2007: At the height the property boom, a Dutch online real estate company called Funda launched in Ireland. We produced a spoof TV ad promoting a new,

The Expert CurveChapter One: What is an Expert?

Chapter One

What Is An Expert?

If you decide (or it is decided for you) to become an expert, you need to know what is in store. I’m hoping that this chapter make it a little clearer to you.

How We Define Expertise

First a warning, this stuff can be dull. So dull that a lot of you might want to skip it, but the information is important as it helps explain my central theme that expertise is a state, the result of environment, a lot of hard focused work and talent. I don’t want this book to read like a user’s manual for a new camera, but becoming an expert is a complicated and very technical process; the science is crucial to knowing what is going on. I hope this information helps you understand it more.

We’ve known for millennia how experts are made and we have known that teaching to expertise is a moving target no matter what the domain. More recently we are finding out why this is so through genome assessments, brain imaging, sophisticated statistical studies by psychologists and various bits and pieces published by other interested parties. This new information, mostly from the past three decades, has changed the way we see the process and engendered more scientific inquiries into how expertise occurs. While it is still up to the student and mentors to make the process work, defining it is the first step.

Experts, whether they are family physicians or plumbers, are characterized by the way they change the environment. Expertise is a social construct that is dependent on the particulars of the situation or expectations of others. You have the knowledge and skills, you do the related work and you interact with others. Something happens and a change takes place. We hope that change is for the good.

The term “expert” is defined differently in various academic and practical fields. A psychologist may delineate an expert in terms of performance and superior ability while a sociologist may look at influence and power without judging competence.

It has long been recognized that such definitions are limited and reflect the needs of those on the receiving end. Dr. Marie-Line Germain (Germain 2006 http://psychology.wikia.com/wiki… ) developed a sixteen point measure of employee expertise that contained five objective items including education and qualifications and eleven more subjective measures that are behaviorally related observations on the part of the employers. These criteria are useful in evaluating candidates for an expert’s job (and they reflect the needs generated by social situation) but are not a specific definition of expertise. They describe how we choose or use experts.

Other definers can include qualifying exams, contests, skills tests and self proclamations.

The way we use experts continue to change and that alters the subjective way we define them. Since the internet has made the body of knowledge more available, we think that we don’t seem to need an expert’s knowledge as much. But at the same time we think that expert knowledge is not needed due to the internet, the opportunities for experts to collaborate and extend the interactions of both knowledge and skills turn the internet into a crowd sourced idea machine. Much of the research on the Higgs boson at CERN was done in this fashion and it is unlikely that it could have been done any other way due to the massive amounts of skill and knowledge needed. The idea that expertise is just specialized knowledge is eroding and is now seen more as communal interactions due to better access to knowledge and input from others. The expert has a special and specific input that is more efficient and more precise, so the interaction is richer for it.

Expertise has always been more than just specialized knowledge. Those who separate experts from specialists, the latter considered a problem solver (such as a physician or plumber) and the former a repository of knowledge, perpetuate an artificial split. No expert has knowledge and does nothing with it. Expertise requires skills, not just knowledge. One of the expert skills is how to use domain knowledge.

A good example is the body of knowledge that medical scientists and physicians share. Because physicians enter into a social contract with patients, they require a different set of skills and a different view of the information than scientists even though both study and know the science behind medicine. Scientists don’t have the need for physician skills, they need tools that are precise and predictable. Physicians have to work with incomplete knowledge regarding their patients and their goal is to help them. Both work with the same knowledge base but for physicians that data is incomplete because of the complex presentations of patient’s illnesses. Medical scientists search or collate knowledge for a better understanding of the why of medicine and look at each disease as a single problem.

Casual inquiries into knowledge and skill are often not that informative. I’ve had discussions with very bright people who use knowledge but not as experts. Commonly they will focus on one small aspect and ignore other important elements. This leads to contradictory ideas about a domain when discussing it with an expert. As we will see, this is due to incomplete knowledge and a novice’s view of the subject.

Those who study experts are clear about one thing. Experts think differently about domains, the acquired field of expertise . Outside of their domains, and it doesn’t have to be too far outside, experts are no better than novices when it comes to making decisions. Expertise is a discreet human state that is different than “normal.”

I used to think that in order to be an expert you had to master the field. And by “master”, I meant that you were at the top of the field. That all changed during my five year experience as the team physician for the United States Shooting Team ending with the 1984 Olympics. Due to my affiliation I met world champions, Olympic gold medal winners and numerous officials and VIPs in the Olympic movement. More importantly, I met a lot of shooters who wanted to be on the Olympic team but who were not at the level of the masters of the game and were struggling to meet those standards. Most never made it but I realized that they were experts nonetheless.

Shooters who are invited to Olympic trials are very good marksmen. They are state champions and are well regarded but they and everyone else knows that few of them would be on the team since there are only a few slots for each discipline. In those days the Olympic movement relied on the military teams to produce the best shooters – they had the money and motivation plus they recruited – and the national governing body had just started raising money for development teams, so there was no general effort to produce master level shooters for the Olympics.

Most of those selected for the Olympic Trials were experts in their chosen sport but nothing more than that. They were not at the level necessary to be an Olympian which required much higher standards. It took me years to understand that being an expert is a state, not a guarantee of success or even of total competence.

This is my definition: Experts are persons who have specialized knowledge, skills and specific motivation to learn more and continue in the performance of whatever constitutes the domain field. The way an expert addresses a problem differs from a layperson because experts use heuristic thinking and have a personal, more efficient, approach called style. In the process of developing a style, skills and knowledge become automated to the point that little internal dialogue regarding method of performance is needed.

While being an expert has been subject to a lot of research, speculation and theorizing for centuries, no one has found a universally acceptable answer to the question of how to define expertise. Mine is a narrow, easy to explain, one that emphasizes functional aspects of style and thinking. Its virtue is that it has a practical value when you explore the hows and whys of expertise.

Traits of Experts

Many things have to happen if you want to be an expert. Experts are competent, well schooled, have good performance skills and are knowledgeable. They put in the time, have the required credentials and fulfill the standards needed to qualify. Since being an expert occurs in a social context, performance is part of the process.

Knowledge itself does not allow a person to be an expert if it is not used in some sort of interactive way. This interaction has to be transformative (a change has to occur) and it affects both the audience and the performer. An expert has to have knowledge and skills that become automated in the learning curve. Without automation, performance suffers or doesn’t happen.

In my introduction, I mentioned that experts are selected, have mentors, develop technical and physical skills, have good mental tools including analytic abilities and have put in the time by practicing. They also have to be qualified by meeting the standards set up by consensus in the field and have the drive to follow through with the work involved in attaining and growing that level of expertise. It helps to start out with talent, intelligence and a strong working memory because these traits make it a lot easier to succeed and are very helpful when that goal is reached. But hard focused work is the key. Environment, with both positive and negative features, is also very important because it includes all of those things we take for granted: teachers, schools, health, finances, policies, opportunities, even weather. It affects talent and the ability to use deliberate practice.

In the end, the most important practical trait needed to become an expert is perseverance. This process takes a lot of work. Take the example of “The Knowledge”, the skills needed to become an All London cab driver.

All London Cab Drivers

“All London” cab drivers are among the most heavily regulated in the world. To become a cabbie, a student has to pass twelve oral examinations proving that they can take one of over three hundred approved routes from two random points in the city while reciting every street, every crossing, and every “point of interest” (which includes virtually every building, statue, restaurant, landmark or shop of note) along the way. To put it in context, there are over twenty-five thousand streets in a six mile radius from Charing Cross to be memorized. It takes three years or more to pass the tests and that’s if you attend one of the Knowledge schools and travel the routes endlessly. It’s a full time job with no pay to learn The Knowledge.

From the guidebook issued by the London Taxi and Private Hire (LTPH):

“To achieve the required standard to be licensed as an “All London” taxi driver you will need a thorough knowledge, primarily, of the area within a six-mile radius of Charing Cross. You will need to know: all the streets; housing estates; parks and open spaces; government offices and departments; financial and commercial centres; diplomatic premises; town halls; registry offices; hospitals; places of worship; sports stadiums and leisure centres; airline offices; stations; hotels; clubs; theatres; cinemas; museums; art galleries; schools; colleges and universities; police stations and headquarters buildings; civil, criminal and coroner’s courts; prisons; and places of interest to tourists. In fact, anywhere a taxi passenger might ask to be taken.”

Brains Change

Not everyone of the thousands who try to achieve this standard passes but of those that do, an interesting thing happens, their brains change.

