“Flirting With Your Possible Selves”, the seventh chapter in David Epstein’s 2019 book “Range”, begins with a description of the remarkable life of Frances Hesselbein:
“She never did graduate from college, but her office is festooned with twenty-three honorary doctorates, plus a glistening saber given to her by the U.S. Military Academy for teaching leadership courses — as well as the Presidential Medal of Freedom, the highest civilian award in the United States. When I visited just after her 101st birthday, I brought her a cup of steamed milk, as I had been advised, and right away asked what training had prepared her for leadership. Wrong questions. ‘Oh, don’t ask me what my training was,’ she replied with a dismissing hand wave. She explained that she just did whatever seemed like it would teach her something and allow her to be of service at each moment, and somehow that added up to training.”
There is nothing exceptional about Hesselbein’s childhood. Born in 1915, it’s likely that today we’d label her a disadvantaged child, having grown up in the steel mills and coal mines of western Pennsylvania. There’s certainly nothing conventional in her experience, education, and training that indicates a deliberate and focused path toward becoming the transformative CEO of the Girl Scouts, of which her first experience with the organization, at 34-years old, was as volunteer leader of her Local Troop 17.
In this way, Hesselbein doesn’t come across as an outlier; as someone who overcame tremendous odds or as a late bloomer who took advantage of a latent but special gift, instead Epstein frames Hesselbein as a “dark horse”, someone who through frequent failure, short term planning, and constant iteration makes their own novel and circuitous path to success in life, despite social pressure to conform to what seems the safest and most certain route. A Harvard computational neuroscientist is quoted about their work on the Dark Horse Project, a formal study of the common qualities of successful nonconformists:
“[Dark horses] never look around and say, ‘Oh, I’m going to fall behind, these people started earlier and have more than me at a younger age. They are focused on, ‘Here’s who I am at the moment, here are my motivations, here’s what I’ve found I like to do, here’s what I’d like to learn, and here are the opportunities. Which of these is the best match right now? And maybe a year from now I’ll switch because I’ll find something better.”
What Epstein gathers from the body of research and expert interviews is what he calls “at once simple and profound: we learn who we are only by living and not before.” That what reads like long-term planning and consistent achievement is simply a mindset of living as the process of learning. Returning to his interview with Hesselbein:
“As [has been] said regarding [Vincent] Van Gogh’s life, some “undefinable process of digestion” occurred as diverse experiences accumulated. ‘I was unaware that I was being prepared,’ [Hesselbein] told me. ‘I did not intend to become a leader, I just learned by doing what was needed at the time.’”
The story of Frances Hesselbein and the subversive puzzle of Dark Horses like her is one of several connections Epstein makes between lived experience and research.
In a world which increasingly demands and incentivizes specialization and hyperfocus at increasingly earlier ages and stages of life, the world also needs “people who start broad and embrace diverse experiences and perspectives while they progress. People with range.”
The ironic marketing of Range is its broad appeal. That it can be read through a variety of lenses, as both an educator and a parent, to use myself as an example; or as a coach looking to broaden the abilities of their athletes, or a business leader wanting to curate more effective teams and understand how to improve collective decision-making. The writing itself is succinct and accessible even as Epstein draws from, again, a tremendous range of disciplines in support of the thesis of the book: that in a world which increasingly demands and incentivizes specialization and hyperfocus at increasingly earlier ages and stages of life, the world also needs “people who start broad and embrace diverse experiences and perspectives while they progress. People with range.”
If dark horses like Frances Hesselbein are the exception, seeming to meander to success by cutting across many fields while never “falling behind”, then who made the rule?
Epstein began his research for Range in response to Malcolm Gladwell’s famous 10,000 hour rule: that “the story of success” isn’t hidden in biographical details or innate talent but is really the cumulative advantage of deliberate practice, that one becomes an expert by amassing their respective 10,000 hours. Epstein also responds to the “Battle Hymn of the Tiger Mom”, a 2011 book claiming that the secret to raising high achieving children was, in his words, “choose early, focus narrowly, never waver”.
By delaying deliberate practice or allowing children to deviate and lose focus, you set your child up for a deficit of time from which they may never recover. To set children up for high achievement in music, for example, start your child on the piano or violin as soon as they can walk, and at 1,000 hours of deliberate practice per year, they will be on track to achieve expertise by junior high when the stakes will be even higher for college acceptance to prestigious universities for which the competition will be as fierce as your tiger parenting style.
