The Six Main Stories, As Identified by a Computer


“My prettiest contribution to my culture,” the writer Kurt Vonnegut mused in his 1981 autobiography Palm Sunday, “was a master’s thesis in anthropology which was rejected by the University of Chicago a long time ago.”

By then, he said, the thesis had long since vanished. (“It was rejected because it was so simple and looked like too much fun,” Vonnegut explained.) But he continued to carry the idea with him for many years after that, and spoke publicly about it more than once. It was, essentially, this: “There is no reason why the simple shapes of stories can’t be fed into computers. They are beautiful shapes.”

That explanation comes from a lecture he gave, and which you can still watch on YouTube, that involves Vonnegut mapping the narrative arc of popular storylines along a simple graph. The X-axis represents the chronology of the story, from beginning to end, while the Y-axis represents the experience of the protagonist, on a spectrum of ill fortune to good fortune. “This is an exercise in relativity, really,” Vonnegut explains. “The shape of the curve is what matters.”

The most interesting shape to him, it turned out, was the one that reflected the tale of Cinderella, of all stories. Vonnegut visualizes its arc as a staircase-like climb in good fortune representing the arrival of Cinderella’s fairy godmother, leading all the way to a high point at the ball, followed by a sudden plummet back to ill fortune at the stroke of midnight. Before too long, though, the Cinderella graph is marked by a sharp leap back to good fortune, what with the whole business of (spoiler alert) the glass slipper fitting and the happily ever after.

This may not seem like anything special, Vonnegut says—his actual words are, “it certainly looks like trash”—until he notices another well known story that shares this shape. “Those steps at the beginning look like the creation myth of virtually every society on earth. And then I saw that the stroke of midnight looked exactly like the unique creation myth in the Old Testament.” Cinderella’s curfew was, if you look at it on Vonnegut’s chart, a mirror-image downfall to Adam and Eve’s ejection from the Garden of Eden. “And then I saw the rise to bliss at the end was identical with the expectation of redemption as expressed in primitive Christianity. The tales were identical.”

Vonnegut, in his ever charming way, was quite pleased with himself for making this connection. And 35 years later, his idea had resonated enough with a group of mathematicians and computer scientists that they decided to build an experiment around it. Vonnegut had mapped stories by hand, but in 2016, with sophisticated computing power, natural language processing, and reams of digitized text, it’s possible to map the narrative patterns in a huge corpus of literature. It’s also possible to ask a computer to identify the shapes of stories for you.

That’s what a group of researchers, from the University of Vermont and the University of Adelaide, set out to do. They collected computer-generated story arcs for nearly 2,000 works of fiction, classifying each into one of six core types of narratives (based on what happens to the protagonist):

1. Rags to Riches (rise)

2. Riches to Rags (fall)

3. Man in a Hole (fall then rise)

4. Icarus (rise then fall)

5. Cinderella (rise then fall then rise)

           6. Oedipus (fall then rise then fall)

Their focus was on the emotional trajectory of a story, not merely its plot. They also analyzed which emotional structure writers used most, and how that contrasted with the ones readers liked best, then published a preprint paper of their findings on the scholarship website <a href="" rel="nofollow"></a>. More on that in a minute.

First, the researchers had to find a workable dataset. Using a collection of fiction from the digital library Project Gutenberg, they selected 1,737 English-language works of fiction between 10,000 and 200,000 words long.

Then, they ran their dataset through a sentiment analysis to generate an emotional arc for each work. “We’re not imposing a set of shapes,” said Andy Reagan, a Ph.D. candidate in mathematics at the University of Vermont and the lead author of the paper. “Rather: the math and machine learning have identified them.”

A Computer Tried (and Failed) to Write This Article

They did this by training the machine to take all the words of the book, section by section, and measure the average happiness of a given bag of words based on how an individual word scored. The researchers assigned individual happiness scores to more than 10,000 frequently-used words by crowdsourcing the effort on the website Mechanical Turk. This portion of the research is fascinating in and of itself: The 10 words that people ranked as happiest were laughter, happiness, love, happy, laughed, laugh, laughing, excellent, laughs, and joy. The 10 words that people ranked as least happy were terrorist, suicide, rape, terrorism, murder, death, cancer, killed, kill, and die. (You can see how all the words ranked by visiting this site.)

There are several theories that say every story known to man can be reduced to one of just a handful of archetypes—a quest, overcoming the monster, rebirth, to name a few—but there’s no consensus on what those stories are. In this case, researchers picked six from a mix of popular lists based on what shapes the computer identified most. And though the researchers were focused on a book’s emotional arc—not the structure of its plot, per se—they found overlap in how plot points reflected emotional highs and lows as measured by the sentiment analysis.

While the plot of Harry Potter and the Deathly Hallows for instance, is “nested and complicated,” they wrote, “the emotional arc associated with each sub-narrative is clearly visible.” (That said, emotional moments discussed briefly—the first kiss between Harry and Ginny, let’s say—didn’t register.)

All in all, “Rags to Riches” stories represented about one-fifth of all the works analyzed. This isn’t surprising. It’s easy to think of examples of such tales in classic literature. The canons of Charles Dickens, Edith Wharton, and Jane Austen are arguably defined by them.

“The ‘Rags to Riches’ emotional arc embodies a story that we all love to believe in, widely popular in the American dream itself,” Reagan said. “It’s a story of hope and fairness, where regardless of beginning in bad times, with effort things will get better and eventually result in good fortune.”

In this case, the prototypical example, according to the researchers, is Lewis Carroll’s Alice’s Adventures Under Ground—which would later be published as Alice’s Adventures in Wonderland. An 1890 novel by the writer Olive Schreiner, Dreams, was another clear match for the “Rags to Riches” model. For both stories, the computer found a near-identical match to “Rags to Riches” with few if any connections to other kinds of emotional arcs. Here’s how the top 20 stories that fit the “Rags to Riches” mode appear on a graph in their paper:

“Rags to Riches” may be popular among writers, but it isn’t necessarily the emotional arc that readers reach for most. The categories that include the greatest total number of books are not the most popular, the researchers found. They examined total downloads for all books from Project Gutenberg, then divvied them up by mode. Measured this way, “Rags to Riches” is eclipsed by  “Oedipus”, “Man in a Hole” and, perhaps not surprisingly, “Cinderella,” all of which were more popular. Reagan told me he and his colleagues now plan to analyze how different arcs are sequenced together in a single story, like in the Harry Potter example above.

Eventually, he says, this research could help scientists train machines to reverse-engineer what they learn about story trajectory to generate their own compelling original works. Already, there are competitions for story-writing bots. (Incidentally, I attempted a similar experiment and it didn’t exactly go as planned.)

“This is an active area of research,” Reagan says, “and there are a lot of hard problems yet to be solved. In addition to the plot, structure, and emotional arc, to write great stories, a computer will need to create characters and dialogue that are compelling and meaningful.”

Vonnegut, of course, always made it sound easy. Consider how he describes the “Man in a Hole” narrative, which is characterized by M.R. James’s Ghost Stories of an Antiquary—or pretty much any 22-minute sitcom:  “Somebody gets in to trouble, gets out of it again,” Vonnegut once said in a lecture. “People love that story. They never get sick of it.”

Read the whole story
1 day ago
16 days ago
Share this story

The web’s best hidden gems


I asked readers to tell me what were the best web sites that they knew about that most people have never heard of. Normally, each of these sites would be worth a blog post on its own, either a longer or shorter one. But this week, we’re…

A few of the sites readers suggested are actually (in my experience) quite well-known. However, if you’ve never been to Daring Fireball,, The Millions, Hacker News, Every Frame A Painting, DuckDuckGo, LMGTFY, xkcd, Nautilus, The Internet Archive, Google Reverse Image Search, or, yes,, you should remedy that as soon as you can.

The rest of the suggestions are really quite something. I’d never heard of most of them. A few others I’ve loved and shared as a secret handshake between friends. In lieu of lengthy explanations, I’m just going to link them all and let you explore.

(This is what most web pages used to be: just long lists of links. We were such dorks back then. But maybe we were on to something.)

Note: The order is not a ranking, but simply an easy way I can count how many links there are.

Also Note: No money changed hands to feature any of these sites on, although in retrospect, that would have been a good idea..

Final Note: If the site required a login, or I couldn’t figure out what it was doing or why it would be interesting to anyone, I didn’t include it. Sorry.

