It's hard not to write satire

Reading the headlines, I am reminded of a line from Juvenal.

Juvenal, Satires, No. 1, Line 30.

Difficile est saturam non scribere
It's hard not to write satire.

Think before you speak

Children's literature is rich in wisdom and the works of Lewis Carroll are especially strewn with nuggets.

Lewis Carroll subtley introduces children to the ideas of path dependency ("Path dependence explains how the set of decisions one faces for any given circumstance is limited by the decisions one has made in the past, even though past circumstances may no longer be relevant") and sunk cost fallacy and the phenomenon escalation of commitment all in a brief exchange with the Red Queen.

From Through the Looking Glass by Lewis Carroll

'Don't let us quarrel,' the White Queen said in an anxious tone. 'What is the cause of lightning?'

'The cause of lightning,' Alice said very decidedly, for she felt quite certain about this, 'is the thunder— no, no!' she hastily corrected herself. 'I meant the other way.'

'It's too late to correct it,' said the Red Queen: 'when you've once said a thing, that fixes it, and you must take the consequences.'

"When you've once said a thing, that fixes it, and you must take the consequences" - Seems like Lewis Carroll had such prescience that he cast the Red Queen as the internet mob.

The English culture of location names

Many years ago, living in England, I came across a charming and humorous book that even then was old, How To Be An Alien by George Mikes.

Mikes was Hungarian born, a journalist and author. He was posted to the UK by his newspaper in the run-up to the outbreak of World War II and never left, becoming a British citizen in 1946. How To Be An Alien was his second book and set the standard for his writing life (he wrote some 40 books). It is the book that made his name and remains in print in the UK. It is a brief, delightfully pointed rundown of the main stereotypes of the British and British customs. Perhaps the most famous chapter is the one sentence Chapter 7 - Sex. In toto it reads -

European men and women have sex lives; English men and women have hot-water bottles.

In that quintessential English way, no-one took umbrage, instead embracing Mikes and his work.

Even today, it remains both humorous and largely still on-mark, particularly outside of London. It is a brief read of less than 40 pages but a delightful 40 pages.

I especially liked Chapter 21 - How to Plan a Town which captures the British genius for whimsey, freedom, eccentricity, and quirkiness. This chapter is as accurate and relevant today as seventy years ago.

The English like to be uncomfortable. They think that this makes them strong. Only weak people from Europe live in comfortable pleasant towns.

People who build English towns want to make everything difficult. In Europe, doctors, lawyers and people who sell books have their houses and shops together in different parts of the town so you can always find a good (or a bad but expensive) doctor anywhere. In England, your address is important. In London, all the doctors live and work in Harley Street, all the lawyers are in Lincoln’s Inn Fields and all the book-sellers are in the Charing Cross Road. The newspaper offices are all in Fleet Street, the people who make men’s clothes are all in Saville Row and the car salesmen are in Great Portland Street. Theatres are near Piccadilly Circus and cinemas are in Leicester Square. Soon all the fruit and vegetable shops will move to Hornsey Lane, all the butchers to the Mile End Road and all the men’s toilets to Bloomsbury.

Now, I want to tell you about how to build an English town. You must understand that an English town is built to make life as difficult as possible for foreigners.

1 First of all, never build a straight street. The English do not like to be able to see two ends of a street. Make bends in the streets or make them S-shaped. The letters L,T,V,Y,W, and O also make good shapes for streets. It would please the Greeks if you built a few 0 or B-shaped streets. Maybe you could build streets like Russian or Chinese letters, too.

2 Never build all the houses in a street in a straight line. The British are free people so they are free to build their houses in circles.

3 Make sure that nobody can find the houses. European people put the numbers 1, 3, 5, 7 on one side of the street and 2, 4, 6, 8 on the other side of the street. The small numbers always start from the north or west. In England they start the numbers at one end of the street, then suddenly stop and continue the numbers on the opposite side going back the other way.

You can leave out some numbers and you can continue the numbers in a side street; you can also give the same number to two or three houses.

And you can do more! Many people do not have numbers on their houses; instead they give their houses names. It is very amusing to go to a street with three hundred and fifty houses and to look for a house called ‘The House’. Or you can visit a house called ‘Orange Tree House’ and find that there are three apple trees in the garden.