So the questions Epstein asks of tiger parents and other proponents of the 10,000 hours rule, what he calls “the cult of the head start”, seem simple: Do specialists actually get better with experience or not, and is narrow, deliberate practice the only way to achieve greatness in a given field?
The analogy for education systems couldn’t be clearer: given that we have limited time and resources, is it better to align them toward narrow, deliberate practice of specific content and skills that improve short-run success as measured by, say, higher test scores — an approach that, if supported, would justify standardization and authoritarian classroom practices — or is it better to spread those resources out across a range of experiences that may push students in different directions and toward less clearly measurable outcomes, but lead to long-term success as students leverage experience into innovation in the world outside of school, as was the case with Frances Hesselbein?
Relaying definitions from the field of psychology, Epstein describes the difference between so-called “kind” and “wicked” environments. “Kind” environments are those in which the rules and patterns are consistent and which feedback is accurate and immediate. Chess is the premiere “kind” learning environment where mastery comes from thousands of hours of deliberate practice, and where the highest achieving chessmasters are the best at recognizing and responding quickly to the familiar patterns unfolding across the board’s 64 squares. It’s also what makes artificial intelligence so good at the game: it’s a predictable closed system defined by a rigid set of rules and a finite amount of moves, and where there is a massive historical data set whose outcomes can be input and analyzed. It’s why computers have been beating human beings at chess since IBM’s Deep Blue AI beat Grandmaster Garry Kasparov in 1997.
Environments in which computers tend to perform poorly are so-called “wicked” environments where “the rules of the game are often unclear or incomplete, there may be or may not be repetitive patterns and they may not be obvious, and feedback is often delayed, inaccurate, or both.” “Wicked” problems are open-ended, involve the influence of many variables, where the rules change and which the familiar patterns of one context may fail when applied to another. “Wicked” environments are the reason autonomous vehicles have been more difficult to develop than chess playing computers, and why AI performance in medicine has been notoriously poor. After IBM’s Watson computer beat Ken Jennings at Jeopardy! In 2011, it was immediately co-opted to the study of cancer, where, in Epstein’s words, it “flopped”: “The difference between winning at Jeopardy! And curing all cancer is that we know the answer to Jeopardy questions. With cancer, we’re still working on posing the right questions in the first place.”
The kind domains of chess, golf, and classical music, Epstein argues, are poor models of most things human beings want to learn but are domains in which hyperspecialization is rewarded with success. He highlights the work of psychologist Barry Schwartz, who studied the ability of participants to discover solutions in a logic puzzle. One group was given a monetary reward for each solution, even if it was the same solution repeated over and over (think: the kind environment of chess, where a winning tactic gets reinforced and used in the next match). When a new group of participants was added to the study and asked to discover all of the possible ways of winning, every student in the new group successfully discovered all seventy possible solutions, while only one student in the previously rewarded group was able to do so.
Rewarding short-term success in the kind environment of the logic puzzle reinforced the use of familiar patterns, but past performance was not predictive of future success when the rules changed, in fact, past performance proved counterproductive: success with familiar patterns made you less likely to think in ways the new situation demanded and rewarded. As Epstein notes, the subtitle of Schwartz’s paper became: “How Not to Teach People to Discover Rules”.
But surely it has to be the case that the mastery of a complex game like chess reaps cognitive benefits across the board, so maybe it’s worth 10,000 hours of training if it improves overall intelligence and cognitive performance? Well not so much. It turns out that while like chess, memory training games, video games, and music instruction do support near transfer of skills, that is, where areas of performance have significant similarities and overlap — say between chess and checkers — however, there isn’t any indication that this specific training directly supports far transfer, that is, improved performance in areas with almost no overlap, between chess and dance for example. To quote a summary of this research from the Singapore Times:
“What all this shows is that it is unlikely chess has a significant impact on overall cognitive ability. So while it might sound like a quick win — that a game of chess can improve a broad range of skills — unfortunately, this is not the case...
The fact that skills learnt by training do not transfer across different domains seems to be universal in human cognition. In other words, you get better, at best, at what you train in — which may just sound like good, old-fashioned common sense.”