  1. Weird Fiction Review: “an ongoing exploration into all facets of the weird, from the classics to the next generation of weird writers and international weird.”
  2. Pink Trombone: You kinda have to play with this one to get it.
  3. Internet Hockey Database: Statistics, Logos, and Trading Cards.
  4. diamond geezer: A lovely blog about trains and architecture in Great Britain.
  5. MLB Uniforms worn on April 19, 2017: You can do this for any date you want.
  6. Nepali: A Beginner’s Primer Conversation and Grammar
  7. Swear Trek
  8. Fuck You, Broccoli: “An in-depth exploration of vegetables and other so-called healthy foods.”
  9. Spacetrawler: A science-fiction webcomic and comics blog.
  10. Extra Ordinary: Another webcomic, about a little girl who has sweetly surreal adventures.
  11. WPC Probabilistic Winter Precipitation Guidance: Tracks snowfall.
  12. The Drawfee Channel: A video blog about drawing with digital tools.
  13. Is it Christmas?: Current answer: NO
  14. BoardGameGeek | Gaming Unplugged Since 2000
  15. Free online Dictionary of English Pronunciation - How to Pronounce English words
  16. Instant No Button! Unhappy Darth Vader on demand.
  17. Seismic Monitor: Recent earthquakes on a world map and much more.
  18. Autostraddle: News, Entertainment, Opinion, Community and Girl-on-Girl Culture. This site is great.
  19. I… I don’t understand this thing I’m seeing.
  20. WWWF Grudge Match - Where useless knowledge breeds champions: If you’ve ever wondered, “What if Boris Yeltsin and Ted Kennedy had a drinking contest?” this site is for you.
  21. The Griddle: A puzzle blog with free puzzles.
  22. Badger Badger! The Original Dancing Badgers!
  23. Futility Closet - An idler’s miscellany of compendious amusements
  24. Skyline - YouTube: “Esports analyst and coach focusing on the game Overwatch.” Two people suggested this.
  25. The Vintagent: A vintage motorcycles blog.
  26. CDC WONDER: The CDC’s online databases. Get as much government data as you can, while you can.
  27. InfiniteLooper: Endless loops of YouTube videos.
  28. The Morning News: A real blog. I love it.
  29. Deadspin’s The Stacks: Old sports stories from out-of-print magazines republished with the author’s permission. I can’t believe I hadn’t been reading this already.
  30. street dog millionaires: Two dogs, one from India, one from Africa.
  31. FL@33 presents What do animals sound like in different countries? (Every language is different!)
  32. - Pat Metheny on Kenny G: Jazz guitarist Pat Metheny takes Kenny G to the woodshed.
  33. McMansion Hell
  34. Scroll Down to Riker: Does what it says on the tin.
  35. Flickr: The Commons
  36. NYPL Digital Collections: This is amazing, and it’s all free to play with.
  37. The Library of Babel
  38. Bowiebranchia: “Nudibranchia or other opisthobranchia compared to the various looks of David Bowie.” Surprisingly compelling!
  39. Play Later: Browse newly released albums and save them for later. Mostly a handy discovery frontend for Spotify.
  40. Simulation of vintage computers and game systems in the browser.
  41. Political status of Western Sahara: So are we going to have a new country in North Africa or what?
  42. Antipope - Charlie’s Diary: Science fiction writer Charlie Stross’s blog.
  43. Interviews with interesting people by journalist/investor Om Malik.
  44. Deuce of Clubs: A Demonstrated Aptitude for Reasonable Mayhem: A site by Godfrey (“Doc”) Daniels, of Mojave Phone Booth fame.
  45. The Well’s annual conversation with author Bruce Sterling.
  46. Beyond the Frame: visual essays about TV and movies by Luis Azevedo.
  47. Artsy Engineering: An information network and open-source software for people in the art world.
  48. Hart Island Project Stories of unclaimed and unidentified people buried in mass graves in an island near New York city beginning in the 19th century.
  49. Damn Interesting: a science, history, and psychology blog.
  50. What’s Noka Worth? (Part 1) — DallasFood. A takedown of a hot, expensive artisanal chocolate maker… over a decade ago.
  51. Mexican Table Salsas: The eGullet culinary institute’s discussion of traditional salsas, and how to make them.
  52. Histography - Timeline of History: A mighty timeline view of every event in Wikipedia. That’s a lot of dots.
  53. People Reluctant to Kill for an Abstraction, a movement: a short story/manifesto by George Saunders.
  54. Langscape: an interactive map of all world languages, put together by researchers at the University of Maryland.
  55. in Bb 2.0 - a collaborative music/spoken word project: twenty musicians, all playing in the same key — you get to mix and match.
  56. Paper Planes: Use your phone to catch and throw paper planes to and from people all over the world.

There! That ought to be enough magic and wonder to last you through the weekend. Thank you, as always, to the readers who submitted links.

Tags: best of the web
Read the whole story
1 day ago
6 days ago
Nashville, TN
Share this story

Court ruling on Zuma's nuclear deal is a marker of South Africa's political health

1 Comment
Celebrations outside the Western Cape High Court after it ruled against the South African government's proposed nuclear deal. Nic Bothma/EPA

The South African government’s plan to bulldoze through a nuclear energy deal has been dealt what might be a fatal blow by the Cape Town High court which has declared the plan invalid. It found that the government had not followed due process in making the decision to pursue a nuclear power option, as well as in other critical areas.

The court’s decision has put paid to President Jacob Zuma’s hopes of clinching the nuclear build programme before leaving office in 2019 if he completes his term.

The case was brought to court by Earthlife Africa and the Southern Africa Faith-Communities’ Environmental Institute. The two NGOs were challenging the way in which the state determined the country’s nuclear power needs. The plan would have seen South Africa purchasing 9,600 megawatts of extra nuclear power.

The judge, Lee Bozalek, ruled the government’s action unconstitutional and found that five decisions it had taken were illegal. These included the government’s decision to go ahead with the nuclear build and the fact that it had handed over the procurement process to the state utility Eskom. The court also ruled that Eskom’s request for information from nuclear vendors, a step taken to prepare the procurement, which ended on 28 April 2017 was invalid.

If it still wants to pursue the nuclear deal the government will have to start all over again. To do so legally it would have to open up the process to detailed public scrutiny. The country’s electricity regulator would have to have a series of public hearings before endorsing what would be its highest ever spend on infrastructure. And any international agreements would have to be scrutinised by parliament.

All this will take time – something Zuma doesn’t have. And it’s unlikely that his successors will be as eager to champion a new deal as he has been. Meanwhile the facts about the deal will become public. This will undoubtedly demonstrate two of the biggest criticisms of the deal to be true: that the country can’t afford it, and that it’s energy needs have shrunk, making the vast investment redundant.

The court’s ruling has turned the nuclear procurement issue into one of the key markers of South Africa’s political health. It’s not yet clear whether the South African government will appeal the Western Cape High Court’s decision, or comply with the judgement. A third option is that Zuma simply ignores the courts and continues to pursue the deal.

Demand and affordability

South Africa currently has more than enough electricity to meet its needs. This wasn’t the case about five years ago when widespread outages hit the country. Since then new electricity generation capacity has been added, through the the rapid roll out of renewables, and the opening up of two new giant coal burning plants. Consumption, particularly by industry, has steadily declined due to faltering economic growth and higher electricity prices. Demand has dropped so much that Eskom plans to close five coal burning power stations.

The argument that the country needs another 9,600 megawatts was identified in documents that produced in 2011. These are now widely acknowledged as being badly out of date. Recent studies by the University of Cape Town’s Energy Research Centre have shown that the country doesn’t need to consider nuclear for another 20 years.

A number of studies have also shot holes in the government’s argument that the country can afford the proposed nuclear build. The Council for Scientific and Industrial Research has developed models showing that new nuclear is likely to be much more expensive than coal or renewables. The price ticket for nuclear – which some estimates put at more than R1 trillion – doesn’t take into account the costs of operation, fuel, insurance, emergency planning or the regulation or decontamination at the end of the life of the reactors.

It would also impose a financial burden on the country’s fiscus which it can ill afford particularly now that the economy has been rated at junk status.

Ulterior motives

So why is Zuma still pushing for the deal to go ahead? One source of pressure might be the Russians. South Africa’s former energy minister, Tina Joemat-Pettersson, had been instructed to signed a deal with the Russian utility, Rosatom to build the reactors. South Africa has also already signed nuclear power cooperation agreements with other countries like the US and South Korea, which the court has declared void.

A more likely reason for Zuma’s zeal is the involvement of the Gupta family with whom he has close ties. The family’s web of interests around the nuclear deal are complex.

What is known is that the Gupta family controls South Africa’s only dedicated uranium mine. The family has developed close relationships with key individuals at Eskom. In November last year the country’s then Public Protector pointed to overlapping directorships between Gupta-owned companies and Eskom.

The report also suggested that Brian Molefe, Eskom’s CEO, had a close relationship with the family. These revelations led to his resignation shortly after the report was published.

Another strand in the complex web is the fact that Zuma’s son Duduzane is a business partner of the Guptas while other relatives are directly employed by them.

Despite his determination, Zuma has become increasingly isolated in his quest for nuclear procurement. The African National Congress is clearly divided on the issue. This is evident from the fact that Zuma has resorted to reshuffling his cabinet to make way for more compliant ministers without reference to party officials as would be the norm.

The private sector has also come out against the idea while the list of civil society organisations opposed to nuclear expansion goes well beyond the environmental lobby and includes a broad spectrum of foundations, faith communities, human rights campaigners and defenders of the country’s constitution.

High stakes

The nuclear judgement in Cape Town indicates that South Africa’s legal system has not yet been “captured” by private interests.

The key question is whether Zuma and Eskom will accede to the verdict, or whether they challenge it while continuing to ignore the rule of law. Not only would this subvert the country’s constitution and its democratic form of government, it would also deny the constitutional right to popular participation in energy democracy.

The stakes are high – for the country as well as for the president. Will he continue to treat the country’s energy future with impunity? Or will this judgement symbolise the rollback of the democratic dispensation envisaged by the authors of the country’s constitution?

The Conversation

David Fig has had a long association with Earthlife Africa, and serves on the steering committee of the African Uranium Alliance.

Read the whole story
1 day ago
Although I am pro-nuclear from a technology standpoint, this political side of it is pretty nasty. Good on the South Africans for putting a stop to it
Share this story

This Is What Musical Notes Actually Look Like - Facts So Romantic

1 Share

A few months ago, I sat poolside with friends in Palm Springs. Amid the quiet desert sublime, we reminisced about all the live music we’ve experienced over the years, just about every big and small act since the mid-80s: Prince, David Bowie, Guns ‘n Roses, Bruce Springsteen, and the Yeah Yeahs Yeahs among the them. But My Bloody Valentine, we decided, was the ultimate experience. The group is reportedly among the top-10 loudest bands of all time, louder than Black Sabbath or Metallica, but still quieter than the Who. My Bloody Valentine’s beautiful, shoegaze noise rolls off the stage and over the crowd like invisible boulders.

