toward
education, is really about conceptual problems that are common to
knowledge businesses and education. Getting people who are
already at home in knowledge-producing enterprises to see schools
as sharing the same problem space with them would be an enormous
step in moving public thinking toward a higher level of educational
objectives and this in turn could do much to draw the thinking of
educators in that direction as well.
Along with the talk there have to be demonstrations. People have
to see instances of real knowledge building going on in schools. In the
work that we have been doing on Computer Supported Learning
Environments (CSILE), we have produced a lot of quantitative
evidence of educational gains (Scardamalia et al., 1992) and this has
been important for credibility, but by far the most influential results
come from simply having young students walk viewers or visitors
through the CSILE database they and their classmates have
constructed, explaining what they have accomplished. One grainy
video alone has been worth millions to us in research support. On it,
an eleven-year-old boy takes you through his investigation of why
chimpanzees can’t speak. He shows hand-drawn computer graphics
comparing the oral cavities and the vocal aparatus of chimps and
humans, explaining the critical differences, and then does the same
with diagrams of their brains, wrapping up his tour with, “And that’s
why we can talk and chimpanzees can’t.” People’s jaws drop. They
have never seen anything like it. Yet it is obvious that the boy got his
information from books, that the neuroscience is simplistic, that the
explanation leaves many questions unanswered. And it is also
obvious that this is not your average eleven-year-old. He is

15
As for imitating business practices, one hesitates to generalize. There are
surely ‘best practices’ here and there that schools could adapt to their needs,
but without better conceptual equipment there is no way to recognize them or
to do more than imitate surface features.
unusually articulate and serious-minded. So what is impressive
about it as an educational demonstration?
Several qualifications are in order before I try to explain. In order
to be seriously impressed—that is, impressed enough to do
something—people need to be assured that what they saw on the
video was not a one-of-a-kind event. They need to see a whole room
full of students doing comparable things. And not everyone is
impressed even then, or they are impressed for the wrong reasons.
We have had educational technology supervisors who ignore what
the students are doing, look at the software, sniff, “It’s just a
database,” or “It’s just a bulletin board,” and go tell their bosses to
forget about it. (This is a version of the “We already do that”
syndrome.) And then there are the educators who see it only as a case
in support of their pet idea, be it constructivism, project-based
learning, or child-centered education. (Yet another version of “We
already do that.”) That leaves, however, a significant number of
people who really seem to get it, and so the question is, what do they
get?
A revealing remark came from a visitor who was a highly-placed
civil servant with the job of promoting knowledge-based industries.
After spending an hour in a CSILE classroom, he said, "I think I have
seen my first learning organization." We have seen companies where
they thought that being a learning organization meant having a
company-run school for training employees. If that is what learning
organization means, then all schools are learning organizations and
the visitor's remark was absurd. In business literature, the term is
used in different ways by different writers and so its meaning in
common usage is far from clear. Business literature relies heavily on
examples rather than definitions, but the examples used to convey
the idea of a learning organization range over practically everything
that could be considered a smart personnel policy or way of making
the most of employee's brain power. What holds the examples
together is some vague idea of the pervasiveness and value of
knowledge in an organization and the importance of nurturing it. As
I argued in Chapter 6, I don't think you can get much farther than
that unless you cash in folk theory for a better epistemology and
theory of mind.
But in the absence of clearer definitions, it helps to have cleaner
examples. The CSILE classroom offered one. There wasn't any
product or service going out into the world. Yet something was being
produced, and about the only thing you could call it was knowledge.
What was it made from? Other knowledge—mostly knowledge
brought in from the outer world and processed into something of
more local value. What was the product good for? For the
production of more knowledge. Thus, in the classroom you had a
model of a learning organization or, as we would prefer, a
knowledge building organization in relatively pure form, unobscured
by the many other functions of a money-making business. I don't
suggest that our visitor saw it in just that way. I don't think you can
see it that way until you have absorbed the idea of conceptual
artifacts. But his experience had sharpened his impressionistic
knowledge enough that in some less articulated way he knew what
he was looking for, and recognized it when he saw it.
There are other approaches to knowledge building in schools,
that may be as effective or more effective than CSILE in what they
are trying to do. Typically, they are less clean as examples because
they either look more like conventional schooling—there is a teacher
up front leading a discussion, for instance—or there is some tangible
project—such as producing a multimedia document—that capture's
the observer's attention. But to the extent that schools can provide
the public with examples of what it means to produce and work with
knowledge, they connect to concerns of the working world at a
higher level than before.
Schools presently connect to the concerns of the working world at
a very low level. The concerns are over why Johnny can't read or
calculate or find Rome on the map. Those are legitimate concerns
and they need to be addressed. But if the connection is only at this
low level and everything beyond it is hand-waving about higher-
order thinking skills and lifelong love of learning, we can expect a
continuation of low-level reforms. To connect at a higher level,
people in knowledge-based organizations need to see the work that
goes on in schools as similar, at a reasonable level of abstraction, to
the work that goes on in their organization. Once they see it that
way, they can not only be receptive to a qualitatively higher level of
educational goals, they may even begin to understand their own
work better. But all of this presupposes a conceptual framework for
knowledge building and learning that enables both business people
and educators to think more penetratingly and constructively about
what they are doing. That is the transformation that will make it
possible to reform educational reform.