Eleanor Maguire PhD of the University College London has studied the brains of London cab drivers since 2000 when she wrote a paper demonstrating a volume increase in the right posterior hippocampal region of those with the Knowledge. This part of the brain is responsible for spatial representation of the environment and her findings were important because they showed that adult brains could change and add neurons if an intense stimulation -in this case learning the streets of London. – was applied. (PNAS vol. 97 no. 8 > Eleanor A. Maguire, 4398–4403, doi: 10.1073/pnas.070039597 Navigation-related structural change in the hippocampi of taxi drivers ) The increase in hippocampal volume was directly correlated with the amount of time spent as a cab driver. https://www.ucl.ac.uk/spierslab/…

Becoming an All London cab driver does not just entail knowing all the streets and finding the best routes. It is expected that you will automatically know the shortest route and not have to consult a map or GPS. In addition you have to be able to answer any questions about London from your fare and be able to discuss the various points of interest that you pass. The reward is a good paying job and some exclusivity in the taxi business. You have both a social component and significant reward for meeting the LTPH demands.

All London cabbies demonstrate the criteria for being an expert. There is a self selection process requiring a commitment of time and money, a long educational period with clear goals that include both pedagogical and practical skills, severe testing and standards set by an outside agency. These conditions have been met by taxi drivers in London since the middle of the nineteenth century, what is new is the information about brain plasticity and hippocampus growth.

Role of the Hippocampus

Dr. Maguire has continued to study the role of the hippocampus as it pertains to memory but her research has touched on a number of aspects that relate to expertise. This part of the brain has a unique role in memory. Without both hippocampuses a person would have no short term memory and no working memory. https://www.researchgate.net/pro… Development of the brain in London cab drivers is one of the type of changes that occur when people become experts. http://www.pnas.org/content/pnas…

The hippocampus does not store memories but it appears to organize and manage them. Scientists have found “Place cells” in the hippocampus which have a specific role in spatial adaptation and awareness. John O’Keefe (neuroscientist) Ph.D, of University College London, won the 2014 Nobel prize in Medicine for the discovery of these cells in 1971 and has been researching them ever since. O’Keefe demonstrated that individual place cells will fire whenever a rat reaches a specific part of a maze and that they do so every time the rat comes back to that spot. In addition these cells are responsible for forming a “cognitive map” that helps people with spatial orientation as they move around by temporarily storing information about the immediate area. The cells are adaptable when the environment changes. https://pdfs.semanticscholar.org…

It also appears that place cells are involved in non-spatial functions and some researchers think that place cells are flexible enough to apply non-spatial memories to similar situations in order to help solve problems and are the basis of the working memory. Studies in rats show that place cells fire in different parts of the hippocampus in discernible patterns when a rat is traveling. The cells fire briefly and seem to be related to both goal-oriented behavior and direction. When Dr. McGuire’s research subjects perform virtual driving under functional Magnetic Resonance (fMR) the hippocampus lights up for a second before the driving begins indicating activity related to planning and organizing. Recalling Routes around London: Activation of the Right Hippocampus in Taxi Drivers

Place cells are involved in pattern completion, the ability to recall an entire memory from a partial cue. They show specific patterns that recur even when only partial signals related to that memory are present and subjects are able to recall the entire memory as a result. These patterns occur without the subject being aware of the process.

Place cells are also involved in pattern separation, which is the ability to discern between competing inputs, and they form specific patterns related to each of these helping to distinguish one memory from another.

Place cells have a role in replaying memories, a process called reactivation. When place cells reactivate, they replay the memories at a rate much faster than the memory was experienced and usually in the same sequence. Functional MR tests of London cab drivers taking virtual reality routes of the city demonstrate activity in their place cells as they rapidly go through the street sequences naming all the points of interest. One cabbie reported that this process was “unconscious” as it just occurred without his having to think about it. This is a typical story that any expert will tell you when they are working within their domain.

Another phenomenon, hippocampal replay, occurs during sleep or while at rest/relaxed but awake. It often happens right before an activity and it occurs in sequence with the activity. It consists of the place cells firing in a known sequence to one that has occurred in the past but adapted to the present.

When hippocampal replay occurs in sleep (in both slow wave sleep and REM sleep), it helps with memory consolidation and when awake, hippocampal replay is not only involved in memory consolidation but in planning and setting up a brain network for future learning on the same subjects. Hippocampal place fields fire when a specific stimulus occurs that can trigger an altered replay separate from the original one if it occurs in a different context than the original stimulus. Other parts of the brain including the visual and motor cortex show similar and simultaneous replay patterns when the hippocampus place cells fire in specific ways.

All of this information shows a vital role of the hippocampus in mapping and processing of memory. It also implies that the patterns place cells generate have important effects on the rest of the brain and that those changes appear to be significant in learning and functioning. This is an important aspect of learning and retaining domain information.

Brains Transform

All London cab drivers are experts in the driving geography of one of the most complex cities in the world. While it is a leap of faith to say that they are this way because a small part of their brains is enlarged by twenty percent over non-cabbies, it can be held that there is a causal relationship between those with the Knowledge and increased posterior hippocampal size. Either way, they are experts by every criterium set by the LTPH, by every passenger and every article done on them since 1865. Something is happening and Dr. McGuire’s studies have given us a monster clue: brains change and grow when you become an expert.

Characteristics of Experts

One of my interests is target shooting, an activity that has a strict measuring system reflecting low variance in the field: you shoot and you score, that’s about it. But while that’s the test for competency in the field, it does not reflect performance one to one because score is not always representative of where a shooter is on the learning curve or whether or not they have the full knowledge and skills that the sport requires. Most American shooting sports have classifications which are more or less arbitrary groupings of similar level shooters who compete against each other in addition to trying to win the whole match. These classifications are there to spread the wealth of prizes and to encourage shooters to improve and keep interest up in the vast majority who never win a match. A typical classification scheme is Marksman (novice), Sharpshooter, Expert and Master, reflecting the average scores over certain periods for each shooter. This particular schema is a reflection of the military awards system for weapons qualification (although they stop at Expert.) The assumption is that at some point a shooter has assimilated enough knowledge, practice and competition that they are truly experts. It also acknowledges that they have not mastered the field.

As you can imagine, the beginners don’t always reach the upper classifications and many either drop out or find themselves content with their place and focus on winning their class. Others try and succeed in reaching the highest classes.

Those who end up in the Master class rarely stay in the Expert class for long. On the other hand those who stay in the Expert class for any length of time rarely get better, even though they tend to have the same equipment, the same knowledge levels and the same desire (if not motivation) as those in the higher classification. They are “experts” by the rules of the governing body but not good enough to win the match. Some people are simply better than others and some are extraordinary.

Being an expert is about making decisions concerning complex problems, usually under some pressure and normally using standard methods that occur without a lot of conscious thinking. Not that experts don’t think, they have to in order to solve the problem before them, but the mechanisms used by them are automated to the extent that they are much more efficient and much more precise than a layman faced with the same problem. This automation allows the experts to think while acting because they don’t have to deal with the conscious thought patterns seen in non-experts Automation and consistency are qualities of experts and while talented novices can have similar results, they tend to be inconsistent and quirky in methods and outcomes. Experts are capable of dealing with a wide range of problems within their field but they may not be universally competent in other domains. That last point is very important because one of the social norms for defining expertise relies on the comparison between laypersons and experts.

Another hallmark of experts is that when they are in the expert mode, they are in an altered state. This allows them to focus on the problems before them and to call up domain knowledge and an enhanced working memory. This process is automatic in experts and it differs from a “normal” state in that focus is more narrow, they literally put on a “game face” (i.e. their posture and facial expressions change), and they are more perceptive and able to make better decisions about the domain.

Laymen vs. Experts

I have a friend, Andrew Madsen (Custom Guitar Luthier Neenah, WI) who is an outstanding luthier. He makes gorgeous guitars that sound terrific and are easy to play. His instruments are superior to most others, so good that I thought he must have some secret techniques that make them so. “I’ll be glad to tell you all my secrets”, he said, “you just have to be able to understand and carry them out.” His point was that there are no secrets, he was just very good at what he did and he attributes it to good mentoring and a lot of practice building instruments.

Building guitars is like a lot of expert fields. It requires the integration of a number of skills including woodworking, metal work, sculpture, a sense of proportion, and so forth. More important is the ability to use these skills to produce an instrument that will please the player and the listener. Most luthiers use manufactured products such as tuners and strings and are able to make them work with their creations. To do this requires thought, precision, a good model and lots of work that has the goal of the final product in mind. Because wood is the most common medium used in lutherie and wood is unpredictable at times, a machine can only approximate what a skilled luthier can do.

