Links 11/18: MayflowURL

In 532, the Byzantines and Persians signed what they called The Perpetual Peace, so named because it was expected to last forever. It lasted eight years. After the ensuing war, the Byzantines and Persians, now less optimistic, named their new treaty The Fifty Year Peace. It lasted ten years.

Patrick Collison and Michael Nielson on diminishing returns from science. Some of you have already seen my thoughts on this, but I’ll post them here in a week or two.

Wikipedia has a page on Armenia/Azerbaijan relations in the Eurovision Song Contest. Highlights include the time Azerbaijan’s secret police rounded up everyone who voted for Armenia, the time Armenia claimed Azerbaijan cut off the broadcast to prevent people from seeing Armenia winning, and accusations from Azerbaijani officials that vapid Armenian love song “Don’t Deny” was dog-whistling a point about the Armenian Genocide.

Everything You Know About State Education Rankings Is Wrong. Most rating systems rank state education success based on a combined measure which includes amount of money spent as a positive outcome, making it tautological to “prove” that more funding improves state performance. See also economists’ Stan Liebowitz and Matthew Kelly corrected ranking table, which also adjusts for some confounders.

The most significant Christian schism of the past five hundred years happened last month, when the Russian Orthodox Church severed ties with the Eastern Orthodox Patriarch of Constantinople due to an argument about Ukraine.

California may allow marijuana, and it may allow alcohol, but at least it’s taking a strong stance against cocktails that include CBD, for some reason.

Recent news in scientific publishing: two statisticians launch RESEARCHERS.ONE (site, Andrew Gelman blog post), a “souped-up Arxiv with pre- and post-publication review”. And Elsevier files a lawsuit forcing a Swedish ISP to ban Sci-Hub; the ISP complies but also bans Elsevier. Also: preregistration works.

Related?: A Chinese barbecue restaurant named itself The Lancet after a top medical journal, and is offering discounts for researchers based on the impact factor of the journals they’ve published in. (h/t Julia Galef)

Experimental archaeology is the practice of doing things we think ancient people might have done to learn more about the details. For example, the Trireme Trust built and rowed a functional Greek trireme to learn more about how triremes worked.

Researchers crack the brain’s code for storing faces (paper, news article), describing it as “a high-dimensional analogy of the familiar RGB code for colors, allowing realistic faces to be accurately decoded with…a small number of cells”.

The Alpine-Himalayan orogenic belt connects the Pyrenees, Alps, Carpathians, Caucasus, Zagros, Tian Shan, and Himalayan ranges.

In what might be the most impressive temper tantrum of all time, the Saudis, angry about Qatar’s support for regional enemy Iran, are planning to dig a giant canal to turn Qatar into an island.

Did you know there are still object-level arguments about libertarianism sometime? It’s true! See Bryan Caplan’s delightfully named Optimality Vs. Fire. Another interesting Caplan: The Triumph Of Ayn Rand’s Worst Idea.

I am always a sucker for the “X as dril tweets” genre, so here is philosophers as dril tweets. EG:


If you want to see all of (someone’s idiosyncratic and dubious selection of what counts as) the rationality-related subreddits in one place, there’s now a Rationality Reddit Feed. Also, gwern has a subreddit now.

Sarah Kliff at Vox is trying to bring transparency to ER prices with a database of what each hospital’s fees are (though it doesn’t look like it’s the kind of transparency where you’re allowed to see the database, apparently for medical privacy law reasons). If you have a recent ER bill, you can submit here, or you can see some of Vox’s reporting on the issue here.

Related: if you missed your previous opportunity to write about effective altruism for Vox, they’re hiring another effective altruism writer/reporter. You can see some of the excellent work by their current EA reporter here.

Science disproves your intrusive thoughts: Most Initial Conversations Go Better Than People Think.

Scandal at meta-analysis producer the Cochrane Collaboration as board members resign en masse. The story seems to go like this: The Collaboration did a meta-analysis showing that HPV vaccines are safe and effective. Cochrane board member Peter Gøtzsche (previously featured here as author of my favorite study on the placebo effect) wrote a savage takedown in the British Medical Journal saying the HPV review did not meet Cochrane standards and should not have been published. The Collaboration’s Board was apparently angry that he took this dispute public and a bare majority voted to expel him. Then the other half of the board stepped down in protest. So much for the one organization we were previously able to trust 🙁

And another academic scandal: Eiko Fried and James Coyne are two of my favorite psychologists and crusaders for high standards in psychology. They’ve recently been having a bad time. As far as I can understand it, Coyne is (by his own admission) well known for being extremely blunt and not afraid of personal attacks on people he thinks deserve it. Fried wrote an article about how a climate of personal attacks and nastiness in the psychology community have gone too far, and most of his examples were of Coyne. Coyne wrote some things accusing Fried of tone policing, but also sued Fried for “cyberbullying” and spread rumors that he was “aligned with racism”. Now Fried has 100% won the lawsuit, the rumors against him have been debunked, various people have come out saying they were harassed by Coyne (and apparently there was also a case of “assault and battery”!) and various institutions Coyne is affiliated with have unaffiliated with him (or said they were never as affiliated as he claimed). I’m really disappointed in this, but it’s helped crystallize some things for me. First, that although cyberbullying is a big problem, mindlessly cracking down on it is dangerous for exactly the reasons shown here – a cyberbully trying to silence their victim by suing them for cyberbullying (and the “aligned with racism” slur is a parallel warning on the dangers of moral panics). And second, that complaints about “tone policing” can often be a smokescreen for just genuinely being a bad actor.

“Superpermutations” is a term for mathematical objects containing every possible permutation of some number of items. The field recently received a jolt when a proof of the lower bound of an important theorem was discovered to have been posted by an anonymous user on a 4chan thread about how many different ways you could watch anime episodes. Now in an equally weird twist of fate, the upper bound of the same theorem has been proven by sci-fi writer Greg Egan, author of Permutation City.

Let’s Fund (description, site) is a crowdfunding site for effective altruism that helps people discover or coordinate campaigns.

This article is called YouTubers Will Enter Politics And The Ones Who Do Are Probably Going To Win, but it focuses on Kim Kataguiri (age 22, the youngest person ever elected to Brazil’s Congress) and other right-wing YouTubers who won positions in the recent Brazilian elections.

The world’s new tallest statue is India’s Statue Of Unity, a 600-foot high (and impressively realistic) depiction of independence hero Sardar Patel.

In 1861, a Tokugawa-era author published the first Japanese book ever on the newly-contacted land of America, called Osanaetoki Bankokubanashi. Although beautifully illustrated, the content was a bit fanciful…

..and by “a bit fanciful”, I mean that this is a depiction of John Adams asking a mountain fairy to help avenge the death of his mother, who was eaten by a giant snake. I assumed the book had to be fake, but Kyoto University seems to endorse it as real. You can find more of Kapur’s commentary here and the rest of the book here.

Karl Friston, previously the subject of a bemused SSC post, is now the subject of an only-somewhat-bemused Wired story. The way this story presents the free energy principle makes it much more of an obvious match for control theory, so much that I’m wondering if I’m misunderstanding it. Related: some computational neuroscience principles used to make a curiosity-driven AI.

Mathematical proofs small enough to fit on Twitter: every odd integer is the difference of two squares.

From the subreddit: the most successful fraudster of all time may have been Jho Low, a financier who offered to manage Malaysia’s $4 billion dollar sovereign wealth fund, took the $4 billion, and walked away with it.

Ever wonder why charities (and other organizations) that say they have enough funding but complain they can’t find enough good employees don’t just raise the salaries they’re offering until they can? Here’s an 80,000 Hours survey on the topic. The main insight is that if a group has 20 employees and can’t find a 21st, then if they want to raise the open position’s salary by X in order to attract more people, they need to raise all their existing employees’ salaries by X or those employees will reasonably complain they’re getting paid less for the same work. So the cost of raising the salary they’re offering for an empty position is less like X and more like 21X.

Sorry, non-Californians, more on the CA ballot propositions – here’s a table of how the state voted on each vs. how SFers voted vs. how LAers voted vs. what the relevant newspapers endorsed. It looks like everyone is pretty much in alignment except the San Francisco Chronicle, which hates everything.

Many Indo-European languages use euphemisms for “bear”, sometime several layers of euphemism, because of a fear that speaking the bear’s true name might summon it. The English word “bear” is a euphemism originally meaning “brown one”. Inside the quest to reconstruct the bear’s True Name. NB: do not read this article aloud or you might get eaten by bears.

This entry was posted in Uncategorized and tagged . Bookmark the permalink.

309 Responses to Links 11/18: MayflowURL

  1. jastice01 says:

    Missed title opportunity: “The PerpetURL Peace”

  2. Slicer says:

    “And Elsevier files a lawsuit forcing a Swedish ISP to ban Sci-Hub; the ISP complies but also bans Elsevier.”

    Hahaha! It really sounds like the Pirate Party needs to make a comeback in Sweden (although it could be argued that that particular country has bigger problems).

    Anyone interested in the state of academic publishing should probably watch Paywall, which shows what happens when people can’t afford access to research.

    • Baeraad says:

      In the sense that the last thing we need right now is yet another party that none of the others want to cooperate with but which is nonetheless occupying a large enough chunk of the vote that no one can form a viable government, yes.

      (each country has its own sort of political dysfunctions, I guess. America has leaders making overheated allegations towards their political enemies. Sweden has leaders that just quietly and sullenly refuses to budge, presumably out of egalitarian zeal – having to budge would be like saying that the person you’re budging for is better than you, and it’s insufferable that anyone should think they’re better than you!)

    • luispedro says:

      What the ISP did was funny, but wasn’t it a clear violation of Net Neutrality?

      Maybe Sweden doesn’t have that type of rules, but I have seen no pushback on the idea that ISPs can block entities they are involved in lawsuits in as a legal strategy.

      • Tatterdemalion says:

        That worried me, too.

      • Chipsa says:

        They aren’t involved in the lawsuit. They’re just the target of a court order.

      • Douglas Knight says:

        EU law applies in Sweden, but I do not believe that this violates it. When people talk about Net Neutrality they almost never talk about DNS. ISPs fuck with DNS all the time and I do not believe that it violates any proposed or existing law, except maybe the Chilean law, which is so vague that everything is whim.

  3. Le Maistre Chat says:

    “The most significant Christian schism of the past five hundred years happened last month, when the Russian Orthodox Church severed ties with the Eastern Orthodox Patriarch of Constantinople due to an argument about Ukraine.”

    !!!
    Come on guys, just no. 🙁

    • theredsheep says:

      Subtext (bearing in mind that I am only an Orthodox layman, not an expert): there’s an ongoing rivalry between the Ecumenical Patriarchate in Constantinople (yes, we know it’s Istanbul, we just don’t care) and the Patriarchate in Moscow over who is the leader of the Orthodox World. The EP is officially our top dog, but in terms of formal power he’s nothing like the Pope; he can call a council, but all he does at it is preside like the American VP over the Senate. He wears the biggest hat and carries the most stuff at liturgy, but really no one person is properly “in charge” of Orthodoxy, which is why we’re still run based on precedents from the first millennium. Everything must be decided by clear consensus. Change comes slowly if at all.

      The EP’s actual flock amounts to a miserable handful of Christians in Turkey proper, since the Turkish government has made a long habit of making the Orthodox as uncomfortable as possible. Leaving aside the turn-of-the-twentieth-century genocides, devshirme, etc., it has policies like outlawing clerical garb for everyone but the EP himself, or requiring the EP to be a Turkish citizen while closing down the last Turkish seminary. Probably they’ll win in the end, because nobody but the Orthodox outside of Turkey cares, and Moscow doesn’t like them enough to tell the Turks to stop being pricks. But, speaking of the Orthodox outside of Turkey, remember how I said we’re run on ancient precedents? One of those is that Constantinople gets jurisdiction over territories not assigned to any other of the big patriarchates. Since America wasn’t a thing in AD 450 or whenever, that means America is under his jurisdiction. Our tithes are a big chunk of what keeps his lights running. And, uh, most of America doesn’t like Russia very much right now …

      After Constantinople fell, Russia took up its mantle (Ivan III married Sophia Palaiologos, niece to the last Emperor), and declared itself the Third Rome, patron and protector of Orthodoxy. It’s been in many respects the heart of the Orthodox world for centuries, even under intermittent state suppression. And they do appear to be in the right here, from what little I’ve troubled to read on the subject. The EP has recognized the independence of a small group of Ukrainian schismatics who aren’t in communion with anyone else, based largely on the political fallout from the ongoing conflict there. Which is itself tied into Russian spiritual identity being linked to Kiev rather than Moscow, but let’s not get into that.

      Long story short (er): this is probably not going to be significant in the long term, simply because the EP is so politically weak and there seems to be general agreement that those Ukrainian rebels do not constitute a legitimate church no matter what he says. In some ways, we Orthodox have it easy; our church scandals tend to involve plain old monetary corruption (“I know I put that retired clergy fund around here somewhere”) or obscure battles over turf based on how one parses a comma splice in the minutes from a council meeting held by Emperor Justinian I in 530. They’re degrading and embarrassing, but have little effect on church behavior at the local level. My parish priest hasn’t even mentioned it (though admittedly we’ve been a bit distracted by Hurricane Michael). I imagine it would be ticklish for a high-level Russian cleric trying to celebrate liturgy in Greece, but the rest of us will probably ignore it. At least in America.

      • Le Maistre Chat says:

        there’s an ongoing rivalry between the Ecumenical Patriarchate in Constantinople (yes, we know it’s Istanbul, we just don’t care) and the Patriarchate in Moscow

        It’s Constantinople not Istanbul, and it’s nobody’s business but the Christians.

        I get that, and it’s really sad. I mean, the Turks trying to destroy Orthodoxy within their borders when the EP is top dog because the Church works on ancient tradition was the whole reason for Greek revanchism (“Great Idea“) and why the Christian monarchy of Russia kept wanting to conquer it from the Ottomans.
        The Russians have a point, as the most casual perusal of history by us non-East-Slavs who have no dog in the fight shows that Kiev is the Ur-Russia. But regardless, it would be hugely stupid for Christians of different Orthodox nationalities to be out of communion when abroad. 🙁

        • theredsheep says:

          I’m not sure to what extent it will matter; it’s entirely possible that if a man goes up to the chalice in Moscow and says his baptismal name is “Ioannis,” the priest will blink, look about furtively, and say loudly “the servant of God Ivan receives the precious body and blood of our Lord, God, and Savior Jesus Christ,” and that will be that. It’s also possible that the priest will stand there and lecture the man about the finer points of the pentarchy right there while the line builds up behind him. It may depend on the individual priest and his attitude towards the schism, nationalism, the Ukraine struggle, and how far he’s willing to stretch his individual right of oikonomia to avoid putting a guy on the spot.

          As an American, I have a poor perspective on how this is taken, because America’s in a weird and non-canonical place right now with overlapping jurisdictions; the only church in my area is Greek by culture, and under the EP, but a few years ago I was in Maryland and went to an Antiochian church (which ultimately reports to Damascus in Syria because history). There was no conversion or anything; they’re all the same church, and they’re all in communion now. In practice we’re free to flip between jurisdictions as we please. And we tend not to take these arguments by overseas hierarchs very seriously, because they usually have little effect on us. A couple of years back there was an embarrassing attempt to call an “Ecumenical Council” on Crete that fizzled over (IIRC) whether Jerusalem or Antioch gets jurisdiction over a handful of churches in Qatar. Half the Orthodox world boycotted the meeting. Most American Orthodox today likely don’t remember it, if they even heard of it in the first place. I wince every time I think of it, but then remind myself that the RCC has to handle much worse.

          In historically and ethnically Orthodox countries, I can imagine that things might be very different. But I’m having a hard time even finding a clear account of whose authority was recognized, or what’s happened in the month since. Anybody else know?

        • vV_Vv says:

          The Russians have a point, as the most casual perusal of history by us non-East-Slavs who have no dog in the fight shows that Kiev is the Ur-Russia.

          So Kiev is the Ur-Russia, Jerusalem is the Ur-Israel, Byzantium/Constantinople/Istanbul is the Ur-different things to different people, and so on. It’s interesting how modern conflicts seem to be driven by thousand years old cultural heritages.

      • Bugmaster says:

        I could be wrong, but I suspect that one reason for the schism is political. The Russian Orthodox Church in Moscow has been steadily strengthening its ties to Putin’s regime; meanwhile, most native Ukrainians hate Putin with a passion, and see his co-option of the Church as yet another symptom of hostile dictatorship.

        Whether or not the schism becomes significant depends on whether or not Ukraine will continue to exist as an independent entity… which is about 50/50 at this point, as I see it.

        • theredsheep says:

          Oh, it’s undoubtedly political; since the Orthodox Church is organized at a national level, almost all our disputes have some political dimension. However, NB that Ukraine was put under Moscow’s authority centuries ago–this is not a novel suggestion of Putin’s, or anything like that.

      • Trevor Adcock says:

        “Constantinople (yes, we know it’s Istanbul, we just don’t care)”

        Isn’t Istanbul just the turkicized version of the phrase “into the city” in Greek and was adopted, because the Greeks used it to refer to the city?

        My understanding was the Ottomans still officially referred to the city as Constantinople until the founding of the Republic of Turkey; which, then adopted the common name as the official one.

        • Joseph Greenwood says:

          My Attic Greek professor told me that was the etymology of “Istanbul,” so I am quite confident it is correct.

        • Mitchell Powell says:

          There’s a form “Stamboul” that was used for a while in Turkish. Personally, it seems a little more likely to me that this comes from the “Stan” and “-ple” in Constantinople, but I’ll admit that I just saw that on Wikipedia and thought it looked more plausible that “eis tan polis”. But I’m not necessarily the guy to judge how plausible etymologies are.

          • keaswaran says:

            The “-ple” in “Contantinople” is actually the same as “polis” of “eis tan polis”, because it’s actually “Constantine-polis”. So it’s just about whether “Stam” is originally “[ei]s tan” or “[Con]stan[tin]” (in either case, the “m” is just how an “n” changes in front of a p/b in Greek).

      • JohnBuridan says:

        It’s worth mentioning the Eastern Catholic Element of this conflict. Since time immemorial, i.e. the 16th Century the Catholic Church has been trying to increase its influence in Western Ukraine.

        Jesuits in the 18th century ensured that the Eastern Catholic Church remained as close as possible to Eastern Orthodoxy in liturgy, privileges, and practice – no rosaries, no Western garb, no Gregorian calendar, and immense autonomy from Rome. Keep the Eastern Catholics like the Orthodox, so the thinking goes, and the Orthodox will be more willing to become Greek Catholic.

        The western part of Ukraine is 30% Greek Catholic and the schismatic group are mostly nationalists in the same region plus in central Ukraine. If you look at a map of Ukrainian seminaries you will see that most are in the West, and there are none in the far east. So, while theredsheep is right that it will not be such a big deal, I would expect these currently Orthodox lay people in the west to drift Greek Catholic over the next two generations.

        The arc of history is bent by 18th century Jesuits.

        • theredsheep says:

          Well, I gather that the main distinction between such Catholics and Orthodox amounts to the filioque and sending money to the Vatican–I don’t think they care if one practices hesychasm, etc.–so it might depend more on how cool Ukrainians are with the Pope. I gather Russians, at least, do not like Pope Francis at all, but for all I know Ukrainians may be totally different.

  4. Johannes D says:

    Mathematical proofs small enough to fit on Twitter: every odd integer is the difference of two squares.

    A nice and intuitive geometric proof. Note that it actually proves a stronger theorem: every odd integer is the difference of two consecutive squares. From the same proof it is also easy to see how the implication is true in the other direction: every difference of two consecutive squares is odd!

    An algebraic proof also fits in a tweet:

    (n+1)² – n²
    = n² + 2n + 1 – n²
    = 2n + 1, qed.

    That is to say, every difference of two consecutive squares is of form 2n+1, that is, odd, and also that for every n, the number 2n+1 (that is, every odd number) is the difference of n² and (n+1)².

    • The Pachyderminator says:

      From the same proof it is also easy to see how the implication is true in the other direction: every difference of two consecutive squares is odd!

