Surpassing Ourselves, by Carl Bereiter and Marlene Scardamalia
My detailed notes on a fascinating study of the nature of expertise, and how to avoid becoming an experienced non-expert.
Surpassing Ourselves is a fascinating examination of expertise in comparison with those who have experience but do not become experts. The authors provide a straightforward explanation of the process of expertise, the nature of creativity, and how it is related to wisdom. There are a lot of great examples from education, which is a primary focus of the authors’ research.
If you enjoy my summary below, I encourage you to buy the book either from Bookshop or Amazon.
Chapter 1: The Need to Understand Expertise
The uniquely human capacity for expertise lies in effortfully - and intentionally - acquired abilities that carry us beyond what nature has specifically prepared us to do. Fish are not “expert” swimmers, and while rats can learn to navigate a maze, they are not developing expertise. Understanding the method of expertise will transform 21st century society just as much as understanding the method of invention transformed 19th century society.
We carry many stereotypes about expertise, especially in assuming it is connected with talent, skill, professionalism, or experience. Importantly, experts do not need to be specialists (even if that makes them easier to identify).
At the high-level, what we do know about experts is that they:
Progressively advance on the problems constituting their field of work throughout their career.
Develop a much deeper understanding of the systems they’re working with. For example, expert surgeons understand the organs and tissues in an intimate, resourceful way much more than someone who is following prescribed procedures.
Are better suited to handle non-routine situations. For example, most people are experienced non-experts when it comes to driving: we can effortlessly do the right thing under normal conditions, most of the time, but may fail catastrophically when faced with high-stress conditions or adverse weather conditions.
Avoid complacency by choosing “to address the problems of their field at the upper limit of the complexity they can handle” and avoid our tendency to simplify (see Herbert Simon’s work on “bounded rationality”).
That said, it is important to observe expertise in a career, rather than a person. Even experts may temporarily lapse into complacency or develop new skills that don't lead anywhere. And not all jobs provide opportunities for an expert career, but many jobs do introduce problems that have the potential for complexity beyond what any person can accommodate.
In order for society to take fuller advantage of the human capacity for expertise, we must a) find ways to make the process of expertise teachable; b) build the process of expertise into complex team dynamics; and c) seek ways to give expertise away to foster broader innovation.
Chapter 2: Experts are Different From Us: They Have More Knowledge
Most earlier research on expertise compared experts with novices. De Groot’s famous research on chess masters (1965) showed that chess masters considered fewer, yet broadly better, chess moves. They also had the ability to remember (sensible) chessboard configurations, an observation about highly significant pattern recognition (related to playing the game) that was later repeated with masters of other domains (e.g. electronics diagrams and sports team configurations). Other studies showed that novices tend to do "means-end analysis” (or reasoning backward, which is incredibly inefficient) whereas experts can focus on the salient elements and reason forward (much easier and straightforward).
The problem with these earlier studies, however, is that they compare experts to novices based on problems that are hard for novices yet trivial for experts. Subsequent studies showed that experts also reason backward when considering challenging problems or problems outside of their own expertise. And experts use 5-6x more words when thinking through complex tasks like writing an essay than non-experts, which indicates they actually think quite a bit more. Non-experts, in comparison, tend to go with whatever first idea that came to mind.
Another problem with these studies is that they compare fluid intelligence in the novices (their ability to think through tasks) against crystallized intelligence in experts (retrieval from previously constructed understanding). This is unfair, and thus the authors identified the importance of comparing experts with experienced non-experts.
Chapter 3: Expert Knowledge and How It Comes About
Deep knowledge is common across all kinds of experts. However, we must consider knowledge in all of its unique forms, beyond the metaphorical “contents of a mental filing cabinet.” There are, for example, two main forms of overt, visible knowledge:
Formal Knowledge (declarative knowledge, or Gilbert Ryle’s “knowing-about”): manifests in explanations. The main benefit here is that it is negotiable and can be written down and shared. Formal knowledge is essential for communication, negotiating truth, teaching and learning, and provides a starting point for the construction of informal knowledge and skills. However, when it fails to transfer into informal knowledge or skills, it remains inert.
Skill (procedural knowledge, or “knowing-how”): manifests in performance. We translate formal knowledge into informal knowledge or skill by interpreting it as a series of actionable subgoals (John Anderson). For example, (pre GPS) we may translate directions into actually driving to your friend’s house for the first time. "Formal knowledge is converted into skill by being used to solve problems of procedure.”
We also have different forms of hidden (or tacit) knowledge:
Informal Knowledge (e.g. guessing which cups, when dropped, will break): This often takes the form of educated or informed common sense. Informal knowledge is often “situated” in that the range of application is limited to certain real-world situations. "Formal knowledge is converted into informal knowledge by being used to solve problems of understanding.”