The first time a novice attempts to make a guitar will be a disaster. That’s a given in the luthier world, even if a kit is used, because of factors that are not well defined and thought to be “secrets” by the uninitiated. There is something about expertise that defies objective measures and that seems to be a problem with defining expertise using only objective methods.

I write a monthly article on mental training for Shotgun Sports Magazine (Mental Training | Shotgun Sports Magazine) and in the process of finding ideas for my column I have talked to a lot of expert level shooters. One of the interesting things I learned was that many had no idea about what they were doing. They would teach the standard techniques but when I videoed them and compared what they did to the standard, there were significant differences. Most of the time they were unaware of these deviations and some denied them even in the face of my video evidence. The same has been true of musicians I know.

This is not to say that these experts don’t know what they are doing in the global sense. They are precise, accurate and consistent and are able to solve the problems offered to them in a swift, confident and decisive manner. Their skills are so automated and efficient at this point that they have stopped monitoring them on a conscious level. This is the basis of expertise, automation of the expert task and integration of the parts needed until there is a synergistic whole greater than the sum of its parts. This integration can lead to unexpected results.

The Power of the Brain: Systems Engineering

Back in the 1940’s the Bell Lab started using the term “systems engineering” to describe their approach to solving problems. Bell Lab members have shared eight Nobel Prizes in physics and chemistry and they invented or improved everything from radar to the transistor and much more. They also developed software, computer languages and published theoretical papers; in fact without Bell Labs our technology would not be anywhere close to what we have these http://days.https://hbr.org/1993/07/how-bell…

One of their seminal ideas was system engineering as defined by the 1995 NASA Systems Engineering Handbook: “…a robust approach to the design, creation, and operation of systems. In simple terms, the approach consists of identification and quantification of system goals, creation of alternative system design concepts, performance of design trades, selection and implementation of the best design, verification that the design is properly built and integrated, and post-implementation assessment of how well the system meets (or met) the goals.” https://spacecraft.ssl.umd.edu/d…

Or as Simon Ramo, the father of the ICBM said, “...a branch of engineering which concentrates on the design and application of the whole as distinct from the parts, looking at a problem in its entirety, taking account of all the facets and all the variables and linking the social to the technological.” (Conquering Complexity: lessons in defence systems acquisition, The Defence Engineering Group, University College London, (ed.) Dr. Alex Weiss, 2005,.10) Conquering Complexity

I bring up the systems engineering model because something similar happens to the human brain and body when a person becomes an expert.

Systems engineering started out as an attempt to develop ways to meet the challenges of incredibly complex proposals including the conception, design, development, production and operation of these projects. A set of concepts, structures and methodologies were developed over the years including broader ideas than just making a machine. In the process they found new ways to do things previously thought impossible. A well known result is the present day portable computer. Prior to the need for a light weight computing device in the Apollo Program, a 4K computer filled a room and demanded a host of acolytes to feed and maintain it.

The work that produced the first portable devices required the input of widely diverse technical and managerial disciplines, specific goals and a lot of work. It demanded the management of an increasing amount of data and an ability to integrate heretofore disparate fields of study. In other words it could have been a disaster of huge proportions had an overarching system not been in place.

One of the hallmarks of systems engineering is the recognition of common patterns or rules that exist in various subsystems. These subsystems have to be integrated since they can have negating effects on one another otherwise.

Take the issue of timing. Several subsystems may use the same resources in their individual cycles and if they clash, there will be delays as the resources are replenished or there may be direct competition between subsystems that degrades the quality and effectiveness of the whole system. Since each of these subsystems have similar rules for the use of resources, there are ways to optimize these cycles through the use of computer programs that take into account all the variables. It is complex and requires some change in the subsystems, but it allows for a more fluid and complete systems process. As a result, unexpected positive ideas and applications appear which would not have happened in a less sophisticated approach.

This is not a treatise on systems engineering, I just want to mention that we are becoming more and more aware of how to deal with what appears to be chaos or impossible issues through the application of managerial methods that are constantly evolving. The reason I point this out is that our brains work like this all the time – we have the most sophisticated super-computers in the world encased in our skulls – and it puts system engineering methods to shame. Experts elevate this ability to another level.

The average human brain has 86 billion cells (2009 Apr 10;513(5):532-41. doi: 10.1002/cne.21974.Equal numbers of neuronal and non-neuronal cells make the human brain an isometrically scaled-up primate brain.J Comp Neurol.Azevedo FA1, Carvalho LR, Grinberg LT, Farfel JM, Ferretti RE, Leite RE, Jacob Filho W, Lent R, Herculano-Houzel http://S.https://pdfs.semanticscholar.org… )and each cell can have tens of thousands of connections. This is a complexity far beyond what the Apollo program or anything else humans have attempted, yet the brain handles it with speed and accuracy.

The concept of systems engineering helps us to see what happens when the brain’s 100 trillion to one quadrillion connections interact and how unexpected answers can result. Without organization, chaos would reign, yet the brain is capable of organizing itself by processes that are not fully understood but have similarities to systems engineering.

Artificial Intelligence and the Brain

Another model may lie in recent concepts of Artificial intelligence (AI).

In the old days, prior to 2006, the concept of artificial intelligence relied on a top down structure. This meant that a programmer would feed rules into a computer and the computer would compare the rules to the situation presented and make a decision. There were specific circumstances in closed systems such as chess in which this paradigm worked well (but not spectacularly), but it required massive input from the programmers and was always restricted to what the rules said. Our brains are much more efficient and clearly do not work that way.

Human brains operate on patterns derived from data and are able to adapt those patterns to the situation at hand. Sometimes this leads to what is known as “cognitive bias”, a deviation from a rational view of an issue but mostly it is a shortcut or heuristic that helps make decisions. In 2006 a new concept of adaptive algorithms for AI were presented by several research groups who developed efficient programs that would learn and change depending on the data offered. This model of AI depends on a layer of input (which could be millions of data points), a layer of output which is the final answer and one or more “hidden” layers which are intermediates where all the work is done.

These are called “hidden” not because they are not seen but because they are not at the beginning or the end of the process. The math surrounding this process is elegant, small and dense but basically the new bottom up process changes the way that the data is looked at and corrects at each level while sending messages back to earlier processes in the hidden layers so that the program learns to be more accurate. For most of us this is like the Underpants Gnome episode of South Park in which the gnomes have a three phase business plan: Phase 1, Collect underpants; Phase 2, ????; Phase 3, Profit!! but it works very well.

This self correcting AI model helps people understand that there are sophisticated actions that occur with each neuron to neuron interaction in the brain and that these interactions can be modified by self regulation to become more efficient.

In computer AI systems, the programmers are not fully aware of what answers their AI systems will give. They can help determine the limits of the modifications, but the strength of the newer models of AI is that the program makes the crucial decisions and comes up with answers that are close to optimal through a process of bottom up analysis of huge amounts of data.

One of the new areas of research into AI is called Generative Adversarial Networks (GANs), another AI technique that introduces a second discriminator network into the process in order to help classify the outcome of the first network and compare it to a standard. This sets up a battle between the two networks that results in more accuracy, precision and efficiency when they agree with each other. The goal is to optimize the outcomes by finding a unique solution.

As people become experts they use mentors and standards to learn and then use both internal and external standards to make expert decisions later on. It is clear that the brain operates this way, at least from the information we have now.

I don’t pretend to understand the math or the entire AI process, but once these new programs came into existence, the practical results (Google Translate, for example) showed revolutionary advances.

In order for AI to work it takes a huge amount of data and a program (which may be very small) that tells the cells how to process it. The answers that come out of this model of AI are very much like those that a human with training would arrive at and in many cases are more accurate because the AI works with a larger data base. These answers can be surprising to the programmers as they look counter intuitive at first but are usually the right one. A similar process in the human brain may account for creativity and heuristic thinking.

There are numerous examples of AI applications that are more accurate than physicians in fields of medicine that rely on pattern recognition such as radiology and pathology. This does not mean that we will soon be bypassing our human physicians, but it does mean that we can develop tools that make them more accurate. The above mentioned doctors should welcome this ability as it helps with the decisions by rendering the field of knowledge a little less incomplete and it changes the way they use data.

The lesson learned here is that there are models for how the brain works and how it develops the expert state. These models are getting to be more and more accurate giving us a better understanding about the process.

One thing is clear, something is happening to the brains and bodies of experts that causes a significant change. From outside observation the process has been known for millennia but the internal details of brain function are only now coming out.