      Isn’t this trivially obvious? An odd number squared is always odd, an even number squared is always even.

      • Johannes D says:

        I guess it is pretty obvious, although that particular argument didn’t even occur to me (not that I spent too much time on it because the proof I presented is already fairly trivial).

    • Luke Perrin says:

      Also every number divisible by 4 is the difference of two squares. We can do a graphical proof

      ■  ■  ■  ■  ■  □
      ◫              □
      ◫              □
      ◫              □
      ◫              □
      ◫  ◩  ◩  ◩  ◩  ◩

      or an algebraic one: 4n = (n+1)² – (n-1)².

    • moscanarius says:

      I really liked this one.

  5. Nabil ad Dajjal says:

    The most interesting part of the bear taboo story, to me anyway, is the connection between the reconstructed PIE word for bear and the word for destruction / demons.

    It just works on a lot of levels. The thing that immediately came to mind for me was the idea of berserkergang. A warrior is possessed by a bear-spirit / demon so that he can unleash that destruction on the battlefield.

    I always forget that long acronym from Unsong about nothing being a coincidence but pretend that I wrote it here.

    • Statismagician says:

      Your having forgotten the this-is-not-a-coincidence-because-nothing-is-ever-a-coincidence acronym is, itself, presumably not-a-coincidence-because-nothing-is-ever-a-coincidence.

    • Salem says:

      A berserker is not one possessed by the spirit of the bear, but one who wears a bearskin shirt (serk).

      • Slicer says:

        The original Brownshirts, then?

        (I’m out the door already…)

        • Le Maistre Chat says:

          Exit, pursued by a bear.

          • Steve Sailer says:

            My impression is that bears were larger and more aggressive in the past. To take a recent example, large, scary grizzly bears, Ursus horribilis, lived in Los Angeles’s San Fernando Valley at least as recently as the 1850s. But grizzlies were extinct in all of California by the 1920s. Black bears have migrated in to fill the grizzlies’ ecological niche, but they are far less frightening to people.

          • albatross11 says:

            There’s a memorable bit in the memoirs of Lewis and Clark where basically the whole party ends up in trees, trying to reload their rifles so they can shoot the extremely pissed-off bear trying to reach them from below.

          • acymetric says:

            Am I wrong to be skeptical that they were able to escape up into trees faster than the bear could follow them? My understanding is that bears are fairly good at climbing.

          • gbdub says:

            I suspect humans being chased by bears are pretty good tree climbers too. In seriousness, my impression of bears’ climbing ability is “surprisingly nimble given how bulky and awkward looking they are“. The problem with climbing a tree is not that the bear can necessarily climb it faster, it’s that they can climb it at all, given the motivation to do so, and at that point you’ll be trapped in a tree with a pissed off bear.

          • The Nybbler says:

            Black bears can climb trees easily. Grizzlies can climb them, probably better than you can, but it’s harder for them (than for black bears) so they might decide to wait for you to come down.

          • Le Maistre Chat says:

            @Steve Sailer: This would be hard to measure scientifically (even anecdotes are thin on the ground), but brown bears vary enormously in… enormity, from the modern Syrian to the Kodiak. Wouldn’t surprise me if the species is divided into “races” some of which have been selected for smaller size over many, many generations of co-existence with human settlements. In Eurasia they get biggest in the Russian steppe, which until recent centuries had nomadic rather than Russian villager populations, and in North America, eg, the extinct California grizzly was noted for being bigger than those that coexisted with the denser Northwest Coast Indians.

          • baconbits9 says:

            The advantage of climbing a tree to avoid a bear is that a 150-200 lb human can squeeze out onto a thinner branch without breaking it than a 400 lb bear can, so you want to climb a tree to the point where it wont handle the bear’s weight.

          • acymetric says:

            @baconbits:

            Assuming (perhaps correctly) that the bear can judge this well and won’t pursue you on a branch that might break. Otherwise things look bleak for both of you.

          • The Nybbler says:

            Or that the bear won’t just rip down the branch you’re on while holding onto the trunk. In which case bad for you, good for the bear.

          • John Schilling says:

            Black bears are very good at both climbing and at being afraid of critters smaller but more confident than they are. But if they do fight, they tend to make it to the death because they can’t afford half-hearted fights. Winning strategy is usually to be more confident than the bear and let him know it.

            Brown bears (including Grizzly and Kodiak) are much less good at climbing, but much more inclined to “let’s see what fighting this thing is like; it’s not like I’m going to get seriously hurt finding out”. Winning strategy is usually to make enough noise that if the bear isn’t up for a fight it will know where to go to avoid one, and if it shows up anyway either shoot it a whole lot, climb a tree, or play dead. If all your experience was with black bears, this is your day to learn something exciting and new.

            Polar bears suck at climbing, but that hardly matters in their native terrain and they will explicitly hunt people with the intent of killing and eating them. Bring enough gun.

            And just to be complete, Moose can’t climb at all but they are are approximately the size and power of a main battle tank and they do not like to be challenged. The nastiness of their bites is both overrated and irrelevant; just back away.

          • acymetric says:

            I suppose at the point where you’re running from a large bear you may not have better options than climbing trees, but it still doesn’t seem like a winning strategy regardless of bear type.

            I think people are confusing “can’t climb as well as a black bear” with “can’t climb better than a human”.

            My main point is that the memoirs of Lewis and Clarke may (are probably) subject to some embellishment, and this certainly seems like a good candidate for that.

          • albatross11 says:

            I suspect grizzlies have a scaling problem there, too–falling ten feet out of a tree is not going to go well for an animal as large as a grizzly.

          • Steve Sailer says:

            So it seems that humans have tended to exterminate bears that were big and aggressive wherever human populations were dense enough and rich enough to undertake the effort.

            In this decade there have been 3 fatal black bear attacks in the lower 48 states, one by a caged bear and two by wild bears, one in Arizona, and one in New Jersey of all places. There have been 8 fatal brown (grizzly) bear attacks in the lower 48 in this decade, even though grizzlies are fairly rare south of the Canadian border.

            As a Californian, I regret that the brown bear on the state flag is extinct. But as a hiker, I’d much rather take my chances with black bears, who have always struck me in my encounters with them as pretty reasonable live-and-let-live creatures.

          • Aftagley says:

            Polar bears suck at climbing, but that hardly matters in their native terrain and they will explicitly hunt people with the intent of killing and eating them. Bring enough gun.

            As someone who’s been hunted (unsuccessfully, thank goodness) by a polar bear, I can confirm. They are terrifying and only sometimes deign to be scared of gunfire.

    • quaelegit says:

      This post made me realize that I’ve been reading/spelling “beRserker” without its first R… so thanks for the correction!

      Also, if anyone wants to read more on taboo deformation, my other favorite blog has a long discussion about it here — includes other hypothesized examples and more discussion of Indo-European sound changes. (The two concepts seem closely connected — discussions of one usually lead to the other.)

    • Unirt says:

      This bear taboo thing stretches broader than just Anglo-Germanic languages. The Finno-Ugric nations inhabiting the same sort of habitats nearby have probably the same thing going on in several layers, though the word roots are different. Bear in Finnish (karhu) and Estonian (karu) means “hairy” I think, which is analogous (but not homologous) to “bear” in neighboring Baltic languages (in which it also means “hairy”, but has a totally different root). And at some later times (18th-19th century?) Estonians started avoiding “karu” when out hunting and using a different nickname “mesikäpp” (similar to the slavic version “honey-eater” in the meaning but not in sound). Looks like the term has gone through an euphemism treadmill for a couple of times. Also, they din’t bother to make up their own euphemisms, but borrowed those of their neighbors.

  6. Le Maistre Chat says:

    “Many Indo-European languages use euphemisms for ‘bear’, sometime several layers of euphemism, because of a fear that speaking the bear’s true name might summon it.”

    And apparently only northern tribes were that terrified of summoning bears, as Indo-Europeanists don’t believe Greek “arktos”, Sanskrit “rkshas”, and possible cognates were euphemisms for an even earlier word. Note though that “rakshas” is Sanskrit for “protector”, and a variant for the class of supernatural beings more commonly called rakshasas that heroes fight in the Hindu epics, sometimes translated as “demons” but which we might call “dark elves” because they’re closely related to yakshas, nature spirits with humanoid bodies.

    EDIT: Ninja ad Dajjal!

    • Scott Alexander says:

      I think I read something else on this speculating that actually, “destroyer” sounds quite a bit like a euphemism for some earlier term, in the same way “honey-eater” and “brown thing” are.

      • Le Maistre Chat says:

        Well that would make total sense. This implies that Balto-Slavic and whatever is ancestral to Armenian (Phrygian?) also belong to the “Greco-Aryan” group that used the “destroyer” euphemism, changing it again to the attested Latvian, Lithuanian, Church Slavonic, etc. euphemisms later (a treadmill!). The reason being that Greek, Armenian and Indo-Iranian have been noted as sharing a number of features, but Greek preserves the earlier “centum” pronunciations while the others, together with Balto-Slavic, share what are called “satemization” and the “Ruki sound laws.”

        • Mark Atwood says:

          I wonder if there was ever an original word at all!

          There is no reason there has to be one.

          It could go back to pre-language days proto-human days, when proto-humans learned and taught each other to stay away from the “the really big really fast thing with teeth that isn’t afraid of us, that also liked to range and roam like we do, in the same ranges we like to hunt in, and loved to find our tribes camps and our tribes caches, nom nom nom”.

          When did bears evolve, and from what?

          • The Nybbler says:

            Within human existence, bears evolved from other bears that were still pretty bearlike. And there was the cave bear (shares a recent ancestor with the brown bears), which the Neanderthals at least were well-acquainted with.

          • Le Maistre Chat says:

            On that note, you know what brown bears are hypothesized to have been too scared of to move south of Beringia and Alaska until it went extinct? A different lineage of bears!

          • Steve Sailer says:

            Prehistoric bears appear to have been bigger than current bears on average. My guess would be that humans have taken a hand in the evolution of bears to be smaller and less terrifying.

    • Johannes D says:

      In Finnish, which is not an Indo-European language, the modern word for bear, karhu, is also an euphemism possibly referring to its fur (karhea meaning rough or coarse). The “true name” for a bear would have been otso or ohto, diminutive forms of ancient Proto-Finnic original oksi. But even karhu became too powerful a word and begat multiple synonyms that Finns have historically used instead.

      Fun fact: the binomial name for brown bear is Ursus arctos, literally just the Latin and Greek words for bear put together.

    • S_J says:

      For extra fun: the old-English poem titled Beowulf may use a euphemism for “bear” as the name of the main character.

      Beo-Wulf is a compound word, which would be rendered in modern English as “Bee-Wolf”. That is, something/someone who would be an attacker or antagonists to the the honey-making insect.

      The character Beowulf has some characteristics which might be subtle references to bear-like behavior. Fights the first monster (Grendel) bare-handed, tears off a limb from the monster, apparently has incredible strength.

      There are other folk-tales of the Bear’s Son: a child who grew up among bears, and had the strength (and love for honey!) of those animals. The Bear’s Son is somewhat integrated into human society, and eventually joins with companions to defend a dwelling against a monster. The companions fail, but Bear’s Son is able to tear the monster’s arm off, and then is able to follow the blood trail to the monster’s lair.

      The first part of Beowulf has some elements in common with this, but the similar with the Bear’s Son tale aren’t all overtly there: Beowulf has a big appetite and incredible strength, but a love for honey isn’t mentioned directly. His name gives a clue to that, though. And he kills Grendel in a style very similar to the outline in the previous paragraph.

      Was there some fear that the Bear would be summoned if the elements of the Bear’s Son tale were mentioned too directly?

      Or was it a stylistic choice, part of embedding Beowulf into the historical background that the story is told in?

      • Le Maistre Chat says:

        Beo-Wulf is a compound word, which would be rendered in modern English as “Bee-Wolf”. That is, something/someone who would be an attacker or antagonists to the the honey-making insect.

        Unrelated but the same psychology: Slavonic medvedi, “devour honey”, which is still medved in Russian (East Slavic), Czech (West), etc.
        Does this mean that a male bear who converted to Christianity would be baptized “Michael Medved“?

      • Aging Loser says:

        I think that Grendel’s “arm” is his penis. First, Beowulf couldn’t have torn off Grendel’s arm because to tear off someone’s arm you have to brace yourself against his body and pull/twist outwards, but Beowulf wouldn’t have been tall enough to brace himself against Grendel’s hip or chest or whatever while pulling/twisting outwards in a way that would have torn off the arm. But by grabbing Grendel’s penis from behind, between Grendel’s leg, the penis being let’s say a yard long, Beowulf could have braced himself with one foot against Grendel’s butt and torn the penis off. Second, Grendel’s “mind was flooded with fear” as soon as Beowulf grabs the limb in question — this rings true if his penis is what Beowulf grabbed, not if his arm is what Beowulf grabbed. Third, tearing off the non-neurotypical outsider’s penis is exactly what alpha jocks are supposed to do. Fourth, hanging a defeated freak’s giant penis overhead is obvious frat-boy humor; hanging an arm up there just doesn’t ring as true.

        The real story is that Grendel wanted to hang out at the beerhall with the fratboys, the fratboys said “Get lost, loser,” Grendel got angry and started shoving people, and the alpha fratboy tore off Grendel’s penis from behind. Then they went and killed Grendel’s alcoholic old mom.

        • Aging Loser says:

          Also, if Beowulf had really been grabbing one of Grendel’s wrists with both hands from in front, as the poem indicates, then Grendel could have been smashing Beowulf in the head with his other fist. But if Beowulf was grabbing Grendel’s penis from behind, between Grendel’s legs, then Grendel might well have been unable to connect with Beowulf’s head with either fist.

      • Conrad Honcho says:

        but a love for honey isn’t mentioned directly.

        The first parts of Beowulf take place mostly in a mead hall. Mead is made from honey.

    • Le Maistre Chat says:

      Indo-European phylogeny, shortened to the point of possible inaccuracy:

      When Sir William Jones figured out in the late 18th century that Sanskrit with all its descendants, Greek and Latin (ergo all Romance languages) were all related, the other living languages quickly deduced to be “Indo-European” included the Persian and its close relatives (Armenian was originally put here), the Germanic group, Slavic group, Gaelic and Welsh, Latvian and Lithuanian, and Albanian.
      Of course, there are also extinct Indo-European languages. One of the most dramatically scientific moments in linguistics came when “Hittite” was discovered in the early 20th century and supported the falsifiable “laryngeal hypothesis“. Hittite turned out to belong to an extinct branch that included its contemporary Luwian and several pre-Greek languages of Anatolia such as Lydian and Lycian. The other branches of the language family all have more in common with each other than with Anatolian, suggesting Proto-Indo-Europeans first split into Anatolian and Everybody Else.
      Another important extinct branch is Tocharian, which is first attested in texts of the 6th century AD… from the Tarim Oasis of Xinjiang, China. This extinct culture is also famous for its light-haired “Caucasoid” mummies, and it’s common to think that pre-Tocharian was brought here in 2000 BC by speakers with this phenotype (but in any other context, calling the PIE speakers “white” or any synonym will be incredibly fraught). Then there’s the continental Celtic languages, Latin’s extinct relatives (Italic branch), Phrygian, Illyrian, Thracian, and Dacian from Classical antiquity…

      After Anatolian splits off, a major division is the centum/satem isogloss. “Centum” is generally believed to be the more archaic pronunciations, so the change in India and Iranian languages (properly one branch, Indo-Iranian), Armenian, Baltic and Slavic (properly Balto-Slavic) is called “satemization.” Every single one of these branches, possibly including Armenian, also went through the “Ruki sound laws“.
      Conversely, a “centum group” would be paraphyletic: Celtic and Italic may be closely related, but are not especially close to Greek, Germanic or Tocharian, nor any of these to each other. As I mentioned, there are shared features that hint Greek may have originally been at the end of a dialect continuum that included Illyrian (when Albanian is taken to be an Illyrian language), Armenian (Herodotus asserted as uncontroversial that the Armenians were Phrygian colonists), Balto-Slavic and Indo-Iranian, esp. the extinct Iranian languages spoken south of Balto-Slavic on the Pontic steppe.

      • kaakitwitaasota says:

        Is satem really monophylatic? The patterning suggests that satemization was a wave shift and not an innovation in a single Proto-Satem: for example, satemization of *ḱ *ǵ *ǵʰ didn’t occur before *l + back vowel in Balto-Slavic (with some analogical exceptions), and there are similar exceptions to RUKI. Also, the Luwic branch of Anatolian shows satemization, and if you consider morphology rather than phonology, the grouping looks quite different (e.g. Greek, Armenian, Phrygian and Indo-Iranian look like they group together, but Greek and Phrygian are centum).

        (epistemic status: the stuff about Balto-Slavic is probably misremembered from two and a half years ago with Sasha Lubotsky)

        • Le Maistre Chat says:

          Is satem really monophylatic? The patterning suggests that satemization was a wave shift and not an innovation in a single Proto-Satem

          It’s not; any implication otherwise was poor terminology on my part. It was almost certainly a wave shift that caught Thracian but Greek and Phrygian were isolated from, among a group of morphologically similar dialects (because Greek, Armenian, Phrygian and Indo-Iranian do look similar). Indeed we should expect wave shift to be the normative form of change; changes would start in one settlement and then spread to the nearest neighbors, then their neighbors. Because Common Slavic existed so late in history, Slavic languages provide a good example of this, and it would only have been stronger in prehistory when states first didn’t exist and then only people who lived in the palace needed to know a “standard” form.

      • aho bata says:

        Interestingly, neither centum languages nor satem languages preserve the original system of back-of-the-tongue stop consonants in PIE, which made a three-way contrast between labiovelars, velars, and what have traditionally been called ‘palatals’ (which were probably produced further back in the mouth than the name implies). Roughly half the languages (centum) have collapsed the palatals with the velars, while the other half has collapsed the velars with the palatals (satem). The only odd one out is Luwian, which preserves the three-way contrast.

        The three-way contrast was already standardly reconstructed before the discovery of Luwian, however. PIE had a rule preventing any two stops with the same place of articulation (of which there were five — besides the three dorsal places, there were bilabial (p, b, bh) and alveolar (t, d, dh) places) from co-occurring in a root. The only apparent exceptions were roots with two velars in the centum languages (deriving from roots with a velar and a palatal), and a different set of roots with two velars in the satem languages (from roots with a velar and a labiovelar). It was also suggestive that an apparently random, non-phonologically conditioned subset of velars in the centum languages corresponded to sibilants in the satem languages, and an apparently random subset of velars in the satem languages corresponded to labiovelars in the centum languages.

        What could the explanation for this pattern of facts be, if not that shortly after IE split into two dialect groups, one underwent satemization and the other centumization? While it’s true that centum languages didn’t just inherit the IE system, the sound change merging ‘palatals’ and ‘velars’ is naturalistic enough that it could have happened multiple times independently in different non-satem subgroups. I noted above that ‘palatals’ is usually thought to be a misnomer, as an unconditioned sound change changing palatals into velars is unattested, as far as I know, in all historical linguistics. But if both sounds were produced further back in the mouth — so that ‘palatal’ sounds were velar and ‘velar’ sounds were uvular — then for them to merge as velars would be very typical, and the sound change that would then have had to take place in the satem languages shifting both uvulars and velars forward would also be naturalistic. So multiple centumizations, affecting all non-satem languages other than Luwian, is a possible alternative to there being a centum group.

        • kaakitwitaasota says:

          For what it’s worth, Kloekhorst thinks *h₂ and *hɜ were originally uvular stops *q and *qʷ, which would mean their stop reflexes in Lycian are archaisms, not innovations. I want to look into how this fits into known root constraints, however.

          A change of palatals into velars is rare, but well-attested in Samoyedic (*š~ś > k) and some dialects of Pashto (š > x), and most prominently in Athabaskan. The *ts series of Proto-Athabaskan became velars /k/ in a few varieties, such as Bearlake Slavey-Hare, and the labialized *čʷ series became velars in Tanacross. (In fact there is a distinction between “satem” Athabaskan, where PAth velars became palatal affricates of some variety, and “centum” Athabaskan, where they did not.)