4 If the road bends, give a different name to the second part of it but, if it bends a lot so it is really two different streets, you can keep the same name. If the street is long and straight, give it many different names (High Holborn, New Oxford Street, Oxford Street, Bayswater Road, Notting Hill Gate, Holland Park, etc.*)

5 Some clever foreigners will find the street that they want, so make it harder for them. Call the street by another name. Don’t just call it a ‘street’, call it a ‘road’, ‘way’, ‘park’, ‘garden’, etc.

Now try this:

(a) Put all the streets with the same name in the same part of town: Belsize Park, Belsize Street, Belsize Gardens, Belsize Way, etc.

(b) Put a number of streets with the same name in different parts of the town. If you have twenty Princes Squares and twenty Warwick Roads, nobody will be able to find the right place.

6 Paint the street name in large letters on a piece of wood. Hide this piece of wood carefully. Put it very high on the wall or very low behind the flowers in someone’s garden, or in a shadow – anywhere where people cannot see it. Even better, take the street name to your bank and ask the bank to keep it for you. If you don’t, somebody will find out where they are.


* These are all parts of one very long, straight street in the centre of London.

This long preamble is brought to mind by a passage in John Keegan's Intelligence in War, an account of the role military intelligence plays in wars and illustrated by accounts of different battles and campaigns. Chapter Three is Local Knowledge: Stonewall Jackson in the Shenandoah Valley. One of Keegan's observations is that, oddly, at the outbreak of the Civil War, no uniform, triangulated national map of the United States existed. There were local maps, some of great precision, but there was no overall integrated map. As a product of historical circumstance, the coasts were well mapped as were the frontier states. The founding colonies made do with local maps. This quirk of history had a significant influence on the circumstances of the Shenandoah campaign led by Stonewall Jackson, who was almost always at a supply and manpower disadvantage compared to the Union generals.

Local knowledge often counted far more than the plates in a shoddy bookshop atlas. It was much more readily available, inside the South, to Confederate defenders than Union invaders. Without it, confusions accumulated. Even quite good maps could be out of date, while there was no guarantee that the mapmaker's choice of place-name was that used by locals. "Cold Harbor, Virginia" (the site of one of General Ulysses Grant's battles in 1864) "was sometimes called Coal Harbor, and there was also a New Cold Harbor and a 'burned' Cold Harbor. Burned Cold Harbor was known by the locals as Old Cold Harbor. Many of the roads were known by one of two names: the Market or River Road; the Williamsburg or Seven Mile Road; the Quaker or Willis Church Road. To add to the confusion, there were sometimes other nearby roads with the same or similar names that ran in completely different directions."

Tell me Keegan isn't channelling Mikes there.

But the Virginians came by their geographical naming proclivities honestly, having been settled by the English.

It's Not Even in The Top 100 Worst Psych Theories of the 1960s - True That

Cognitive pollution - ideas which are so attractive that we want them to be true when they are not. Also known as Normative Sociology.

Early in my career I read up on the Kübler-Ross model of grief as an issue related to working with high performance teams. I soon discovered that, much like the vaunted Myers-Brigg tests, it was conjured from thin air with no empirical basis. Yet, decades later, we still hear people describe the five stages of grief.

At Language Log, I came across this pleasantly meta mocking of Kübler-Ross.

This makes it hard to find a new novel, sometimes

I make the effort to read fiction but it remains a trivial portion of my reading diet. Too much contemporary literary fiction appears to me to be self-absorbed comfortable middle class authors writing about the tribulations of self-absorbed comfortable middle class people. The inclination to insert emotional dysfunction in order to generate even a spark of tension and energy seems too great to resist. These just aren't the people you would want to share time with. Or at least, I wouldn't. And don't.

I am glad to find that I am in good company. I spent hours reading Le Guin books to my children when they were young. From an interview with Ursula K. Le Guin in The New York Times.

What genres do you especially enjoy reading? And which do you avoid?

I read mostly novels, any kind of novels, and poetry, and all kinds of nonfiction, especially some kinds of science, biographies, some history, and books about and by Native Americans, and Tierra del Fuego, and Darwinian adaptation — oh, give me a book and if it’s interesting, I’ll read it. Avoidance? At the moment, I tend to avoid fiction about dysfunctional urban middle-class people written in the present tense. This makes it hard to find a new novel, sometimes.