Impressionistic Knowledge (things that register impressions of being good or bad, strong or weak, active or inactive, etc.). This a) tends to persist longer than declarative knowledge (our long-term memory of books tends to be captured as impressionistic knowledge); b) can be developed (e.g. expert wine tasters have a consistent set of interpretations for their impressions); c) informs complex decision making; and d) contributes to the enrichment of life. Creative activities often depend on an impression of promisingness.
Self-regulatory Knowledge (knowledge of how to manage yourself to attain your goals). This is highly personal (e.g. a writer’s daily habits), and much more than metacognition. Important self-regulatory knowledge “has to do with rhythms of work and relaxation, production and reflection, concentration and incubation, and so on.”. Other forms include: habits of planning and checking, drawing back to take stock of a difficult situation, etc. Importantly, this is generalizable to different fields.
The classic non-expert response to a puzzling question (e.g. of how is it possible for a satellite to remain fixed above one spot on earth) is to acknowledge it as puzzling and then to not think about it any further. This approach minimizes growth and revision of knowledge. We must learn to work at the edge of our own competence and understanding, and cultivate an awareness of the complexities that we’re not yet dealing with.
Informal knowledge also develops through observation, working under the direction of others, and by trying to solve practical problems. People who attend more closely to what they observe, participate more directly with others, or try hardest to solve those practical problems develop the greatest amount of informal knowledge. And we must not forget that our knowledge may need to be translated into the relevant real-life situation before our performance can be judged. For example, a doctor in a remote village may struggle to adjust their formal knowledge to the fact that a newspaper may be more sterile than an available towel.
Chapter 4: Expertise as Process
We all engage in learning and problem solving. The important difference is that "experts tackle problems that increase their expertise, whereas non-experts tend to tackle problems for which they do not have to extend themselves.” The authors explore this through the example of two elementary school teachers: the expert teacher (Margot), after mastering basic classroom management skills, moved on to progressively identify and solve problems of understanding; the experienced non-expert (Cynthia) stopped learning after mastering the basic classroom management skills, despite the fact her placid students were struggling to learn.
Problem solving consists of searching or deliberation toward a goal. The complexity of the problem we’re attacking often depends on the number of constraints that must be honored. Given we can generally consider four things actively in our mind at a time (see Herbert Simon’s work on “bounded rationality”), we simplify in order to not be overwhelmed. We can circumvent our mental resource limitations through a) pattern learning (particularly patterns that cover the majority of realistic cases); and b) procedural learning (shortcut habits we create, though with the cost of losing conscious access to the action).
As we develop automaticity (through pattern learning and procedural learning), whether we develop expertise depends on whether we a) choose to fall into complacency; or b) lean into our new mental bandwidth to engage in reinvestment (motivational aspect) and progressive problem solving (cognitive aspect).
Reinvestment: is where we go beyond normal learning (which is rife with satisficing, per Herbert Simon). We reinvest in learning (through lessons, expert forums, etc.), seek out more difficult problems (tackling more complicated projects, learning harder choreography), or tackle more complex representations of current problems. This requires paying attention to the nuance of the problem and not getting stuck in the rut of your original simplification.
Progressive Problem Solving: is where we continue to work on problems just beyond our reach, layering in complexity from the endless complexity available. This is particularly true when we consider the constitutive problems of a domain, for example, the elimination of ignorance (within the domain of education). Or the achievement of health for everyone (in the domain of medicine, though the field is still rooted in the framing of the “elimination of disease”). This is the opposite of problem reduction, where we instead reduce problems to tasks that can be handled with routine procedures. Problem reduction is important in many (peripheral) aspects of life, but will not lead to true expertise.
This process can be seen in many domains. For example, non-expert writers tend to mimic learned patterns (i.e. the writing of others), whereas expert writers use learned patterns as a vocabulary to produce original constructions. Also important is how this process means that expert performance is no longer limited to an individual, yet can be reflected in expert teams that similarly follow the rules of reinvestment and progressive problem solving. For example, expert subcultures of jazz musicians are a form of second-order environment in which everyone grows as a result of each other’s successes. Similarly, Csikszentmihalyi’s concept of flow came from observing experts pursuing activities at optimal levels of difficulty (as opposed to ways that led to boredom or anxiety).
Chapter 5: Creative Expertise
Popular notions associate creativity with many things, and sometimes expertise is interpreted as unrelated, or even a threat, to creativity. Often when we actively try to be creative, we apply creativity as a constraint to problem solving and end up with a mishmash of incoherent ideas. If we instead focus on solving significant problems in a progressive way, creativity will emerge from our work. This is particularly true of design problems, in which the goal itself isn’t totally clear until the task is finished (it emerges from the work).