          Is there any particular reason to think that Tocharian forms a subgroup with western Europe (other than the r-passive)? In many respects it seems to pattern best with Anatolian–not that they form a subgroup (they don’t) but that both must have split off very early and developed independently (there are a number of shared lexical items, and some grammatical patterning, such as the feminine, which is nonexistent in Anatolian and half-developed in Tocharian). My presumption is that in the very early stages of Core IE, after Anatolian split off, there were early exodi to the west of Italic, Celtic, Phrygian/Greek and Germanic and to the east of Tocharian. Satemization then developed and spread through the middle of what must still at the time have been a dialect continuum stretching from Germany or France to the Indus, failing to reach Tocharian (too isolated) or the languages of western Europe (too far west? unintelligible? who can say?)

    • aho bata says:

      The words for bear in Greek (arktos), Sanskrit (rkshas), Latin (ursus), Old Irish (art) and Hittite (hartaggas) are uncontroversially recognized as cognates, going back to a PIE word reconstructed as *h2rtkos (with h2 being a consonant whose pronunciation can only be speculated on).

      The sound correspondences are somewhat unusual — Sanskrit sh normally corresponds to Greek s, Greek coronal stops normally correspond to Sanskrit t — but ‘bear’ isn’t the only word that shows this set. Others include ‘carpenter’ (Greek tekton, Sanskrit takshan) and ‘earth’ (Greek khthon, Sanskrit ksham). In the 19th-early 20th centuries these were thought to reflect a phoneme that would have merged with different existing phonemes in the daughter languages, but nowadays this correspondence is generally reconstructed in PIE as *tk, with an s being inserted in between (as is already known to happen between two successive coronal stops) and the k moving to the front of the cluster. (I can explain why it is reconstructed this way in a separate comment, if anybody is interested.) The s was then lost in Greek (as it regularly was between consonants) and the t in Sanskrit.

      The other sound changes are simply the regular outcomes. h2 becomes a in Greek, and its presence is assured by the Hittite cognate. All o become a in Sanskrit (except in open syllables, where they become long a)., and s always becomes sh after k. Hittite orthography, inherited from the Akkadian writing system, only has characters representing a single vowel, CV, VC, and CVC — nothing for a single consonant — so any cluster of three or more consonants has to be broken up by a (probably) purely orthographic a. Indo-European voiceless consonants are rendered with doubled consonants. The Akkadian script made a distinction between voiced g and voiceless k, but Hittite uses them indiscriminately, leading some linguists to conjecture that the voicing distinction in PIE become one of length or tenseness in Hittite. The Old Irish reflex of tk/kts, t, seems unlikely at first glance, but a transitional form is attested in Cisalpine Gaulish dewogdonion (‘of gods and men’; cf. Sanskrit deva for ‘gods’ (not an exact cognate, but from the same root, div); the latter half of the compound is derived from ‘earth’ (khthon, ksham) by the adjectival suffix -i-, — so literally ‘earthlings’ — and -on is the genitive plural suffix).

      Unfortunately this means rkshas ‘bear’ can’t be connected to rakshas ‘demon’. Rakshas is derived from the root raksh ‘protect’; early sources gloss it as ‘that from which the sacrifice is protected’, although the original meaning may have been ‘protector’. Raksh is the regular outcome of PIE h2leks, incidentally cognate with Greek alex-, as in Alexander (protect-man).

      And to be a spoilsport twice over, the fact that the ‘bear’ etymon isn’t attested in Germanic is no evidence that it was tabooed. It is by far the exception for the same meaning to survive in cognate words across all the IE daughters, and the pattern of replacement often doesn’t map onto any known subgroups. Sometimes words just fall out of use. Also worth noting that the connection to ‘brown’ is contested, as the root *gwher- ‘animal’, the source of Greek ther (cf. English theria), would also regularly give rise to ber- in Germanic (in spite of the surface implausibility — again, I can lay out the argument in a separate comment), and the semantic connection is much less problematic.

      • quaelegit says:

        I’m not the person you responded to but thanks for the detailed explanation! I have no background in linguistics (just read about it on the internet), and every time I read these explanations of PIE reconstructions and sound changes in daughter languages I feel like I understand a bit more 😛

        I am very interested in the the change from *gwher- to ther and ber if you don’t mind typing it out. The gwh>b seems related to Grimm’s law? But I don’t really understand that, and I’m not sure where th fits in.

        • kaakitwitaasota says:

          The fate of *gʷʰ in Germanic is not entirely clear; it was a fairly rare sound in PIE, which muddies the waters. We know that *kʷ became /hʷ/ (compare Latin quod with English what). *gʷ sometimes gave /kʷ/ (English queen, Greek γυνή, Old Prussian genno from *gʷénh₂) but sometimes, apparently, /w/ (English womb, Greek δελφύς, Sanskrit garbha, though why it’s not wolb isn’t clear to me–possibly not actually from *gʷelbʰ-?).

          *gʷʰ sometimes gives /w/: English warm, Greek θερμός, Persian garm. Other reflexes are not entirely clear. The “slay” root *gʷʰen- (Sanskrit hanti, Hittite kuenzi, Greek θείνω), and it is thought, at least by Wiktionary, to give both /g/ and /b/ in English (Old English gūþ ‘battle’ and modern English bane). The jury is still out.

        • aho bata says:

          On the Germanic side, the shift from gwh to b partly involves Grimm’s Law, which would deaspirate it to gw while changing kw to hw and gw to kw. The other part of the development (which can’t be dated relative to Grimm’s, but assuming Grimm’s came first for simplicity’s sake) involves shifting the resulting gw to b in word-initial position. Evidence for this comes from a limited number of words. One example (on top of kaaki’s ‘slay’) is *gwhedhyeti ‘ask for’ to OE bitt, Old Norse biðr, Gothic bidjiþ. (‘warm’ is sometimes associated with Greek thermos which would make the IE labiovelar gwh, not gw, but there is alternative root *wer- with descendants meaning ‘boil, cook, burn’.) Developments of PIE gwh word-internally were much more complex. Generally it becomes w, but the velar stop articulation is preserved after nasals (where in Old Norse, original labiovelars trigger u-umlaut, or labialization + backing, in the singular past indicative of -va stem verbs: syngva ‘to sing’, sǫng ‘sang’ (the ǫ being an u-umlauted a), søkkva ‘to sink’, sǫkk ‘sank’), and possibly after other resonants as well – the evidence is difficult to interpret.

          There are also a couple isolated instances of other labiovelars in Germanic possibly becoming labials, although that’s more controversial. Examples include *penkwe ‘five’, PGmc fimf (cf. Sanskrit panca, Greek pente) and *wlkwos ‘wolf’, PGmc wulfaz (cf. Sanskrit vrkas, Lithuanian vilkas).

          A few more words about ‘slay’. The b of OE bana ’cause of death’ putatively continues the original gwh, whereas in guþ ‘battle’ (PGmc gunþiz) it is delabialized by the following u, as also happens in several other words, e.g. Gothic kaurus (PGmc kuruz), cf. Sanskrit gurus, Gk barus. The un in gunþiz comes from syllabic n; this is the regular development, cf. the Germanic negative suffix un- beside Sanskrit, Greek a(n)-, Latin in-. The syllabic n is in turn triggered by the suffix -þiz (cf. Sanskrit –tis, Greek -sis), whose PIE ancestor co-occurred with the full grade of the root (i.e. with e) in the nominative, accusative and vocative cases (gwhen-) but with the reduced grade in the other cases (gwhn-); apparently the latter root form was generalized in Germanic, as often happened. So yes, we should be cautious to accept that gwh regularly becomes b word-initially on the basis of such a small number of examples, but a change from labiovelar to labial is not unusual and it opens up the possibility for an unintuitive but plausible connection between these two words.

          On the Greek side, the regular outcome of labiovelars is coronals before front vowels (i and e), labials elsewhere. So, related to the verb theino that kaaki cited there is the noun phonos ‘murder’. But the labial-coronal alternation is most striking in the interrogatives and other correlatives, where corresponding to Latin quis, Sanskrit kim (whose m ending originally belonged to the accusative singular, which eclipsed the other forms) are stems in ti-, while the po- stems correspond to forms with non-front vowels, e.g., poteros ‘which of two’ corresponds to Sanskrit kataras, Germanic hwaþeraz (English whether) with the same meaning.

  7. arbitraryvalue says:

    Regarding the state education rankings: that the commonly-accepted rankings are so totally skewed by Simpson’s paradox is remarkable if true. They can’t be that wrong… Can they? I almost hope that there’s a mistake in the Liebowitz and Kelly article, because I doubt that conventional wisdom will change even if they’re right. (Their results do match my personal beliefs.)

    That said, I wonder if there are other important but neglected subcategories beside race. One obvious possibility is, of course, wealth. I expect that students from wealthy families will do better, even with lower school spending, than students from poor families…

    • Jacob says:

      Reason.com changes a whole bunch of things in their analysis and they don’t show their work at all. All they provide is a final ranking. They don’t show disaggregated their data by race, beyond the one comparison between Texas and Iowa. And they change how spending is counted, among other things. So there’s no way to say how much of the change in rankings is due to which factor.

      • If sufficiently curious you could email one of the authors to ask if they have a more detailed account in a longer paper.

      • MoebiusStreet says:

        They do give a rough outline, including

        we removed factors that do not measure K–12 student performance or teaching effectiveness, such as spending per student (intentions to raise performance are not the same as raising performance), graduation rates (which often indicate nothing about learning, since 38 states do not have graduation proficiency exams), and pre-K enrollment.

        Most of it makes good sense to me, but I’m concerned about removal of the graduation rates – that does seem like a terminal goal to me, or at least a pretty good proxy of one.

        In my mind, that leaves me with two ranking schemes that are probably both flawed. That at least a step forward because there’s something to point to other than what has been the standard party line.

        • The problem with graduation rate is that you don’t know if a high graduation rate means they are doing a good job of teaching kids or have a low standard for graduation.

          • MoebiusStreet says:

            It is a terminal goal in its own right – at the very least it’s keeping the kid of the streets.

            Other criteria have similar problems. Like, are the high test scores just the result of “teaching to the test”, or do they really represent a broader foundation being built?

            I’d speculate that in the presence of other positive clues, we could at least view the graduation rates as tending to support the other evidence. But I don’t know how to work that into an objective model.

          • cassander says:

            @MoebiusStreet

            Other criteria have similar problems. Like, are the high test scores just the result of “teaching to the test”, or do they really represent a broader foundation being built?

            Assuming the test is even minimally well designed, what’s wrong with teaching to it? How is teaching to a test different than teaching to a syllabus?

          • idontknow131647093 says:

            Replying because I want to remember to check in to see an answer to cass’s q

          • Murphy says:

            @MoebiusStreet

            It is a terminal goal in its own right

            it’s one of those “once metric has become a measure it becomes useless as a metric”

            Situation 1: Graduating high school is a demonstration of some basic level of proficiency and as a result high school graduates do better in life because people are more willing to hire them.

            Someone sees this situation and misses the bit about proficiency but see the correlation between graduation and doing better.

            So they make a bad law saying that the school can’t fail anyone. Surely this will mean prosperity for all!

            Situation 2: Graduating high school now demonstrates no level of proficiency. It now means nothing except that you were present. Students who can’t read or write, who spent their time cutting class and going to a local abandoned lot to burn things….. get the same diploma as students who worked hard and did well.

            The data still shows that students with diplomas did better because now the only ones without are the kids who ended up in jail or a coma and couldn’t physically be in the building to count as attending high school.

            Clearly the problem is still that there’s not enough graduation! We can see that kids without diplomas do terribly! Most of them end up sick or in jail!!!

            So another bad law is passed: all citizens are to by mailed an official high school diploma.

            Surely now nobody shall end up in jail or sick because they have highschool diplomas!

            The “terminal goal” should be learning useful skills. everything else is symbolism and all you do by making it easier to get the diploma is to make the symbolism worthless as a symbol and strip it of information content.

          • TRAVIS MYERS says:

            @Cassander

            Because tests are supposed to be random samplings of your knowledge. If you don’t know what’s on the test, then the random sampling is a good indicator that you have a wide range of knowledge on the subjects. If you do know what’s on the test, then it indicates that you only know the very narrow range of things that are on the test.

          • Conrad Honcho says:

            Travis, that’s only true if you’re talking about “teaching to the test” as essentially having the questions beforehand and cheating. If you know that “addition” is going to be on the test, then “teaching to the test” just means teaching the kids how to add, which is good. Now if you know the question is going to be “what’s 3+5” and you teach them “8” but they have no idea how to figure out what 7+9 is, that’s a different story.

            And graduation rates are like the #1 go to example of Goodhart’s Law.

          • SamChevre says:

            @cassander, Re: “teaching to the test”

            I’m an actuary: actuarial credentials are defined by passing tests. So I have no inherent problem with teaching/studying to the test. However, there are a major potential problem; while it is really a problem of test design, it is a real problem with typical K-12 tests.

            The problem is tests so designed that learning how to answer the test questions with an entirely inadequate understanding of the subject being tested is easier than learning the material. In a poorly-performing classroom, it’s not atypical for a teacher to spend a several week teaching algorithms like “if there are two numbers and the word altogether appears somewhere in the problem, type “n1 + n2 =” and check whether the result is one of the options for the multiple-choice. If it is, select it and go on. To say this doesn’t teach arithmetic usefully is an understatement.

          • MoebiusStreet says:

            @Murphy – I agree with the principle of your objection – this could potentially happen. However, in today’s situation we’re probably somewhere in the neighborhood of your “situation 1”. For measurements taken in past decades, we know that we’re not going to be as far off as situation 2.

            Also, nobody’s reacted to the latter part of my comment, where I speculated about being able to use the graduation rate as a confirmation when other positive clues are in evidence, but not necessarily as a factor in its own right.

          • cassander says:

            @SamChevre

            it’s not atypical for a teacher to spend a several week teaching algorithms like “if there are two numbers and the word altogether appears somewhere in the problem, type “n1 + n2 =” and check whether the result is one of the options for the multiple-choice. If it is, select it and go on. To say this doesn’t teach arithmetic usefully is an understatement.

            Even in this extreme example, to solve the problem the person solving it would still have to know how to add n1 and n2, so as an arithmetic test, it’s far from useless.

          • Mr. Doolittle says:

            @cassander

            Since he used the word “type” for entering the n1+n2, I take it he means using a calculator of even a phone or tablet. In that case, there may be no learning at all involved, or very little.

            At that point, I would blame the teacher for allowing a calculator for basic addition, rather than teaching to the test.

      • albatross11 says:

        I suspect one of the most important factors is the fraction of kids who take the standardized tests. ISTR that some states manage to justify excluding a fair fraction of kids from testing, while others make just about everyone take the test. You can get quite a boost from excluding the bottom 10% of kids from testing.

    • idontknow131647093 says:

      If anything, I think many of these adjustments were underdone. Massachusetts still being at #2 for quality doesn’t pass the smell test. Ditto New Jersey. Illinois schools being ahead of Wisconsin by that big a margin is also silly. I don’t know anyone who would move south from southern Wisconsin to get to the “better” IL schools. Not that Wisconsin schools are that great (indeed comparing Wisconsin to Texas as the Reason writers did TX and Iowa is a traditional way of showing Simpson’s paradox in education in Republican writings).

      Their findings are hardly novel. See, e.g. https://www.educationnext.org/are-wisconsin-schools-better-than-those-in-texas/

      • The Nybbler says:

        Massachusetts still being at #2 for quality doesn’t pass the smell test. Ditto New Jersey.

        Why? Aside from the “Abbott” districts (which tend to be poor-and-minority), NJs schools are generally considered of very good quality (unlike, say, Pennsylvania). Since states are heterogenous in education quality, the relative populations (or weights, if population is not used) of good and bad areas matter; Maryland minus Baltimore would do a lot better for instance, whereas if DC were retroceded to Maryland it would do a lot worse. NJ wins here by not having any major cities (which tend to be terrible in education, Philadelphia being no exception and NYC being mixed).

        • idontknow131647093 says:

          The point of the adjustments is to eliminate the fact that Baltimore sucks, because it is supposed to suck. This is why I think the adjustments were under-done.

      • idontknow131647093 says:

        As someone from the area, you are talking about the difference of moving from a poor, rural, white SE Wisconsin district to a rich, suburban, white NE IL district. There are plenty of equivalent poor, rural IL districts on the IL-Wisc border.

        What you’ve helped me do is improve my critique: This survey doesn’t adjust enough for rich white people being incredibly easy to teach. You could take some of those districts you are talking about and staff them with only gym teachers and they would still end in the top 5% in math/english on the SATs.

    • Jiro says:

      If you think that there is no such thing as a racial group that has low scores for reasons independent of education quality, then there’s no Simpson’s Paradox. Having more of a low-scoring group is entirely your fault because you’ve failed to teach more people, so it can legitimately be part of the comparison.

      • notpeerreviewed says:

        I think there’s a better argument that the analysis should control for *more* characteristics rather than fewer – controlling only for race is arbitrary. Does the Hispanic population of Florida consist largely of fully-assimilated Cubans and does the Hispanic population of Iowa consist of migrant workers? Not saying that particular example is true, but it points to why the comparison is difficult.

        • albatross11 says:

          It would indeed be interesting to see how much better your predictions get as you add stuff. My guess is that language spoken at home, race, parental IQ/education, and poverty are going to get most of the benefits available from looking at group membership of students. But it would be very interesting to see what really comes out of that analysis.

        • Mark V Anderson says:

          I think there’s a better argument that the analysis should control for *more* characteristics rather than fewer

          Yes indeed. This study did a good job showing that the traditional ratings that fully aggregate state rankings are very deficient, exemplified by their example of Iowa being much better than Texas in aggregate, but worse than Texas for EVERY racial group when they are racially dis-aggregated. I don’t know if dis-aggregating by race is the best way to look at these rankings, but it is better than nothing. The racial dis-aggregation should only be the beginning. I bet there’d be plenty more changes in the rankings if they dis-aggregated by income or wealth or parents education. The more the better, if you truly want to judge each state. Maybe it gets a bit ridiculous after a while. But certainly the study dis-aggregated by race is a valuable addition to knowledge because it shows that the traditional method doesn’t hold water.

        • sharper13 says:

          The more you add in additional controls for various factors, the more you probably end up with the result demonstrating the null hypothesis of the “schooling” not really mattering in terms of making a difference, defined in education as the results of these steps showing no real change at the end of the process:
          1. Take any pedagogical innovation or educational intervention.
          2. Subject it to a controlled experiment.
          3. Evaluate the experiment’s outcome several years later.
          4. If the experiment works, attempt to replicate the experiment in more situations.

          That said, it does seem really dishonest in retrospect to have certain groups loudly proclaiming they need more money in order to improve their system’s ranking on results which are largely based directly on how much money is spent. I mean, that makes more money=better rankings a truism, but not in the way the people they were arguing at so publicly would consider a positive outcome for spending the money.

    • owleabf says:

      I think the disconnect between the two rankings can be explained in a more simple fashion:

      The commonly accepted rankings are functionally there to let upper middle class parents determine what school they should send their kids to. That has as much to do with signalling as anything, so it’s more about sending your kid to a school that’s perceived as a good school as it is about actually increasing their learning.

      The Reason.com rankings seem to make some reasonable adjustments to try and actually measure quality of instruction. That said I’m very skeptical of any research that, by seeming coincidence, manages to confirm ALL of the writers’ priors in the way this does. As others have mentioned it would be interesting to look at their data.

      • Mr. Doolittle says:

        So you’re suggesting a meta-level approach to finding schools is in effect? Rich[er] families look for highly funded schools as a proxy for class in the district?

        Then it does make some sense to look at money alone at a higher meta-level, and to give money to failing schools. If your goal is to get rich white kids into a minority-majority school, pumping money into it and raising its ranking will provide a false positive for prospective families.

        Of course, being a false positive will result in a higher meta level where those parents investigate the area more thoroughly to determine if that’s actually the case.