In practical terms all this knowledge of neurophysiology has not made a lot of difference at the basic level of education, but knowing what is happening can make the process more efficient and help parse out some of the noise that occurs when a training program is being developed.

Patterns and Chunks

We know from numerous studies that experts are able to see patterns much better than novices when these patterns are related to the domain field. For example, one study asked novices and baseball fanatics to recall what happened in a half inning of a baseball game. Each group rendered about the same amount of information but the baseball experts answers were much more focused on the game itself and not the weather, the size of the crowd or other information that was not germane. It is accepted that experts develop “chunks” (the official term) of domain information that are larger and more pattern orientated than non-experts and that their working memories are able to function better when they use these chunks. The Science of “Chunking,” Working Memory, and How Pattern Recognition Fuels Creativity Consciousness and the Prefrontal Parietal Network: Insights from Attention, Working Memory, and Chunking

Social Context

Becoming an expert is still dependent on social context, so improvement in the competence and performance of experts is evolutionary rather than a revolutionary. The development of expertise also depends on the field being discussed. Some sports, for example, showed a significant difference between the amateur and professional ranks in the past but in the last thirty years or so the differences diminished as the money improved and those changes have filtered down into the amateur ranks. These days participants are separated not by physical or mental development but by skill levels and the needs of the game. At the highest levels in many sports, athletes are no longer being self selected but are subject to objective models that eliminate anyone who does not meet the criteria. This has lead to a change in the standards for expertise in these sports and a narrowing of the skill continuum.

For example, aspiring professional baseball players that don’t meet certain foot speed qualifications are usually eliminated before selection to the major leagues no matter what other attributes they may have. (The exceptions are pitchers and catchers who have their own special qualifications.) Sabermetrics, the mathematical analysis of players output, now has a much greater part in choosing who is on a major league team than the gut feelings of scouts and coaches. This is not a change in the level of expertise as much as it is in the definition of expertise. Players are becoming more fit, more skilled, more knowledgeable and more aware of the social and environmental aspects of selection as they are being chosen based on more precise qualifications. This has lead to a change in work load and skill types that reflect these adaptations in selection standards. It’s also a good example how technology revises expertise.

This model (of choosing the participants with the best numbers) has always been true of other fields such as medicine. Candidates for admission to medical school are chosen not on their abilities as doctors – that doesn’t exist at this level – but for their ability to be good students and having passed a number of other filters that have little to do with being a good physician. The assumption is that if you can handle the fire hose of information that comes with medical school, the school will be able to transform you into a doctor of some kind. While in medical school students get a chance to try out various aspects of medicine and the school gets to evaluate them and help direct them into specialties. At least that’s the way it is supposed to work and since it doesn’t happen that way all the time, the levels of competence vary.

Medicine is a field with a knowledge base that is constantly growing and changing. In fact, the field is so vast that no one can master it and few are able to reach the highest levels of knowledge without becoming so narrowly focused that they are not able to fully understand much beyond what their sub-field offers. Because complete knowledge is impossible, it is not surprising that medical students are selected on the basis of their skills as a student in the hope that they will have a fighting chance in the pursuit of a medical career. The main purpose of medical school is to transform a student into a physician, a person who thinks and acts in a certain manner that is conducive to healing others. This requires the ability to understand and parse the constantly changing body of medical information, but it also requires someone who is capable of dealing with disease and the incomplete information that diagnosis and treatment bring. In other words someone who is willing to literally bet a patient’s life on their own judgment and who is able to deal with the inevitable mistakes that are made while trying to minimize them.

All of this requires years of dedication, study, hard work, motivation, mentoring, focus and above all exposure to things that most of us are not able to handle well. Just to be selected takes intelligence, a strong work ethic, a good working memory and the ability to pass tests and even then half of those who could make it through medical school fail to get in due to the competition.

This is a typical scenario on those fields where there are formal standards for experts. Candidates are selected based on aptitude, trained in a standard fashion, tested on a regular basis and then approved in a separate test or ritual to practice. The key component is selection and the process is repeated in professions and jobs that have a guild structure.


Guilds have been around for all of recorded history in one form or another – the shreni of India, medieval guilds, etc. – and they serve many purposes. The main ones are to select and train apprentices and journeymen, to develop a collective model of the profession that includes specific (and often secret) knowledge and skills, to enforce a set of rules to work by (including a code of ethics) and to ensure that society honors and understands the role of the members. Guilds are responsible for the formation of universities in Europe in the eleventh century and the basic system of teaching continues today.

The most important things that guilds do is act as a vessel of collective knowledge and control the environment of learning. While guilds themselves have evolved into a number of other entities, collective knowledge remains in the form of unions, schools, and other forms including the internet and technologies that allow anyone to interact with the knowledge base and others who share an interest.

Guilds shape the perception of experts not only by setting standards, but by devising an atmosphere of exclusivity that allows the guild to determine who belongs while persuading society that they are the sole dispenser of their expertise. That perception is not only for the masses but it also shapes the mindset of those who are under their tutelage. This is not by accident.

While there are economic and social reasons for guilds to keep the secrets of their skills to themselves, it is also the nature of becoming an expert that shapes this behavior. Much of this has to do with the state of the novice brain.

Novice vs. Expert Brains

A novice brain is not the same as an expert brain. Almost all the imaging studies that have looked at expert brains show volumetric increases or specific changes in the way they operate compared to novice brains. These studies use controls, persons who don’t have the same skill sets as the experts, to show these differences. One of the main aspects of my thesis is based on these studies showing transformation in the brain and while these changes differ among experts (depending on the field of expertise and other factors) there is a predictive aspect, namely that the parts of the brain that change are linked to the skill set or the function of the performance demanded by the field in question. In other words their brains transform as a result of the training they undergo. These alterations change the way experts understand the world vis a vis their domain and separates their knowledge from laymen in a fashion that appears secretive because of the assumptions that experts make and laymen don’t when confronted by domain knowledge.

What is interesting about novice brains is that they don’t show these changes except in rare cases. Instead, novice brains are the products of genes, epigenetic influence, environment, language and a host of other factors and are optimized as best as possible for their own lives. Expert brains are different due to the stimulus of training and education.

Because of these brain changes there are significant differences between how experts and novices act. Novice brains are not wired to be precise or consistent about domains, instead they are designed to take on a myriad of situations and be adaptable. As a result of that adaptability we make assumptions, run from danger, and react with a combination of emotion and logic to circumstances. All of this is beneficial to normal life.

We tend to ignore intense situations that don’t directly affect us through a process of denial and automatic thoughts in order to avoid unpleasantness and we mildly focus on tasks so we can monitor our environment. We concentrate on those aspects of our lives that are pleasant or neutral and we make decisions based on the emotional set we bring to the situation. And in general these are good things.

Note that I used the term “we” because when not in expert mode, experts act this way too. But when in expert mode, things change.

Experts use domain knowledge in a different way. They don’t just accept it, they absorb it and are constantly upgrading or comparing it in order to make sure that it is precise and accurate. They do this by practicing their skills and changing them as needed based on an analysis of how optimal that skill set is in a changing knowledge base. Experts are consistent, precise, accurate, and they correct any mistakes much faster than those who are not experts. When they correct mistakes, it is in a very specific way that reflects the mistake that was made. They learn to analyze issues in a way that novices don’t. As a result they are confident of their skills and their ability to maintain them.

Humans are extremely adaptable and this trait is reflected in the way our brains work. For example, it is well known that most people (about 70% McGuire Royal Society talk, see references below) demonstrate the so-called “boundary error” which occurs when a person is exposed to a picture for a very short period of time and then asked to reproduce it. They will extend the borders of the picture in their drawings so when the two pictures are compared, the original seems to be larger than the one imagined by the subject. In general we have a tendency to see things in a broader aspect than they exist and this is a positive trait because it allows us to expand our expectations and concerns, some of which may involve danger. Because we are not literal in our observations, we have the capacity to change and create. The boundary error is not a bug, it is a feature.

The Human Brain

When human brains are measured against other animals in relation to body size, with one exception this quotient is greater than found in any other mammal. (The exception is the treeshrew, a southeast Asian relative of primates and hardly a competitor on the intelligence scale.) Our brains are larger because of the cerebral cortex which includes parts responsible for reasoning, planning, language, abstract thought and various executive functions – all of which separate us from the rest of the animal kingdom. Scientists have found a unique gene, ArhGAP11B (Human-specific gene ARHGAP11B promotes basal progenitor amplification and neocortex expansion Science 27 Mar 2015: Vol. 347, Issue 6229, pp. 1465-1470 DOI: 10.1126/science.aaa1975 Consciousness and the Prefrontal Parietal Network: Insights from Attention, Working Memory, and Chunking ), that plays a role in how our brains develop. This gene occurred some time after our evolutionary line split off from chimpanzees and before the advent of Neandertals and Denisovans who shared it. It is responsible for neocortex folding, a process that greatly increases the surface area of the brain, especially the cerebrum, and results in our unique qualities.