What truly matters for creativity is knowledge of promisingness. This knowledge can follow a variety of different paths. It can involve recognizing that an idea has a direct match to your overall goal, or that the idea matches your own capabilities. An idea could also be promising because it points to further possibilities and is highly generative of new ideas. Katchadourian’s sense of a “starting point” could simply be where you most easily find promising ideas. Dostoyevsky famously started with a promising character, Agatha Christie with a promising setting, and creative scientists with a promising question. Promisingness is an informal prediction based on variable knowledge. It can be applied in a sustained, accretive way over the duration of a long-term design project or reflected within the split-second decisions of a jazz pianist playing his next note.
According to this perspective, "creative experts are simply people who take greater risks than other experts — and succeed.” Success is important to develop an accurate sense of what is promising or not (and not just in hindsight, but when you’re mired in the uncertainty of everything). When you tackle problems just beyond your level of competence, you are forced to make serious judgments of promisingness.
“If your process of expertise consists of many small ventures beyond existing competence, you should steadily increase your competence — this may, in fact, be the best way to build solid expertise in most fields — but you will not have had to worry very much about promisingness. It is as if you were camped in a desolate region and were venturing forth without a compass to explore the terrain. If each trip extends but a short distance beyond where you have gone before, you can rely on accumulating knowledge of landmarks to find your way. Only if you make larger ventures, beyond the range of already known landmarks, will you need to learn how to read the stars, the moss on tree trunks, and the like, in order to find your directions. Otherwise, you may master the vicinity very well, qualifying as an expert in its geography, but you will not have acquired the abilities of an explorer. Only by venturing more, by employing a larger step size, would you acquire those skills.”
How do we foster creativity and our ability to recognize promisingness?
Pursue creative goals and experience occasional success. If you only experience success, you’re not tackling big enough challenges.
Seek mentors who have developed a sense of promisingness through their own successes
Study promisingness explicitly, especially in moments when you find yourself most critical (i.e. while considering a new development in your field).
Work on projects where solutions emerge through a series of drafts or versions, where feedback is focused on issues of promisingness.
Remember: “Bad teaching [...] focuses on the merits and shortcomings of the present version, whereas good teaching focuses on [promisingness and] what the present version might eventually become."
Chapter 6: Expert-like Learners
Expert-like learners engage in “learning as progressive problem solving” and are more likely to go into “theory building” mode where they take all of the facts and try to construct an explanation that accounts for all aspects.
Tackling a learning project can involve three types of goals:
Task accomplishment goals. Seeking to finish the task. Any learning is a by-product, and you may not even be aware of the underlying educational goals. You’re done when you’ve finished the task. These goals are often quite explicit. This is “learning as a routine.”
Instructional goals. Seeking to achieve the knowledge and skills the project was intended to cultivate. You’re done when you’ve mastered the concept. These goals can be explicit, but are not often made so.
Knowledge-building goals. Working toward goals held by the student, which may include, but not be limited to, the instructional goals. You’re never quite done, since these goals lead to progressive problem solving. These goals are rarely explicit, and are often emergent.
Interestingly this mindset of "learning as problem solving" (versus as a routine) applied across different contexts, when experts from one domain were presented with challenges well outside of their field of expertise. Depth of planning was similarly correlated to performance across all contexts.
In contrast, non-expert learners use the “best-fit” approach, where they directly assimilate new information into existing schemas regardless of how faulty or inaccurate those may be. Non-expert learners tend to make subjective judgments of importance, jump to conclusions based on scant evidence, quickly construct simplistic interpretations (and retain them in the face of contraindications), and dismiss whole topics as boring. Thus, they “suffer from the consequences of inadequate prior knowledge — faulty concepts, oversimplification, disconnectedness, dysfunctionality.” For example, children who interpret a newly introduced color “olive” as the equivalent of green.
Schools tend to be challenging places to nurture knowledge-building activity. Schemas “specialized for rapid pick-up and retrieval of knowledge” are typically favored. Good teachers can be models for learning the process of expertise, and good classes can save children from “a great deal of floundering about in ignorance.” Rich intellectual experiences outside of school are more common, even in examples where the overall structure of progressive problem solving is likely being driven by a parent in a kind of joint production (with the young child supplying the talent).
Chapter 7: Schools as Non-Expert Societies
Most schools seem to be designed to produce non-experts, no matter how much we value expertise in the rest of the world. The three serious defects in schooling tend to be: 1) only dealing with the visible parts of knowledge (formal knowledge and demonstrable skills); 2) leaving invisible the knowledge objectives that are pursued, which reduces activities to task accomplishment; 3) only giving teachers access to the process of expertise and progressive problem solving.