Pharmakon, whose performance I recently attended, wasn’t anything like this. Pharmakon is a lone woman, Margaret Chardiet, who tortures tones and pulses out of mixers, loops, and other assorted electronics, and she also screeches into the mic. It wasn’t exactly music to me, more a physical experience. At the climax of her performance at Gray Area, a space dedicated to art and technology in San Francisco’s Mission district, she pulled out an A/V cord, taped it up her forearm, slapped her head and slid her arm all over her body,…
Read More…

Read the whole story
2 days ago
Share this story

Book Review: The Hungry Brain

1 Comment and 3 Shares

[Content note: food, dieting, obesity]


The Hungry Brain gives off a bit of a Malcolm Gladwell vibe, with its cutesy name and pop-neuroscience style. But don’t be fooled. Stephan Guyenet is no Gladwell-style dilettante. He’s a neuroscientist studying nutrition, with a side job as a nutrition consultant, who spends his spare time blogging about nutrition, tweeting about nutrition, and speaking at nutrition-related conferences. He is very serious about what he does and his book is exactly as good as I would have hoped. Not only does it provide the best introduction to nutrition I’ve ever seen, but it incidentally explains other neuroscience topics better than the books directly about them do.

I first learned about Guyenet’s work from his various debates with Gary Taubes and his supporters, where he usually represents the “establishment” side. He is very careful to emphasize that the establishment doesn’t look anything like Taubes’ caricature of it. The establishment doesn’t believe that obesity is just about weak-willed people voluntarily choosing to eat too much, or that obese people would get thin if they just tried diet and exercise, or that all calories are the same. He writes

The [calories in, calories out] model is the idea that our body weight is determined by voluntary decisions about how much we eat and move, and in order to control our body weight, all we need is a little advice about how many calories to eat and burn, and a little willpower. The primary defining feature of this model is that it assumes that food intake and body fatness are not regulated. This model seems to exist mostly to make lean people feel smug, since it attributes their leanness entirely to wise voluntary decisions and a strong character. I think at this point, few people in the research world believe the CICO model.

[Debate opponent Dr. David] Ludwig and I both agree that it provides a poor fit for the evidence. As an alternative, Ludwig proposes the insulin model, which states that the primary cause of obesity is excessive insulin action on fat cells, which in turn is caused principally by rapidly-digesting carbohydrate. According to this model, too much insulin reduces blood levels of glucose and fatty acids (the two primary circulating metabolic fuels), simultaneously leading to hunger, fatigue, and fat gain. Overeating is caused by a kind of “internal starvation”. There are other versions of the insulin model, but this is the one advocated by Ludwig (and Taubes), so it will be my focus.

But there’s a third model, not mentioned by Ludwig or Taubes, which is the one that predominates in my field. It acknowledges the fact that body weight is regulated, but the regulation happens in the brain, in response to signals from the body that indicate its energy status. Chief among these signals is the hormone leptin, but many others play a role (insulin, ghrelin, glucagon, CCK, GLP-1, glucose, amino acids, etc.)

The Hungry Brain is part of Guyenet’s attempt to explain this third model, and it basically succeeds. But like many “third way” style proposals, it leaves a lot of ambiguity. With CICO, at least you know where you stand – confident that everything is based on willpower and that you can ignore biology completely. And again, with Taubes, you know where you stand – confident that willpower is useless and that low-carb diets will solve everything. The Hungry Brain is a little more complicated, a little harder to get a read on, and at times pretty wishy-washy.

But listening to people’s confidently-asserted simple and elegant ideas was how we got into this mess, so whatever, let’s keep reading.


The Hungry Brain begins with the typical ritual invocation of the obesity epidemic. Did you know there are entire premodern cultures where literally nobody is obese? That in the 1800s, only 5% of Americans were? That the prevalence of obesity has doubled since 1980?

Researchers have been keeping records of how much people eat for a long time, and increased food intake since 1980 perfectly explains increased obesity since 1980 – there is no need to bring in decreased exercise or any other factors. Exercise has decreased since the times when we were all tilling fields ten hours a day, but for most of history, as our exercise decreased, our food intake decreased as well. But for some reason, starting around 1980, the two factors uncoupled, and food intake started to rise despite exercise continuing to decrease.

Guyenet discusses many different reasons this might have happened, including stress-related overeating, poor sleep, and quick prepackaged food. But the ideas he keeps coming back to again and again are food reward and satiety.

In the 1970s, scientists wanted to develop new rat models of obesity. This was harder than it sounded; rats ate only as much as they needed and never got fat. Various groups tried to design various new forms of rat chow with extra fat, extra sugar, et cetera, with only moderate success – sometimes they could get the rats to eat a little too much and gradually become sort of obese, but it was a hard process. Then, almost by accident, someone tried feeding the rats human snack food, and they ballooned up to be as fat as, well, humans. The book:

Palatable human food is the most effective way to cause a normal rat to spontaneously overeat and become obese, and its fattening effect canot be attributed solely to its fat or sugar content.

So what does cause this fattening effect? I think the book’s answer is “no single factor, but that doesn’t matter, because capitalism is an optimization process that designs foods to be as rewarding as possible, so however many different factors there are, every single one of them will be present in your bag of Doritos”. But to be more scientific about it, the specific things involved are some combination of sweet/salty/umami tastes, certain ratios of fat and sugar, and reinforced preferences for certain flavors.

Modern food isn’t just unusually rewarding, it’s also unusually bad at making us full. The brain has some pretty sophisticated mechanisms to determine when we’ve eaten enough; these usually involve estimating food’s calorie load from its mass and fiber level. But modern food is calorically dense – it contains many more calories than predicted per unit mass – and fiber-poor. This fools the brain into thinking that we’re eating less than we really are, and shuts down the system that would normally make us feel full once we’ve had enough. Simultaneously, the extremely high level of food reward tricks the brain into thinking that this food is especially nutritionally valuable and that it should relax its normal constraints.

Adding to all of this is the so-called “buffet effect”, where people will eat more calories from a variety of foods presented together than they would from any single food alone. My mother likes to talk about her “extra dessert stomach”, ie the thing where you can gorge yourself on a burger and fries until you’re totally full and couldn’t eat another bite – but then mysteriously find room for an ice cream sundae afterwards. This is apparently a real thing that’s been confirmed in scientific experiments, and a major difference between us and our ancestors. The !Kung Bushmen, everyone’s go-to example of an all-natural hunter-gatherer tribe, apparently get 50% of their calories from a single food, the mongongo nut, and another 40% from meat. Meanwhile, we design our meals to include as many unlike foods as possible – for example, a burger with fries, soda, and a milkshake for dessert. This once again causes the brain to relax its usual strict constraints on appetite and let us eat more than we should.

The book sums all of these things up into the idea of “food reward” making some foods “hyperpalatable” and “seducing” the reward mechanism in order to produce a sort of “food addiction” that leads to “cravings”, the “obesity epidemic”, and a profusion of “scare quotes”.

I’m being a little bit harsh here, but only to highlight a fundamental question. Guyenet goes into brilliant detail about things like the exact way the ventral tegmental area of the brain responds to food-related signals, but in the end, it all sounds suspiciously like “food tastes good so we eat a lot of it”. It’s hard to see where exactly this differs from the paradigm that he dismisses as “attributing leanness to wise voluntary decisions and a strong character…to make lean people feel smug”. Yes, food tastes good so we eat a lot of it. And Reddit is fun to read, but if someone browses Reddit ten hours a day and doesn’t do any work then we start speculating about their character, and maybe even feeling smug. This part of the book, taken alone, doesn’t really explain why we shouldn’t be doing that about weight too.


The average person needs about 800,000 calories per year. And it takes about 3,500 extra calories to gain a pound of weight. So if somebody stays about the same weight for a year, it means they fulfilled their 800,000 calorie requirement to within a tolerance of 3,500 calories, ie they were able to match their food intake to their caloric needs with 99.5% accuracy.

By this measure, even people who gain five or ten pounds a year are doing remarkably well, falling short of perfection by only a few percent. It’s not quite true that someone who gains five pounds is ((5*3,500)/800,000) = 98% accurate, because each pound you gain increases caloric requirements in a negative feedback loop, but it’s somewhere along those lines.

Take a second to think about that. Can you, armed with your FitBit and nutritional labeling information, accurately calculate how many calories you burn in a given day, and decide what amount of food you need to eat to compensate for it, within 10%? I think even the most obsessive personal trainer would consider that a tall order. But even the worst overeaters are subconsciously managing that all the time. However many double bacon cheeseburgers they appear to be eating in a single sitting, over the long term their body is going to do some kind of magic to get them to within a few percent of the calorie intake they need.

It’s not surprising that people overeat, it’s surprising that people don’t overeat much more. Consider someone who just has bad impulse control and so eats whatever they see – wouldn’t we expect them to deviate from ideal calorie input by more than a few percent, given that this person probably has no idea what their ideal input even is and maybe has never heard of calories?

And as impressive as modern Westerners’ caloric balance is, everyone else’s is even better. Guyenet discusses the Melanesian island of Kitava, where there is literally only one fat person on the entire island – a businessman who spends most of his time in modern urbanized New Guinea, eating Western food. The Kitavans have enough food, and they live a relaxed tropical lifestyle that doesn’t demand excessive exercise. But their bodies aren’t making even the 10% error that ours do. They’re essentially perfect. So are the !Kung with their mongongo nuts, Inuit with their blubber, et cetera.

And so are Westerners who limit themselves to bland food. In 1965, some scientists locked people in a room where they could only eat nutrient sludge dispensed from a machine. Even though the volunteers had no idea how many calories the nutrient sludge was, they ate exactly enough to maintain their normal weight, proving the existence of a “sixth sense” for food caloric content. Next, they locked morbidly obese people in the same room. They ended up eating only tiny amounts of the nutrient sludge, one or two hundred calories a day, without feeling any hunger. This proved that their bodies “wanted” to lose the excess weight and preferred to simply live off stored fat once removed from the overly-rewarding food environment. After six months on the sludge, a man who weighed 400 lbs at the start of the experiment was down to 200, without consciously trying to reduce his weight.

In a similar experiment going the opposite direction, Ethan Sims got normal-weight prison inmates to eat extraordinary amounts of food – yet most of them still had trouble gaining weight. He had to dial their caloric intake up to 10,000/day – four times more than a normal person – before he was able to successfully make them obese. And after the experiment, he noted that “most of them hardly had any appetite for weeks afterward, and the majority slimmed back down to their normal weight”.