The brain weighs, on average, two and a half to three pounds and contains around 86 billion neurons and non-neuronal cells. The largest number of neurons is in the cerebellum which is a more “primitive” part of the brain (I like to call it the lizard brain) even though it is much smaller than the cerebrum. The cerebellum (which means “little brain” in Latin) is involved in movement and motor control, but it also has known cognitive functions and regulates coordination, precision, accuracy and motor learning. There are extensive connections with the cerebral cortex that have not been totally explored.

Because of the unique (or at least advanced) qualities of the human cerebral cortex – language, planning ability, abstract thought, etc. – humans are much more adaptable than other mammals. From its complex structure we derive our ability to think, to create, to move, to love, to remember, to plan and execute our plans. Over the years the brain has evolved to be a jack of all trades while retaining some seemingly “hard-wired” functions such as language, social instincts and the ability to automate actions and responses. A combination of genetic, epigenetic, and environmental factors can change the brain or start it out with certain traits from birth.

The most intriguing thing we are learning is that brains have functions that are not obvious. Take the above mentioned boundary error, it seems like this is a problem, but that is mostly because it was labeled as an “error” which is the wrong thing to focus on. It’s actually a good example of near time prediction, the ability to determine before hand what is going to happen while an experience is occurring. This function has great adaptive value as it allows you to anticipate and plan in the short term.

A simple example is tying your shoes in which you can envision the sequence and be prepared to move on to the next part of your dressing routine. If you are watching someone else tie their shoes, you can use your experience to predict their sequence with good accuracy. If there is an error in your prediction due to some other circumstance not related to your first set of memories, you update it and add this experience to your working memory for future reference. This response is automatic and involves both conscious and unconscious parts of brain function.

When experts watch other experts in the same field they show activity in the same areas of the brain that the observed show even though these observers are quiet. They also send subliminal signals to the same muscle groups that the observed are using. Novice observers don’t show those changes and experts in similar fields don’t either. This appears to be a clue about how the brain adapts when it is stimulated by a focused and intense amount of information and skill acquisition, the very things found in the education of experts.

Another example of hard wiring is the fight or flight response to stress which is an automatic function triggered by the brain to help us survive dangerous situations. In general, once a danger is perceived – something that may or may not be obvious to us – there is an automatic sequence of events that includes the generation of adrenalin, shunting blood to brain and motor structures, increased perception and a sense of danger. This process changes the way we perceive the world and how we respond to it. Because the fight or flight mechanisms occur at times we may not logically think are dangerous, they can have a profound affect on performance. Expert brains modify this response to have less effect when appropriate.

Before people become experts they start out with with a complex and unique set of brain functions that affect performance and perception. When we begin training we retain that adaptability but we develop significant changes in both body and brain that allow us to become more precise and to view our domain in a different way. We continue to be subject to things such as the fight or flight response but the response is not the same.

Other qualities of the brain such as working memory, personality expression, executive functions, etc. are used differently in the expert domain. These are more or less hardwired in (although they develop with age and thus are influenced by environment and other factors) but are clearly different in expert situations.

Performance and Performance Anxiety

So far I’ve talked about the internal changes that occur when someone becomes an expert. But there is much more to it than just a change in brain structure and the addition of new brain cells. Experts are defined by societal forces so they have to deal with human interactions and have to meet outside expectations. As I mention above, many of the definitions of expertise focus on the effects an expert has on the environment, but environment has an equally important effect on experts. These effects can significantly change the outcome before, during and after the interaction.

The best term that describes what an expert does is “performance.” Most of us think of this in terms of entertainment or sports but an expert’s stock in trade is producing some kind of result. This action is performance.

While performance implies an audience, an audience doesn’t have to be present in order for an expert to perform, it just has to exist. Since the actions of an expert are transformative and things change, it affects others no matter where or when it occurs.

If we hark back to the idea that experts have to meet standards, we realize that there is an internal pressure to do well or at least to meet the criteria set up. But there are other stresses that have to be dealt with, namely those that occur in the social sphere. The most common one is performance stress, the physiological and psychological arousal that occurs before and during performance.

Unlike the response to standards of care or common models or dictates of the profession, performance anxiety is a product of our hard wiring. Many experts, especially those in fields such as individual sports or music, don’t perform that much. When the time comes, not only is it an unusual event (compared to practice or preparation time) but it is also threatening due to the perceived expectations of the audience. As a result, an automatic response occurs. Our lizard brains react as if there is a danger, setting off the fight or flight response, even though logic tells us that it does not exist. This is something that occurs to everyone, including the most successful and masterful in the field, and needs to be accounted for. Part of becoming an expert is learning to deal with performance anxiety.

The most successful responses to this arousal have to do with training. Professions with indeterminate expectations such as medicine (because of the wide variety of problems and incomplete knowledge) or team sports (because of the wide variability; the more participants, the more variability) train their students to expect and to react to stressful situations. This training takes place in environments that are pressured and determined by coaches. In professions like medicine there is a gradual increase in responsibility until the expert is considered independent of immediate supervision, often earlier than the student expects. Either way, learning to deal with stress is built into the system.

Stress training helps the student counter anxiety by predicting problems and offering solutions that can be transferred to similar situations. It develops a sense of confidence and self assurance, qualities that can be enhanced and strengthened with continued experience. It triggers the “fight” part of the stress response which has none of the hormonal issues that “flight” has and thus is more useful in these stress situations.

Some experts don’t have the opportunity to deal with performance anxiety because they are not aware of how it affects them or the initial training does not take it into account. A good example is the shooting sports that I mentioned above.

Shooting is a mental and physical skill that can be used in competition. There are several types of shooting sports but for the sake of demonstration I’ll talk about Olympic Free Pistol shooting. The rules are simple, you have an hour to shoot sixty shots at a dime sized bullseye at fifty meters. There are concentric circles on the target of various values and the center is worth the most points. A perfect score is 600 and in the old days the person with the highest score at the end of the match was the winner.

Theoretically you could isolate yourself and shoot the best score after you perfected your technique. But shooting in a Free Pistol match is not just about good technique, it is a competition. In order to do well you have to have shooting skills and competition skills. A lot of shooters can shoot but can’t compete because the majority of them are self taught so they don’t take the competitive stress into account until it is too late. Because their training regimen fails to address stress, it is likely that stress will not be addressed during the match.

This brings up a question: are you really an expert shooter if you are not prepared to be a good competitor? I think that you are, but you are not a competent expert because there is a disconnect between the expectations (to be a good shooter) and the reality (to be a good competitor) that is a result of the training (or in many cases, lack of) offered to want-to-be experts. If those who set the standards fail to take into account such things as stress, experts in that field will have a hard time being competent.

The training most experts get has to take into account all the factors I’ve mentioned (and a whole lot more) if it is to be successful. The anatomical, psychological, perceptive and social changes that occur as a result of training should address all these issues. Training does not guarantee success or competency if one succeeds in finishing a course, but it does cause change. The more intense, the more complete, the more precise and the more focused the training, the more likely that a competent level of expertise will be met.

Let’s look at the steps that lead to becoming an expert.


The very first step is being selected. In most guild type training selection mirrors the needs and tenets of the guild. If you want to become a US Navy SEAL, you have to be screened for mental and physical fitness in order to even qualify for the prep school that leads to the Basic Underwater Demolition/SEAL (BUD/S) school (which is 24 weeks of hell.) About 6% of those who apply to be SEALS meet the fitness criteria and of those who finish the prep school, only a quarter qualify to be SEALS.

In comparison, when I was running a Navy Drug Counselor’s School in the early 1970s, anyone could be admitted, they just had to get orders to go. For a while there my school had the second highest school attrition rate in the Navy (second to BUD/S) because we were often sent the worst candidates. At that time the need for drug counseling, which was acute from a medical standpoint, was not considered to be crucial by those military chiefs who set the standards for selection. Selection plays a crucial role in the development of experts and it usually attests to the standards of the guilds. As you can see, sometimes the guilds have no idea what they are doing and it reflects in the training and other considerations later on.

Many people are self selected before they present themselves to an entity in the formal sense. This is most likely part of the benefits of talent since talent doesn’t manifest itself until a person is exposed to a domain. It seems that the more talent asserts itself, the more likely that a person develops an attraction to it, especially if this occurs in childhood or adolescence. Many athletes will tell you that as soon as they started their sport they knew that this was going to be an important if not overwhelming part of their lives.