        • owleabf says:

          I’m suggesting that the US News rankings are performing their actual intended purpose (which differs from Reasons intended purpose.) The US News rankings are essentially a ranking of how many high scoring students come from a given state, not how well that school did in improving their scores.

          That matters from a cultural signaling and college entrance standpoint, which is ultimately what matters to the consumer of the US News rankings.

          To put it in terms of colleges/universities: it’s generally accepted that Harvard/Yale are some of the best schools because they output lots of high producers. But are they actually the best at raising the level of education of their students? Or do they just start with students that have a higher baseline? I’d argue the latter.

          But even while arguing the latter, if I’m hiring I’ll give an applicant from Harvard extra consideration because historically that’s a signal that they are already a high producer.

          A couple asides on this that I didn’t mention before:

          -I wonder if there’s an upper bound on the ability to make gains in scoring, which would give schools that start with lower baseline students more space to improve than a higher baseline would.
          – The NAEP test is only one of many tests given to students. Many states give the ACT/SAT to all students. Would be interesting to do the same analysis with that data.

  8. baconbits9 says:

    Everything You Know About State Education Rankings Is Wrong. Most rating systems rank state education success based on a combined measure which includes amount of money spent as a positive outcome, making it tautological to “prove” that more funding improves state performance. See also economists’ Stan Liebowitz and Matthew Kelly corrected ranking table, which also adjusts for some confounders.

    This reminds me of the old WHO country rankings for medical care. They consistently had measures like equality of distribution and fairness of pay. I believe at one point it was so skewed in favor of “socialist” medicine that a country which denied all access to medical care to all people and taxed the rich to fund it would have gotten an equivalent score to the US at the time.

    • cassander says:

      The commonwealth fund puts out studies like this all the time. They include both “equity” and “access” as categories with equal weight, along with both “process” and “administrative efficiency”. Even when you drill down into healthcare outcomes, the pick some bizarre choices to measure, and some that are known to be problematic, like infant mortality. One begins to suspect that they’re gunning for a certain outcome…..

      • Murphy says:

        You appear to have mangled a quote.

        “The only serious black mark against the NHS was its poor record on keeping people alive. On a composite “healthy lives” score, which includes deaths among infants and patients who would have survived had they received timely and effective healthcare, the UK came 10th. The authors say that the healthcare system cannot be solely blamed for this issue, which is strongly influenced by social and economic factors. Although the NHS came third overall for the timeliness of care, its “short waiting times” were praised. “There is a frequent misperception that trade-offs between universal coverage and timely access to specialised services are inevitable. However, the Netherlands, UK and Germany provide universal coverage with low out-of-pocket costs while maintaining quick access to speciality services,”, the report added.”

        “The problem with health outcomes is that it is notoriously difficult to work out to what extent they are really attributable to the health system, and to what extent they are attributable to lifestyle, environmental or socioeconomic factors.”

        So if in country A there’s some cultural thing where people with chest pains don’t rush to the doctor they may have worse outcomes than a country where they do. Or sometimes Britons are just more alcoholic and fat.

        That’s the problem with comparing total outcomes.

    • Murphy says:

      “access” seems like a damned important measure.

      if you ignore “access” then a dictatorship that taxes all the serfs to the bone and the only person who gets medical care in the country is the Dear Leader …. but that medical care is unbelievably spectacular… that dicatorship would do spectacularly well because every single patient in the country getting medical care has a team of 20 consultants dedicated to their care and the best of the best of every piece of equipment they could ever need. #1 world ranking!

      “equality of distribution” is a bit more debatable but a hypothetical country that harvests the organs of the disfavored ethnic group and ceases their property to pay for the medical system while denying them any medical care … might have amazing metrics for a large fraction of the population, no shortage of kidneys at all! but it has kind of sucky equality of distribution.

      • albatross11 says:

        This makes me think of how Cuba avoided much of the impact of the AIDS epidemic, by putting a whole bunch of HIV+ gay men into concentration camps. Good outcome, nasty methods.

        • baconbits9 says:

          This sounds like a Texas sharpshooter excuse. Its only a “good outcome” if you separate it from all the bad outcomes that came along with a government having the power to imprison segments of the population at its whim.

      • gbdub says:

        The argument is not that access or equal distributions are bad things to have. The argument is that if you’ve weighted “access” and “equality” so heavily that it swamps every other factor like “provides actual useful care”, then your rankings aren’t very good and were probably motivated.

      • baconbits9 says:

        Your complaints are (at least partially) built into the other metrics used for outcomes. The hypothetical organ stealing would lead to very low life expectancy outcomes for the abused subgroup, and the infant mortality rate for the population with the dictator would be high, pulling down their scores (as well as a bunch of other outcomes).

        So while “access” is important it automatically shows up in many of the other health care adjacent metrics that were already used. The WHO report took access as an “egalitarian” measure which is not the same as a health measure, it stated that a rich person with great access to health care and a poor person with good access to health care in the same country was bad (ie received a low score) where as a country that was entirely poor and had broadly bad access was good (ie received a high score). This reflects not a measure of the health care system, but a measure of a political desire.

        This was also not a hypothetical situation, but one that was reflected in their rankings. I’m going on 15 year memory so the details won’t be correct but one of the last reports that they put out using these metrics (2005ish) had the US ranked in the top 5 for health outcomes and at the bottom for things like “access” and “fairness of distribution and payment” and ended up in the low 30s in the ranking. One or two spots behind them was a country near the bottom of health outcomes (I think it was Colombia), very low in infant mortality and life expectancy etc but was ranked near the top for the egalitarian measures.

        • Murphy says:

          Might you mean the 2000 health report?

          I mean you can drill down into the tables but the US didn’t do all that great even when you get to the level of looking at chances of kids dying before 5, chances of dying in teens, life expectancy…. etc

          http://www.who.int/whr/2000/en/whr00_en.pdf

          Some of it’s likely down to the high rate of obesity and various other wealth-linked health issues.

          15 for “attainment of goals” which is pretty respectable.

          • baconbits9 says:

            Yes, probably that one.

            I mean you can drill down into the tables but the US didn’t do all that great even when you get to the level of looking at chances of kids dying before 5, chances of dying in teens, life expectancy…. etc

            There was a composite score made up of 5 attributes for “overall health system performance”. The US came in 37th place and Colombia came #22.

            Columbia’s life expectancy was 74th, distribution of life expectancy 44th, the US 24th and 32nd.

            Colombia’s responsiveness measures were 82nd and 93rd the US #1 and #3-38

            Colombia’s goal attainment was 41st the US 15th

            Colombia’s fairness in financial contribution was #1 the US was 55th

            Spending per capita it was #49 for Colombia and #1 for the US.

            How is it that Colombia ended up 15 spots ahead of the US then? You have to follow to a different paper to get to the breakdown here where they say

            Then the following weights were used to
            construct the overall composite measure: 25% for health (DALE), 25% for health inequality, 12.5% for the level of responsiveness, 12.5% for the distribution of responsiveness, and 25% for fairness in financing.

            So inequality of outcomes counted for 62.5% of the total index. Everyone dying at age 20 is worth the same as half the population dying at age 20 and half at age 120. The first would get bottom ranking for DALE and top ranking for distribution, and the second the inverse. ‘Fairness’ in financing has as much influence as life expectancy (their primary measurement for actual health outcomes).

  9. Nornagest says:

    ..and by “a bit fanciful”, I mean that this is a depiction of John Adams asking a mountain fairy to help avenge the death of his mother, who was eaten by a giant snake.

    I want to live in the country that Tokugawa-era Japan thinks I live in.

    • Le Maistre Chat says:

      Which could be the same country where Virginia has cliffs and tree fairies.

      • Nornagest says:

        As seen in Pocahontas? Virginia’s got a few cliffs, but they’re pretty far west of Jamestown. And not as impressive as Disney set dressing usually looks, though the area’s very pretty in autumn.

        • Le Maistre Chat says:

          Yes, Pocahontas. I didn’t know that; I’ve been to eastern Virginia and it is very pretty, because of the forests on flat-ish terrain.

    • Phigment says:

      Speaking as a person with living parents he is on good terms with, I have to say I’m OK with living in a country where mothers seldom, if ever, get eaten by giant snakes.

  10. sourcreamus says:

    Isn’t the cost of a new person not 21x but x + 20(x-y) where y is the current salary?

    • baconbits9 says:

      The cost of increasing the new hires wages is 21x where X is the amount they have to increase their offer over the previous wages to attract a candidate.

  11. DragonMilk says:

    Regarding the Perpetual Peace, Extra Credits has a good series on Khosarau

  12. Salem says:

    What is your take on the Cochrane situation? As someone with no particular knowledge of the field, it’s hard for me to have an opinion on the specifics, but this is probably the most prestigious name in health recommendations, that as non-experts we all rely on. Should we be downgrading our opinion of them?

    • Scott Alexander says:

      I downgraded slightly, based on them apparently taking criticism very badly, but I won’t claim to know anything about what’s going on.

      • Jacob says:

        Many organizations have a norm that insiders are supposed to keep their criticisms internal and not go public. Gøtzsche broke that (possible) norm. What this says to me is that we can’t trust what high-level employees of the Cochrane Collaboration say in public about the Cochrane Collaboration, which we should’ve known already.

        • Douglas Knight says:

          Gøtzsche points out that Cochrane has an annual prize for criticism. It is named in honor of Bill Silverman, whom it describes as a ‘troublemaker.’ (Not that that’s evidence about norms.)

      • Cheese says:

        Hi Scott.

        I think if you are going to alter your priors on the actual science put out, it may pay to at least dig a bit deeper. The board statement was fairly emphatic about the expulsion and it’s prior investigation being unrelated to criticism in question, which matches up with the timeline. The timing is unfortunate, but the criticism of the Cochrane publication doesn’t really bear out scientifically. Both sides I think have handled it badly and both groups have since retreated a bit; with respect to statements made about the board processes, not the actual changes to the review – which are minimal and don’t affect the outcome.

        I think it may be fair to make a criticism about the expulsion coming at a bad time and requiring public justification. The problem here is that Cochrane won’t release the report because of legal issues – a statement which I tend to take on face value given the potential confidential nature of complaints of bad personal behaviour. Gøtzsche has of course made some rather selective quotes public but unsurprisingly they are the ones positive towards him.

        Hilda Bastian has some pretty decent summaries of the events – I tend to agree with most of her points if not some of her broad philosophical ones.

        https://blogs.plos.org/absolutely-maybe/2018/09/18/boilover-the-cochrane-hpv-vaccine-fire-isnt-really-about-the-evidence-but-its-critical-to-science/

        http://hildabastian.net/index.php/67-the-hpv-vaccine

        I probably am a little biased against people like Gøtzsche because he has a history of making rather grandiose statements about a wide variety of medical treatments sometimes with and sometimes without evidence. The latter is problematic given how he can be viewed in popular culture. In some cases his contrarian position may be right, in some cases it may be wrong. I don’t think the former gives him a pass on the latter, a perspective which he seems to lack, from my reading of his statements.

  13. Sniffnoy says:

    Note that Greg Egan doing original work in combinatorics is not that surprising, I’m pretty sure has before. You can certainly find him hanging around and commenting on math blogs such as Shetl-Optimized. 😛

    The 4chan part, though… yeah, that’s pretty damned surprising. Note that the proof as posted was originally pretty sketchy but other mathematicians have since expanded it into an actual proof.

    Also note that the problem isn’t solved — you make it sound like we’re talking about matching upper and lower bounds, two parts of the same proof, that solve the problem, but that’s not the case. It’s just, y’know, a lower bound and separately an upper bound. But, you can read Robin Houston — one of the people working on the problem since then — talking about this more here on Hacker News. In a comment below, he speculates that the correct answer may be n! + (n-1)! + (n-2)! + (n-3)! + (n-4).

  14. moridinamael says:

    How come the explanation for all neologisms isn’t just “a cute baby adorably referred to a bear as a ‘brown’, everyone agreed this was hilarious and started calling bears ‘browns’ as an inside joke, within two generations everybody in the family/village/social unit just called bears ‘browns’ and couldn’t remember that they had any other name.”

    Because this seems pretty plausible to me, as someone who speaks basically a different language when I’m at home, which is 20% based on the mispronunciations of toddlers, some of which are two or more generations old.

    • dndnrsn says:

      There’s also the compromise: a baby called a bear that and everyone was like “all hail the baby, who has driven the horrible [redacted] into the night!”

    • Randy M says:

      Our daughters would talk about their “chuthers” when they were young. It cracked me up when I figured out they were mis-hearing “Each other”.

      • baconbits9 says:

        My 5 year old today asked to listen to the “dasta pasta song again”, I had no idea what he meant, and clarified that he really was saying “dasta pasta” and only figured out that it was “another one bites the dust” after going through all the songs we had heard together recently.

      • SamChevre says:

        My oldest used to ask for the “thumpthump thump” song when he was about 2.

        It took forever (well, it seemed like forever given the amount of fussing) to figure out that it was “A Fifth of Beethoven”

      • Conrad Honcho says:

        My high school gym teacher would encourage students to use their “musular skrempf” (muscular strength). To this day, in my family one uses their musular skrempf to, say, open stuck jars.

      • gbdub says:

        Apparently, when he was a student, one of my professor’s professor’s had a heavy accent and pronounced “chaos” as “cows”. So the class had a running gag about “taking the cows for a random walk”

    • Nootropic cormorant says:

      It is often asserted that linguistic change is driven by adults and not children’s speaking patterns, but I haven’t reviewed the evidence.

      • Aging Loser says:

        But every stage in the emergence of language as a whole must have come from small children playing noise-games. For example, small children climb onto a rock chanting “a-hump-a-bump, a-hump-a-bump!” and when they’re on top of it they shout “Ka-pumba!” and then when they’ve jump down they shout “Ha-balla!” And so, as they grow up, “Ahumpabump” comes to mean “Rising,” “Kapumba” comes to mean “Above,” and “haballa” comes to mean “Descending”.

    • b_jonas says:

      And then they invited a whole fucking ritual for how to avoid offending the spirit of the bear after the hunter succeeds to kill it, by inviting the dead bear to dinner in the house, and simultanously insisting that it’s not dead, and that it’s died from a natural accident by falling that the hunter had nothing to do with, with twenty pages of description eventually edited into Kalevala? I like adorable babies as much as any other person, but taking it that far seems unlikely to me. No, I don’t buy it, not in the case of the bear.

  15. Jon S says:

    Some of the survey responses re:talent constraints sound really myopic to me. “I don’t really know anything we could reasonably do with money to get more qualified people to join. They just don’t really exist.” I’m having a hard time conceiving of how this could possibly be true.

    If you can’t find people with the right skills, you can train them. Offer intelligent/conscientious college graduates piles of money to come work in your training program. For those who take sufficiently well to the training, offer them bigger piles of money to stay.

    “Raising salaries… is unlikely to be very helpful for attracting top talent since the most suitable candidates are also the most altruistic ones.”
    This seems very wrong to me, but I can’t put my finger on the heart of the problem. If the market rate for the talent you need is X, just hire people for a salary of X – if they’re altruistic and they want to donate 50% of X back to you, that’s great, but it doesn’t have much bearing on the cost of the talent. If somebody whose skills are worth X in the open market is very altruistic, why should they prefer working for you at a discount rather than earning to give elsewhere?

    • baconbits9 says:

      Some of the survey responses re:talent constraints sound really myopic to me. “I don’t really know anything we could reasonably do with money to get more qualified people to join. They just don’t really exist.” I’m having a hard time conceiving of how this could possibly be true.

      The labor market does get tight sometimes, there are lots of issues matching up employee skills with employer needs and they get amplified from the employer’s perspective when the UE rate is very low (as it is now).

      If you can’t find people with the right skills, you can train them.

      This assumes you are a large enough outfit to already have a quality training system in place.

      • Jon S says:

        A tight labor market definitely makes it tougher to find/retain people, but to me that mainly means [significantly] more expensive. The people still exist (or don’t, if that’s the case), you just need to incentivize them enough to beat out their next best opportunity.

        • baconbits9 says:

          If you take the literal interpretation then yes, those people do exist, but the reality in a tight market is that employers are often constantly fighting to retain workers or replace those who are taking those better offers elsewhere (or have some other reason for leaving). They are also often seeing a much higher ratio of under-qualified applicants to qualified applying for their open positions for a variety of reasons, which adds to the presumption that they can’t find a qualified person with reasonable effort.

    • Aapje says:

      @Jon S

      People can have salary requirements regardless of their altruistic nature. For example, if they have substantial college debts. Another example is that the male gender role puts higher demands on men to earn (so they can provide), so a lower salary reduces well being more for men than for women (also since women are more likely to be able to find a provider). So a woman who accepts a salary below what they can get will on average make less of a sacrifice than a man who accepts the same salary.

      Then picking the person who accepts the lower salary can mean selecting against people with poor parents, selecting against people who went to expensive colleges, selecting against men, etc.

      I also question whether the most altruistic candidate is the best. Firstly, the goal is not for the workers themselves to win an altruism competition for their own behavior, but presumably to influence others to act effectively altruistic. A worker can convince many thousands of people to give more effectively is a better effectively altruistic employee than someone who can convince no one, but is more effectively altruistic in their own lives.

      Secondly, effective altruism is not the same as just altruism. A person who gives 1% of his income, but puts a lot of effort into making sure his money is used effectively, is much more effectively altruistic than a person who gives 10%, but spends it on giving cocaine to the homeless. I’d pick the former person for an EA company, not the latter person.

  16. baconbits9 says:

    Re: the 21x to raise incoming salary.

    This shouldn’t be a real thing unless you have a, uh, unusual corporate structure. First you really shouldn’t have 20 people in a 20 person company making the same wage, and secondly you should still be able to hire entry level people and train them well, and only a small fraction of your employees should be at entry level wages at any one time.

    Beyond this there are plenty of well worn ways around these issues, you can offer signing bonuses for new hires or other compensation (offering to pay for moving expenses etc) that are specific to the new hire and don’t seem to cause systemic demand for higher wages. You can (or you could pre ACA) target certain classes of people by offering or not offering competitive health benefits that wouldn’t scale across the whole company (i.e. in a company largely filled with straight out of college grads without kids you offer medical coverage for kids which would be a benefit for some potential employees but wouldn’t be for most of your current ones).

    • panloss says:

      You can also hire them as an independent contractor, which obfuscates the difference in salaries. It also probably doesn’t matter that much. Most people are reluctant to discuss salary, and even at large, banded software companies where big pay gaps aren’t supposed to happen they often do.

      This is based on my experience in startups, not charities, but I’ve found that “talent gaps” are one of the best excuses for a team to have in its back pocket in the case of underperformance.

    • Urstoff says:

      It’s not uncommon, at least, at public institutions like universities where pay scales are public and advancement slow.

  17. smocc says:

    Whoa whoa whoa, Armenia’s Eurovision song “Don’t Deny” was definitely, definitely about the genocide. It was their entry on the 100th anniversary of the genocide, and look at the official music video.

    Which makes the arguing even more fun to watch!

  18. Johannes D says:

    “Superpermutations” is a term for mathematemical

    Was this intentional, by the way? 🙂

  19. benquo says:

    Interesting that the more realistically labeled peace lasted longer. Perhaps they should have iterated downwards to see if they could get a yet better peace out of it?

    • Le Maistre Chat says:

      “Confusingly, the Fifteen Year Peace has lasted twenty. I hope to find a mathematical formula for this phenomenon before another Persian War calls me away from my studies.”

      • benquo says:

        Quick! Sign a Zero Years’ Peace!

        • Slicer says:

          We actually kind of did that with North Korea, and it’s worked middlingly well so far.

        • Statismagician says:

          I think that’s the one that causes Apocalypse IV: Return of the Son of the Apocalypse. Safer to take Zeno’s approach.

          • albatross11 says:

            The enemy’s missiles can’t possibly reach us, since first they’d have to traverse half the distance….

          • Statismagician says:

            That’s the ABM application of the same general principle – I was thinking that since length of actual peace varies inversely in some proportion with the length of the name of the peace, we could make a series of peaces with shorter and shorter units of time in the name and thus increase the length of actual time indefinitely without dividing by peace and/or war by zero and causing Armageddon (IV: Electric Bugaloo^2).