In fact, to turn back to music instruction and training, the evidence seems to support the notion that buying your toddler a violin and sticking to a rigid practice regimen — in the Tiger parenting mantra choose early, focus narrowly, and never waver — might even inhibit their future level of performance. For example, John Sloboda, a researcher in the psychology of music, studied British boarding school students and concluded that, “It seems very clear, that sheer amount of lesson or practice time is not a good indicator of exceptionality…the strong implication is that too many lessons at a young age may not be helpful.”
The study found that less skilled students stuck with only a single instrument and that every single student in the study who spent the most amount of time early in their development on structured lessons fell under the study’s definition of average, “not one was in the exceptional group”. However, what exceptional students had in common was a distribution of effort across different instruments — three seemed to be a key number — and a so-called sampling period: “often lightly structured with some lessons and a breadth of instruments and activities, followed only later by a narrowing of focus, increased structure, and an explosion of practice volume.”
This sampling period — a time of loosely structured or unstructured discovery and experimentation across a range of activities in a given area — seems vital, and leads what Epstein calls the “eventual elites” in their fields to draw from a range of abilities and experiences upon which to focus later technical practice.
Out of this a general pattern emerges: early focus leads to early plateaus and performance ceilings when familiar tools break down in “wicked” conditions or simply reap diminishing returns over time, while a sampling period of initial unfocused exploration lets you do more later and go further by building a more robust and responsive tool kit to wield against the changing conditions of the “wicked” world.
“What is the computer program going to reward? Is it going to reward some vapid drivel that happens to be structurally sound?”
While 10,000 hours of deliberate practice undoubtedly trains human beings to better recognize and respond to the particular patterns of a “kind” environment — to get better at the game the more you play it — is school a “kind” domain like chess that lends itself to the 10,000 hour approach? If it is, does our preparation for students in the “kind” world of school really prepare them for life outside its walls or does it prove counterproductive, or even maladaptive, when the rules change and those hours of training fail to transfer? And to what extent does this “kind”/”wicked” framing help us understand the divide in our approaches to education?
The notion of “wickedness” seems to confirm what we understand about the narrowness of writing instruction in schools, for example, in which students master the patterns and rules of, say, the 5 paragraph essay or an AP free response rubric as a proxy for authentic writing.
Writing is a wicked problem that in many cases is formed by convention but ultimately comes down to the questions a writer must ask about themselves, their content, and their audience. When school attempts to circumvent this essential interrogation, it reduces writing to a “kind” problem that author John Warner calls a “writing-related simulation”, which sounds a lot more like chess than authentic communication, and I’ll quote from Warner at length:
“The five-paragraph essay is more avatar for the problem than the problem itself. There’s nothing troubling about essays with five paragraphs, but the “five-paragraph essay” comes coupled with some very troubling things.
The primary problem is the practices which attach to the teaching of the five-paragraph essay, and the totalizing system of accountability which privileges the teaching of the five-paragraph essay. Prescriptive rules such as: a thesis must be the last sentence of the first paragraph, the last paragraph must start with “in conclusion” and restate the body paragraphs, and each paragraph should have between five and seven sentences do not help students learn the basic skills of structure and argument. They help them create what I call “writing-related simulations” which pass very basic muster on surface-level assessments, but don’t actually help students learn to make better arguments or think in the ways we expect in college.
Effective argumentation is about learning to make choices consistent with audience, purpose and message (the rhetorical situation). The way the five-paragraph essay is employed as prescriptive practice actually prevents students from practicing those far more vital and complicated skills.”
Training students in the “kind” domain of “writing-related simulations” for the purpose of schooling leads to long-term failure when the tools of the five paragraph essay or the AP rubric exceed their purpose in the truly wicked world of authentic writing and communication. These “writing-related simulations”, much like the chess games analyzed by IBM’s Deep Blue AI, are becoming more easily assessed and “graded” by computer scoring technologies, and to mixed reviews.
In a 2018 NPR piece, Massachusetts Department Of Education Deputy Commissioner Jeff Wulfson said there have been “huge advances in artificial intelligence in the last few years” adding, “I asked Alexa whether she thought we’d ever be able to use computers to reliably score tests, and she said absolutely.”
Teachers interviewed for the article, on the other hand, sounded skeptical. One English teacher is quoted saying, “The idea is bananas, as far as I’m concerned. An art form, a form of expression being evaluated by an algorithm is patently ridiculous.” Another asks “What about original ideas? Where is room for creativity of expression? A computer is going to miss all of that,” adding, “What is the computer program going to reward? Is it going to reward some vapid drivel that happens to be structurally sound?”
