Review of David Perkins’ “Outsmarting IQ: The Emerging Science of Learnable Intelligence,” 1995, Free Press.
Link to review on Amazon --> http://www.amazon.com/review/R3AYGZV7G7AUTO/ref=cm_cr_rdp_perm
Profound Thinking By Example
This is the single best book I’ve come across on the potential for improving human thinking ability. I give it my highest recommendation; I think it should be read by everyone interested in problem solving, decision making, and human abilities in general. It is amazingly broad in its coverage of data, profoundly deep in its treatment of specific lines of relevant evidence, and ingenious in its vision of the future.
What impressed me most about this book is that the author, David Perkins, demonstrates the power of deep reflective thinking by his own example in the organization and treatment of evidence throughout this book, in his critical treatment of his own evidence and ideas, in his creative original ideas, and in his effective consolidation and filtering of massive amounts of research. Showing how asking the right questions can help us understand seemingly contradictory data about intelligence, Perkins gives an engaging plausibility proof for the kind of reflective intelligence he argues for in this book.
The Concept of Realms of Thinking
To give away the ending, the book culminates in a model of problem solving ability based on the metaphor of a map. Human thinking ability results from learning our way around. Navigation is fundamental to all sorts of human thinking. Perkins suggests that all intelligent human thinking results from navigation of various kinds, which can be thought of in terms of levels of realms. Perkins organizes the realms in an overall map or “mindscape” from the lowest level of specific contexts of thinking to the highest level dealing with thinking itself.
In learning to solve problems we not only learn our way around physical realms geographically, but we learn our way around specific contexts we find ourselves in such as the realm of buying a house or the realm of choosing a career. We learn our way around different situations like resolving conflicts or making purchases in general. We learn our way around professional fields like law, physics, and mathematics, and areas of technical expertise such as probability and statistics, game theory, and business. We learn our way around the use of tools. We learn our way around various basic kinds of challenges like problem solving, decision making, planning, and learning. Finally, at Perkins’ top level, which he calls thinking dispositions, and we learn our way around thinking itself in terms of the qualities and attitudes that make it more or less effective.
Perhaps the central thrust of this book is that in organizing human problem solving areas into navigational realms, Perkins is not just providing a training map for learning problem solving skills a million different areas, he is also making a case for the learning the critical skills of navigation itself.
Perkins’ realms are very similar to the traditional concept of domains of expertise, but different in one critically important way: realms emphasize the central skills of navigation rather than just the use of repetition or rote memorization or even just the use of deliberate practice. The concept of realms makes it more explicit that all areas of ability that we learn share some commonality in terms of key skills and attitudes we need for navigation itself.
It is learning to be a better navigator; in all realms of human thinking and not just certain subset of them; that is the central message of Perkins’ book. This is encapsulated in his concept of “reflective intelligence.” Reflective intelligence is the aspect of intelligence that can be most improved for the greatest effect across the range of all realms of thinking. Perkins reviews a number of different attempts to improve human thinking and makes various suggestions based on their results regarding specific kinds of changes that can be made to educational curricula in order to teach children to be better navigators in all areas.
Getting Perspective on Intelligence through 3 Dimensions
In giving away Perkins’ final model, I’ve skipped over two very important and interesting aspects: his argument for the model he uses and for the prospect of learnable intelligence through better navigation, and his predictions for important areas of the evolution of learnable intelligence.
The bulk of Outsmarting Intelligence deals tightly with the subject of the title, the legacy of how intelligence has been envisioned and researched so far. Perkins deals in equally deep, reflective, careful, and often fascinating manner with: (1) the evidence for a single common problem solving ability from psychometric data, (2) the evidence showing us how novices think differently from experts, and (3) the evidence showing us what happens when we try to learn general skills and rules for solving problems in general and how computers solve problems.
From these three bodies of evidence, Perkins derives three corresponding dimensions of human intelligence: (1) a neural intelligence dimension which respects what psychometric data gets right and is most closely associated with what we typically assume IQ tests are measuring, (2) an experiential intelligence dimension which respects what expertise research data gets right, and (3) a reflective intelligence dimension which respects what we have learned about metacognition and from the various programs that have tried to teach thinking skills in general.
Neural intelligence, Perkins concludes, is a real dimension of human ability and very important in some situations especially, but it is simply the wrong target for attempts at improvement for various reasons.
Experiential intelligence represents most of our actual problem solving abilities in practice.
Faced with novel and complex situations where we have no relevant experience, our neural intelligence gives us our best chance at solving the challenges presented. But once we have been acquiring experience in an area, a difference in expertise will make people better problem solvers in that area than will a difference in general intelligence.
So experiential intelligence and neural intelligence work together to make us the generally good problem solvers that we are in most situations: neural intelligence helps us deal with novelty and complexity, and experiential intelligence helps us acquire the knowledge and skills we need to deal with specific domains.
So the obvious question is: what role does reflective intelligence play and why does Perkins consider it so important?