What is going on here? Like so many questions, this one can best be solved by grotesque Frankenstein-style suturing together of the bodies of living creatures.

In the 1940s, scientists discovered that if they damaged the ventromedial hypothalamic nucleus (VMN) of rats, the rats would basically never stop eating, becoming grotesquely obese. They theorized that the VMN was a “satiety center” that gave rats the ability to feel full; without it, they would feel hungry forever. Later on, a strain of mutant rats was discovered that seemed to naturally have the same sort of issue, despite seemingly intact hypothalami. Scientists wondered if there might be a hormonal problem, and so they artificially conjoined these rats to healthy normal rats, sewing together their circulatory systems into a single network. The result: when a VMN-lesioned rat was joined to a normal rat, the VMN-lesioned rat stayed the same, but the normal rat stopped eating and starved to death. When a mutant rat was joined to a normal rat, the normal rat stayed the same and the mutant rat recovered and became normal weight.

The theory they came up with to explain the results was this: fat must produce some kind of satiety hormone, saying “Look, you already have a lot of fat, you can feel full and stop eating now”. The VMN of the hypothalamus must detect this message and tell the brain to feel full and stop eating. So the VMN-lesioned rats, whose detector was mostly damaged, responded by never feeling full, eating more and more food, and secreting more and more (useless) satiety hormone. When they were joined to normal rats, their glut of satiety hormones flooded the normal rats – and their normal brain, suddenly bombarded with “YOU ALREADY HAVE WAY TOO MUCH FAT” messages, stopped eating entirely.

The mutant rats, on the other hand, had lost the ability to produce the satiety hormone. They, too, felt hungry all the time and ate everything. But when they were joined to a normal rat, the normal levels of satiety hormone flowed from the normal rat into the mutant rat, reached the fully-functional detector in their brains, and made them feel full, curing their obesity.

Skip over a lot of scientific infighting and unfortunate priority disputes and patent battles, and it turns out the satiety hormone is real, exists in humans as well, and is called leptin. A few scientists managed to track down some cases of genetic leptin deficiency in humans, our equivalent of the mutant rats, and, well…

Usually they are of normal birth weight and then they’re very, very hungry from the first weeks and months of life. By age one, they have obesity. By age two, they weigh 55-65 pounds, and their obesity only accelerates from there. While a normal child may be about 25% fat, and a typical child with obesity may be 40% fat, leptin-deficient children are up to 60% fat. Farooqi explains that the primary reason letpin-deficient children develop obesity is that they have “an incredible drive to eat”…leptin-deficient children are nearly always hungry, and they almost always want to eat, even shortly after meals. Their appetite is so exaggerated that it’s almost impossible to put them on a diet: if their food is restricted, they find some way to eat, including retrieving stale morsels from the trash can and gnawing on fish sticks directly from the freezer. This is the desperation of starvation […]

Unlike normal teenagers, those with leptin deficiency don’t have much interest in films, dating, or other teenage pursuits. They want to talk about food, about recipes. “Everything they do, think about, talk about, has to do with food” says [Dr.] Farooqi. This shows that the [leptin system] does much more than simply regulate appetite – it’s so deeply rooted in the brain that it has the ability to hijack a broad swath of brain functions, including emotions and cognition.

It’s the leptin-VNM-feedback system (dubbed the “lipostat”) that helps people match their caloric intake to their caloric requirements so impressively. The lipostat is what keeps hunter-gatherers eating exactly the right number of mongongo nuts, and what keeps modern Western overeaters at much closer to the right weight than they could otherwise expect.

The lipostat-brain interface doesn’t just control the raw feeling of hunger, it seems to have a wide range of food-related effects, including some on higher cognition. Ancel Keys (of getting-blamed-for-everything fame) ran the Minnesota Starvation Experiment on some unlucky conscientious objectors to World War II. He starved them until they lost 25% of their body weight, and found that:

Over the course of their weight loss, Keys’s subjects developed a remarkable obsession with food. In addition to their inescapable, gnawing hunger, their conversations, thoughts, fantasies, and dreams revolved around food and eating – part of a phenomenon Keys called “semistarvation neurosis”. They became fascinated by recipes and cookbooks, and some even began collecting cooking utensils. Like leptin-deficient adolescents, their lives revolved around food. Also like leptin-deficient adolescents, they had very low leptin levels due to their semi-starved state.

Unsurprisingly, as soon as the experiment ended, they gorged themselves until they were right back at their pre-experiment weights (but no higher), at which point they lost their weird food obsession.

Just as a well-functioning lipostat is very good at keeping people normal weight, a malfunctioning lipostat is very good at keeping people obese. Fat people seem to have “leptin resistance”, sort of like the VMN-lesioned rats, so that their bodies get confused about how much fat they have. Suppose a healthy person weighs 150 lbs, his body is on board with that, and his lipostat is set to defend a 150 lb set point. Then for some reason he becomes leptin-resistant, so that the brain is only half as good at detecting leptin as it should be. Now he will have to be 300 lbs before his brain “believes” he is the right weight and stops encouraging him to eat more. If he goes down to a “mere” 280 lbs, then he will go into the same semistarvation neurosis as Ancel Keys’ experimental subjects and become desperately obsessed with food until they get back up to 300 again. Or his body will just slow down metabolism until his diet brings him back up. Or any of a bunch of other ways the lipostat has to restore weight when it wants to.

(and I would be shocked if the opposite problem weren’t at least part of anorexia)

This explains the well-known phenomenon where contestants on The Biggest Loser who lose 200 or 300 pounds for the television camera pretty much always gain it back after the show ends. Even though they’re highly motivated and starting from a good place, their lipostat has seized on their previous weight as the set point it wants to defend, and resisting the lipostat is somewhere between hard and impossible. As far as I know, nobody has taken Amptoons up on their challenge to find a single peer-reviewed study showing any diet that can consistently bring fat people to normal weight and keep them there.

And alas, it doesn’t seem to work to just inject leptin directly. As per Guyenet

People with garden variety obesity already have high levels of leptin…while leptin therapy does cause some amount of fat loss, it requires enormous doses to be effective – up to forty times the normal circulating amount. Also troubling is the extremely variable response, with some people losing over thirty pounds and others losing little or no weight. This is a far cry from the powerful fat-busting effect of leptin in rodents. [Leptin as] the new miracle weight-loss drug never made it to market.

This disappointment forced the academic and pharmaceutical communities to confront a distressing possibility: the leptin system defends vigorously against weight loss, but not so vigorously against weight gain. “I have always thought, and continue to believe,” explained [leptin expert Rudy] Leibel, “that the leptin hormone is really a mechanism for detecting deficiency, not excess.” It’s not designed to constrain body fatness, perhaps because being too fat is rarely a problem in the wild. Many researchers now believe that while low leptin levels in humans engage a powerful starvation response that promotes fat gain, high leptin levels don’t engage an equally powerful response that promotes fat loss.

Yet something seems to oppose rapid fat gain, as Ethan Sims’ overfeeding studies (and others) have shown. Although leptin clearly defends the lower limit of adiposity, the upper limit may be defended by an additional, unidentified factor – in some people more than others.

This is the other half of the uncomfortable dichotomy that makes me characterize The Hungry Brain as “wishy-washy”. The lipostat is a powerful and essentially involuntary mechanism for getting weight exactly where the brain wants, whether individual dieters are cooperative or not. Here it looks like obesity is nobody’s fault, unrelated to voluntary decisions, and that the standard paradigm really is just an attempt by lean people to feel smug. Practical diet advice starts to look like “inject yourself with quantities of leptin so massive that they overcome your body’s resistance”. How do we connect this with the other half of the book, the half with food reward and satiety and all that?


With more rat studies!

Dr. Barry Levin fed rats either a healthy-rat-food diet or a hyperpalatable-human-food diet, then starved and overfed them in various ways. He found that the rats defended their obesity set points in the expected manner, but that the same rats defend different set points depending on their diets. Rats on healthy-rat-food defended a low, healthy-for-rats set point; rats on hyperpalatable-human-food defended a higher set point that kept them obese.

That is, suppose you give a rat as much Standardized Food Product as it can eat. It eats until it weighs 8 ounces, and stays that weight for a while. Then you starve it until it only weighs 6 ounces, and it’s pretty upset. Then you let it eat as much as it wants again, and it overeats until it gets back to 8 ounces, then eats normally and maintains that weight.

But suppose you get a rat as many Oreos as it can eat. It eats until it weighs 16 ounces, and stays that weight for a while. Then you starve it until it only weighs 6 ounces. Then you let it eat as much as it wants again, and this time it overeats until it gets back to 16 ounces, and eats normally to maintain that weight.

Something similar seems to happen with humans. A guy named Michel Cabanac ran an experiment in which he put overweight people on two diets. In the first diet, they ate Standardized Food Product, and naturally lost weight since it wasn’t very good and they didn’t eat very much of it. In the second diet, he urged people to eat less until they matched the first group’s weight loss, but to keep eating the same foods as normal – just less of them. The second group reported being hungry and having a lot of trouble dieting; the first group reported not being hungry and not having any trouble at all.

Guyenet concludes:

Calorie-dense, highly rewarding food may favor overeating and weight gain not just because we passively overeat it but also because it turns up the set point of the lipostat. This may be one reason why regularly eating junk food seems to be a fast track to obesity in both animals and humans…focusing the diet on less rewarding foods may make it easier to lose weight and maintain weight loss because the lipostat doesn’t fight it as vigorously. This may be part of the explanation for why all weight-loss diets seem to work to some extent – even those that are based on diametrically opposed principles, such as low-fat, low-carbohydrate, paleo, and vegan diets. Because each diet excludes major reward factors, they may all lower the adiposity set point somewhat.

(this reminds me of the Shangri-La Diet, where people would drink two tablespoons of olive oil in the morning, then find it was easy to diet without getting hungry during the day. People wondered whether maybe the tastelessness of the olive oil had something to do with it. Could it be that the olive oil is temporarily bringing the lipostat down to its “bland food” level?)