The sea has swallowed men who have spoken a thousand different languages

From A Sailor's Pay by Jack Cady in the anthology, Sea-Cursed, edited by T. Liam McDonald.

I like this combination of the speaking in sounds and languages.

The sea speaks with the sounds of thunder, or it is susurrous, or it hisses, or it murmurs. It is nearly as ancient as the earth. The sea has swallowed men who have spoken a thousand different languages: it has taken into its restless maw Persians, Phoenicians, Romans, Spaniards and Englishmen.

Root causes of demographic inversions

Hmmm. Empirics of Strategic Interdependence: The Case of the Racial Tipping Point by William Easterly

The abstract is:

The Schelling model of a “tipping point” in racial segregation, in which whites flee a neighborhood once a threshold of nonwhites is reached, is a canonical model of strategic interdependence. The idea of “tipping” explaining segregation is widely accepted in the academic literature and popular media. I use census tract data for metropolitan areas of the US from 1970 to 2000 to test the predictions of the Schelling model and find that this particular model of strategic interaction largely fails the tests. There is more “white flight” out of neighborhoods with a high initial share of whites than out of more racially mixed neighborhoods

The fact that neighborhoods can dramatically change their ethnic makeup in relatively short spans of time is long established. In this paper, which looks at 1970-2000, nearly 10% of urban neighborhoods switched from majority white to majority non-white. As an aside, I have never seen a comparable study that looks at the reverse phenomenon which we usually refer to as gentrification, i.e. where urban neighborhoods switch from majority non-white to majority white. The absence of academic papers on that aspect of gentrification is interesting because it implies an odd asymmetry; it is bad when whites leave and it is bad when they return. But that is more an ideological issue than an epistemological one.

I found this paper interesting in that it tests a longstanding assumption that has never been rigorously tested before and finds the data inconsistent with the model forecasts. As always, why did we accept the model as true for so long without actually testing it? This is not to say that Easterly's research is the final word. Rather, why did it take fifty years to even question whether the model was true?

Easterly covers all the technical details. We know already that it takes only small levels of positive affiliation (by race or class or religion or other attribute) to tip neighborhood majorities strongly one way or the other and that negative affiliation doesn't even need to be a factor for high degrees of segregation to occur.

What Easterly's work begins to get at is something I have often wondered about in earlier such studies. In general, the hypothesis has been something along the lines of "what percentage of neighbors need to be non-white before the neighborhood experiences white flight and switch to majority non-white?" My question is about whether we are framing this correctly in the first place. What are the reasons people elect to leave a neighborhood and does racial constitution play an independent role among the reasons for leaving?

We look at neighborhood racial inversion from a very constricted perspective, framed as why whites dislike blacks. That frame has a very large range of assumptions that should be tested. We are assuming the answer into place. The question, I believe, isn't "Why do whites flee from blacks" as most commonly assumed. I think the question is instead, "why do neighborhood demographic inversions occur?" With gentrification, the inversion of a neighborhood from majority black to majority white, we don't make the assumption that blacks are fleeing whites and we don't try and figure out what the tipping point is in terms of how many whites there can be before blacks start fleeing. Posing it in that way exposes the biased assumptions in our current thinking. We know that gentrification is an economic phenomenon, not a racial one. So why might not other inversions have different root causes than racial aversion as well?

What we really want to know are the root causes of demographic inversions, not assume the answer into place. Even Easterly's paper is caught in the current paradigm of looking only at white to black inversions. I would argue that we ought to be looking at all ethnic/racial inversions - white to black, black to white, black to Hispanic, white to Asian, etc. And we shouldn't be looking only at race but at all observable affiliative patterns; why do some neighborhoods switch from rich to poor, from native born to immigrant, from immigrant to native born, from older to younger. There are some pretty obvious answers to some of these as well as some less obvious ones as well. By assuming whites are racist and will always leave a neighborhood when there are more than X% black, we miss what is really going on, coarsen our discourse and handicap our capacity to craft beneficial policies.