      • The move from infinite to 50 years only brought two extra years of peace so I’m guessing that you have to put in a lot of decimals before you get anything that lasts a while.

  20. benquo says:

    Seems like there’s also good news in the Cochrane scandal – Peter Gøtzsche seems trustworthy and we should assume that future organizations he’s involved with are basically honest, until and unless he blows the whistle. Also, Cochrane board had a solid enough moral foundation that when it became clear that the board had just barely taken over by thugs, the minority faction immediately raised the alarm rather than try to accommodate evil.

    On the other hand, yeah, pretty sad and scary that Cochrane’s been taken over by thugs. Seems like maybe people should update towards sounding the alarm and demanding decisive action when the board’s only 33% thugs or maybe even much less than that. (In other words, pretty much the opposite of what Bryan Caplan said.)

    • Radu Floricica says:

      When politics is involved, you can almost never split the vote as “thugs” and “good guys”. You may have quite a few principled people swinging the vote the way it did.

      It’s fairer to say that overall the organization failed to handle it well.

      • benquo says:

        Almost never, perhaps – but why not in *this* case? It seems like an exceptionally clear case.

        • benquo says:

          PS I didn’t say that the dissenters were pure “good guys”. I just called them the minority faction, which they evidently were since they lost a vote!

          What I care about is what sort of behavior I can expect from an organization dominated by one or the other type, not whether there’s a principled rationalization on either side.

    • Douglas Knight says:

      3 of the 4 people who resigned in protest had voted to censure Gøtzsche. They only objected to the punishment, not the verdict. This does not seem consistent with your black-and-white reading of the events. These board members were taken by surprise, from one vote to the next. Surely they did not think that even 1/3 of the board consisted of thugs. Maybe the second vote changed their minds and now they endorse a black-and-white reading. Maybe it proved that Gøtzsche was right and that they censured him for being right. But they haven’t said that; maybe they still hold that the behavior for which they censured him was bad. Gøtzsche (now) says that Cochrane is authoritarian and corrupt. Surely they now agree that it is authoritarian, but maybe they censured him for saying that it is corrupt and maybe they still deny that. The quote from Mintz seems to say that Cochrane is authoritarian but not corrupt, though he could certainly be clearer.

      On a slightly different topic, the board voted 6-5-1 to expel Gøtzsche. Presumably the 4 people who resigned in protest all voted against this motion. What about the 2 people who were asked to resign to maintain the elected-appointed balance? Were they the remaining 2 that did not vote for expulsion? That would be pretty suspicious.

      • benquo says:

        Thanks for the added detail, which wasn’t apparent to me in either Scott’s summary or the linked article.

        The statement of the remaining Cochrane board is a bit confusing, asserting the following:

        The Cochrane Collaboration’s Governing Board of Trustees voted unanimously on 25th September to terminate Professor Peter Gøtzsche’s membership of the organization, and his present position as a Member of the Governing Board and Director of the Nordic Cochrane Centre.

        But also:

        The final decision follows an initial Governing Board meeting on 13th September 2018. At this meeting:

        * Eleven members of the Board voted that Professor Gøtzsche had breached the Trustees’ Code of Conduct as a result of these behaviours;
        * a majority voted that Professor Gøtzsche should accordingly resign as a trustee; and
        * a majority concluded that Professor Gøtzsche has acted counter to the best interests of the charity as a whole and therefore voted to serve notice to terminate Professor Gøtzsche’s membership of the charity, in accordance with Cochrane’s Articles of Association.

        These are of course reconcilable if you know the additional fact that between those two events the dissenting minority resigned, but the omission of that fact doesn’t really inspire confidence that the main relevant facts are being included. The unanimity of the subsequent decision sure suggests that all dissenting members resigned, and the board took another vote so they could say the decision was unanimous.

      • benquo says:

        BTW where did you get the vote #s? They weren’t evident in any of the linked material I found.

        • Douglas Knight says:

          6-5-1 comes from the resignation letter, hosted by Gøtzsche.
          I said that at least 3 of those who resigned must have voted to censure him based on the claim that 11 voted to censure him, out of a board of 12+Gøtzsche. The number 11 seems to come from the rump board, as you found.

      • benquo says:

        On whether this is black-and-white, it seems more likely that there was a prior pattern of trying to “responsibly” keep things quiet (and writing the rules trustees have to follow in a way consistent with this), which was voted for by a combination of people who just wanted to keep bad news quiet and people who had a vague sense that it’s the responsible thing to do. Probably members of each group were unaware of the existence of the other group because of the illusion of transparency. Then this vote exposed that things had “gone too far” from the dissenters’ point of view.

        • Douglas Knight says:

          Gøtzsche, while involved at a high level from the founding, only joined the board last year in an insurgent campaign to clean up corruption. Presumably he joined the board to lobby other board members. What could he have done to be more divisive and break the illusion of transparency?

          Do written rules matter? Gøtzsche claims that he had been campaigning against the spokesperson policy (which he claims was instituted to harass him, years before he joined the board). He claimed that a number of board members supported his proposal to rescind the policy. So why did those board members vote to censure him? Maybe it wasn’t about that policy, but some more generic policy, like not being abrasive, or something not actually written, like airing dirty laundry.
          Mainly I’m confused.

  21. ADifferentAnonymous says:

    The Friston piece references a Doom-playing AI based on free-energy minimization. This seems like a chance to get a concrete answer to “why don’t you just minimize prediction error by standing still?”

    Unfortunately, this seems to have been a conference paper whose text isn’t available.: https://news.ycombinator.com/item?id=18470266.

    The author (Rosalyn Moran) has another paper based on free-energy-minimizer learning Doom: https://sci-hub.tw/10.1016/j.bpsc.2018.06.010 . There’s a ‘supplemental information’ section that’s useful as to the general free-energy approach.

    And it states, “Utility, or the agent’s sense of reward is quite literally the absence of prediction error; ‘The
    agent will find this outcome rewarding’ and ‘the agent believes this outcome is likely’ are
    equivalent statements.” Apparently optimistic prior + minimization of prediction error just works, even if it seems kinda crazy.

    • aho bata says:

      I’m still trying to wrap my head around this. What if your prior is not optimistic? Are you doomed to look on helplessly as your predictions of negative outcomes are brought to pass by your own self-destructive actions? I bet the answer has something to do with increased precision-weighting of “evidence” conducive to positive outcomes, but I wonder if that account can be cashed out in a non-circular way.

      • herculesorion says:

        So it’s the Zero System from Gundam Wing? (Or, even nerdier, how Sauron corrupted the Palantirs to show only tragic futures, so that any will too strong for him to dominate would go mad from looking in them?)

    • AlexanderTheGrand says:

      Optimistic prior is important because it encourages exploration (getting around the “staying still” situation you said).

      Imagine that you start out thinking every move gives you 100 points. You learn a lot about your environment, and discover that from the states you’ve visited you can gain at best 80 points per move. Then, whenever you have a chance to explore a state you haven’t seen before, acting “greedily” will tell you to go to the new one, not one you’ve seen.

      There’s a lot of ways to encourage exploration, this is just one of them. This one is nice because you can follow the strategy of “do what you think is best at all times”. Another thing you can do is move randomly one out of every 100 actions.

  22. Galle says:

    In what might be the most impressive temper tantrum of all time, the Saudis, angry about Qatar’s support for regional enemy Iran, are planning to dig a giant canal to turn Qatar into an island.

    Not mentioned in this summary for some reason: they also plan to build a nuclear waste dump on the Qatar side of the canal.

    • Conrad says:

      Thus adding radioactive insult to watery injury.

      Although whether the Saudis will ever build their proposed nuclear plants is an open question I guess.

  23. Tenacious D says:

    An interesting example of experimental archaeology was in the Negev. Some researchers tried to recreate the techniques for runoff agriculture used by the Nabateans. Each cultivated plot requires a micro-catchment an order of magnitude or two larger than its own area.

    Also, does speaking the bear’s true name count as experimental archaeology?

  24. cassander says:

    On State education rankings there’s also this, which adjusts NAEP scores based on what you’d expect given the state’s racial composition. The results are quite interesting.

    I accidentally posted this on the openthread, meant to put it here.

    • mdet says:

      Tell me if I’m reading their chart right: the light blue dots are how well the state’s schools performed on NAEP measures, and the dark dots are performance adjusted for 2013 demographics. Therefore, New Mexico, Louisiana, Texas, and Mississippi are the states whose schools do best (on NAEP measures) relative to their demographics, while New Hampshire, Vermont, Minnesota, and Utah do worst.

      • cassander says:

        that is correct.

      • idontknow131647093 says:

        It depends on whether you are counting the size of the adjustment or overall quality when taking the adjustment into account. Mass has a large negative adjustment, but according to this chart you’d still rather be a black kid studying in Mass than Arkansas which has a large positive adjustment.

    • Steve Sailer says:

      Most of the federal NAEP test scores by state are plausible (such as Massachusetts usually being at or near the top). But I’ve often wondered about Texas’s relatively high scores.

      Texas does very well on the low stakes federal NAEP test relative to its demographics, especially compared to California. But I haven’t seen much evidence that Texans outscore Californians on high stakes college admissions tests. (Unfortunately, it’s tricky to compare admissions scores by state since different states have different percentages of students taking the SAT and ACT.)

      Perhaps Texas officials just want high scores on the NAEP more than California officials do?

      • educationrealist says:

        Texas got slammed a while back for hiding their sped kids. They say they don’t anymore, but the scores suggest otherwise.

  25. Wolpertinger says:

    > on diminishing returns from science

    The article spends several of its motivating paragraphs on analyzing nobel prize distributions. That seems to be questionable since the amount of nobels per year is fixed. For chemistry and biology the rate looks flat, which would not be unexpected if the amount of important discoveries go up but the peak importance stays the same. I.e. one would not expect the best discovery of the 90s to have a larger importance than the one from the 80s or 70s, there are simply more with near-ideantical importance which are not measured.

    • BlindKungFuMaster says:

      Even worse: If the density of the most important discoveries goes up, they don’t stand out as much. You might by more inclined to give the prize to the one big thing from the eighties than to one of the ten big things from the nineties.

    • herculesorion says:

      There’s also the filtering of time. Important discoveries from eighty years ago look very important because eighty years to forget about all the less-important ones. Right now there’s plenty of things that all look equally important, and thus our attention is divided and nothing looks as shatteringly significant as, say, E = mc2

    • AlexanderTheGrand says:

      If I had to define “what makes a discovery important”, I would say it’s “how much understanding it created, and how much future understanding it enabled.”

      For example, we all think the wheel is important — but the first wheel probably didn’t do all that much. We care about its future rewards, and attribute all wheel-benefits from then on in some part to the initial discovery (sounds a bit MLM-like).

      Discoveries from the past have an innate advantage in understanding-creation, because they’ve been around long enough to have been fully exploited. Furthermore, every discovery now has to split its credit between all of its building blocks (NMR is amazing, but splits credit between understanding of electricity, superconductors, and nuclear physics.)

  26. Bugmaster says:

    The “Optimality vs. Fire” article says,

    Public choice economists have long argued that conventional economists hold markets to far higher standards than they hold government. Markets “fail” unless they’re optimal. Governments “succeed” unless they’re on fire.

    But I think this all comes down to the claims of each economic model’s adherents, as well as rhetoric.

    Libertarians claim that the free market is basically the perfect solution to every conceivable problem, including coordination problems, security, liberty, etc. Statists (i.e., everyone who is not some form of a radical anarcho-capitalist) claim that the government is a decent solution to some of these problems; it’s pretty bad, but it’s the best we’ve got for now.

    Thus, it seems fair to hold each proposition to its own standards.

    • Libertarians claim that the free market is basically the perfect solution to every conceivable problem

      Not true of libertarians in general. Some of us have been arguing for a long time that market failure exists but is a more serious problem for political markets than for private markets.

      And the Statist “the best we’ve got for now” requires showing not only that the market solution is imperfect but that it is more imperfect than the government solution. Hence it faces just the problem that the piece points out–applying different standards for the two.

      • Bugmaster says:

        You are right in implying that I shouldn’t lump all libertarians in one bucket; they range from relatively mild cases who believe that taxes should generally be lower; to extreme radicals who live in secure bunkers high up in the mountains, constantly scanning the skies for black helicopters. Most libertarians fall somewhere between these two extremes.

        That said, I have never seen a convincing Libertarian solution for market failure; in fact, many of them (arguably excluding yourself) refuse to even acknowledge the possibility of market failure. In the end, it usually boils down to saying either “surely some sort of a private corporation will emerge to solve this problem, though I’m not sure how”, or “people will naturally become nicer and more cooperative without the Big Bad Government breathing down their necks”.

        Thus, as I see it, the contest is not between two imperfect solutions (the Statist one and the Libertarian one); but rather, between an imperfect solution with a proven real-world track record (which, admittedly, does include failures), and nothing at all. Hence, the different standards do make sense to me.

        In fact, I could go even further: the Libertarian solution does, in fact, have a real-world track record; it just isn’t very good. Whenever official government collapses, strong-man warlords usually arise to take its place. You could view their actions as libertarian in some sense, and they are in fact competing with each other on the free market — but I would prefer democracy to “might makes right”, any day.

        • cassander says:

          That said, I have never seen a convincing Libertarian solution for market failure;

          (A) “market failure” is not one thing. there are many types of market failure, and no one solution to all of them.

          (B) Libertarians are not anarchists. By definition, they believe in a state. Assuming that by “no solution to market failure” you actually mean they all reject all non-market solutions, this is not accurate.

          (C) If I said “I have never seen a convincing statis solution for government failure” I doubt you would accept that as a slam dunk argument in favor of libertarianism. I feel similarly about your statement as an argument against it.

          in fact, many of them (arguably excluding yourself) refuse to even acknowledge the possibility of market failure.

          That’s not even close what davidfriedman said.

          Thus, as I see it, the contest is not between two imperfect solutions (the Statist one and the Libertarian one); but rather, between an imperfect solution with a proven real-world track record (which, admittedly, does include failures), and nothing at all.

          Again, you are lumping an awful lot of things together here. And as for the idea that the alternative to the state is “nothing at all”, well, don’t be ridiculous. There’s not a single good anywhere in the world that hasn’t been produced by both state and non-state actors. The question of “should the state control X” ought to be litigated on an issue by basis, not all lumped into a single question.

          Whenever official government collapses, strong-man warlords usually arise to take its place.

          At best, this is a criticism of anarchism, not libertarianism, and not a very good one. Libertarians believe in the state.

          • (B) Libertarians are not anarchists. By definition, they believe in a state. Assuming that by “no solution to market failure” you actually mean they all reject all non-market solutions, this is not accurate.

            Someone always brings this up and it’s wrong because there are a non-trivial minority of libertarians who don’t believe in the state. And those people punch above their weight on SSC.

          • Bugmaster says:

            “market failure” is not one thing. there are many types of market failure, and no one solution to all of them.

            True.

            Libertarians are not anarchists. By definition, they believe in a state.

            Wait, what ? I’m sure some Libertarians do believe in a state; but most of them practically define themselves by their opposition to government control over the economy and private property. They also tend to believe that markets provide better solutions to e.g. reducing violence than governments do… so, what’s left of the State, at that point ?

            That’s not even close what davidfriedman said.

            Yeah, that’s why I excluded him.

            And as for the idea that the alternative to the state is “nothing at all”, well, don’t be ridiculous.

            That is not what I meant. I meant that the Libertarians specifically provide no viable alternatives; this is different from claiming that alternatives cannot exist in principle.

            The question of “should the state control X” ought to be litigated on an issue by basis…

            Again, in practice, most Libertarians tend to just answer “no” to this question, regardless of X. The problem with answering “yes” to X is that X requires Y, and Y requires Z, and before long you’re just advocating for lower taxes, not Libertarianism. There’s nothing wrong with lower taxes, mind you.

          • cassander says:

            @wrong species

            Someone always brings this up and it’s wrong because there are a non-trivial minority of libertarians who don’t believe in the state. And those people punch above their weight on SSC.

            Those people are called anarchists, not libertarians.

            @Bugmaster says:

            Wait, what ? I’m sure some Libertarians do believe in a state; but most of them practically define themselves by their opposition to government control over the economy and private property. They also tend to believe that markets provide better solutions to e.g. reducing violence than governments do… so, what’s left of the State, at that point ?

            Libertarians, by definition, believe in a state that enforces law. Their preference for law might be limited to “No stealin or killin”, but wanting a small state is very different than wanting no state.

            Yeah, that’s why I excluded him.

            My apologies, I misread your comment as “arguably including you”.

            That is not what I meant. I meant that the Libertarians specifically provide no viable alternatives;

            I’m not sure what you mean by this. Libertarians believe that markets are a better solution than the state for most issues, so they advocate market solutions. What, exactly, do you think they should be advocating instead?

            Again, in practice, most Libertarians tend to just answer “no” to this question, regardless of X. The problem with answering “yes” to X is that X requires Y, and Y requires Z, and before long you’re just advocating for lower taxes, not Libertarianism. There’s nothing wrong with lower taxes, mind you.

            Well, yeah. If you think the state is a good solution to a lot of problems, you don’t tend to become a libertarian.

          • Wait, what ? I’m sure some Libertarians do believe in a state; but most of them practically define themselves by their opposition to government control over the economy and private property.

            Insofar as we have data–I’m thinking of the Liberty Magazine polls from a fair while back–only a minority of libertarians are anarchists.

          • I meant that the Libertarians specifically provide no viable alternatives

            I don’t know how you define “viable.” I spent a third of a book (the link is to the second edition) published more than forty years ago sketching how a stateless society might replace the government in its most fundamental function, making and enforcing legal rules. One chapter of that book discussed the harder problem of defending against states without a government, a topic I expanded on in the third edition.

            Since you are making a general claim about what libertarians don’t do, you have presumably read the libertarian proposals to solve those problems and found them wanting. Have you published your explanation of why mine couldn’t work somewhere?

            For the more general issue of market failure, you can find a chapter in the third edition of Machinery discussing it. The central argument is that market failure is real but that the conditions that create it are the exception on the private market, the norm on the political market, hence shifting activities from the former to the latter is likely to make the problem worse, not better.

            Again, in order to defend the claims you are making you ought to have ready an explanation of why that argument is wrong, why, for a considerable range of activities, we can expect government provision to be more nearly optimal than private provision. It is not sufficient, as Bryan points out, to merely show that the latter is, for predictable reasons, not perfectly optimal. It is obviously no sufficient to simply assume that government provision will be optimal–you should apply the same assumptions about human behavior to both systems, try to predict not what you want each to do but what each will do.

          • Murray Rothbard called himself a libertarian, was active in the Libertarian Party, and wrote a book called For A New Liberty: The Libertarian Manifesto. Regardless of what you think is the “right” definition of libertarianism, there is a sizeable group of people who don’t want a government who identify with libertarianism. And many self-described anarchists think that anarcho-capitalism isn’t really anarchists, so if anything, calling them anarchists is more confusing. Ancaps certainly have more in common with a minarchists than they do with left wing anarachists.

          • Bugmaster says:

            @DavidFriedman:
            I will of course defer to your expertise on Libertarian demographics; I don’t move in Libertartian circles, after all. But still, in my personal experience (which may be anomalous) most Libertatians I’ve met (online as well as offline) advocate for the abolition of the State in some form or another.

            I spent a third of a book published more than forty years ago sketching how a stateless society might replace the government

            I confess, I haven’t read your entire book; I’ve only skimmed a few sections.

            you have presumably read the libertarian proposals to solve those problems and found them wanting. Have you published your explanation of why mine couldn’t work somewhere?

            Well, no — I’m a software engineer, not a political scientist, after all. I know that I have argued with you (and others) on this blog regarding Libertarianism in the past, but I’m not sure if that counts as “publishing”.