Why should some food make the lipostat work better than other food? Guyenet now gets to some of his own research, which is on a type of brain cell called a POMC neuron. These neurons produce various chemicals, including a sort of anti-leptin called Neuropeptide Y, and they seem to be a very fundamental part of the lipostat and hunger system. In fact, if you use superprecise chemical techniques to kill NPY neurons but nothing else, you can cure obesity in rats.

The area of the hypothalamus with POMC neurons seem to be damaged in overweight rats and overweight humans. Microglia and astrocytes, the brain’s damage-management and repair cells, proliferated in appetite-related centers, but nowhere else. Maybe this literal damage corresponds to the metaphorically “damaged” lipostat that’s failing to maintain weight normally, or the “damaged” leptin detector that seems to be misinterpreting the body’s obesity?

In any case, eating normal rat food for long enough appears to heal this damage:

Our results suggest that obese rodents suffer from a mild form of brain injury in an area of the brain that’s critical for regulating food intake and adiposity. Not only that, but the injury response and inflammation that developed when animals were placed on a fattening diet preceded the development of obesity, suggesting that this brain injury could have played a role in the fattening process.

Guyenet isn’t exactly sure what aspect of modern diets cause the injury:

Many researchers have tried to narrow down the mechanisms by which this food causes changes in the hypothalamus and obesity, and they have come up with a number of hypotheses with varying amounts of evidence to support them. Some researchers believe the low fiber content of the diet precipitates inflammation and obesity by its adverse effects on bacterial populations in the gut (the gut microbiota). Others propose that saturated fat is behind the effect, and unsaturated fats like olive oil are less fattening. Still others believe the harmful effects of overeating itself, including the inflammation caused by excess fat and sugar in the bloodstream and in cells, may affect the hypothalamus and gradually increase the set point. In the end, these mechanisms could all be working together to promote obesity. We don’t know all the details yet, but we do know that easy access to refined, calorie-dense, highly rewarding food leads to fat gain and insidious changes in the lipostat in a variety of species, including humans. This is particularly true when the diet offers a wife variety of sensory experiences, such as the hyperfattening “cafeteria diet” we encountered in chapter 1.

Personally, I believe overeating itself probably plays an important role in the process that increases the adiposity set point. In other words, repeated bouts of overeating don’t just make us fat; they make our bodies want to stay fat. This is consistent with the simple observation that in the United States, most of our annual weight gain occurs during the six-week holiday feasting period between Thanksgiving and the new year, and that this extra weight tends to stick with us after the holidays are over…because of some combination of food quantity and quality, holiday feasting ratchets up the adiposity set point of susceptible people a little bit each year, leading us to gradually accumulate and defend a substantial amount of fat. Since we also tend to gain weight at a slower rate during the rest of the year, intermittent periods of overeating outside of the holidays probably contribute as well.

How might this happen? We aren’t entirely sure, but researchers, including Jeff Friedman, have a possible explanation: excess leptin itself may contribute to leptin resistance. To understand how this works, I need to give you an additional piece of information: Leptin doesn’t just correlate with body fat levels; it also responds to short-term changes in calorie intake. So if you overeat for a few days, your leptin level can increase substantially, even if your adiposity has scarcely changed (and after your calorie intake goes back to normal, so does your leptin). As an analogy for how this can cause leptin resistance, imagine listening to music that’s too loud. At first, it’s thunderous, but eventually, you damage your hearing, and the volume drops. Likewise, when we eat too much food over the course of a few days, letpin levels increase sharply, and this may begin to desensitize the brain circuits that respond to leptin. Yet Rudy Leibel’s group has also shown that high leptin levels alone aren’t enough – the hypothalamus also seems to require a second “hit” for high leptin to increase the set point of the lipostat. This second hit could be the brain injury we, and others, have identified in obese rodents and humans.

And he isn’t sure exactly what aspect of the normal rodent diet promotes the healing:

I did do some research in mice suggesting that unrefined, simple food does reverse the brain changes and the obesity. I don’t claim that it’s all attributable to the blandness though– the two diets differed in many respects (palatability, calorie density, fiber content, macronutrient profile, fatty acid profile, content of nonessential nutrients like polyphenols). Also, we don’t know how well the finding applies to humans yet. One of the problems is that it’s very hard to get a group of humans to adhere strictly to a whole food diet for long enough to study its long-term effects on appetite and body fatness. People are very attached to the pleasures of the palate!

But all of this together seems to point to a potential synthesis between the hyperpalatability and lipostat models. Modern society has been incentivized to produce hyperpalatable, low-satiety food as superstimuli. Overeating this modern food in the short term raises the lipostat’s set point (for some reason, possibly involving brain damage and leptin resistance), causing us to gain weight in the long term, in a way that is very difficult to reverse.


But I still have trouble reconciling these two points of view.

A couple of days ago, I walked by an ice cream store. I’d just finished lunch, and I wasn’t very hungry at the time, but it looked like really good ice cream, and it was hot out, so I gave in to temptation and ate a 700 calories sundae. Does this mean:

1. Based on the one pound = 3500 calories heuristic, I have now gained 0.2 lbs. That extra weight will stay with me my whole life, or at least until some day when I diet and eat 700 calories less than my requirement. If I were to eat ice cream like this a hundred times, I would gain twenty pounds.

2. My lipostat adjusts for the 700 extra calories, and causes me to exercise more, or ramp up my metabolism, or burn more brown fat, or eat less later on, or something. I don’t gain any weight, and eating the ice cream was that rarest of all human experiences, a completely guiltless pleasure. I should eat ice cream whenever I feel like it, or else I am committing the sin of denying myself a lawful pleasure.

3. My lipostat will more or less take care of the ice cream today, and I won’t notice the 0.2 pounds on the scale, but it is very gradually doing hard-to-measure damage to my hypothalamus, and if I keep eating ice cream like this, then one day when I’m in my forties I’m going to wake up weighing three hundred pounds, and no diet will ever be able to help me.

4. The above scenario is impossible. Even if I think I just ate ice cream because it looked good, in reality I was driven to do it by my lipostat’s quest for caloric balance. Any feeling of choice in the matter is an illusion.

I think the reason this is so confusing is because the real answer is “it could be any one of these, depending on genetics.”

Note the position of the grey squares representing BMI

Right now, within this culture, variation in BMI is mostly genetic. This isn’t to say that non-genetic factors aren’t involved – the difference between 1800s America and 2017 America is non-genetic, and so is the difference between the perfectly-healthy Kitavans on Kitava and the one Kitavan guy who moved to New Guinea. But once everyone alike is exposed to the 2017-American food environment, differences between the people in that environment seem to be really hereditary and not-at-all-related to learned behavior. Guyenet acknowledges this:

Genes explain that friend of yours who seems to eat a lot of food, never exercises, and yet remains lean. Claude Bouchard, a genetics researcher at the Pennington Biomedical Research Center in Baton Rouge, Louisiana, has shown that some people are intrinsically resistant to gaining weight even when they overeat, and that this trait is genetically influenced. Bouchard’s team recruited twelve pairs of identical twins and overfed each person by 1,000 calories per day above his caloric needs, for one hundred days. In other words, each person overate the same food by the same amount, under controlled conditions, for the duration of the study.

If overeating affects everyone the same, then they should all have gained the same amount of weight. Yet Bouchard observed that weight gain ranged from nine to twenty-nine pounds! Identical twins tended to gain the same amount of weight and fat as each other, while unrelated subjects had more divergent responses…Not only do some people have more of a tendency to overeat than others, but some people are intrinsically more resistant to gaining fat even if they do overeat.

The research of James Levine, an endocrinologist who works with the Mayo Clinic and Arizona State University, explains this puzzling phenomenon. In a carefully controlled overfeeding study, his team showed that the primary reason some people readily burn off excess calories is that they ramp up a form of calorie-burning called “non-exercise activity thermogenesis” (NEAT). NEAT is basically a fancy term for fidgeting. When certain people overeat, their brains boost calorie expenditure by making them fidget, change posture frequently, and make other small movements throughout the day. It’s an involuntary process, and Levine’s data show that it can incinerate nearly 700 calories per day. The “most gifted” of Levine’s subjects gained less than a pound of body fat from eating 1,000 extra calories per day for eight weeks. Yet the strength of the response was highly variable, and the “least gifted” of Levine’s subjects didn’t increase NEAT at all, shunting all the excess calories into fat tissue and gaining over nine pounds of body fat…

Together, these studies offer indisputable evidence that genetics plays a central role in obesity and dispatch the idea that obesity is primarily due to acquired psychological traits.

These studies suggest that one way genetics affects obesity is by altering the tolerance level of the lipostat. Genetically privileged people may have very finicky lipostats that immediately burn off any extra calories they eat, and which never become dysregulated. Genetically unlucky people may have weak lipostats which fail to defend against weight gain, or which are too willing to adjust their set point up in the presence of an unhealthy food environment.

So, given how many people seem to have completely different weight-gain-related experiences to each other, the wishy-washyness here might be a feature rather than a bug.

One reason I’ve always found genetics so exciting is that there are all these fields – nutrition is a great example, but this applies at least as much to psychiatry – where everyone has wildly different personal experiences, and where there’s a large and vocal population of people who say that the research is exactly the opposite of their lived experiences. People have tried to shoehorn the experiences to fit the research, with various levels of plausibility versus condescendingness. And for some reason, it’s always really hard to generate the hypothesis “people’s different experiences aren’t an illusion; people are genuinely really different”. Once you start looking at genetics, everything sort of falls into place, and ideas which seemed wishy-washy or self-contradictory before are revealed as just reflecting the diversity of nature. People who were previously at each other’s throats disputing different interpretations of the human condition are able to peacefully agree that there are many different human conditions, and that maybe we can all just get along. The Hungry Brain and other good books in its vein offer a vision for how we might one day be able to do that in nutrition science.


Lest I end on too positive a note, let me reiterate the part where happiness is inherently bad and a sort of neo-Puritan asceticism is the only way to avoid an early grave.

