Fail Better — ล้มอีก ล้มให้ดีกว่าเดิมFail Better — Beckett, Edison, and Confucius on Perseverance

There is a hardback notebook lying next to the oven. The cover is smudged with butter, the spine swollen from steam, and in the top corner, in marker pen, someone has written: Things that don't work yet. Imagine a small rented kitchen on the edge of the city, half past ten at night. A woman opens the oven and finds that tray forty-one of this month's macarons has cracked again. The LINE group where her orders come in has been silent for three days. The next kitchen payment is due at the end of the month. A faint smell of burnt butter hangs in the air like a question nobody wants to ask out loud.

What she does next is the part worth watching. She does not tip the tray straight into the bin. She photographs it, opens the notebook, and writes one short entry: "No. 41 — cracked like no. 40? No. Finer cracks. Oven probably still too hot. Tomorrow: down five degrees, and rest the tray in front of the fan before it goes in." Then she washes the tray, turns off the lights, and goes home.

In the previous essay in this series (on craft) we talked about doing the work in front of you with your whole hand. But there is a truth that essay left unsaid: every craft is built on a heap of attempts that didn't work — and not every failure teaches. Some people fail a hundred times and collect a hundred pieces of knowledge; some fail a hundred times and collect the same bruise a hundred times over. This essay is about perseverance — not in the gritted-teeth sense the posters like to sell, but in a much narrower and more useful one: what exactly separates trying again smarter from merely trying again?

Fail again, fail better — darker than the poster

"Ever tried. Ever failed. No matter. Try again. Fail again. Fail better."

Samuel Beckett, Worstward Ho (1983)

Someone has probably sent you this line. It hangs on startup meeting-room walls and turns up in onboarding decks; the tennis champion Stanislas Wawrinka has the whole sentence tattooed along his left forearm; Richard Branson has quoted it; and when the moderator of a Dublin technology conference wanted a stirring close in front of Elon Musk, this was the line he reached for.^[7] In the poster version it is a promise: go ahead and fail — failure is a rung on the ladder to success.

The man who wrote it promised nothing of the kind. Samuel Beckett was an Irish writer and Nobel laureate, the author of Waiting for Godot, a play in which two men wait for someone who never comes. Worstward Ho is one of the last prose works of his life, written in 1983, only a few dozen pages long, the language pared to the bone. The title mocks the old sailors' cry "Westward Ho!" — but Beckett is not heading west. He is heading for the worst. The whole book is an attempt to keep speaking where language cannot reach, to keep going where there is nowhere to go.^[1]

And if you read past the point where the poster stops, the text continues: "Try again. Fail again. Better again. Or better worse. Fail worse again. Still worse again. Till sick for good. Throw up for good."^[7] These are not the words of a touchline coach. The writer Mark O'Connell has observed, with some relish, how neatly the startup world misread this polished shard: Beckett cared about failure as a permanent condition of being human — the necessary defeat of every endeavour, of communication, of language itself — not as a stage in anyone's progress toward wealth.

So why keep the sentence at all? Because the dark version consoles in a way the poster version cannot. Beckett's line releases us from the heaviest burden of trying: the obligation to win. If failure is the normal condition of every meaningful kind of work — art, science, raising a child, opening a small bakery — then the question changes from "when will I finally stop failing?" to "was this failure of mine finer than the last one?" Beckett's "better" does not measure your distance from victory. It measures the quality of the fall — and quality, unlike victory, is genuinely in your hands.

Several thousand things that won't work

"Results! Why, man, I have gotten a lot of results! I know several thousand things that won't work."

Thomas Edison, as recalled by Walter S. Mallory, in Dyer & Martin, Edison: His Life and Inventions (1910), ch. XXIV

You probably know this quote in a different body: "I have not failed. I've just found 10,000 ways that won't work." That version is sharper, more quotable, better suited to a poster — and there is no evidence Edison ever said it. The Quote Investigator website traced its lineage and found what happens to almost every famous quotation: the numbers swell with each retelling. The 1910 account says "several thousand"; a 1946 retelling makes it 999; by the 1980s it has become a round ten thousand, with the phrase "I have not failed" — absent from the original — bolted on for good measure.^[8] To this day, quotation databases such as Wikiquote file the ten-thousand version under "disputed."

The real thing sits in a 1910 biography written while Edison was alive and cooperating. The teller is Walter Mallory, a close associate of many years. Summoned to the laboratory in the middle of the battery work, he found Edison at a long bench crowded with hundreds of little test cells, and learned that the past five months had produced more than nine thousand experiments without a single one that came near an answer. Pity got the better of Mallory's judgment, and he blurted out: wasn't it a shame that so tremendous an amount of work had produced no results? Edison "turned on me like a flash, and with a smile," and gave the reply above.^[2]

Two small details in that scene matter more than the sentence itself. The first is the smile — this was not a prepared aphorism but a live answer, from a man in the middle of failing, to a friend looking at him with pity. The second is the bench. There was no free-floating "effort" on that bench; there were hundreds of test cells, each one a designed question, distinct from its neighbours, with a corps of chemists assembling them and logging the outcomes, every trial numbered in a notebook. Edison's failures became "results" not because he thought positively, but because each one was written down: what was tried, what was changed, and exactly how it broke.