The Significance of Reflective Intelligence
Perkins reviews various lines of research into the wide variety of situations where otherwise powerful problem solving abilities seem to fail us in systematic ways. He looks at social psychological effects, cognitive shortcuts, and so on, similar to other reviews of blind spots in human thinking by many other authors except that Perkins attempts to characterize these foibles specifically in terms of side effects of our experiential intelligence.
Perkins suggests that the human mind is mostly akin to a pattern seeker and pattern-driven problem solving engine and as a result its weaknesses are also those we would expect from a pattern-driven process. The human mind often tends to be hasty, narrow, fuzzy, and sprawling.
HASTY. The goal of a pattern seeking intelligence is to find the right response that most closely matches the current situation rather than making an exhaustive search. As a result, our experiential intelligence tends to mislead us to jump to hasty conclusions when the situation is an unusual variation of a known situation.
NARROW. As a result of efficiently seeking patterns we have already seen, the domain-specificity of expertise tends to make us think in narrow ways when we think we have grasped the situation rather than to broaden our thinking.
FUZZY. Part of the power of pattern-matching is that we can so often generalize the lessons from one situation to another similar one. In situations where the appearance is very similar but the underlying principles are different, again our pattern matching effectiveness leads to mistakes: we overgenerallize from our experience.
SPRAWLING. When a pattern-seeking process does not have a single clear path to follow, as often happens in very complex situations, it will tend to follow one path after another and keep switching back and forth rather than working toward an overall goal.
Experiential intelligence, Perkins concludes, is an elegant system for long-term moderate success. When situations are new to us or complex, we get help from our neural intelligence and we have also learned various tricks for getting around our weaknesses, and these are largely accounted for in reflective intelligence. Reflective intelligence represents realms where we think about our own thinking in order to avoid settling on hasty conclusions, to broaden our thinking beyond the initial scope we assumed, to use precision to distinguish similar looking but different things, and to stay on track when notice we are sprawling.
This explains why reflective intelligence is so important to us in tricky situations where we have inadequate experience and where experience misleads us. But it also helps explain, in Perkins’ view, why reflective intelligence is so important for us to learn to be better thinkers in general. Neural intelligence does not replace experiential intelligence, it tends to reinforce it.
When we don’t have experience, neural intelligence helps us grasp the situation, but when we do have experience, we tend to use our neural intelligence to reinforce what our experience already tells us. That’s one big reason why genius is not simply high IQ. That’s why reflective intelligence is so important, it is the tool we use to remind us of the weak points in our own thinking and help us compensate for them regardless of our experience and general intelligence. The abilities and traits we need in order to overcome our blind spots are learnable. A large and crucial aspect of intelligence is learnable.
Existing Approaches: How they Compare
There are various approaches to teaching reflective intelligence, and Perkins reviews the best known and the best studied among them such as Project Intelligence, Reuven Feuerstein’s Instrumental Enrichment, Edward de Bono’s CORT, and Matthew Lipman’s Philosophy for Children, and others, reviewing their approaches and their results and comparing and contrasting them in order to get a sense of what it takes to enhance reflective intelligence.
One of the things that distinguishes Perkins as a deep reflective thinker himself is that he anticipates, researches, and deals fairly with opposition to his arguments. The very idea of learnable intelligence has in the past come under attack from several angles such as past failures of various programs which tried to teach improved thinking, the implications of expertise and psychometric research data, the apparent weakness of general methods for problem solving, and the challenge of transfer between learning domains. Perkins addresses each of these concerns in turn, resulting in a very persuasive case for the very real improvability of intelligence through changes in education.
The Future of Learnable Intelligence
Toward the end of the book, Perkins reveals the ingenuity of his vision through his discussion of several areas for the future evolution of reflective intelligence: areas which ended up being (remarkable for a book written in 1995) accurate predictions of areas that have since become central areas of interest for science and human improvement in general:
1. Intelligence can become distributed -- good thinking depends upon artifacts to offload the limitations of our attention and memory, and we can use our symbol systems and tools to help us keep track of things we could not track individually. This is a wonderful general description of how we are attempting to use computer networks to help us manage complexity (as opposed to some of the more superficial books in recent years which imply that networks somehow replace rather than enhance individual thinking).
2. Intelligence can embrace complexity -- through information visualization tools, effective use of classification, tagging, and finding things by meaning, consolidation, filtering, the mathematical tools for finding large scale patterns in complex phenomena, and by eliminating narrow information silos, we can use our intelligence to solve increasingly complex problems.
3. Intelligence can be dialectical -- this means raising the level of thinking from lower level more concrete concerns to higher order patterns by recognizing the properties specific to complex systems. Perkins offers Peter Senge’s “The Fifth Discipline” and Murray Gell-Mann’s “The Quark and the Jaguar” as exemplifying ways of understanding dialectical intelligence.
Perkins covers a massive amount of data about intelligence and problem solving, summarizes it effectively, and applies it to a practical, powerfully supported, and exceptionally understandable approach to improving human life by teaching ourselves to be more intelligent. Thinking well in general is an unnatural act but we can learn to do it. All that is left is for us to overcome the ideological and political barriers. This book would make a wonderful, gentle manifesto for that grand effort.