There’s an underlying fatalism to the discourse around “food reward”. If the enemy were saturated fat, we could just stick with the sugary sweetness Coca-Cola. If the enemy were carbohydrates, we could go out for steak every night. But what do we do if the enemy is deliciousness itself?

A few weeks ago Guyenet announced The Bland Food Cookbook, a collection of tasteless recipes guaranteed to be low food-reward and so discourage overeating. It was such a natural extension of his philosophy that it took me a whole ten seconds to realize it was an April Fools joke. But why should it be? Shouldn’t this be exactly the sort of thing we’re going for?

I asked him, and he responded that:

If I thought enough people would actually be capable of following the diet, I would consider making such a cookbook non-ironically. The second point I want to make here is that there are many ways to lose weight, and deliberately reducing food reward is only one of them. You could also exercise, eat a low-calorie-density diet, eat a high-protein diet, restrict a macronutrient, restrict animal foods, restrict plant foods, eat nothing but potatoes. Most approaches overlap with a low-reward diet to varying degrees, but I don’t think the low reward value encapsulates everything about why every weight loss strategy works. BTW, low-carb folks often have a knee-jerk reaction to the low-reward thing that goes something like this: “I eat food that’s delicious, such as steaks, bacon, butter, etc. It’s not low in reward.” But it is low reward in the sense that you’re cutting out a broad swath of foods, and an entire macronutrient, that the brain very much wants you to eat. Eating more of a particular category of rewarding food doesn’t completely make up for the fact that you’re cutting out a whole other category of rewarding food that you would avidly consume if you weren’t restricting yourself.

So things aren’t maximally bad. And hunter-gatherers enjoy their healthy diets just fine. And certainly there are things like steak and wine and so on which are traditionally “good food” without being hyperprocessed hyperpalatable junk food. But if you really enjoy a glass of Chardonnay, is that “food reward” in the sense that’s potentially dangerous? Is anything safe? What about mongongo nuts? Is there anywhere we can get them?

Overall I strongly recommend The Hungry Brain for everything I talk about here and for some other good topics I didn’t even get to (stress, sleep, a list of practical real-world diet advice). I would also recommend Guyenet’s other writing, especially his debate with Dr. David Ludwig on the causes of obesity (Part 1, Part 2, Part 3. I also recommend the list of diet tips that Guyenet gives at the end of the book. I won’t give them all away here – he’s been nice enough to me that maybe I should repay him by not reprinting the entire text of his book online for free. But it’s similar to a lot of standard advice for healthy living, albeit with more interesting reasoning behind it. Did you know that exercise might help stabilize the lipostat? Or that protein might do the same? Also, one piece of advice you might not hear anywhere else – potatoes are apparently off-the-charts in terms of satiety factor and may be one of the single best things to diet on.

And speaking of good things to diet on…

(note that this next part is my own opinion, not taken from The Hungry Brain or endorsed by Stephan Guyenet)

Slate Star Codex’s first and most loyal sponsor is MealSquares, a “nutritionally complete” food product sort of like a solid whole-foods-based version of Soylent. I’m having some trouble writing this paragraph, because I want to recommend them as potentially dovetailing with The Hungry Brain‘s philosophy of nutrition without using phrases that might make MealSquares Inc angry at me like “bland”, “low food reward”, or “not hyperpalatable”. I think the best I can come up with is “unlikely to injure your hypothalamus”. So, if you’re looking for an easy way to quit the junk food and try a low-variety diet that’s unlikely to injure your hypothalamus, I recommend MealSquares as worth a look.

Read the whole story
2 days ago
Boil 'em, mash 'em...
Share this story

Learning To Love Scientific Consensus


[Related to: Contrarians, Crackpots, and Consensus, How Common Are Science Failures?. Epistemic status is “subtle and likely to be misinterpreted”.]


There’s a list of scientific mavericks who were ridiculed by hidebound reactionaries but later vindicated that’s been going viral. I examined the first ten mavericks on the list to see if its claims held up. Overall I wasn’t too impressed. Let me go over them in more detail.


His idea that electrolytes are full of charged atoms was considered crazy. The atomic theory was new at the time, and everyone “knew” that atoms were indivisible (and hence they could not lose or gain any electric charge.) Because of his heretical idea, he only received his university degree by a very narrow margin.

Sure, the professors who were judging his PhD thesis weren’t too convinced. So Arrhenius sent his proposal to the world’s top chemists at the time, and they were super-interested and started fighting among themselves to work with Arrhenius on it. Top chemist Wilhelm Ostwald received the paper the same day his daughter was born, and suggested that the paper was the more exciting of the two events. He journeyed to Arrhenius’ hometown of Uppsala, Sweden to try to convince Arrhenius to work with him; Arrhenius refused for personal reasons but later got a scholarship and worked with the top physicists in Europe. Arrhenius became a professor in a prestigious university about ten years after presenting his “ridiculed” paper, and won the Nobel Prize ten years after that.


Astronomers thought that gravity alone is important in solar systems, in galaxies, etc. Alfven’s idea that plasma physics is of equal or greater importance to gravity was derided for decades.

This isn’t a great description of Alfven’s conflict with the establishment, but the list seems basically right insofar as Alfven’s ideas were ignored for thirty years before being proven mostly correct. I will give them this one.


When the first television system was demonstrated to the Royal Society (British scientists,) they scoffed and ridiculed, calling Baird a swindler.

I can’t find any reference to this in various Baird articles and biographies. The closest I can come is this article by someone who was there at the demonstration, who said “They didn’t believe it…the pictures were a bit of a blur but it was amazing, they were all absolutely flabbergasted by it.” It looks like he is using “they didn’t believe it” in the colloquial way of “they thought it was amazing”. A TIME magazine article from the time described the same scientists as “deeply impressed”, though the wording is kind of unclear and they might have been referring to a different demonstration a year later.

In any case, it seems very clear that within a year everyone agreed he was legitimate and overcame their initial shock.


Everyone knows that dinosaurs are like Gila monsters or big tortoises: large, slow, and intolerant of the cold. And they’re all colored olive drab too! 🙂

Bakker did help produce the paradigm shift in paleontology from cold-blooded dinosaurs to warm-blooded dinosaurs. But he was not a lone maverick being ridiculed by everyone else. He learned that dinosaurs were warm-blooded from his professor at Yale, who was also part of the minority-but-totally-existing faction that believed dinosaurs were warm-blooded. He himself got a PhD at Harvard from professors who were apparently sympathetic to the same theory. And within seven years of his first paper being published, Scientific American was calling his ideas “the dinosaur renaissance”, which doesn’t leave a lot of time for him to be ridiculed and ignored in.


Not ridiculed, but their boss W. Shockley nixed their idea for a non-FET “crystal triode” device. When they started investigating it, he made them stop. They were supposed to be working on FETs instead.

ARG, I GOT THIS WRONG, THIS PART BELOW IS A BELL LABS STORY REGARDING ZONE REFINING OF SILICON, NOT THE BJT TRANSISTOR PROJECT: So, they assembled their ZONE REFINING experiment on a wheeled cart and continued. Whenever the boss was scheduled to check up on them, they could shove it into an adjacent unused lab.

Okay, it looks like the guy compiling the list admits he was wrong on this one. Moving on…


Endured decades of scorn as the laughingstock of the geology world. His crime was to insist that enormous amounts of evidence showed that, in Eastern Washington state, the “scabland” desert landscape had endured an ancient catastrophy: a flood of staggering proportions. This was outright heresy, since the geology community of the time had dogmatic belief in a “uniformitarian” position, where all changes must take place slowly and incrementally over vast time scales. Bretz’ ideas were entirely vindicated by the 1950s. Quote: “All my enemies are dead, so I have no one to gloat over.”

This one is basically right and I’ll give it to them.


Chandra originated Black Hole theory and published several papers. He was attacked viciously by his close colleague Sir Arthur Eddington, and his theory was discredited in the eyes of the research community. They were wrong, and Eddington apparently took such strong action based on an incorrect pet theory of his own. In the end Chandra could not even pursue a career in England, and he moved his research to the U. of Chicago in 1937, laboring in relative obscurity for decades.

Sort of true, but he was hardly shunned by the scientific community. He made his discoveries about black holes in the early 1930s, was well-received by many people, and won a Bronze Medal in some physics competition. In 1935, Eddington attacked his theory, possibly because Eddington was racist and didn’t like Indian people. But many other scientists, including Niels Bohr and Wolfgang Pauli, continued to support him (quietly, so as not to offend Eddington, which will be a recurring theme in these kinds of situations). Chandrasekhar was made a Fellow of the Royal Society in 1944, won the Royal Astronomical Society Gold Medal in 1953, and generally led a long and prestigious life. His theories were resurrected once people had better evidence that black holes existed. I’ll give this one half a point.


The scientific community regarded Meteorites in the same way that modern scientists regard UFO abductions and psychic phenomenon: quaint superstitions only believed by peasant folk. All the eyewitness reports were disbelieved. At one point the ridicule became so intense that many museums with meteorites in their geology collections decided to trash those valuable samples. (Sometimes hostile skepticism controls reality, and the strongest evidence is edited to conform to concensus disbeliefs.) Finally in the early 1800’s Ernst Chladni actually sat down and inspected the evidence professionally, and found that claimed meteorites were entirely unlike known earth rocks. His study changed some minds. At the same time some large meteor falls were witnessed by scientists, and the majority who insisted that only ignorant peasants ever saw such things were shamed into silence.

As the quote points out, this is a kind of weird one as meteorite work was ridiculed for a long time, but Chladni was taken seriously and helped change minds. Looking at Wikipedia, a lucky meteorite fall two years after Chladni first published his theory helped turn the tide in his favor, and by ten years after publication Chladni’s meteorite theories were pretty well-regarded. Even when people disagreed with him about meteorites, Chladni remained widely respected for some of his other work in acoustics.

There is a story here, but it’s probably not right to center it around Chladni, and his work was only scorned for a few years before everyone agreed it was true. I’ll give this another half a point.