This is the dividing line we have been hunting since the first paragraph. Nine thousand failures with a notebook are nine thousand pieces of knowledge. Nine thousand failures without one are a single failure — replayed nine thousand times.

The last basket of earth

"If there want but one basket of earth to complete the work, and I stop, the stopping is my own work… Though but one basketful is thrown at a time, the advancing with it is my own going forward."
「譬如為山，未成一簣，止，吾止也；譬如平地，雖覆一簣，進，吾往也。」

Confucius, Analects 9.19 (Lau's numbering; Legge 9.18), tr. James Legge

Twenty-five centuries before Beckett and Edison, Confucius said the same thing with the plainest image imaginable: one basket of earth. He was teaching his students about self-cultivation — work that lasts a lifetime and offers no finish line. The first picture is a man building a mound who quits one basketful short of the top. The second is a man levelling a plain who has tipped only a single basketful and keeps walking. Notice that both pictures end on the same word: myself.^[3]

Two things in this passage are usually read past. First, it does not forbid stopping. It says only that stopping has an owner. Confucius does not threaten heaven's punishment for the one who sets the basket down; he simply refuses to let us blame heaven — or luck, or the market, or the algorithm. If I stop, it is I who stop. And the knife cuts the other way too: the credit for the one basketful just tipped is entirely mine as well. The responsibility in this sentence runs in both directions. It confiscates our excuses, and hands us back our agency in the same motion.

Second, the unit of measure is startlingly small. Confucius does not talk about the mountain — no mountain has ever fit in anyone's hands. He talks about the basket, because the basket is the only thing a pair of hands can actually hold. Edison counted his perseverance one test cell at a time; Beckett counted it one sentence at a time; the woman in the rented kitchen counts it one tray at a time. Perseverance, as Confucius sees it, is not an epic. It is the arithmetic of small countable units — plus an unblinking admission of whose hands are on the basket.

And a few passages away in the same book, Confucius speaks of sprouts that spring up but never flower, and of flowers that never set seed.^[3] It is as if he wants us to know he has seen both sides of the truth: perseverance guarantees nothing — the rain is not in our basket. Sometimes setting the basket down is even the right decision. Confucius asks only one thing of us: do not call that setting-down fate. Call it by its real name, which is our choice.

Ten years of a battery that wasn't good enough

Edison's story is usually told in jump cuts: ten thousand failures, the light bulb glows, the end. The real thing is longer, darker, and teaches a different lesson than the poster summary. The story worth telling is not the light bulb at all. It is the battery — and it begins with a ruin.

In the late 1890s Edison poured more than two million dollars of his own money — a fortune in that era — into an iron-ore concentrating plant in New Jersey. The technology worked. Nature played a joke anyway: rich, cheaply mined ore was discovered elsewhere, and his plant lost its economic reason to exist almost overnight. Mallory, who was in the venture with him, recorded the closing scene. Riding home one day, the two men saw the price of the General Electric shares Edison had sold off to fund the project — worth over four million dollars had he kept them. Edison went quiet for a moment, tugged at his right eyebrow the way he did when thinking hard, and then his face lit up: "Well, it's all gone, but we had a hell of a good time spending it."^[2]

What he did with the wreckage was choose a new problem immediately. In 1900 he resolved to build a new kind of storage battery, one that would use no sulphuric acid, and set to work on the schedule Mallory documented: into the laboratory at seven in the morning; the carriage came at midnight and frequently waited until two or three, sometimes returning empty because he had decided to work through until dawn. Five months of that, seven days a week, produced over nine thousand experiments — and nothing that worked. That is the scene in which Mallory blurted out his question and got back the smile and the immortal sentence. A few weeks later, with the count past ten thousand, Edison found "the missing link in the combination sought for."^[2]

If the story ended there, it would just be another poster. The real story was only starting. The first battery went on sale and sold faster than the factory could make it; the newspapers were full of praise. Then the field reports began to blink: some cells — not all, just a scattering — were losing capacity, and nobody could say why. Edison chased the fault down to the graphite in the positive electrode. The fork in the road was stark. Most customers were still satisfied; some were literally queuing with orders in hand, asking to buy more. He could keep selling at a handsome profit — or admit that a product the whole market called a success had failed his own standard.