 Conclusion
In broadest terms, the problem as I have tried to formulate it in
this book is to get educators and others out of the two-dimensional
world of folk theory and into a three-dimensional world in which it is
possible to do fuller justice to the role of knowledge in a knowledge
society. There is no way that this can happen overnight. The left-
right swings of educational reform and counter-reform are likely to
continue for years. But if Peter Drucker is right that education will be
the most important factor in a nation’s prosperity in the 21st century,
then the future belongs to that society that can fasten on to and
achieve qualitatively higher standards. The successful society will
steer past back-to-basics movements and the periodic revivals of
child-centered education. It will try to steer education to the same
place it is trying to steer itself.
The problem is that, in whatever sense a society may be said to
know or intend, society does not know where it is trying to steer
itself. The envisioned knowledge society that Drucker and many
others are now talking about is not a place where any of us have
been. We may have experienced intimations of it on a small scale, in
our own families or in groups we have worked with, but we have not
seen it on an institutional scale. We can make visionary statements,
but we cannot say with confidence that any particular reform is a
step in their direction.
Schools will continue. Will they continue to be the battleground of
old pedagogies or will they become Internet cafes? If those are the
only choices, it is a safe bet that they will continue to be
battlegrounds. There is a third possibility, however, and that is that
they become laboratories for testing designs for a knowledge society.
Schools have many limitations but for this third possibility many of
those limitations are advantages. Schools are places where
knowledge creation can go on, but where it does not have to be
market driven or competitive. That has been the virtue of research
universities and they continue to demonstrate its value, despite
alarming intrusions of the marketplace. But knowledge creation in
universities is the work of a minority, and most of it takes place out
of eyeshot of the undergraduates, who go about their business much
as they did in centuries before the research university was invented.
Knowledge creation that goes on within the educational process is a
different matter, and while it is possible in undergraduate university
programs, it is easiest to implement in schools.
 Knowledge creation in schools is the creation of knowledge by
students for their own use. It is thus like subsistence farming.16 The
school is like an agricultural village. Unlike knowledge-creating
companies, its members are not selected because of their skills. They
belong because of where they live. Thus the school is more a
miniature society than a miniature enterprise, and accordingly it has
the potential to be a model for a knowledge society rather than only
a model for knowledge-creating organizations within some perhaps
quite different kind of society. If a school is to be a humane and
successful knowledge society, everybody in it must have a part,
everyone must find it meaningful and rewarding. A vast range of
talents and temperaments must be accommodated. And the
knowledge that is produced must be good knowledge. It must be
effective in producing more knowledge and it must be worth
students’ carrying with them when they leave school. Transforming
schools in this way presents big and mostly unsolved problems, but
they are problems society as a whole must solve if the knowledge age
is to be a good age for humanity. Schools will have to tackle them
too, sooner or later. I am suggesting they do it sooner, thus in certain
ways leading social change rather than following in its wake.
In order for any of this to make sense, however, people have to be
able to see knowledge creation by students as genuine productive
work, not fundamentally different in kind from the knowledge-
creating work that goes on out in the world where people produce
knowledge of various kinds for a living. They must not confuse the
work of producing knowledge with the learning that inevitably
accompanies it or with the media objects and performances that may
grow out of it. Such confusion exists to some extent in the world at
large, but in education it is endemic. The result is to undermine any
effort to make the classroom into a genuine rather than a pretend
knowledge creating community and to give it a role in the
functioning of a knowledge society beyond that of merely furnishing
students with knowledge and skills that may be of use in the future.

16
The analogy can be carried further. To the extent that knowledge created in
schools has value beyond the classroom where it is created, it enters into a
barter economy. No one is likely to pay money for it, but it may be exchanged
with other classrooms or with other social groups such as museum curators
or teacher trainees or more advanced students, who find value in it for their
own purposes and who have knowledge of their own to trade (Scardamalia &
Bereiter, 1996).
 The main point of this closing chapter has been to elaborate on
the idea of a qualitatively higher standard of educational outcome.
Quantitatively higher standards are easy to understand. Simply take
the scales by which we currently measure educational results and
point somewhere higher up than the current level. Qualitatively
higher standards require that you imagine a different kind of
outcome. You need a different conception of what it means to be a
good reader or writer, of what it means to understand a scientific
theory, a historical event, or a poem. Folk theory of knowledge and
mind fails at this point. Conventional standards are grounded in the
folk conception of the mind as a container and in commonsense
notions of skill. Raising standards means demanding more items in
the mental container and moving the bar a notch higher in tests of
skill. A notion like deeper understanding, which implies a qualitative
shift, becomes either an empty slogan or it gets translated into tests
of skill.
Envisioning qualitatively higher standards does not require
exceptional imagination. We can point to people who are exemplars
of superior kinds of reading and writing ability, of scientific,
historical, or literary understanding. What we need are conceptual
tools for translating exemplars into goals and goals into pedagogy.
The two-dimensional world of folk theory and folk pedagogy does
not provide the tools or the space in which to use them. We need, I
have argued, a third dimension that allows conceptual artifacts to be
distinguished both from the mental states of the people who create
or deal with them and from the physical and social world that these
conceptual artifacts relate to. Qualitatively higher standards of
cognitive skill imply skills in working with conceptual artifacts.
Qualitatively higher standards of understanding imply qualitatively
different relationships between the knower and conceptual artifacts.
There are plenty of educational problems that can be recognized
with our old folk theoretic conceptual equipment, and available
knowledge sometimes suggests a solution. But, as we are seeing in
the case of reading reform, the equipment is so crude that reformers
tend to get it wrong, can’t distinguish a solution from a subterfuge,
and the counter-reformers get it wrong as well. And so we get
pendulum swings instead of progress. The society that moves ahead
will not be one that fixes the pendulum at one position or another. It
will be a society in which there is enough conceptual change that
something begins to happen that has not been seen before: the
disciplined production of new educational ideas.