It turns out that robo-scoring computers love “writing-related simulations” and score them very favorably. One MIT researcher even invented an essay generator called the Basic Automatic BS Essay Language generator, or BABEL, designed to exploit AI algorithms and give perfect scores to perfect nonsense. See if you can understand the main idea of this excerpt from a 500-word BABEL essay based on a GRE practice prompt:
“History by mimic has not, and presumably never will be precipitously but blithely ensconced. Society will always encompass imaginativeness; many of scrutinizations but a few for an amanuensis. The perjured imaginativeness lies in the area of theory of knowledge but also the field of literature. Instead of enthralling the analysis, grounds constitutes both a disparaging quip and a diligent explanation.”
Were you able to make sense of the piece? Don’t worry, neither could the researcher who designed the generator: “It makes absolutely no sense, There is no meaning. It’s not real writing.”
When fed to the robo-grader, this not-real-writing of course earned a perfect 6 out of 6 on the GRE scale, and according to their criteria represents “a cogent, well-articulated analysis of the issue and conveys meaning skillfully.”
Writing is communication. When I think of great speeches, essays, and works of world literature, accessing their meaning depends on understanding historical context and culture. Writers demand that readers share a similar vernacular and set of idioms, analogies, and symbolism — or they induct the reader into the club of Shared Language. Communication fails when this complicated handshake between writer and audience breaks down. And what are we communicating to the robo-graders? Writing for the audience of an algorithm necessarily reduces the wicked task of writing to a chess game; where the goal is not communion, but rather competition; where communication fails and BABEL prevails.
What about the rest of school? I think the “kind”/”wicked” framing might actually be one of the more useful frameworks for understanding the divide in our approaches to education. Believe it or not, there are defenders of the robo-graders. The GRE’s automatic scoring software is made by Education Testing Service or ETS. A senior research scientist at ETS is quoted in the NPR piece saying, “If someone is smart enough to pay attention to all the things that an automated system pays attention to, and to incorporate them in their writing, that’s no longer gaming, that’s good writing. So you kind of do want to give them a good grade.”
As I’ve covered on our podcast, Re:Teaching, PISA, the Programme for International Student Assessment, is an international standardized assessment that functions in much the same way as the GRE in both supposing to measure the quality of education and the ranking and sorting of education systems. The same logic of the research scientist applies, if education is a kind game where feedback is accurate and where there is a direct relationship between inputs and outputs, then if systems are smart enough to pay attention to all the things that PISA pays attention to, and to incorporate them into their schooling, that’s no longer gaming, that’s good learning. So you kind of do want to give them a good score.
…And very quickly, entire education systems are faced with bending toward the values of PISA — with all of the side-effects — or risk being left behind in the rankings.
Systems of efficiency, control, and accountability are inexorably linked. Turning school into a “kind” game in which those 10,000 hours matter a great deal makes us feel better about controlling their use and content. It would also demand holding accountable those who would “misuse” instructional time to less focused ends. And of course there are institutions whose political and financial interests are in maintaining that control. But ultimately, if our values as educators align with the mission that we need to create a generation of changemakers (or call them solutionaries) — people who actually understand the world around them, want to make a difference, and feel empowered to do so (and are supported along the way) — we not only need to rethink what “accountability” means within the education system but what kind system we actually have. If our assumptions about the “kindness” of school are incorrect and our response to the feedback of standardized testing and high stakes assessment have mislead us — when we train our students to outsmart the robo-grader — we risk misrepresenting to our students the assumptions and measures of the very real and very “wicked” world.
To quote Epstein in the closing pages of the book,
“Compare yourself to yourself yesterday, not to younger people who aren’t you. Everyone progresses at a different rate, so don’t let anyone else make you feel behind. You probably don’t even know where exactly you’re going, so feeling behind doesn’t help…Approach your own personal voyage and projects like Michaelangelo approached a block of marble, willing to learn and adjust as you go, and even to abandon a previous goal and change directions entirely should the need arise…Even when you move on from an area of work or an entire domain, that experience is not wasted.”
After all, there will always be a computer waiting to outsmart the next Grandmaster of chess, but there will only ever be one irreplaceable Frances Hesselbein.