The next step is to have a mentor. Mentor, teacher, coach, sensei, guru, master, no matter what you call them, experts always have some sort of model to emulate or guide them. Most of the time it is more than one person who is in an authoritative role or who has an emotional and logical affect on the student. This doesn’t have to be a memorable personality, a mentor just has to be able to change you in some way, mostly by making you work.

Mentors are important for a variety of reasons. The most obvious is that they monitor and affect change, but they also serve as a model, impart knowledge, plot training schedules, act as a repository of experience and set standards. Being an expert is about social interaction, but so is becoming one.

Face to face interaction with mentors doesn’t have to happen very often for the magic to work; just consider how much time a person spends in class versus the time they spend practice or studying. The process occurs because there is a relationship, something that does not happen (or happens very slowly) if there is no mentor.

Mentors evoke non-logical responses in students, which is important because becoming an expert involves non-logical change. You may hate your teacher, love your teacher or be diffident, but your response will never be totally indifferent. An emotional response will translate into motivation, focus, memory and commitment at some point. Mentors also give the outlines of how to become an expert and how to train; in fact they have a lot more to do with success than most people believe.

Early on, students tend to use rules to solve problems, rules gained as dictums from teachers or as a result of modeling. It’s not unusual for students to mimic the language, attitudes or even dress of their teachers, a process that sets the students on a course towards their domain. When students use the rules set by their mentor, problem solving becomes more efficient since the rules are generally based on knowledge and experience.

Of course, poor mentors can lead to just the opposite, not because they discourage you (which they might) but because they can influence the process of becoming an expert by leaving huge gaps in training and knowledge.

Ideally a mentor should be knowledgeable, experienced, interested, empathic, organized, challenging, trustworthy, helpful and honest. This set of qualities and others like them are not only hard to find, but they are also difficult to continue with consistency unless the mentor puts as much into the process as the mentee.

While becoming and expert and maturing in the process, most students have more than one teacher. The advantage is that a series of many teachers gives the student a chance to be a critical thinker, a quality that is needed if one wants to be able to self monitor. As an expert progresses, there are different needs that have to be fulfilled if you want to have a complete experience and different mentors can help in every step along the way.

Because the selection process can be faulty – being selected because you can pass the tests and tolerate the program is not a good predictor of success later on – mentors have to test the student and help them become successful. This may involve greater exposure to stressors, guidance in the “real world” aspects of the field or even discouraging the student if warranted. Job performance evaluation (as opposed to meeting certain criteria) is very important if a mentor wants a competent student.

A good example of this is the Education Dominance (Acuitus.com – DARPA – Education Dominance) program designed by the Defense Advanced Research Agency (DARPA) for the US Navy. The Navy had a problem: there were hundreds of ships with IT networks including 300,000 nodes connected to 200 different networks. Many of these networks were on ships that were out to sea and if the networks went down, they needed onboard specialists to fix them immediately. The problem was that the Navy didn’t have the time or teachers to supply expert technicians for the systems they had.

DARPA recognized that with proper mentoring, a novice could become an expert in computer networks in a short period of time but that one to one mentoring was not cost effective on a large scale. In the first phase of their study, fifteen new sailors were trained by twenty-four highly skilled mentors producing students who were exceptional in their skills and performance and who won numerous competitions against veteran IT specialists in the Navy. This tutelage was carefully monitored and the techniques and individual training data were put into an AI program called Digital Tutor with the purpose of replicating the exceptional tutors of the original group. A study was done in which Digital Tutor students were compared against students trained in the standard fashion. Those taught by the Digital Tutor program, which mimicked exceptional one to one tutors and took less time, were head and shoulders above the others when tested in difficult real world problems. This study showed that gifted mentors, be they humans or AI, make a difference.


The third factor is learning the body of knowledge in a chosen field. This is not just learning all the facts about a certain subject; it’s also learning how to interact with domain knowledge . With the exception of some closed systems, the amount of knowledge in a field is massive and new facts are coming in all the time, altering the way it is used or experienced. You have to keep on top of things and use critical thinking to evaluate each bit of information. More important is the interaction between student and the environment of facts. As a student progresses, automatic insight occurs with familiarity with the facts. In open systems (systems that have no limits because they have too many outside influences – e.g.medicine or politics), most practitioners work with incomplete information and have to make decisions based on optimization rather than a simple logical equation.

Optimization is the process of finding the best available answer to a situation or problem given a limited amount of time, an answer that works because there is no best answer. The greater the body of knowledge that an expert commands, the more likely that the optimal answer will be discovered and the more likely it will be effective.

Studies have shown that experts with more specific and greater knowledge of a subject have a better chance of making good decisions in situations where there is ambiguity. As students become experts and experts develop further there is a shift from rules based decisions to “exemplar” (which means comparing situations to similar experiences) decision making. This has to do with working memory retrieving other examples and is a very efficient way to work with typical domain situations. When the problem is atypical, experienced experts go back to the rules but because they have greater knowledge, the rules are used in a more sophisticated manner.

In medicine there is the concept of “evidence based treatments” which are rule based treatments that have strong statistical and scientific knowledge dictating them. Evidence based decisions are considered the gold standard of treatment but the problem is that there are very few treatments that meet that high level of confidence. The next level, consensus of experts, is the most commonly used standard because it reflects what seems to work even though it has not met the strict criteria of being scientifically based. One of the reasons that there are few evidence based medical decisions available (other than the cost of research) is that most conditions physicians encounter are mixed; they have multiple causes and presentations. There is no one answer that works all the time. Most of the time consensus based treatments are successful and often they are the only treatments available, which is the reason why experienced physicians use them. Evidence based treatments are vetted in a very rigid manner and there are clear rules as to how to apply these treatments, but life is not always that simple.

Doctors are not scientists, they are applied scientists at best. They have extensive knowledge of their fields and they are constantly finding new things to help their patients. Their focus is on the individual patient so it is imperative that they keep up even though the knowledge that they have changes and is always incomplete, both from a diagnostic and treatment perspective. Patients never present with a pure diagnostic problem to solve because they have many things going on at the same time. As a result, physicians have to constantly be aware of the individual needs of patients so knowledge is paramount. Learning is one of those life long skills that has to be continued if a doctor wants to remain competent.

A crucial aspect of learning is the ability to parse out what is useful and what is noise. If your area of expertise has a huge knowledge base, there will be plenty of useless facts flying around. It is important to develop a critical mindset regarding information. Since gaining knowledge is not isolated from the rest of the learning process, students find that knowledge dovetails well with their skills. Eventually there is an automatic filter that picks out helpful information. This happens as the brain becomes more focused on tasks and it becomes easier to see what is domain knowledge and what is not.

Learning as much as possible about an area of expertise also gives a student the chance to develop a style based on “standard practice”, or the consensus of what is the best way to do things. Standard practice does several things: it allows those in the field to compare results, it helps a student learn in a more efficient way that eliminates dead ends and it forms a base for techniques that are needed later on. Standard practice changes but only after there are reasons to do so. In golf, techniques have changed due to new materials in clubs and balls and exploration by the masters of the game to improve their scores. Over the years these changes have been accepted and tweaked as players want to have better performance and as outside influences (TV, money, fame) increase motivation.

Learning is a life long process and experts who continue to be successful develop a strong desire to learn more and more. Most experts take advantage of access to the internet, social media (to find out what is going on in real time), coaching and print media in order to keep up.

Acquiring knowledge is not just a matter of memorizing stuff. The environment of learning to be an expert is loaded with cues, hints and direction that steer the student towards domain specific knowledge. Students are taught to emphasize information that is limited in scope but is vital to the technical and reasoning skills needed for their area. Students often form specific biases as a result of their knowledge base and the influence of mentors. A study of physicians who have different specialties show that each specialist tends to diagnose the same difficult problem as an aspect of their field. An interesting observation in this study is that those physicians who are considered among the elite are able to include other possibilities in their differential diagnoses even though they tend to favor their own interpretation. These high level experts are able to see patterns that are more complete than students or others who are not trained as well.

As experts progress, a fundamental change in their cognitive abilities occurs. Most people have working memories (the function of memory that retrieves long term memories in a way that helps solve problems) that bring up patterns of thoughts in chunks when they solve a problem. This process, which is mediated by the hippocampus, helps people with planning and near time prediction based on the size of the chunks. Studies show that experts and talented people have chunks that are larger and more comprehensive than novices, and that there is a direct correlation between levels of knowledge and the ability to make useful inferences from partial information. This ability is not based directly on intelligence as much as specific domain related knowledge.