            In general, though, the reason I find most Libertarian proposals wanting is because they are unstable: they either collapse into some sort of feudalism, or reduce to a more conventional de-facto state. For example, it is easy to say, “in the absence of a central government, freely competing security companies will enforce peace”; but there’s no mechanism that prevents one security company from achieving a market monopoly and becoming a de-facto dictatorship. Or, you could say, “we will still keep the government around, but its only job would be to maintain a standing army against external threats”; but armies require a lot of money to run, which means that someone has to pay the government every month; which means that someone needs to collect and log those monies somehow; and now you’ve re-invented taxes.

            The central argument is that market failure is real but that the conditions that create it are the exception on the private market, the norm on the political market

            I think I’ve read that chapter before; I’ve just re-read it, and I still wasn’t convinced. You say that voters have less influence over the government than customers have over corporations, but I don’t believe this is true; try calling Comcast’s support line for a demonstration. You also “assume away problems such a public goods and externalities”, but those are practically the entire point ! (although, admittedly, you do mention that this paragraph is a simplified summary).

            Yes, I agree that “voters, politicians, lobbyists, judges, policemen— almost never bear much of the cost of their actions or receive much of the benefit”. However, “not much” is not the same as “zero”. Furthermore, you are lumping some very different actors into a single category; for example, policemen bear a much larger proportion of the cost of their actions than the average voter. Perhaps more importantly, the average customer in a capitalist system likewise has very little effect on the market; you can personally boycott McDonalds, but they aren’t even going to notice unless you get thousands of people to go along with you.

            I would personally argue that government is much better than corporations at solving tragedies of the commons; and at producing goods and services whose benfits are either uncertain, or very diffuse. Pure scientific research (as opposed to applied engineering research) is one easy example: it has virtually no short-term economic benefits, and is thus an empty waste of resources from the capitalist point of view; yet its long-term benefits are obviously helpful to everyone.

            With all that said, I’m curious about something else: would you agree with cassander that “Libertarians believe in the state” ? Do you personally “believe in the state” ? If so, what does that mean, given that you claim (if I read you correctly) that the state is terrible at virtually everything it does ?

          • idontknow131647093 says:

            I think I’ve read that chapter before; I’ve just re-read it, and I still wasn’t convinced. You say that voters have less influence over the government than customers have over corporations, but I don’t believe this is true; try calling Comcast’s support line for a demonstration. You also “assume away problems such a public goods and externalities”, but those are practically the entire point ! (although, admittedly, you do mention that this paragraph is a simplified summary).

            I mean, this seems more like a meme than an actual reality. I call Comcast’s customer support all the time and they quickly either fix my connection issues or they give an explanation related to lines being down for repair/due to weather/etc. OTOH you can complain about a pothole every day for a month and you still won’t even have a formal response. Everyone knows the DMV memes, which are truer than comcast memes by an order of magnitude.

            So???

            Yes, I agree that “voters, politicians, lobbyists, judges, policemen— almost never bear much of the cost of their actions or receive much of the benefit”. However, “not much” is not the same as “zero”. Furthermore, you are lumping some very different actors into a single category; for example, policemen bear a much larger proportion of the cost of their actions than the average voter. Perhaps more importantly, the average customer in a capitalist system likewise has very little effect on the market; you can personally boycott McDonalds, but they aren’t even going to notice unless you get thousands of people to go along with you.

            Yes, that is very true single people in the marketplace have very little power. But they share that with single people in the political space. The advantage of the marketplace over the political space is that people actually get “extra votes” based on their actual involvement in the sector. Like if a grocery store isn’t getting very good carrots and another grocery store is, and there are a few people who really love carrots, that will actually affect each store’s carrot sales. But if there is a slight majority of people who care about cabbage, but not much and the rest don’t care, the cabbage doesn’t affect much. Whereas, the voting system is typically the opposite (except in certain very bad ways I’ll discuss later), where the 51% majority will vote for the cabbage proposal and the carrot proposal will fail.

            And going back, public choice shows us how, even if the carrot people succeed in the political space, it will probably because they have a strong vested interest in concentrated benefits whereas the rest of the voters have hidden costs (in other words a bunch of organic carrots are going to rot driving up prices for everything else). And, even worse, since its the political sphere, the carrot people haven’t even expended everything on carrots. It cost them nothing, now they can, with the same vigor, pursue a beet campaign.

          • Le Maistre Chat says:

            it is easy to say, “in the absence of a central government, freely competing security companies will enforce peace”; but there’s no mechanism that prevents one security company from achieving a market monopoly and becoming a de-facto dictatorship.

            Yeah, this always seemed to me the best pragmatic reason not to be a dogmatic libertarian. “Private security” erroneously assumes that no company will ever violate the Non-Aggression Principle, act like a company of soldiers, and run the fief it gains like a state. You’d need such a totalitarian society that the private police/soldiers are mentally incapable of thinking non-libertarian thoughts, an obvious paradox.

          • cassander says:

            @Le Maistre Chat says:

            “Government security” erroneously assumes that no government will act like a company of soldiers, and run the fief it gains for its own good. You’d need such a totalitarian society that the government police/soldiers are mentally incapable of thinking selfish thoughts, an obvious paradox…..

          • Le Maistre Chat says:

            @cassander:

            “Government security” erroneously assumes that no government will act like a company of soldiers, and run the fief it gains for its own good.

            Er? Some governments have originated as companies of soldiers and run the land they conquered only for their own good. I don’t know what point you think you’re making; some cutesy equivalence between utopian anarcho-capitalist suggestions and the 5,000+ year history of states? Because these things aren’t equivalent.

            Or if we’re being pithy, I believe in the state and not anarcho-capitalism just as I believe in monkeys and not Bigfoot.

          • idontknow131647093 says:

            Er? Some governments have originated as companies of soldiers and run the land they conquered only for their own good. I don’t know what point you think you’re making; some cutesy equivalence between utopian anarcho-capitalist suggestions and the 5,000+ year history of states? Because these things aren’t equivalent.

            Or if we’re being pithy, I believe in the state and not anarcho-capitalism just as I believe in monkeys and not Bigfoot.

            The point is easy to see if you care to try. He is saying your critique is invalid because the same applies to government. And, to be honest, if you want to invoke 5000 years I don’t think it turns out as well as you think as we are running the gamut from Ghengis Khan to Stalin to the Pharoh. The norm is a government that engages in the parade of horribles trotted out by those who fear a monopolistic “protection company” (because there is no difference between the two).

          • albatross11 says:

            Bugmaster:

            As a recommendation, maybe if you want to talk about what anarchists believe, you should use the word “anarchists” instead of “libertarians.” T

          • Jiro says:

            I mean, this seems more like a meme than an actual reality. I call Comcast’s customer support all the time and they quickly either fix my connection issues or they give an explanation related to lines being down for repair/due to weather/etc.

            The horrors with Comcast customer support are not uniform among all types of customer problems. If your issue can be fixed by following a script, it may get fixed. If it involves a billing problem, or if it involves the company saying one thing and not remembering it, or if it involves something the company has failed to fix once, or if you have to describe any special circumstances that are not in the script they will utterly fail.

          • sentientbeings says:

            @Le Maistre Chat

            “Private security” erroneously assumes that no company will ever violate the Non-Aggression Principle, act like a company of soldiers, and run the fief it gains like a state.

            That’s a common misconception. It is assumed that sometimes private security will violate the NAP, or breach contract, or behave unethically. The difference is that they will face higher economic and legal liability, rather than being shielded by things like qualified immunity or justifications based their role in society.

          • “Private security” erroneously assumes that no company will ever violate the Non-Aggression Principle, act like a company of soldiers, and run the fief it gains like a state.

            The argument for competing rights enforcement agencies concludes that, under some but not all circumstances, agencies will not find it in their interest to violate the rights of their customers.

            The analysis might be wrong–human societies are complicated, making it hard to predict the outcome of a set of institutions that has not existed in any modern developed society. But it’s an assumption only if you haven’t looked at the argument.

          • Le Maistre Chat says:

            @DavidFriedman: No, I get that. But even when the stars circumstances are right for self-interested security companies to not violate their customers’s rights, they have no motivation not to violate what their non-customers believe to be their rights.
            So say AnCaps get enough representatives to dissolve by a razor-thin margin the government of… let’s say France. Several companies pop up to provide private security for AnCaps, while the Social Security Cooperative pops up for the sort of people who voted Socialist before the State was dissolved and the Muslim minority buys Sharia police. None has an incentive to respect what non-customers think their rights are.

            Furthermore, I have no memory of what your argument was that poorer people would be able to afford private law enforcement. In the absence of that, I’d see “customers” being analogous to “chiefs” surrounded by rights/taboos while poorer people have none.

          • baconbits9 says:

            None has an incentive to respect what non-customers think their rights are.

            Yesterday’s non customer is tomorrow’s potential market base, and the non customer has friends, relatives and neighbors who are customers or potential customers. To state that there is no incentive to respecting non-customers rights is simply incorrect.

          • Furthermore, I have no memory of what your argument was that poorer people would be able to afford private law enforcement.

            Only a very small fraction of government spending goes to the function of police and courts, so the cost of private rights enforcement would be a small fraction of the current cost of government. Poor people presently pay taxes–most obviously sales taxes directly, various other taxes indirectly. So for most of them, the savings from not having to pay for government would be larger than the cost of paying for private rights enforcement.

            Further, while poor people presently pay for government rights enforcement, quite a lot of them, most obviously in the inner city areas, don’t get it.

            My picture of how you get A-C isn’t a razor thin majority voting to abolish the government. Its a continuous shift of a society towards more and more being done privately, less and less by government.

          • The Nybbler says:

            My picture of how you get A-C isn’t a razor thin majority voting to abolish the government. Its a continuous shift of a society towards more and more being done privately, less and less by government.

            History seems to point the opposite way. Gradual change moves towards more and more being done by government. Sudden change usually does also, but sometimes moves the other way.

          • albatross11 says:

            Counterexamples off the top of my head:

            The end of the draft.

            Abolition of Jim Crow laws.

            The current slow-motion legalization of pot.

            Substantial deregulation of industry, including the elimination of the CAB.

          • Le Maistre Chat says:

            @The Nybbler: Yeah. Upon David’s clarification, his goal seems low-harm but low-probability. I mean, no matter how bad a country might get without a monopoly on violence, with gradualism the allies of the dogmatic AnCaps could always abandon them once a government like pre-1917 USA, or whatever, is reached. So the only catch is “How likely?”

          • The Nybbler says:

            @albatross11

            End of Jim Crow counts as “sudden”, so does the end of the draft.

            I guess pot counts. Deregulation for airline and telephone, perhaps, but in general regulation has been increasing.

          • @Nybbler:

            You are looking at only very modern cases. England in the early 19th century shifted away from government control, with the abolition of the corn laws the most obvious example. I think the early medieval period had generally weaker government than the late Roman, although I could be wrong.

          • The Nybbler says:

            But it took the fall of Rome to make that happen. While Rome existed, did it tend to become more or less authoritarian?

          • Aapje says:

            @baconbits9

            Yesterday’s non customer is tomorrow’s potential market base, and the non customer has friends, relatives and neighbors who are customers or potential customers. To state that there is no incentive to respecting non-customers rights is simply incorrect.

            In a free market, they are two basic strategies: universal appeal and niche appeal.

            For the first strategy, the best marketing approach is usually spectacular blandness. This is obviously relatively hard to pull off.

            The other strategy is generally much easier because you can appeal to tribal identities. You on purpose align yourself with a divide in society, flattering one group while slagging off another group. Then people will use your product to signal their identity. They will even become defensive of your product as much as they are defensive of their identity.

            The latter strategy is especially useful when there is already a divide that you can tap into. For example, imagine a purely theoretical scenario where there is a societal divide between those who believe in long sentences, tough policing, disallowing drug use, etc vs those who want the opposite. Then a logical outcome is at least two policing and judicial companies, where one is much tougher than the other. Each company can then align with a tribe.

            Due to toxoplasma of rage, you will get very toxic controversies where people will pick sides and refuse to compromise.

            Then the logical outcome is that you get neighborhoods policed by company A or company B. Of course, crime doesn’t necessarily stop at a neighborhood border. What happens if a B person smokes a joint in a neighborhood governed by warlord company A and they get arrested?

            IMO, this is a recipe for civil war or the country splitting up in two nations.

          • @Aapje:

            What your analysis is missing is the fact that, since violent conflict is expensive, an important feature of a rights enforcement agency is its ability to get agreements with other such agencies on what arbitration agency they will accept for disputes between their customers. The more an agency tries to insist on rules that fit the preferences of its customers, the harder it will be to get other agencies whose customers have different preferences to agree.

          • Aapje says:

            @DavidFriedman

            Many times, coercive behaviors don’t actually seem that expensive. Quite a bit of bullying on the personal to international level gets results and is never punished.

            Furthermore, it seems quite common for people to underestimate the expense. How many wars would be “over before Christmas?”

            IMO, one of the main goals of the legal system is to discourage vigilantism by:
            – providing a fairly well functioning system, so using that instead of a vigilante solution generally results in better outcomes
            – punishing vigilantism even if the victim is rather powerless to defend themselves, to increase the costs of that choice

            International law and agreements between nations tries to do something similar.

            It all seems rather fragile to me, however. It often seems to not take that much for people to decide that the legal system is not working and vigilantism is needed or countries to decide that war is the best choice.

            Much of the work to preserve faith in justice is quite expensive and not very cost effective if one just looks at the case under consideration & only pays off in the long run. I see private organizations acting short-sighted to profit in the short term, destroying faith in the progress.

            Note that many people seem quite content with someone of the other tribe being punished. A black man (perhaps) raped your daughter? Better lynch a black man. Add a little cognitive dissonance resolution and he was actually guilty anyway.

            By your reasoning, The Troubles should never have happened, because it was so expensive.

            The more an agency tries to insist on rules that fit the preferences of its customers, the harder it will be to get other agencies whose customers have different preferences to agree.

            When people can choose policing A for their own neighborhood and have become convinced that it is quite just for criminals to be treated like A, it becomes very hard for them to accept that a rapist from neighborhood B suddenly will get treated according to policy B.

            The same differences exist between nations, but then nation B is ‘the other’ who get excused for being ‘the other’. If people within the nation have to become ‘the other’ to make people accept such differences, then how can they simultaneously accept these people as ‘ours?’

          • Guy in TN says:

            @David Friedman
            If you have two governments in competition with each other, over who will rule the people of a given area, there seems to be a mountain of historical evidence that they will not strictly compete in a non-violent manner. In fact, the outcome of this is typically one “rights enforcement agency” violently crushing the other into submission.

            You can come up with a logical deduction for why this violent conflict should rarely happen. But whatever reason you come up with, will necessarily be grinding against the empirical evidence that it almost always happens this way, with rare exception.

          • baconbits9 says:

            If you have two governments in competition with each other, over who will rule the people of a given area, there seems to be a mountain of historical evidence that they will not strictly compete in a non-violent manner. In fact, the outcome of this is typically one “rights enforcement agency” violently crushing the other into submission.

            How do you explain Germany AND France currently existing? Plus a whole bunch of other European states? There is a ton of evidence that sometimes one state crushes an other, but also a ton of evidence that states can have relatively peaceful relations with their neighbors for long stretches.

          • But whatever reason you come up with, will necessarily be grinding against the empirical evidence that it almost always happens this way, with rare exception.

            I think you have it backwards. Almost all countries, almost all of the time, are at peace with each other–it’s war that is the rare event.

            Beyond that, the incentives of a territorial sovereign are different from those of a firm competing for customers who, in the initial situation, are free to take their business elsewhere.

            The argument here is about equilibrium–if the sort of system I have argued for comes into existence, will it stay in existence. Getting to that point is a different problem.

          • Guy in TN says:

            @baconbits9
            I don’t think anyone is arguing that you can’t have separate rights agencies, with the limits of their authority demarcated by geographic boundaries.

            If that’s an acceptable situation to you, with each rights enforcement agency maintaining a violent monopoly over its geographic area (while maintaining peaceful relations with neighboring rights enforcement agencies), what is the objection to states again?

          • John Schilling says:

            but also a ton of evidence that states can have relatively peaceful relations with their neighbors for long stretches.

            Peaceful relations with their neighbors, yes. The question was one of governments competing for the power to rule people in the same area. Germany is not the same place as France, definitively so, and Germany does not try to rule people in France or vice versa.

            Nations insisting that they have the power to rule people in the territory of other nations, has much less relevant history behind it and most of that is of a basically paternalistic colonial or imperial variety that often ends in violence.

          • Le Maistre Chat says:

            I can’t help thinking of Mystery Science Theater 3000‘s send-up of Hamlet now.

            Claudius: Will you be ruled by me?
            (as Laertes): Well, I have an attractive offer from the king of Flanders…

          • Nations insisting that they have the power to rule people in the territory of other nations, has much less relevant history behind it and most of that is of a basically paternalistic colonial or imperial variety that often ends in violence.

            The Catholic church maintained such a claim for many centuries, for the most part peacefully.

            You are taking it for granted that the claim is of territorial sovereignty, in which case there is an obvious conflict. But the institutions I am proposing are ones in which nobody is making that claim, and anyone who does faces costly conflicts that all other agencies avoid. Polylegal systems, systems in which different people in the same country are under different legal rules, are fairly common in the historical record.

          • Aapje says:

            @DavidFriedman

            Catholic history includes a lot of prosecution of ‘heretics.’ It is one of the major reasons why my country exists.

            It’s also one of the major reasons why American Jews are mostly concentrated around New Amsterdam York.

            The Catholic church has since actually moved to weaken its interpretation of the bible to clash less with ‘Caesar.’ For example, Paul says that transgressors should be shunned, but the Catholic church removed shunning/vitandus from canon law.

            A conflict over the authority of the Catholic church vs local rule was one of the major reasons for the English Civil War, English Reformation and eventually the Glorious Revolution which resulted in establishing parliament as the ruling power of England.

            Your proposal seems to me to be a regression that can only work with heavy segregation, while segregation is becoming less viable.

          • Nornagest says:

            It’s been relatively common historically for different law codes applying to different people to coexist in the same geographical area, without strict segregation of the populations they applied to (some informal self-segregation usually happened), but they’ve usually been class- or religiously stratified and had fairly rigidly defined relationships between them.

        • albatross11 says:

          If you’re going to list Somalia-style warlordism as a failure of libertarianism, okay, I guess I can go along with that, as long as I can list Mao’s China, Stalin’s USSR, Hitler’s Germany, Pol Pot’s Cambodia, North Korea under the Kims, etc., as failure modes for statism.

          Most libertarians want to keep a state around, they just want it less powerful and intrusive and large than the current version we see in most Western democracies. Assuming that any state smaller than the current one = Somalia is exactly as valid as assuming that any state larger than the current one = North Korea.

          • If you’re going to list Somalia-style warlordism as a failure of libertarianism, okay, I guess I can go along with that, as long as I can list Mao’s China, Stalin’s USSR, Hitler’s Germany, Pol Pot’s Cambodia, North Korea under the Kims, etc., as failure modes for statism.

            You don’t need to go that far. Just look at conditions under the military dictatorship of Mohamed Siad Barre before the government that had been created by the exiting powers–for a society that had never had a government–collapsed. Estimates of the Isaaq genocide alone under Barre range from 50,000 to 200,000, with nearly a million more either fleeing the country or being internally displaced. That was in a country with a total population of about six million.

            Since Barre was overthrown, Somalia has been kept in turmoil by the attempts of outside powers, mostly the U.S. and U.N., to reimpose central government–using for the purpose the military of Ethiopia, Somalia’s traditional enemy. Somaliland, the northern part of Somalia, has its own functioning “Republic of Somaliland” based mostly on the traditional institutions, but the rest of the world refuses to recognize it, since that would require admitting that Somalia isn’t a country instead of trying to force it to become one.

          • Aapje says:

            @albatross11

            Most ‘statists’ are actually in favor of a mixed economy with many freedoms to make (economic) choices, not full control by the government. Totalitarian failures are not mixed economy failures.

          • baconbits9 says:

            Arnold Kling puts it similar to this.

            Statists: Markets fail, use governments to fix it.

            Libertarian: Markets fail, use markets to fix it.