Curriculum guides tend to consist of a) a list of facts/etc. to be covered; b) a list of skills (a.k.a. processes) to be developed; c) a catalogue of suggested activities to go along with them. Curriculum for the lower grades often only consists of (c) with labels associated with each activity that purport to cover (a) and (b). Other goals, - like to instill a love of reading - are regarded as aspirations. “In every area of the school curriculum, the activity replaces the objective and eventually obliterates it.” The common operative metaphor is that schools are factories where students are ‘semi-skilled’ workers who can master certain procedures yet never come close to becoming managers or engineers.
The two typical (yet still positive) calls for school reform are insufficient. The typical didactic response - “We need to update (or extend) the conventional curriculum” - still leaves the setting of goals, monitoring of progress, and solving of problems blocking progress out of student hands. The typical child-centered response - “We need to focus on creating a nurturing environment to enable children to develop naturally” - relies too heavily on a child’s curiosity and desire to grow. Most successful child-centered environments, when you look closely (and ignore their high entrance requirements), are actually successful second-order expert communities.
Classrooms can function as legitimate knowledge-building communities. There is no need to “play-act” in roles of meanintful knowledge building (the “cognitive apprenticeship” movement), or pretend to discover information that is easier to find in books (as with the “learning by discovery” movement). However, it is challenging to make these communities both satisfying (i.e.somewhere we want to belong) and dynamic (where we continually reinvest resources). This can be achieved by leaning into our common desires for: a) recognition and respect from the people we regard as peers; b) having impact; and c) participating in significant discourse (far beyond the typical “ teacher initiates, student responds, teacher verifies” discourse of schools).
Schooling according to the knowledge-building community model would have the following characteristics:
Sustained study of topics in depth, sometimes over a period of months, rather than superficial coverage.
Focus on problems rather than on categories of knowledge… i.e. “how does the heart work?”
Inquiry is driven by student questions, and the teacher helps them formulate better questions and encourages them to reformulate questions at higher levels as inquiry proceeds.
Explaining is the major challenge, and students are encouraged to produce their own theories to account for facts and criticize one another’s theories by confronting them with facts and evidence.
Day-to-day focus is progress toward collective goals of understanding and judgment rather than on individual learning and performance.
Little schoolwork of the traditional kind (individually working on the same things), and more work in small groups on distributed tasks related to the common goals.
Discourse is taken seriously, and students are expected to respond to one another’s work and are taught how to do so in helpful, supportive ways.
Teacher knowledge does not curtail what is to be learned and investigated. Teachers are one (of many) sources of information.
Teacher remains the leader, but their role shifts toward modeling the expert learning process (rather than being outside of the learning process and guiding it).
Chapter 8: Toward an Expert Society
So how do we encourage broader development of human expertise, either within a workplace or in society at large?
Formulate shared ideal goals that are simultaneously not attainable yet in which progress can be demonstrated.
Frame individual work as contribution to collective progress on these higher-level goals (e.g. individuals get to participate in the higher goal of building a cathedral rather than laying stones).
Ensure rewards (recognition, compensation) are aligned with contribution to collective progress (to reinforce #2).
Allow emergent goals to be discovered along the way as a result of progress. These may be particular interpretations or elaborations on the shared ideal goals.
Share expertise broadly. Avoid locking it up in a closed subculture.
Doing this allows us to improve quality of life and bring wisdom back into the notion of progress (rather than blindly believing that progress will sort out our ills). True expertise is a form of active wisdom, with wisdom - particularly when choices are shown in retrospect to have been wise with respect to social outcomes that could not have been foreseen - relying heavily on an intuition for promisingness. Whether that comes in collectively seeking excellence or quality, the following are the six essential ideas to lend coherence to our efforts to make better use of the human capacity for expertise:
Nurture Second Order Environments: social environments in which progress or growth is a continuing requirement of adaptation to the environment.
Reinvest Mental Resources: as we practice and things become easier, reinvest mental resources or energy into new challenges. This is the psychological basis for progress.
Engage in Progressive Problem Solving: tackle new challenges at higher and higher levels. This is the dynamic element in the process of expertise, and keeps experts from becoming habit-bound and makes genuine progress possible.
Work at the Edge of Competence: this is the natural consequence of progressive problem solving, and allows growth of expertise.
Develop Creative Expertise: this happens when we engage in adventurous problem solving and are required to make decisions on the basis of judgments of promisingness.
Cultivate Active Wisdom: this is the special kind of expertise marked by the ability to make decisions that turn out to have been wise ones. Based on judgments of promisingness with respect to human values and far-reaching consequences.