Other studies show that once an expert is outside of their domain knowledge, their ability to work with partial knowledge is no better than a novice in their field. Basically you are smarter in your domain because the types of information that the working memory acquires are in larger chunks and the patterns are more obvious.

One of the clues for this is the use of jargon. Every specialty has words that have specific (and not so obvious to outsiders) meanings. This specificity is one of the ways chunks manifest themselves, and while these words may be obscure or have different meanings to laypersons, they immediately conjure a particular image to experts in the field. Law is loaded with terms that befuddle non-lawyers who can’t understand why a term doesn’t mean the same to them as it does to a judge. But these are precise terms that have a long history of litigation defining them and are often arcane due to being in the legal system for centuries.

Comprehensive specific knowledge is important in any field of expertise. In the process of learning, there are significant changes in cognitive functioning, points of view, and the ability to process information. As an expert progresses there is the challenge of metacognition, the ability of knowing what you know and what you don’t know. An expert needs to know how to distinguish information that is pertinent from information that is noise. This selectivity is hard in the beginning (which is why rules are so important early on) but a competent student will not only learn to determine relevance but also be able to go back and examine earlier information and incorporate it into the expert mode. This allows the chunking process to improve because when information in long term memory is pertinent, the ability to solve problems using the patterns engendered by these memories is much more useful. This process is learned and automated when a student enhances specific long term memory by the use of many different tools (such as mnemonics) and the brain automates the use of this memory.

Knowledge is useless without some practical challenge. As I mentioned earlier, experts work in a social context so they have to perform. In order for knowledge to be useful, it has to be available in a form that meets the needs of the audience. Knowledge is the servant of the expert and an expert has to learn how to use it. We know that the expert brain adapts to work situations – something all brains do – but that adaptation is deeper, more specific and more efficient than a layperson who read a Wikipedia entry. An expert scans the problem for patterns of regularity and irregularity, fields pertinent clues, thinks abstractly about the situation, finds any oddities and makes a decision. Most of the process is automated with the exceptions of abstract thinking and metacognition (and even that might just be a “bad feeling”.) Automation allows the more conscious aspects of problem solving to be efficient because there are no distractions.

The effects of learning are not just an accumulation of a pool of facts; there are significant brain changes that occur and part of the process is automating how knowledge is used. Deep knowledge puts the expert on familiar ground in situations that are unique to a layperson. That familiarity is the reason an expert can function without the lizard brain firing off and interfering with performance. Having a working memory that can pull up patterns and offer solutions immediately makes things a lot easier. The experience of learning allows an expert to have longer attention spans vis a vis the subject of their expertise and it lets them work within the constraints of the situation because it gives them internal patterns to use when the environmental cues are partial.

Domain knowledge is special knowledge that may overlap with a general knowledge set, but will evoke other responses in an expert. I call this response (for lack of a better term) domain heuristic which is an emotional and automatic framework used by experts to solve problems. It is related to chunks and how working memory uses them, but it can bleed over into non-domain knowledge. Domain heuristics are often used like an algorithm by experts – an algorithm is a logical well defined set of instructions based on a deterministic set of values that don’t vary – but unlike algorithms these instructions are nearly instantaneous and come from a non-logical base. Domain knowledge is powerful because it is not just an accumulation of facts, it is tied to non-logical cues (read emotion, among other things) that reinforce it and make it easier to obtain and use. Experts are bound to their domain by more than logic, that’s why they stick to it even in the face of failure.

While there are many reasons why novices are attracted to an expert domain, one of the important reasons that they continue is this emotional attachment which is made richer and more binding by learning and knowledge accumulation.


Skills acquisition, learning how to do the things your field demands, is a much more down to earth part of becoming an expert. Virtually anyone can learn a skill if given enough time, and studies show that the brain becomes efficient very quickly when a simple skill is learned. But skills are subject to loss if not practiced and most expert fields demand the use of many skills.

Because expertise involves interaction with others and it requires transactions of some sort, there are skills that most of us take for granted or don’t even consider when discussing expert domains.

Hard skills are those skills needed to complete a task and are specific to a situation or circumstance. They are easily described and have specific outcomes that can be measured in some way. Examples include such things as shooting free throws, installing a water heater or flying an airplane – skills that have specific and measurable outcomes.

Because hard skills are so quantifiable, they can be transferred to other fields of expertise. The qualifying aspect of this transfer is that they don’t work in isolation, as domain skills are an organic part of the expertise experience. In shotgun shooting, for example, the ability to hit a moving target can be learned in a few hours but shooting in a competitive situation is completely different. The latter involves many other skills that affect how you use the moving target skill. Even among the various shotgun sports the goals and perceptions are very different, requiring that you use your moving target skill in ways that are optimal for the specific sport. Domains are specific and the optimal skills for that domain are too.

Soft skills are less quantifiable but no less important. They include various people skills, communication skills, and things like common sense. These skills add to productivity, especially in a group situation, and they are important because experts interact with others. The US Army (which invented the term “soft skills”) lists them as: leadership, organization, management, communication and team building. They teach these skills throughout the careers of soldiers. starting in basic training, and all soldiers are expected to develop and improve soft skills along with their military skills.

One thing is clear, if you are going to learn a skill, you have to put in the time. As I mentioned above, simple skills can be learned over time, but the types of skills experts use are usually not simple and many people never learn them.

Golf is a good example. In 2013 the National Golf Foundation reported that in the United States there were 25.3 million golfers above the age of six who played golf at least once on a golf course. About ten million of them played golf on a regular basis and, remarkably, there were 32 million people who said that they wanted to play golf but couldn’t find the time. (Number of golfers steady, more beginners coming from millennials – Golf Digest April 23 2015 ) In that group there are 28,000 professional golfers of which 241 men and about 200 women members are eligible to play in the top tournaments. The vast majority of golfers never even come close to these elite players except to watch them play. In spite of the popularity of the game, golf skills are not easy to learn.

Golf is a simple game … to describe. The object of the game is “to hit a very small ball into an even smaller hole with weapons singularly ill-designed for the purpose.” (Winston Churchill) It’s called golf because all the other scatological names had been taken.

The legendary Ben Hogan, in the first chapter of his 1957 book Five Lessons, recounts a shot he took at the US Open in 1950 when he was 38 years old. He needed the shot in order to tie for a playoff (which he won) the next day.

“It was all I could have asked for. I got down in two putts for my 4, and this enabled me to enter the playoff for the title which I was thankful to win the next day.

I bring this incident up not for the pleasure of retasting the sweetness of a “big moment” but, rather, because I have discovered in many conversations that the view I take of this shot (and others like it) is markedly different from the view that most spectators seemed to have formed. They are inclined to glamorize the actual shot because it it was hit in a pressurized situation. I don’t see it that way at all. I didn’t hit that shot then – that late afternoon at Merion. I’d been practicing that shot since I was 12 years old.” (Page 14, Ben Hogan’s Five Lessons Golf Digest 1957, 1985 edition)

Ben Hogan, a child prodigy who knew that he wanted to be a professional golfer as soon as he was introduced to the sport, states later on in that chapter that he had spent the majority of his life developing a “A CORRECT, POWERFUL, REPEATING SWING” (his capitals) and wants to impart this knowledge to other golfers to help them break a score of 80, one of those magic numbers golfers strive for.

I’m not a golfer, but I love Five Lessons . The first chapter, “Fundamentals”, concisely sums up the total knowledge of skill acquisition. Hogan’s insights, his knowledge of the game and his obsessive desire to be perfect are all evident in this chapter. It is a model that experts should strive for when they learn skills. Ben Hogan was a genius at his game with an unusual understanding of what made golf and golfers tick.

The beginning and the end points of learning skills have a lot in common. Perhaps the most obvious is that the skills seem “simple” when discussed by novices and masters but for different reasons. In the beginning you see the big picture – just hit the ball and it will fly towards the green – but after a long and often painful period of improvement (this is golf, after all), the skill becomes automated and part of the game for you. It is simple because you are no longer thinking about what you are doing but are using the swing as a tool to play the game well.

It’s what happens in the middle that makes the difference.

Hogan talks about how difficult it is for him to watch a golfer who never improves because he repeats his mistakes over and over. He sees these golfers rationalizing their lack of improvement by touting the “exercise and companionship” instead of getting better. “The greatest pleasure is gained by improving”, he states. He opines that these golfers will only get worse as they reinforce their errors and more creep in.