Not ridiculed. But they were instructed to drop their research. They continued it as “bootleg” research.

The list admits they were “not ridiculed”. They were told to stop their research because there was all sorts of academic politics around who was going to be the first to discover DNA, and the guy in charge of their university was rooting for another team.


Proposed a theory of the optical Doppler Effect in 1842, but was bitterly opposed for two decades because it did not fit with the accepted physics of the time (it contradicted the Luminiferous Aether theory.) Doppler was finally proven right in 1868 when W. Huggins observed red shifts and blue shifts in stellar spectra. Unfortunately this was fifteen years after Doppler had died.

I haven’t been able to find anything about this in various short online biographies of Doppler (1, 2). Doppler tested the effect himself by having someone play a trumpet on a train (really), someone else successfully tested it in 1845, and it was independently rediscovered in 1848. Doppler himself was made the head of the Institute For Experimental Physics in Vienna and died about as prestigious and beloved as a physicist can get.

So my impression is that only a third of these people really fit the pattern. Most of them were doubted for very short periods, continued to be respected in their fields for their other accomplishments even during those periods, or were part of medium-sized movements rather than being lone geniuses. After a few years – maybe an average of ten, very rarely as long as thirty – their contributions were recognized and they assumed their rightful place in the pantheon. Science isn’t perfect. But it is darned good.

[EDIT: Bill Beatty, author of the original list, responds here. My response to the response here.]


I bring this up in the context of my last post on progress in the rationalist movement. There used to be a stereotype that rationalists were too quick to challenge scientific consensus. I think that was exaggerated, but based on a core of truth. Given that we’re interested in the ways that bias can prevent people from accepting truth, it’s unsurprising that we would focus on cases like these.

But I personally have changed my thinking on this a lot. Not in any way that I can explain explicitly – I’ve always thought something like:

Scientific consensus is the best tool we have for seeking truth. It’s not perfect, and it’s frequently overturned by later scientists, but this is usually – albeit not literally always – the work of well-credentialed insiders, operating pretty quickly after the evidence that should overturn it becomes available. Any individual should be very doubtful of their ability to beat it, while not being so doubtful that nobody ever improves it and science can never progress.

– and I still think that. But I’ve shifted from being the sort of person who shares viral lists of maligned geniuses, to the sort of person who debunks those lists. I’ve started emphasizing the “best tool we have” part of the sentence, and whispering the “isn’t perfect” part, rather than vice versa.

I’ve changed my mind on this because of personal experience. Rather than trying to describe it, it might be more helpful to give the most salient examples.

1. The Replication Crisis: I previously thought the scientific consensus was flawed because it failed to take the replication crisis seriously enough. I later learned that everyone else took the repliaction crisis exactly as seriously as I did. A poll in Nature shows that 90% of scientists believe reproducibility issues constitute a “crisis”, compared to only 3% (!) who don’t. For every person complaining about “methodological terrorists”, there are a dozen who are very concerned and trying to change the way they practice research.

This is especially impressive because as far as I can tell the whole shift happened in about ten years. I would date the beginning of the crisis from Ioannidis’ original 2005 paper, although it was only aimed at medicine. It got into high gear in psychology sometime around 2011 with Simonsohn’s False Positive Psychology. A Google Trends analysis suggests people only started searching the relevant keywords around 2013.

I started thinking about this sort of thing in 2009 after reading this LW post. At the time I thought this was some sort of exciting failure of modern science that I alone had figured out. But this was well after sharp people like Ioannidis were talking about it, and only a few years before everyone was talking about it. Framing this as “I was right and scientific consensus was wrong” seems grandiose. Better might be “I started betting on a winning horse about a quarter of the way between the beginning of the race and when its victory became blatantly obvious to everyone”.

2. Nutrition: The Bad Old Paradigm of nutrition says that obese people just have poor impulse control, that weight is a simple matter of calories in vs. calories out, and that all calories are equally good except fat, which for some inexplicable reason is the Devil. Anybody who’s read a few good books about nutrition science knows that the Bad Old Paradigm is woefully inadequate. I read a few of those books and became convinced that I was right and scientific consensus was wrong.

Unfortunately, this whole issue exploded when Gary Taubes published Good Calories, Bad Calories, which as best I can tell combined the first publicly available good critique of the Bad Old Paradigm with a flawed and basically false attempt at a new paradigm. There were lots of confused attacks against Taubes’ bad information which did collateral damage to his good information, and lots of confused defenses of his good information which inadvertently shielded his bad information from criticism. I previously focused on defend the good parts, but recently shifted more towards criticizing the bad parts.

After reading some more good books here (one of which I hope to review soon), my impression is that most nutrition scientists don’t believe in the Bad Old Paradigm and haven’t for a while. At the very least, most of them seem to believe in the lipostat and think it’s important, which is my proxy for “basically has their heart in the right place”. Insofar as the Bad Old Paradigm continues to be popular wisdom, it’s because of the diet industry, the government, social inertia, and nobody really having a good new paradigm to replace it with. I’m gradually seeing popular wisdom shift, and nutrition scientists themselves seem to be helping this process rather than hurting it.

Maybe somebody in this area has discovered the new paradigm and is a maverick being persecuted by hidebound reactionaries. But it isn’t Gary Taubes. And it certainly isn’t me.

3. Social-Justice-Related Issues: Another narrative I used to believe was that a lot of sketchy ideas were being flattered because they spoke to left-leading academics’ biases in favor of social justice. Implicit association tests, stereotype threat, the idea of zero meaningful psychological differences between men and women, et cetera.

When I started worrying about implicit association tests, I thought I was defying some kind of broad scientific consensus. But the meta-analyses showing the Implicit Association Test didn’t do what people thought had been around since 2009 and have only gotten more numerous since then, with broad media coverage. Problems with stereotype threat research are getting mainstream coverage and even airtime on NPR.

The problem here is that there was no equivalent of the Nature poll on the replication crisis, so I didn’t realize any of this was happening until just recently. For example, in 2016 this Voxsplainer made it sound like there was a monolithic consensus in favor of Implicit Association Tests that no sane person had ever disagreed with, even though by that point there were already several big meta-analyses finding they weren’t practically useful. The correct conclusion isn’t that this is really what scientific consensus thinks. The correct conclusion is that Vox shouldn’t be trusted about any science more complicated than the wedge vs. inclined plane. Once I realized that there was all this intelligent analysis going on that I’d never heard about, my claim to be boldly defying the scientific consensus evaporated.

Yes, Cordelia Fine is still around and is still writing books arguing against gender differences. But she’s starting to sound really defensive, basically the literary equivalent of “I know I’m going to be downvoted to hell for this, but…”. Meanwhile, other scientists are doing a good job pointing out the flaws in her books and conducting studies like this biggest-ever look at male vs. female brain differences, this magisterial look at personality differences, et cetera – not to mention great and widely-accepted work on how intersex people take on more characteristics of their hormonal than their social gender (honestly, we should probably thank transgender people for making this field socially acceptable again). People talk a lot about how Larry Summers was fired from Harvard for talking about male vs. female differences, but Steven Pinker did a whole debate on this and remains a Harvard professor.

Even things about genetic psychological differences between population groups are less bold and maverick-y than their proponents like to think. The relevant surveys I know trying to elicit scientific consensus (1, 2, 3) all find that, when asked anonymously, most scientists think these differences explain about 25% – 50% of variance.

I hate to bring that up, because it’ll probably start a flame war in the comments, but I think it’s important as a sign of exactly how hard it is to politicize science. Global warming skeptics talk about how maybe the scientific consensus on global warming is false because climatologists face political pressure to bias their results in favor of the theory. But scientists studying these areas face much more political pressure, and as long as you give the surveys anonymously they’re happy to express horrendously taboo opinions. This is about the strongest evidence in favor of the consensus on global warming – and scientific consensus in general – that I could imagine.

4. Nuture Assumption and Blank Slatism: The prologue of the first edition of The Nurture Assumption is Judith Rich Harris telling her “maverick genius kept down by hidebound reactionaries” story. But the prologue of the second edition is her being much more hopeful:

To some extent at least, times have changed…there is now more acceptance of the idea that behavior is influenced by genes and that individual differences in behavior are due in part to differnces in genes. People are more willing to admit that children can inherit behavioral quirks and personality characteristics…was it this cultural shift that led to greater acceptance of my theory? Or was it the fact that new findings, consistent with the theory, kept turning up? Over time, the early, angry response to The Nurture Assumption has softened noticeably, both within and outside of academia. Today, the book is widely cited in textbooks and journal articles. It’s assigned and discussed in courses in many colleges and universities; it shows up in exams…in his foreward to the first ediction of The Nurture Assumption, Steven Pinker made a rash prediction about the book: “I predict it will come to be seen as a turning point in the history of psychology”. Perhaps it is too soon to judge whether psychology has rounded a bend; perhaps it will take the perspective of twenty or thirty years. Even at this point, though, there are signs of a slight shift in direction. Within developmental psychology, I’ve noticed that descriptions of procedures and results are beginning to sound a bit defensive. Greater progress has been made in other areas of psychology. And the email I receive from students gives me high hopes for the younger generation coming up.

There were ten years between the first and second editions of The Nurture Assumption. In the almost ten years since the publication of the second edition, my impression is that its ideas have become even more widely-accepted. This month’s edition of the American Journal of Psychiatry, onbe of the top journals in the field, has a great study showing that child abuse does not cause cognitive disability, in contrast to several previous studies in the area. It cites Deary, Plomin, and Ioannidis, hits all of the talking points about genetic confounding of developmental outcomes, and receives glowing endorsement in the journal’s editorial section, which says that “if our causal explanations are wrong, we may be wasting our effort or even doing damage”. Every single psychiatrist in the country is getting exposed to this way of thinking.

And this has real results. I got to present a summary of behavioral genetics to a meeting of psychiatrists, including a lot of psychoanalysts, and I was shocked that most of them were at least a little receptive. I think they misunderstood it. I think they carefully raised caveats in exactly the right places to ensure they didn’t have to change anything they were doing. But the overall response was “Oh, yeah, we’ve heard stuff like that, it seems plausible, good thing that for various hard-to-explain reasons none of it applies to us.” This is what the first stage of progress looks like.