        • John Schilling says:

          That said, I have never seen a convincing Libertarian solution for market failure

          If “market failure” means that the market’s performance was less than perfect, isn’t the solution usually going to be a giant “Meh”?

          Attempting to force perfection onto good enough is one of the great mistakes that causes great harm almost everywhere, and it is one seen far more often in affairs of state than in the market. “Solving” most market failures, ought to be way down on anyone’s list of ways to make the world a better place.

          • albatross11 says:

            So, some libertarian solutions to market failures (imperfect, but so are the alternatives):

            a. For positive externalities (where people don’t do enough of X because some of the benefit spills over onto others):

            Social forces–prestige, standing in the community, etc., gained by being seen to do pro-social things. This exists just about everywhere.

            Charity–plenty of libertarians give money to soup kitchens, for example. 100% voluntary, no coercion, and an attempt to provide things that are pro-social.

            b. Negative externalities: (People do too much of X because they don’t pay all the costs.)

            Nuissance law/lawsuit/private right of action

            Contract (restrictive covenants banning drumming in Nodrumlandia)

            c. Information asymmetry:

            Seller reputation, brand reputation, third-party rating organizations

            d. Monopoly

            Not letting the government set up / defend a monopoly.

            Competition

            Now, I’m surely leaving things out, and every one of these is deeply imperfect. But so are the alternative government solutions. Government attempts to do things with positive externalities are often successful, but they also often turn into machines to hand out taxpayer money to cronies of the people in power, or permanent sinecures that can’t ever go away even when they stop making sense. Government attempts to prevent negative externalities can end up choking industries with regulation, or protecting incumbents because they got grandfathered in. In the worse case, you get the war on drugs.

            [ETA] Just to be clear, I’m not claiming that libertarian approaches to solving these problems work great, or are always adequate. I’m saying that libertarians do, in fact, have a fair number of ways of addressing known market failures. Indeed, this is something some libertarian thinkers spend a lot of time thinking about.

        • gbdub says:

          Very rarely in the realm of policy debate is the choice between “full anarcho-capitalism” and “totalitarianism”.

          Usually it’s “privatize/nationalize this” or “regulate/deregulate that”, where “this and that” are incremental things rather than huge sweeping changes.

          And in that realm, it absolutely is wrong to hold “the state” and “the market” to double standards. If one way or the other is more efficient / makes more people happy on net / whatever your metric is, just do it. Wasting money or causing needless grief is just as bad when the government does it as when a corporation does it.

          • Le Maistre Chat says:

            Yeah, this. I know from experience that there are lots of things the government is bad at, so let’s just try to get it out of them if we can. “Any tax is theft! Non-Aggression Principle!” is not an interesting discussion to have, because if the State was ever abolished by a thin majority, everyone who’s not a dogmatic AnCap would be at liberty to recreate it.

          • because if the State was ever abolished by a thin majority, everyone who’s not a dogmatic AnCap would be at liberty to recreate it.

            1. Creating a state faces the public good problem–if you spend resources doing it on the theory that it will make everyone better off, everyone else shares the benefits. So even if it is worth doing it may not be done. One of those “market failures” people have been talking about.

            Whether it happens depends on how well the people trying to do it manage to find a private solution to that particular market failure, and how well the people who don’t want it done manage to solve their side of the same problem.

            We know that it doesn’t always happen because there have been stateless societies that remained stateless for extended periods of time, although no modern developed ones.

          • Le Maistre Chat says:

            @DavidFriedman: Yeah, I know from studying anthropology that primitive anti-authoritarianism is a thing. The standard sequence we learn about settled village societies is tribe -> chiefdom -> state. And a chiefdom doesn’t seem like a better deal for the people living in it, as members of chiefly lineages have far more rights and which chief controls a village has a high probability of violently changing after ~3 generations.

            One obvious problem with a modern developed population trying to revert to a stateless, chief-less society is how difficult it would be to live at urban densities with total sovereignty over each little piece of property. The stateless period of medieval Iceland, the famous late example, had practically 0% urbanization.

          • albatross11 says:

            Growing your new state from a criminal gang into a warlordom probably manages to get around the public goods problem.

          • Aapje says:

            @Le Maistre Chat

            I think that chiefdoms can work very well when they are small scale and little opportunity for rent seeking exists. Basically, a system where the chief is heavily dependent on approval by his subjects and where his power heavily depends on his subjects voluntarily complying.

            As soon as you get things like standing armies that can give their allegiance to the chief and can suppress (minor) revolts, the chief can start to ignore the will of the people too much.

        • Garrett says:

          As others have noted, there are several flavors of libertarian. I tend to come via the Ayn Rand-aroid route as a minarchist. That having been said, I believe that there are essential a roles of the State:
          * Police, courts, national security and public health (mostly in a communicable disease manner).
          * These roles should be limited to protecting the people against violations of their individual rights via violence, theft, fraud and the like.

          There are issues where market failure does exist. In addition to the usual economic failures typically addressed, these usually happen for reasons such as:

          * Public property, such that costs can be costlessly externalized.
          * Topology: Running a road on the ground is cheap – running a competing road elevated above an existing road is extremely expensive.

          • albatross11 says:

            I’m not exactly a libertarian, but that’s the closest label to what I am, so let’s run with it.

            I think it’s useful to distinguish between people who hold to some political ideology in the sense of an endpoint (this is the ideal society I’m wanting to get to) or a direction (I’d like to move society in this direction). I’m a libertarian in the second sense, but not the first–I think redesigning society is extremely hard and unlikely to go well, and I don’t think even very smart people (hi, David!) can work out how that will turn out in detail. We have a pretty high-functioning society now, with institutions that work and a high standard of living for most people. I don’t want radical change, because I think it’s likely to break stuff we will find hard to fix, in much the same way we broke marriage and family norms and now have no idea how to fix them.

            I’d like to try some incremental moves toward more liberty, less centralization of power, less intrusion by the state in private lives and decisions and markets, smaller governments, lower taxes, smaller deficits, less aggressive and interventionist foreign policy, fewer bailouts of politically important industries, less occupational licensing, etc. We have worked examples of our own and other socieities functioning passably well under such conditions, so this isn’t roll-the-world-back-to-the-year-zero stuff. This all seems doable, and I suspect that now, at the margins we would benefit from becoming more libertarian overall. This is true even though I don’t really think we can run a whole functioning society on the non-aggression principle.

          • Tenacious D says:

            @albatross11

            We have a pretty high-functioning society now, with institutions that work and a high standard of living for most people. I don’t want radical change, because I think it’s likely to break stuff we will find hard to fix, in much the same way we broke marriage and family norms and now have no idea how to fix them.

            I’d like to try some incremental moves toward more liberty …

            I agree with both parts of this. At the same time, I’d like to see some test tubes for people who want to try designing new institutions from first principals. Maybe charter cities or seasteads or even–a long way down the road–lunar/martian habs could provide such opportunities some day. New wineskins for new wine.

        • Guy in TN says:

          @Bugmaster
          Be precise with your word choice there- I’m not sure, from your post, if you want to talk about “failure of the market” in the plain-English sense of the phrase, of “Market Failure” in the neoclassical-economic sense.

          (The latter which I don’t see why non-libertarians should be too concerned about, in the first place)

      • rahien.din says:

        the Statist “the best we’ve got for now” requires showing not only that the market solution is imperfect but that it is more imperfect than the government solution.

        David.

        I’m sympathetic to a lot of things about libertarianism. I like coming here to indulge my curiosity.

        But my biggest concern about libertarianism is that it is basically a power vacuum. And human beings in a power vacuum operate in reliable zero-sum, winner-take-all ways.

        Libertarians, in my experience (too), largely discount that feature of human behavior, while simultaneously espousing a zero-sum, winner-take-all approach to political discourse.

        Instruct me in which one I should actually believe?

        • You should believe that, under either system, individuals try to take the actions that best serve their objectives. They are not perfectly rational—sometimes they are wrong about what actions have what results—but insofar as there is a pattern it is that they tend to choose the actions which best achieve their objectives.

          The difference between the political marketplace and the private marketplace as systems is that people with exactly the same preferences and abilities find it in their interest to act in quite different ways in the two systems.

          If that isn’t obvious, the long explanation would be a book on price theory and a book on public choice theory. You can find shorter explanations in a variety of places, including my first book.

        • sharper13 says:

          One thing you should distinguish is between “no government” and “no monopoly government.” Anarcho-capitalists, voluntarists, whatever you want to call that subsection of small-l libertarians who aren’t really Libertarians, are in favor of the latter, not the former.

          Firms organized to provide competing services to their potential customers have historically provided an overabundance of whatever services the market desired. The government services market isn’t an exception to the economic natural laws based on human beings which drive grocery markets, or ski instructor markets or whatever is being transacted.

          There are well-understood public choice (and other) economic reasons the current monopoly government services tend to look less like an American grocery store and more like a Soviet grocery store.

          A power vacuum assumes no one provides a particularly needed government service, but there are people (customers) available. That’s just not our experience with any other sort of market. Typically, as quickly as someone can “invent” a new need, or a technology which fills a need no one realized they even had yet, in a free economy people pop up all over the place trying to fill that need because of the incentives involved.

          Yes, if your only experience with shoe stores was a government monopoly store where they had exactly two styles of shoes, each made in the sizes estimated to be needed that year and a “fashion” committee of bureaucrats who were supposed to keep the costs low in order to be able to reduce the annual government free single pair of shoe costs, then you’d have a difficult time imagining how if the government stopped monopolizing shoes and giving them away for free, anyone would ever manage to have decent footwear, but the example only sounds ridiculous because of your real life experience of how markets in shoes work currently in a freer economy. The market doesn’t provide “less” shoes to people just because the government monopoly is shut down.

          Oh, but shoes aren’t essential to life. What about the power of providing life-saving stuff like crime protection and defense. Wouldn’t that give firms too much power because everyone has to have those? Well, see also grocery stores and restaurants again. Everyone needs to eat to live, right? Yet somehow there is plenty of competition and customers responsiveness.

          • The problem with your initial point is that it depends on how you define government. Anarcho-capitalism, at least my version, has private providers of what we usually think of as the most central function of government–law and law enforcement. Does that mean they are governments?

            I don’t think so. Given that pretty much everything governments do has at one time or another been done by something not thought of as a government, I don’t think the right definition is in terms of function.

            For my answer to the question, see this chapter from the third edition of Machinery.

          • sharper13 says:

            Obviously I wasn’t being as clear as I would want to be. I’ll need to rethink how I can make the point clearer.

            I was using the term government as a short-cut for the idea of provision of traditional government services such as law and police, not for a specific organization. So in your version, I wouldn’t call them “governments”, but I would call them providers of government services, in the sense that someone can provide accounting services, or any other kind of services, if you see the difference in emphasis. I can see in retrospect that for most people the word government is too closely aligned with an organization which “governs” people, rather than providing a certain type of services to people.

            For now I’ll awkwardly rephrase my point as:

            One thing you should distinguish is between “no [services currently performed by the government] being available” and “no monopoly on providing those same services”. Anarcho-capitalists, voluntarists, whatever you want to call that subsection of small-l libertarians who aren’t really Libertarians, are in favor of the latter, not the former.

          • Anarcho-capitalists, voluntarists, whatever you want to call that subsection of small-l libertarians who aren’t really Libertarians

            In the usage I’m familiar with, the distinction between “Libertarian” and “libertarian” isn’t beliefs, it’s that a Libertarian is a member of the Libertarian Party, a libertarian need not be. The former describes membership, the latter ideology. Some libertarian anarchists are party members. Some minarchist libertarians are not.

            How are you defining the distinction?

          • sharper13 says:

            I distinguish the same way you do, clarifying I was talking about a subset of political classification/beliefs and ideology using the term, not referring to the LP by using the term.

          • Then I don’t understand how you were using the libertarian/Libertarian distinction. If “Libertarian” doesn’t mean LP member, what does it mean, how does it differ from “libertarian?” What defines the subset?

          • sentientbeings says:

            @sharper13

            One thing you should distinguish is between “no government” and “no monopoly government.”

            If I understand your meaning in this distinction, I think there is a better way to phrase it, which is a distinction between no governance and no government; i.e. some set of institutions relating to enforcing rules among humans versus the coercive state.

          • sharper13 says:

            @DavidFriedman,
            I think you misunderstood my most recent clarification. I was using the capital L Libertarian to refer to the Libertarian Party, originally primarily to say that I was talking about a subset of libertarians as a whole, not about a subset of the LP. Hope that makes more sense.

            @sentientbeings,
            Governance is a good word to use. Possibly even better to be more specific about what is meant by breaking it down into a few examples. So not in favor of no provision of legal, protection, charitable welfare, road, etc… services, but rather in favor of no organization monopolizing the provision of those services.

  27. In general, though, the reason I find most Libertarian proposals wanting is because they are unstable: they either collapse into some sort of feudalism, or reduce to a more conventional de-facto state. For example, it is easy to say, “in the absence of a central government, freely competing security companies will enforce peace”; but there’s no mechanism that prevents one security company from achieving a market monopoly and becoming a de-facto dictatorship.

    I spent parts of both the first and third edition of Machinery discussing under what circumstances that would or would not happen.

    Yes, I agree that “voters, politicians, lobbyists, judges, policemen— almost never bear much of the cost of their actions or receive much of the benefit”. However, “not much” is not the same as “zero”.

    In the particular case of a voter in a national election, “not much” is roughly 3×10^-9, since the effects are spread across the entire population of the U.S. That is very close to zero.

    Perhaps more importantly, the average customer in a capitalist system likewise has very little effect on the market; you can personally boycott McDonalds, but they aren’t even going to notice unless you get thousands of people to go along with you.

    If I choose to not eat at McDonalds the result is that I don’t eat at McDonalds. McDonalds sells one fewer of their hamburgers, gets one hamburger’s less revenue, I pay one hamburger’s less money and eat one fewer McDonalds hamburger. If ten people do that the same thing happens ten fold. Assuming McDonalds is paying attention, ten fewer hamburgers get produced. Our action is having a direct effect which has nothing to do with a boycott. You are trying to squeeze the mechanism of the market system into your mental model of the mechanism of a political system—that isn’t how market systems work.

    Pure scientific research (as opposed to applied engineering research) is one easy example: it has virtually no short-term economic benefits, and is thus an empty waste of resources from the capitalist point of view; yet its long-term benefits are obviously helpful to everyone.

    Terence Kealey, a British scientist, has an interesting book discussing the indirect mechanisms by which basic research is funded on the market and trying to measure the real world effect of government subsidies for such research. You might find it worth reading. Your mistake is taking it for granted that if the government intervenes it will make the right choices.

    One of the reasons I left theoretical particle physics after a couple of years as a post-doc was concluding that because the success of the A-bomb had convinced the general public, including the government, of the importance of physics, far too many resources were being diverted into it. The result was that a lot of smart people were working in that field who would be of more use doing something else. The accomplishments of the field in the forty-some years since do not strike me as evidence that I was mistaken.

    Do you personally “believe in the state” ?

    I believe that in many, but not all, circumstances a stateless society would be stable and superior in its results to any likely state.

    This started with your writing:

    I meant that the Libertarians specifically provide no viable alternatives

    It is reasonable enough to hold opinions on subjects one knows little about–we can’t all be experts on everything. It isn’t reasonable to state such opinions as facts.

    Let me be, for a moment, immodest. In recent decades there have been two prominent anarcho-capitalist theorists. Murray Rothbard is the other one. The fact that you didn’t know that is some measure of how carefully you investigated libertarian ideas before making the quoted statement.

    • rahien.din says:

      Bugmaster : I haven’t encountered a libertarian solution other than “The market will magically fix everything.”

      David Friedman : That’s because you haven’t paid enough attention to me and this other genius.

      Bears. Beets. Battlestar Galactica.

      • Bugmaster : I haven’t encountered a libertarian solution other than “The market will magically fix everything.”

        David Friedman : That’s because you haven’t paid enough attention to me and this other genius.

        Bears. Beets. Battlestar Galactica.

        I’m afraid the Battlestar Galactica reference went over my head, since I haven’t watched it. But assuming “this other genius” is Rothbard, I am not suggesting that people should pay more attention to him. Huemer maybe.

    • Aapje says:

      @DavidFriedman

      It’s rather senseless to calculate the short-term benefit of a single vote to the individual, because the operative mechanism of democracy is collective behavior, not individual behavior. You don’t vote to get the government to meet your own exact needs, voting creates change due to collective behaviors.

      Furthermore, the main benefit of democracy is not to nudge leaders slightly in the direction of your desired policy, but to be able to get rid of leaders who act badly. This is extremely beneficial.

      So the proper calculation is not: my vote has 0.000001% influence on nudging the minimum wage down/up, but: if many people in society participate in elections, there is a high chance that leaders with really bad ideas get kicked out of office, greatly reducing the risk of everyone ending up oppressed by Adolf Stalin during my life.

      Voting is like a Dutch people paying for dike maintenance. It’s hard to see the benefit when you do it and it may be tempting to spend the money on stroopwafels, but if we do that, we will come to regret it.

      • sentientbeings says:

        Furthermore, the main benefit of democracy is not to nudge leaders slightly in the direction of your desired policy, but to be able to get rid of leaders who act badly. This is extremely beneficial.

        Quite nearly 100% of the set that I consider “leaders who act badly” are either retained or replaced by others for whom I can find no meaningful difference. That number could be lower for other people, depending on their viewpoints, but incumbency rates and very long list of politicians that survive major scandals speak to my point. If democracy only protects us from the very worst while enabling a great many who generally act badly, that’s only a weak point in its favor, since there are other ways to protect against the very worst.

        If I don’t want a product or service, I can usually immediately stop buying it (and I can do so without the provider or competitors threatening to lock me in a cage).

      • So the proper calculation is not: my vote has 0.000001% influence on nudging the minimum wage down/up, but: if many people in society participate in elections, there is a high chance that leaders with really bad ideas get kicked out of office, greatly reducing the risk of everyone ending up oppressed by Adolf Stalin during my life.

        What happens if many people do X is not relevant to what I should do since I control only one person. If everyone on the ship moves over to the starboard side the ship capsizes. That’s not a reason why I shouldn’t.

        The proper calculation is what the effects are of my action–obtaining the information to decide who to vote for and then doing so. If the effect of my deciding which candidate for president is better and then voting for him is to increase the probability that the better candidate wins by about 10^-7, a calculation that optimistically assumes I am likely to get the right answer, then the question is whether increasing that chance by that amount is worth the cost to me of doing so.

        If I am sufficiently altruistic to count a benefit to three hundred million strangers as worth at least a million times as much to me as a similar benefit to me, then the answer is probably “yes.” If much less altruistic than that, probably “no.”

        I think observed voting behavior is consistent with that argument. Many people vote because they enjoy the feeling of participation in an important ritual, but few of them bear the costs needed to make a rational judgement on how they should vote–as suggested by the fact that most elections are reasonably close. If a hundred million people are rationally calculating the answer to a question with only two answers, it would be surprising if about half of them got it wrong.

        • Aapje says:

          What happens if many people do X is not relevant to what I should do since I control only one person.

          This is overly individualistic. Your actions influence the norm, which influences the behavior of others. So you do control other people to some extent.

          Studies have shown that people heavily take their cue from others (although SSC readers are probably relative outliers in this respect).

          Many people vote because they enjoy the feeling of participation in an important ritual, but few of them bear the costs needed to make a rational judgement on how they should vote

          It is actually very rational to participate in a vote between Tweedledum and Tweedledee and yet not put too much effort in because the two are not hugely different, just to preserve democracy as a working institution.

          That lazy voter can just copy the choice of a friend who doesn’t seem insane and would probably not vote for a truly absurd candidate, vote for whom the media portrays as decent, etc.

          • I wrote:

            What happens if many people do X is not relevant to what I should do since I control only one person.

            Aapje responded:

            This is overly individualistic. Your actions influence the norm, which influences the behavior of others. So you do control other people to some extent.

            And such effects should be taken into account, but that doesn’t change the conclusion.

            In the case of voting we have a secret ballot, so how I vote cannot influence others. How I say I am going to vote might, but unless I am a very high profile person the net effect on the outcome is still tiny.