He stopped the sales. He shut down his own production line, under pressure from every direction, and went back into the laboratory for five more years — a stretch one of his lab men called "record-breaking experiments… I might almost say heart-breaking, too." The answer, when it came, was nickel flake thinner than paper in place of the graphite, and slender tubes in place of flat pockets. The factory started up again in the early summer of 1909. One assistant recalled that they numbered the experiments from 1 to 10,000 and then wrapped around and started again, several times over — "it was not far from fifty thousand" in all, filling more than one hundred and fifty notebooks.^[2]

Notice what actually makes this a story about perseverance. Not the fifty thousand — a bare number is, at best, stubbornness. What makes it perseverance is its shape: every experiment designed to differ from the one before; every outcome logged where it could be traced; and, when the evidence said the current version could not go on, the nerve to declare a failure in something the whole market still called a success. That is what "fail better" looks like off the poster. And one fact deserves honest mention beside it: Edison had capital, a corps of chemists, and a cushion under his falls that most people do not have. The lesson we can actually take home is not the size of his effort. It is the way the effort was shaped.

What the research says — and doesn't

The nearest thing to perseverance in contemporary psychology is called grit. In 2007 Angela Duckworth and her colleagues defined it as perseverance and passion for long-term goals, and found that it helped predict which cadets would survive West Point's brutal summer training and which children would go deep in the National Spelling Bee.^[4] The finding launched bestsellers and grit-building curricula around the world. But the number in the original paper that rarely gets retold is this: on average, grit accounted for about 4 percent of the variance in success — and it correlated strongly with conscientiousness, a personality trait psychologists had been measuring for decades.

Ten years later, a large meta-analysis by Credé, Tynan and Harms pooled 88 samples and 66,807 people, and the picture cooled considerably: grit correlates with academic performance at only about ρ = .18, with retention at about ρ = .12–.18, and overlaps with conscientiousness at ρ = .84 — so high that the authors asked whether this was a new trait at all, or an old shirt with a new price tag. The half of the construct that still earns its keep is "perseverance of effort," which clearly outpredicts the other half, "consistency of interest"; and the abstract's blunt conclusion is that programmes designed to build grit "may only have weak effects" on performance and success.^[5] If you have ever felt guilty because you finished the grit book and your life did not change, the evidence is on your side, not the advertising's.

The second line of research is more remarkable, because the people who corrected the theory were the people who built it. In 1967 Martin Seligman and Steven Maier found that dogs subjected to electric shocks they could not control would later lie down and accept their fate even when an escape route stood open. The phenomenon was named learned helplessness, and it served as a pillar of modern psychology for half a century. Then, in 2016, the two men returned with a sentence very few scholars ever dare to write about their own work: the original theory "got it backward." Fifty years of neuroscience indicate that passivity is not learned at all — it is the mammalian brain's default response to prolonged, uncontrollable hardship. What has to be learned is control: when the prefrontal cortex detects that our actions genuinely change outcomes, it reaches down and quiets the circuits of surrender.^[6]

For anyone in the middle of failing repeatedly, the implication runs deep. The urge to raise the white flag is not a moral defect, and it is not proof that you "lack grit" — it is the factory setting of every one of us. And the only argument against it the evidence supports is the small, repeated experience of control: proof, delivered to your own brain, that your hands still move the outcome — Confucius's one basket at a time, Edison's one variable at a time. The customary caution of this series applies: the 2016 work is built mainly on animal experiments, and extending it to the tangle of a human life calls for humility. Nor does any study above prove that you must never retreat — research describes the averages of many lives, and you have exactly one, with a context all its own.

My own view is that the most evidence-honest translation of "fail better" is "fail with data." I do not believe perseverance is a gift that some people have and others lack — the evidence above suggests it is closer to a procedure plus a set of conditions: one notebook, one variable at a time, wins small enough to be had but real enough to convince your brain it still controls something, and a cushion thick enough that a single fall is not a catastrophe. And I believe that deliberate quitting — retiring a version, not a dream — is perseverance's sibling skill, the one the poster culture never teaches anyone.

Try it this week

The equipment list has one item: a blank notebook. (A notes file on your phone will do, but paper makes the ritual slightly more sacred.) Name it the way the woman in the rented kitchen named hers: Things that don't work yet.

Ask yourself again on Sunday

Back to the rented kitchen at half past ten. Tray forty-one is in the bin, but the line she wrote is still in the notebook, and tomorrow tray forty-two goes into an oven five degrees cooler. Nobody on earth can promise her which tray will be the one — Beckett promises nothing except a finer quality of falling; Edison himself had to wait for attempt ten thousand; Confucius says it plainest of all: the mountain was never in anyone's hands. The only thing truly in hers is this — that tray forty-two must not be tray forty-one replayed. And if the day comes when she chooses to close the shop, may it be a basket set down with her eyes open, her own decision — not a decision made for her by exhaustion on the darkest night of the week.

In the next essay, we turn to something that sounds like the opposite of fifty thousand experiments but is in fact their precondition: cutting away everything that isn't essential. Because often the reason we keep failing the same way is not a shortage of perseverance — it is that we are carrying too many variables to see which one is the culprit.