What Ben Hogan recommends is a simplification of the swing into five fundamentals that can be practiced and examined in an objective manner. These five fundamentals are standard practice now, but at the time Hogan proposed them they were revolutionary. He felt that they were the essence of the golf swing, that they were so fundamental that ignoring any of the elements would negate the benefits of the others and that a golf student should learn them in a way that lets the student find flaws and repair them. “I have never seen a great player whose method of striking the ball did not include the fundamentals we will emphasize. Otherwise – it is as simple as that – that golfer could not be a great player.” (Ben Hogan’s Five Lessons)

The important insight for all of us is that a skill can be broken down into fundamental quantifiable parts and these parts need to be mastered, often in order. The skill itself is a unitary function that becomes automatic, but when acquiring it you have to make sure that you do everything well and that you know when things are right or wrong in order to adjust and fix your skill. Training for a skill often takes an incremental knowledge and practice of fundamental parts that eventually meld into an automatic whole.

The first chapter in Hogan’s book neatly sums up the decades of research and opinion on skill acquisition that occurred mainly after 1957. Mastery of skills requires time, direction, an ability to critique oneself either through a mentor or ability to objectively see how one is doing and a layered learning of fundamentals that leaves no gaps.

The latter point is important and often overlooked, especially by persons with high levels of talent. If you are going to be an expert, you have to learn skills – that’s skills with an “s” – and those skills become more and more sophisticated as they integrate with one another. I mentioned earlier that the human brain has system engineering skills, it can join many things into a whole, and learning to join layered fundamentals is a good example of that. But it takes time and a lot of practice to happen. To master a field each skill has to be perfected and maintained and if the fundamentals of each skill are not present, the more advanced parts will be weakened or never be achieved.

This is why practice is so important.


British English makes a distinction between the noun “practice” and the verb “practise”, the latter referring to repeated systematic exercise in order to improve. This is an interesting difference as many professions called their work a practice, but as we will see there are reasons that practice can be more than just work.

The deliberate practice model (DPM) states that what it takes to become an expert is lots of hard work focused on making changes that specifically direct a person towards the expert domain. The founder of behaviorism, John Watson, stated if you gave him a dozen healthy infants, he could take one at random and train him to become any kind of specialist “regardless of his talents.” (Watson, Behaviorism, University of Chicago Press page 104, 1930)

Watson’s idea is that people are a blank slate and all it takes is the correct training to make someone an expert, you just have put in the time and do it the right way. We know now that this is not a complete view of what happens when someone becomes an expert. Hard work alone is not the answer and, as Ben Hogan points out, the “right way” often isn’t either.

Don’t get me wrong, the idea behind DPM is correct, you have to put in the time and you have to practice well. Many expert domains are extremely complex (maybe all of them if you want to master them) and the competence level that an expert reaches is substantially due to the amount of work put in; but DPM is not the total decider and in many cases is but one factor.

In the last decade or so there have been a number of academic papers that have examined the issue of DPM and how much it contributes to the development of expertise. Using sophisticated statistical methods and controlling for reliability of information these papers have more precisely determined how much variance (the amount of competence that is achieved by experts) can be attributed to deep practice. The results vary according to the field studied and can range from very small (4%) to being the majority determinant depending on the type of expertise studied. (Hambrick, et al Intelligence 45 (2014) 34-35 (PDF) Accounting for expert performance: The devil is in the details )

This is not surprising. While the deliberate practice authors claimed that work is the one factor that accounts for individual differences in performance, the experience of others belies that. Ericsson and others relied on memory and retrospective analysis for most of their findings and virtually no prospective studies were done and duplicated, the gold standard in other fields. The results of their studies became immensely popular due to a number of best selling books by authors such as Malcolm Gladwell and Daniel Coyle who touted the meme that hard work alone will get you where you want to go. This egalitarian viewpoint is very seductive but in the long run, not proven.

These authors are both wrong and right. Hard work will bring on worthwhile rewards, especially if it is directed in the way that Ericsson’s work recommends. But DPM is not the only ingredient. Without practice, performance will not be improved, but the positive effects of DPM would not happen without other factors such as talent, environment, personality (which Ericsson acknowledged in 1993), intelligence, working memory capacity or early starting age. These are all traits that are innate but with DPM will increase considerably. As deliberate focused training continues, these become integrated into the process and after a while are inseparable parts of the expert mode as each changes the other. The whole becomes greater than the sum of its parts.

DPM is the ladder to success in this scenario. Hardly anyone is selected with the assurance that they will succeed. It is well known that of the number who start on the pathway to expert success, only a few reach the pinnacle of mastery. The fact is that even if you have the talent, the motivation and the drive to succeed, if you don’t have the right type of training that includes the teaching of standard proven techniques by teachers who know what they are doing, using well defined goals that stretch your abilities, learning to monitor your progress and changing as needed, practice with full concentration and focus and building and modifying layered skills in order to improve, you will not be able to be the very best. (Ericsson Peak pp 99-100)

As you can tell, that is a lot to expect of a student and a lot to do, but it is doable.

One of the common issues of society is that many are chosen to learn even if they are not called. Expertise is a social construct and as such people are called upon to become experts by necessity. The military draft and military service are a good example. Training new recruits and draftees is a significant problem for societies and the results of attempts at training constitute a natural experiment in various methods.

By the end of World War One 2 million men volunteered and 2.8 million were drafted. Of those drafted about thirty percent did not meet criteria, mainly due to physical defects and chronic disease. The rest were sent to National Army divisions, trained for six to eight months and sent overseas. That training was thorough as far as basic skills went, but troops sent into battle had little practical experience with war.

My grandfather went to the 83rd Division, trained for eight months at Camp Sherman in Chillicothe, Ohio and shipped to France in May of 1918. His regiment (about 4000 men with attached units) was selected to go to Italy as part of a political decision on the part of President Wilson and General Pershing. All of the other regiments were depot troops who filled in as casualties mounted on the Western Front.

The 332nd Infantry Regiment (United States) was unique in WWI. Not only did they get to pick the best soldiers from the other regiments (and leave behind others) but when they arrived in Italy, they were trained by Arditi from the XXIII Reparti d’assalto medaglia d’oro 23esimo, a highly decorated unit of commandos who used live ammunition and hand to hand combat techniques to teach the 332nd to assault enemy trenches, cross rivers, and use shock tactics on the enemy. Six American soldiers died during this training and fifty were wounded. They would march up to 50 kilometers a day and were in live fire trenches facing the Austrian and German forces as part of their training. When battle came, they were far better off than the other regiments in the division and their performance in battle and their peace keeping role after the war showed it.

This was no accident, the unit had the best soldiers from the 83rd Division (many were college graduates and/or professional athletes), their colonel was a highly regarded combat veteran, and most of the officers were regular army, which meant that those who were in charge had a clear vision of what they wanted. The regiment underwent basic training but then were trained, under realistic war conditions, by the best troops that the Italian army had.

This was an early use of the DPM style of training that is common in the US military these days. The result was a highly effective force that was flexible and well trained. In fact, the 332nd Infantry had to transition to peacekeeping forces in the Balkans right after the war ended and were so successful that they were studied as a model for US troops during the Balkan crisis in the 1990s.

The lesson here is that a group of draftees (albeit the best of the bunch) were able to be molded into an elite fighting force through very hard work with experienced mentors guiding that work to mimic the demands of battle.

Practice is important and is the most crucial element needed to become an expert. No one becomes an expert without the brain and body changes needed and practice is the part that makes these changes happen. But practice is not mindless repetition, it has to be directed and meaningful if it is going to make a difference. Experts are made and you have to put in the work if you are going to progress from layman to expert.


1) To become an expert you have to practice in a way that causes change in brain, mind and body. These changes occur so you can automate the skills and knowledge needed to be an expert.

2) Automaticity is the sign of an expert. It does not confer competence, only efficiency in thinking and acting.

3) Knowledge and skill are very important but they have to be gained in a way that lets you perform as an expert. If you can’t make a difference, you are not much of an expert. (Even though, by my criteria, you are one.)

4) Your expertise is defined by your social context. This means that your performance is transformative – what you do effects others – and others will judge if you are competent or not. The same is true of your training. Others set the standards even if you are self selected and trained.

5) Experts never stop learning or training. Things change, knowledge changes, skills change, and expectations change. You have to stay up to date and if you don’t continue to improve you will be left behind.

6) How do you get there? Practice, practice, practice.

Further References

Marie-Line Germain

(PDF) Expertise: A Practical Explication


Royal Academy Talk by Eleanor McGuire PhD

United States Navy SEAL selection and training

Copyright © Michael Keyes 2018