            If effects of that sort are sufficient to solve the public good problem in voting, where the external benefit (U.S.) is about three hundred million times the private benefit, solving the various market failure problems commonly offered as a justification for government should be easy.

        • nimim.k.m. says:

          If everyone on the ship moves over to the starboard side the ship capsizes. That’s not a reason why I shouldn’t.

          But that example does not describe elections at all.

          Are you aware of planned collective action to capsize the ship, loudly lead by the First Mate who has been telling everybody who listens to move starboard on tomorrow midday, and it seems he has gathered quite plenty of supporters (one wonders why). And there’s also this the grand opposing plan to capsize the ship better by getting everyone to run to port side, proposed by His Majesty the Captain’s Loyal Opposition, also to take place on tomorrow midday, and they also are not small in their numbers?

          While you have talked to other passengers and maybe have some idea what they say they will do, however, you are not absolutely sure what everyone will do, but have the knowledge of your own preference of the outcome. And for the sake of further dreadful torture of the metaphor, let us assume that for some reason everyone has agreed to stick their heads in a ballot box to obscure their identities when the crucial moment comes to alleviate peer pressure.

          Of course you can still loiter on the deck randomly and not contribute to the result except by the chance, but that is not everyone else will be doing.

  28. The Nybbler says:

    So the cost of raising the salary they’re offering for an empty position is less like X and more like 21X.

    That’s _a_ reason, and one which firms get around in various ways (like giving the new position a new name or level, or keeping the salary confidential). But it’s not the only reason. Another one is low elasticity of supply. It’s possible everyone in position available between the current prevailing salary and the highest salary which would be economic to pay is already employed. If that’s the case, raising the salary may poach someone from another firm, but the other firm could just raise to get them back. It just starts a bidding war which brings salaries up but provides no new talent. If the firms figure out this is the situation, they’ll be very reluctant to raise salaries.

  29. nameless1 says:

    Bear: this is the second time I find Hungarian language is doing something IE languages are supposed to do despite not being IE. Wolf = farkas = the tailed one, deer = szarvas = the horned one. Ancient totem animals whose name was taboo, so euphemisms were invented. How is Finnish in this regard?

    First time was more a cultural custom than language. Someone casually mentioned “ancient Germanic custom of double kingship, one sacral, one military”. What? This was supposed to be a feature of ancient Hungarian culture being markedly different from the IE cultures and more in common perhaps with the Tengri religion cultures in Inner Asia.

    • spkaca says:

      There are intriguing hints of dual rulership structures in the early Roman traditions. The story of Romulus and Remus, for instance, or the fact that the highest office in the Republic (the consulship) was dual i.e. held by two men at a time. This may be a coincidence – the desirability of separating religious and military leadership is sufficiently clear that the practice might have been independently developed more than once.

      • Salem says:

        Sparta also had dual kingship. (And occasionally duel kingship…)

      • The original Mr. X says:

        The evidence for dual kingship in early Rome is pretty tenuous, IMHO. For one thing, Romulus and Remus weren’t actually kings together (Romulus killed Remus before they’d finished founding the city). As for the consulship, it wasn’t introduced until around 250 years (traditional dates) or 150 years (dates suggested by modern archaeology) after the foundation of the city, and was preceded by a long period of sole rule. Plus, the consulship did not in any way imply a separation of religious and military leadership — both consuls had full religious and full military authority.

        As for the Spartan monarchy, this also didn’t imply a division of labour — whilst it was usual in Classical times for one king to lead the army and the other to stay at home, this was apparently a relatively late (late 6th century) development, at least according to Herodotus, and each king anyway retained full military authority, even if only one of them was actually in the field.

    • Ketil says:

      Wolf = farkas = the tailed one

      Hard to believe this isn’t related to Scandinavian “varg”. From WP:

      In Old Norse, vargr is derived from the reconstructed Proto-Germanic *wargaz, ultimately derived from the reconstructed Proto-Indo-European (PIE) root *werg̑ʰ- “destroy”

      The standard name for wolf in Scandinavian (the Swedish prefer “varg”, I think) is “ulv”, which is

      From Old Norse úlfr, from Proto-Germanic *wulfaz, from Proto-Indo-European *wĺ̥kʷos.

      But I also find:

      From Proto-Indo-Aryan *wŕ̥kas, from Proto-Indo-Iranian *wŕ̥kas (“wolf”), from Proto-Indo-European *wĺ̥kʷos (“wolf”)

      …so while there are probably good reasons for deriving etymology like this, it’s peculiar that the same root (wĺ̥kʷos) gives rise to a word for wolf that is suspiciously close to “varg”.

  30. AlphaGamma says:

    I remember trying to reconstruct what the “real” English word for bear would be based on sound changes between PIE and English. I think we arrived at *orth.

  31. nameless1 says:

    So, Friston. And this: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3347222/

    > Agents that predict rich stimulating environments will find the “dark room” surprising and will leave at the earliest opportunity. This would be a bit like arriving at the football match and finding the ground empty

    So. I agree that my depression is predicting negative emotions with high certainty. I don’t go to the football match because I think I would not enjoy anyway. This results in self-fulfilling prophecies, like avoiding having fun, this way my brain reduces surprise: I don’t do fun stuff and just as predicted. Won’t have fun. But doing fun stuff like going to that football match would not be helpful, because my brain would strongly resist the updating that this is actually fun. I don’t think “just do fun stuff even if you don’t want to” was ever a working treatment, it is too obvious. Besides many depressed people do fun stuff out of habit or social obligations, just stop enjoying them.

    Could you have so much fun stuff that basically you force your brain to overcome the resistance and update? Well that may be why psychedelics look promising for depression treatment. It is like forcing pleasure down the brain’s throat until it stops resisting and accepts that yes this feels good. But apparently normal pleasures like sex, food or music don’t really work for it.

    Or can one directly address the prediction? The depressed brain predicts negative emotions with high certainty, and negative emotions are themselves a prediction of an uncontrollable state of things. Friston calls it hyperprediction, but that is weird, metaprediction would be more understandable. Predicting the ability to predict which equals the ability to control. So depression is like “I am sure I cannot control my life, whatever I do does not matter.”

    Which suggest retraining the brain by doing things repetitively that you can predict and control the outcome. Just play a videogame on easy mode. But it does not seem to work. If it worked we would have noticed it. Our brain does not really care that much if we can beat this monster or have to reload that quicksave from two minutes ago. We don’t care about videogames in the first place that much, and they got easier, 25 years ago you had to replay a whole level if you died or the whole game, now you reload the last quicksave or e.g. the Assasin’s Creed the last checkpoint in the mission. They are less challenging now. Whether we can control not having to replay the last two minutes or not does not matter as much for us as replaying the last 30 minutes.

    We have to find stuff that 1) matters 2) is actually kind of hard to control enough 3) become good at controlling it and repeat it to change the metaprediction of uncontrollability. Then after enough repetitions it stops working because the brain partially updates and thinks “okay this thing is controllable, not most things though” and becomes a bit less depressed and then we have to learn another new thing. This way we could train our brain out of depression, perhaps.

    But! The problem is that a depressed person finds it very hard to find things that matter, precisely because predicting that nothing will be fun. So what do? Do what other people do? Like if other people enjoy going to the cinema with friends, go to the cinema with friends even though it is not fun, and repeat until it becomes fun? And if it doesn’t? Is that even supposed to work?

    But the cinema is maybe a bad example. Your control ends at choosing the movie. The rest is passive consumption. We know exercise helps, but I think most people tend to think exercise equals boring jogging or weight lifting. What if you force yourself to learn to play tennis? The exercise helps with the bad mood anyway, you care about scoring a point because humans are competitive (you don’t enjoy it while you suck at it but maybe later when you get some success yes) and you learn to control it, you can win half your serves, you get to predict at least with 50% chance that I am getting a score, yeah? Is this a good idea?

    EDIT: is there something like tennis but more oriented towards muscle building, dunno, heavy medicine ball sport or whatever? Something competitive, something where you get to predict similarly often – in tennis about every 3 seconds – if you succeed on doing something or not, but builds muscle mass?

    • spkaca says:

      How about archery? Competitive but not violently so, strong predictive/ feedback element, exercises/ builds upper-body muscle, more interesting than jogging/ weightlifting. There are satisfying quick wins in the early phase of learning, and you get to make informed critique/ commentary on historical movies/ TV.

      • nameless1 says:

        This actually sounds good. I was thinking about doing something historic, like longswords, but I have slow reflexes. But there are satisfying quick wins in the early phase? It is a common historical truism that training archers took super long, and that is why firearms took over…

        • theredsheep says:

          Uh-oh, you said “longsword.” And you didn’t specify that you meant something two-handed? Take cover.

          • So far as SCA combat is concerned, most commonly sword and shield, there may be some fun in the early stages but mostly the early stages consist of feeling clumsy and ending up with sore muscles. I suspect that’s true of most competitive sports.

            One advantage of archery is that it doesn’t have to be competitive. Being able to hit the target when you couldn’t before is some satisfaction even if other people are hitting the gold.

        • Protagoras says:

          As I understand it, that historical truism is exaggerated; firearms, especially primitive firearms, took a lot of training. Firearms mostly took over because of their greater lethality and armor penetration ability. And what truth there is to it is partly because historic longbows took a tremendous amount of strength. Modern archers do not use bows with that kind of pull.

        • spkaca says:

          “One advantage of archery is that it doesn’t have to be competitive. Being able to hit the target when you couldn’t before is some satisfaction even if other people are hitting the gold.”
          This is quite true, in my experience.
          “historic longbows took a tremendous amount of strength. Modern archers do not use bows with that kind of pull.”
          Also true. Longbows are lighter than modern recurve bows (the usual kind you’d use to begin with) but harder to use. Getting to a militarily useful level of proficiency would take a while.

          • Longbows are primitive weapons–basically bent sticks, although the yew version takes advantage of a sort of natural composite structure.

            Modern recurves are a modern tech version of an ancient high tech bow, the wood/horn/sinew composite. I think I’ve seen a description of a Chinese version with a draw weight in the range of the reconstructed Mary Rose bows, but for ordinary military archery, largely on horseback, nothing that extreme was needed.

            And they outranged longbows. By a lot.

            But I agree that getting to a militarily useful level of proficiency would take a while.

    • gbdub says:

      Indoor rock climbing? Obviously good exercise. The routes are all rated so you can literally “level up”. You can compete with a friend or against yourself to see who can successfully climb the toughest routes, climb them fastest, get a new route on the first try, etc.

    • Simulated Knave says:

      The problem with failing to predict and building muscle mass is that exerting your muscles in the wrong direction is a great way to strain them. Exert a large force in the wrong direction and you get hurt, not stronger.

      If you live in the country, split wood by hand. But you probably don’t.

      Medieval martial arts with weighted swords might work (and is arguably easier to do on your own than tennis).

      • Why “weighted swords?” Do you mean heavier than the real swords were?

        • Simulated Knave says:

          Because if he’s talking about throwing medicine balls around, a “mere” two or three pound weight isn’t going to seem like enough.

          EDIT: Seem like enough to him, to be clear.

          • Using a three pound sword to throw blows at a rate of better than one a second is pretty good exercise. And the shield being used to block blows aimed at you is likely to be several times that heavy.

    • Creutzer says:

      Could you have so much fun stuff that basically you force your brain to overcome the resistance and update? Well that may be why psychedelics look promising for depression treatment. It is like forcing pleasure down the brain’s throat until it stops resisting and accepts that yes this feels good.

      I don’t think that can be right. For one thing, psychedelics aren’t necessarily that pleasurable, and in fact you can have antidepressant aftereffects even after a bad trip.

      Maybe they reduce your confidence in your priors so that your prior on things not being fun isn’t so strong anymore, and then you can update on things being fun afterwards.

      • Aapje says:

        @Creutzer

        Perhaps some depressed people have been disappointed so much that they do almost no updates on success anymore, writing everything off as a fluke.

    • Aapje says:

      @nameless1

      The most pleasurable for most people seems to be a level of challenge where you mostly succeed, but only if you put the effort in.

      A way to tap into this might be to first measure ability and then to set the difficulty just below the maximum ability. Or to have a game that adapts to people’s performance (quickly at first and more slowly later).

      However, perhaps one cause of depression is that some people are extremely sensitive to failure, causing very many successes to be undone by a single failure. So then it makes sense to not try, unless the chance of failure is almost zero, but at that point there is probably no challenge and the successes become meaningless.

      Perhaps what works is a game where unpredictable success happens regularly but unpredictable failure never happens. So it can look like you are losing, but you then get lucky to win, but you never get unlucky to lose when you started off winning.

      • A way to tap into this might be to first measure ability and then to set the difficulty just below the maximum ability.

        A real world example would be my father’s solution to the problem of making it fun for me to play ping-pong with him–we had a table in the basement. The game is played to 21 points. I started with some number of points. Every time I lost a game, my starting number went up one. Every time I won, it went down one. The result was a handicap that almost exactly balanced his advantage in skill.

        Eventually I got the handicap down to zero.

    • Luke G says:

      We know exercise helps, but I think most people tend to think exercise equals boring jogging or weight lifting…

      is there something like tennis but more oriented towards muscle building

      “Boring” exercise I take to mean very predictable outcomes. Unpredictability could come from either an opponent (whose actions you can’t predict), or such a high skill requirement that success is uncertain.

      With opponents, full-contact sports can be good for building strength–though lifting is still necessary to reach your potential. The major downside is higher injury risk than most other options. Consider grappling martial arts: Brazilian jiu-jitsu, wrestling, etc. In addition to unpredictable opponents, they have a very high skill ceiling which keep them interesting for a long time. Also great for building self-confidence as you’ll know you can beat everybody else up.

      Without opponents, Olympic weightlifting has a very high skill requirement and is obviously good for muscle building. The technique is much more advanced than other lifts, increasing chance of misses and making it more mentally engaging. I might compare a lift to a tennis serve: quick, precise, explosive movement.

  32. JPNunez says:

    Postmodernism: there’s no relationship between the signifier and the signified

    Actual languages: yes, this is by design, to avoid being eaten by “the brown ones”

    • Machine Interface says:

      [missing the point of the joke] The absence of relationship between the signifier and the signified is a structuralist idea that predates postmodernist views by half a century.

      • JPNunez says:

        I remember hearing about it on a philosophy podcast, but couldnt’ find who was who actually said it, so I went with the good old postmodernism.

        • Statismagician says:

          Wittgenstein, yes?

          • Urstoff says:

            The idea that the relationship between the signifier and the signified is arbitrary (not that there is no relationship; the relationship is established in a linguistic community, otherwise words wouldn’t mean things) goes back at least to 19th century linguistics, but has probably been around much longer. A lot of people credit Saussure with bringing it to wider attention, but he didn’t come up with the notion.

      • Protagoras says:

        If you’re trying to determine priority, Plato’s Cratylus discusses this issue, and makes it sound like this was not a new discussion.

  33. b_jonas says:

    > Now in an equally weird twist of fate, the upper bound of the same theorem has been proven by sci-fi writer Greg Egan, author of Permutation City.

    Scott, if you achieve anything anything useful and quotable that isn’t directly connected to medicine, such as research about education or effective altruism, I think we’ll now have to right to spread the word that “in an equally wierd twist of face, … by clinical psychiatrist Scott Alexander”.

  34. arch1 says:

    All (perhaps especially b_jonas:-), if you even think you *might* understand the referenced Erica Klarreich article on shortest superpermutations, you should read it – it’s so well written there’s a good chance you *will* understand it (and as bonuses you’ll see a good example of how mathematicians can leverage work done on one problem to make progress on another, *and* you’ll more fully appreciate the “Permutation City” coincidence:-)
    PS. Scott, your descriptive paragraph might be clearer if you replace the first “theorem” with “superpermutation-related quantity”, and the second “theorem” with “quantity”.

  35. Sniffnoy says:

    On the subject of “Ayn Rand’s Worst Idea”, also worth remembering Eliezer’s old essay from the sequences: Tolerate Tolerance.

    Key excerpt:

    Cooperation is unstable, in both game theory and evolutionary biology, without some kind of punishment for defection. So it’s one thing to subtract points off someone’s reputation for mistakes they make themselves, directly. But if you also look askance at someone for refusing to castigate a person or idea, then that is punishment of non-punishers, a far more dangerous idiom that can lock an equilibrium in place even if it’s harmful to everyone involved.

    […]

    I am not generally a fan of “tolerance”. I certainly don’t believe in being “intolerant of intolerance”, as some inconsistently hold. But I shall go on trying to tolerate people who are more tolerant than I am, and judge them only for their own un-borrowed mistakes.

    • j r says:

      I honestly cannot really conceive of what it means to be tolerant v. intolerant in the contemporary context. What do these words mean given the superficial, sound-bite-y nature of most current public discourse? What does it mean to be tolerant v. intolerant within the echo chambers in which most of these conversations take place? After all, intolerance does way more to police inter-group disputes than it does to settle disputes between different groups.

      Maybe there is no such thing as ethical discourse under social media.

      • christopher hodge says:

        It’s the simplest thing in the world, j r: we love everyone, we accept everyone, we detest violence, we hate authoritarianism, and we love peace and the rule of law, and if you don’t agree with us then you’re a nazi and you should be hit in the head with a bike lock. I mean, what could you possibly fail to grasp?

      • The Nybbler says:

        In modern discourse, Yonatan Zunger’s essay is probably the best answer: Tolerance is how you handle wrongthinkers if you can’t actually crush them.

      • Sniffnoy says:

        Eliezer is pretty clear about what he means. What these words might mean in the “public discourse” doesn’t have that much relevance to this piece.

  36. arancaytar says:

    In 532, the Byzantines and Persians signed what they called The Perpetual Peace, so named because it was expected to last forever. It lasted eight years. After the ensuing war, the Byzantines and Persians, now less optimistic, named their new treaty The Fifty Year Peace. It lasted ten years.

    I wonder if there is a point where the peace lasts longer than expected, or a local maximum beyond which shortening the estimate also shortens the actual duration. Got to be one or the other.

    And Elsevier files a lawsuit forcing a Swedish ISP to ban Sci-Hub; the ISP complies but also bans Elsevier.

    If the ISP was ordered to block a specific set of domains that belong to SciHub, I wonder if it would violate the letter of the order to do so but also redirect Elsevier’s domain to SciHub’s server.

    • Protagoras says:

      Actual peace treaties often last longer than expected; resuming the war usually happens because there is some new excuse, which if it doesn’t happen before the peace expires (usually leading to somebody violating the peace) is unlikely to happen exactly at the point the peace officially ends.

  37. herculesorion says:

    RE: Coyne

    Maybe what’s going on here–in #metoo as well as this–is less about equality for the oppressed and abused, and more about weighing “he’s a bastard” more heavily than “he’s OUR bastard”. We’re deciding that we’d rather have fewer results than have assholes who get results.

  38. Pete Michaud says:

    re: the hiring thing, it’s worse than that. The assumption that you can just raise your bid in order to find good employees is based on the idea that people with the right mix of skills and knowledge and goal alignment, etc, exist at all, and further that the deciding factor those people are considering when choosing their employment is their salary. Sometimes that’s true, and it’s roughly true in the average case… but in sufficiently weird fields and/or positions, sometimes the problem really is just finding anyone who could possibly do the thing, and in my experience money isn’t generally the limiting factor for people who are “maybe”s.

    • Edward Scizorhands says:

      How about finding someone who is a close fit and training them?

      • Pete Michaud says:

        I mean, it’s not like we’re looking for FORTRAN developers. I mean like finding someone with deep expertise in ML, who also has a wide network in both the industry and academic circles because that’s the type of personality they have, and they care about X cause, and they are willing to work with us both from the compensation and branding perspective. That person may not exist.

        Of course we try lots of angles–training, persuasion, etc–but these are sometimes not tractable.

    • herculesorion says:

      Training is a form of salary. “Raising your bid” might also mean saying “if you seem like a good fit personally but don’t quite have the experience we need, we will pay for you to take courses in what we want, or allocate some non-job-functioning ‘play time’ each day for you to build the skills we want”.