5. Intelligence Explosion And AI Risk: This was another place where I and many of my friends thought we were right and the consensus was wrong. It was another place where a lot of self-appointed defenders of the consensus told us we were crackpots and needed to listen to what real scientists thought. And again, when I looked into it, there was no consensus against the idea and lots of prominent researchers were in favor. Going to the Asilomar Conference and seeing a bunch of people from MIT and Harvard talk about how concerned they were really opened my eyes on this. Google now has an AI Ethics Board, Berkeley, Oxford, and MIT have foundations working on it, and people like Elon Musk and Bill Gates are involved. Bostrom’s survey of AI researchers and some more recent and rigorous not-yet-published surveys I’ve heard about confirm the impression. Nobody would ever say there’s a scientific consensus in favor of Bostrom’s theories. But at this point I think it’s also indefensible to say there’s a consensus against.

Bostrom first started writing about these sorts of things extensively in the early 2000s, so there was really only a ten-year gap between entering the intellectual environment and it becoming a (mostly) accepted part of the established field. Those ten years felt pretty long while we were in them, but the ability of a field to accept an on-the-face-of-it completely-insane-sounding theory within ten years seems to me a very strong argument against the hidebound-reactionaries theory and a very strong argument for considering scientific consenses to be unreasonably effective.

6. IQ: Another case where I worried about apparent failure of scientific consensus due to politically bias. I certainly encountered a lot of falsehoods around this when I was younger. My high school psychology textbook included a section claiming that all IQ tests were biased towards rich white people because they were based entirely on questions like “how many shots below par is a bogey?” Then it presented an “alternate IQ test” which “proved” that poor minorities had higher IQs than rich whites by asking some other questions with the opposite bias (I think they were about slang for drugs – certainly an interesting way to fight stereotypes). This kind of thing naturally made me assume that nobody had any idea what was actually in IQ tests and scientists were idiots.

But more recently I’ve been reading actual surveys, which find that about 97% of expert psychologists and 85% of applied psychologists agree that IQ tests measure cognitive ability “reasonably well”. And 77% of expert psychologists and 63% of applied psychologists agree IQ tests are culture-fair (with slightly different numbers depending on how you ask the question, but always about 50% of both groups).

This seems like less of a problem with expert consensus, and more of a problem of nobody else (including textbook writers!) listening to experts who are continually trying to beat reality into people’s heads. But I have a vague memory of having recently seen a survey (which I can’t find) that even experts in softer fields like sociology are generally in favor of IQ and admit that it has its uses. And even some left/liberal sources like Vox and Freddie deBoer are aware of the consensus and willing to respect it.

At the same time, I’ve encountered some people like Borsboom and Nostalgebraist who have relatively sophisticated (and limited) critiques of IQ, and who have allowed me to round off other people’s less-well-framed critiques to something more like what they are saying and less like the stupid things my high school textbook said.

So it seems to me that generally experts agree with reasonable statements about IQ, and where they seem to disagree they may hold reasonable disagreements rather than unreasonable ones. Again, where this fails is not in the experts but in the ability of people who don’t listen to the experts to get disproportionate social power and hide the existence of the expert consensus.


Last week I wrote about universally-known criticisms of economists, like “they’re silly for assuming everyone behaves perfectly rationally”:

My impression is that economists not only know about these criticisms, but invented them. During the last few paradigm shifts in economics, the new guard levied these complaints against the old guard, mostly won, and their arguments percolated down into the culture as The Correct Arguments To Use Against Economics. Now the new guard is doing their own thing – behavioral economics, experimental economics, economics of effective government intervention. The new paradigm probably has a lot of problems too, but it’s a pretty good bet that random people you stop on the street aren’t going to know about them.

The same pattern explains a lot of my concerns above. I knew some criticisms of a scientific paradigm. They seemed right. I concluded that scientists weren’t very smart and maybe I was smarter. I should have concluded that some cutting-edge scientists were making good criticisms of an old paradigm. I can still flatter myself by saying that it’s no small achievement to recognize a new paradigm early and bet on the winning horse. But the pattern I was seeing was part of the process of science, not a condemnation of it.

Most people understand this intuitively about past paradigm shifts. When a creationist says that we can’t trust science because it used to believe in phlogiston and now it believes in combustion, we correctly respond that this is exactly why we can trust science. But this lesson doesn’t always generalize when you’re in the middle of a paradigm shift right now and having trouble seeing the other side.

I realize I’m (ironically) risking making my narrative of scientific success unfalsifiable. Suppose someone wants to argue that scientific consensus is wrong. If they point to something it used to be wrong about, I can respond “Yes, but it self-corrected and it’s correct now, so that’s fine.” If they point to something where cutting-edge scientists say it’s wrong but nobody else agrees, I can respond “Yes, this is what the beginning of a paradigm shift looks like, so that’s fine”. And if they point to something where nobody in the field thinks it’s wrong, I can say “You’re a crackpot for going against all reputable scientists; the problem is with you.” And if later they turn out to be right, and everyone acknowledges it, I can say “Yes, but it self-corrected and it’s correct now, so that’s fine.”

(and I’m making it even easier for myself in that I say “scientific consensus for” when I probably mean “no scientific consensus against”. I don’t claim that 90%+ of scientists always believe true things, only that there are very few cases where 90%+ of scientists believe things which smarter people know to be false.)

Against this I can only offer a personal narrative: the only light I have by which to judge scientific consensus is my own Inside View assessment of what seems correct. Again and again I have tried to defy scientific consensus. And every time, I either find that I am wrong, find that I am a few years ahead of a trend that most scientists eventually agree with, or find that what I thought was “scientific consensus” was actually a fiction peddled by biased industry or media sources slandering a scientific community which actually had a much more sophisticated picture. My history of trying to fight scientific consensus has been a Man Who Was Thursday-esque series of embarassments as I find again and again that my supposed enemy agrees with me and is even better at what I am trying to do than I am.

Scientific consensus hasn’t just been accurate, it’s been unreasonably accurate. Humans are fallible beings. They are not known for their ability the change their mind, to willingly accept new information, or to put truth-seeking above political squabbles. And our modern society is not exactly known for being an apolitical philosopher-kingdom with strong truth-seeking institutions completely immune from partisan pressure. I feel a deep temptation to sympathize with global warming denialists who worry that the climatological consensus is biased politicized crap, because that is exactly the sort of thing which I would expect to come out of our biased politicized crappy society. Yet again and again I have seen examples of scientific fields that have maintained strong commitments to the truth in the face of pressure that would shatter any lesser institution. I’ve seen fields where people believe incredibly-bizarre sounding things that will get them mocked at cocktail parties just because those things seem to be backed by the majority of the evidence. I’ve even seen people change their minds, in spite of all the incentives to the contrary. I can’t explain this. The idea that scientific consensus is almost always an accurate reflection of the best knowledge we have at the time seems even more flabbergasting than any particular idea that scientists might or might not believe. But it seems to be true.

(note that I’m talking about “scientific consensus” to mean a very high-level pattern, consisting of hundreds of scientists over the space of decades evaluating a broad body of work. Any individual study is still probably total garbage.)

Given how weird all of this is, I realize there’s another possible bias here that should be taken very seriously – which is that I’m wrong about one or both sides of this. Which is more likely: that Science always agrees with Truth? Or that one guy’s perception of Science always agrees with that same guy’s perception of Truth? The latter gives me two degrees of freedom: I can either cherry-pick experts who agree with me and declare them to be Consensus, or I can conform my opinions to consensus so slavishly that I end up discovering only that Consensus agrees with itself. I don’t feel like I’m making this kind of mistake. But then again, nobody ever feels like they’re being biased.

But if I’m making this mistake, I think it’s at least a better mistake than the one where people dream up stories about being mavericks persecuted by hidebound reactionaries. This mistake at least sets the terms of debate as “let’s try to ascertain what the scientific community thinks” and forbids me from believing completely crackpottish things. And it encourages trust in one of our more trustworthy public institutions, always a prosocial sort of thing to do. I would rather have a world of people debating who agrees with scientific consensus or not, than a world of people debating whether scientific consensus is even valuable.

There are two caveats to the above. First, I think it’s dangerous to promote a norm of agreeing with scientific consensus, insofar as that helps encourage exactly the mistakes about the nature of consensus that I discussed above. When poorly-informed diet industry gurus support the Bad Old Paradigm, their rallying cry is usually “You’re a stupid crackpot, bow to the scientific consensus which agrees with me”. I gave three examples above of cases where I would have gotten the scientific consensus 100% wrong if I didn’t have access to a formal survey of scientific experts. In a world where these surveys had never been done – or some existing field without these surveys – or some field where these surveys have been done inaccurately or in a biased manner – people will often believe the consensus to be the opposite of what it really is. In those cases, demands that people respect consensus can be used to shut down people who are actually right – the field-wide equivalent of calling true facts you don’t like debunked and well-refuted. I see this happening all the time and I worry that waxing too poetically about the unreasonable effectiveness of scientific consensus will only serve to empower these people. Goodhart’s Law says that a measure which becomes a target ceases to be a useful measure, so we should be reluctant to target scientific consensus too strongly.

And second, I think that even when the Outside View tells you that the consensus is correct, you should continue pursuing your Inside View hunch that it isn’t. This avoids awkward situations like every individual scientist doubting the consensus, but suppressing their doubts because the “scientific consensus” has to be right.

So maybe the things I’m saying about scientific consensus aren’t very actionable. But respecting scientific consensus in a non-actionable way is a lot less exhausting than believing yourself to be against it, and talking about how you’re against it, and taking flak for being against it. And in the same way it’s helpful to believe that God is good, even if He never really gets around to doing much about it, so it’s reassuring to be able to have faith in our institutions every so often.

Read the whole story
2 days ago
Share this story
Next Page of Stories