The reason I have wondered this is based on the specific examples of particular neighborhood inversions. Take my city for a contemporary example. Sure, there were numerous neighborhoods that switched from white to black back in the sixties and seventies. But a good number of those are now switching back to white in the 2010s. In the meantime, in the 1990s and 2000s, numerous black neighborhoods became Hispanic neighborhoods. In the 1980s, one historically Jewish neighborhood, in the space of ten years, went from majority Jewish to overwhelmingly gentile. There is a lot of demographic flux going on all the time and with relative rapidity. But why is it happening and does racial affiliation/animus have much to do with it? How many of these changes might be driven by contextual economic and other factors?

For example, the neighborhood changes in the 1950s and 1960s occurred at a time of rural migration to the cities (black and white but probably disproportionately black). At the same time, with post-war prosperity, many urban whites now could afford cars in a way that they had not been able to before and therefore suburban living became a real option. Were urban whites moving to the suburbs because that is where they could now afford to live, with cheaper cars and improved transportation infrastructure, and that simply coincided with rural black migration into cities? Closely related to that, was neighborhood racial inversion more a function of economic opportunities (whites moving to the faster growing suburbs) than of race? Did rapid inversions occur more because of signalling than because of race per se?

In other words, there are a host of alternative reasons for neighborhood demographic inversions and it is unclear to me to what degree and how often race played a factor independent of and greater than alternative explanations. I am trying to get at an explanation that encompasses what we have seen in the past twenty years in my city. It does not appear to me that blacks are fleeing neighborhoods where Hispanics are becoming more common, nor does it appear that Hispanics are seeking to live where there is a majority of African-Americans. I am skeptical that gentiles were simply eager to live among Jews and Jews wanted to get away from that incursion. It appears to me that there are many more demographic inversions occurring than we acknowledge and that those inversions occur for a wide range of reasons, mostly economic. Racial animus might be, and in some instances likely is, in the mix, but I think the root causes are different than are usually acknowledged.

Easterly's work doesn't answer those questions but does provide evidence to support that those might be the more pertinent questions.

The cognitive biases you carry are independent of your intelligence and of one another.

From Scott Alexander.

Individual Differences In Cognitive Biases – Evidence Against The One-Factor Theory Of Rationality. Unlike IQ where lots of different kinds of intelligence tests correlate with each other pretty well, rationality does not appear to have a general factor and people who do well in avoiding one kind of cognitive bias aren’t much more likely to do well at avoiding another. Doesn’t look like a super well-cited paper, which makes me wonder whether Stanovich et al have a response to this.

From the paper's abstract:

In this paper we seek to gain an improved understanding of the structure of cognitive biases and their relationship with measures of intelligence and relevant non-cognitive constructs. We report on the outcomes of a study based on a heterogeneous set of seven cognitive biases — anchoring effect, belief bias, overconfidence bias, hindsight bias, base rate neglect, outcome bias and sunk cost effect.

[snip]

This pattern of results suggests that a major part of the reliable variance of cognitive bias tasks is unique, and implies that a one-factor model of rational behavior is not plausible.

I have often wondered whether anyone has prepared a mapping of the type and prevalence of biases to which populations might be prone.

60:15:15:10

From Genes influence academic ability across all subjects, latest study shows by Hannah Devlin. Devlin is reporting on two separate issues. Notable is that this is in the UK's Guardian which tends to be inimicable to such findings.

The findings add to growing evidence that school performance has a large heritable component, with around 60% of the differences in pupil’s GCSE results being explained by genetic factors.

and

You may feel you are just not a maths person, or that you have a special gift for languages, but scientists have shown that the genes influencing numerical skills are the same ones that determine abilities in reading, arts and humanities.

The study suggests that if you have an academic Achilles heel, environmental factors such as a teaching are more likely to be to blame.

The evidence keeps piling up and there are all sorts of outliers and counterfactuals that have to be vetted on an on-going basis, but it appears to me that individual academic outcomes are 60% inherent capability from heritable IQ and 40% of the variance is due to external circumstances. From other research, my best estimate is that overall life outcomes are 60% IQ, 15% general culture, 15% familial culture and 10% experiential circumstances.

That breakdown is more a stake in the ground than a hard claim. The research is so voluminous, often initially contradictory, so full of cognitive pollution and advocacy, etc. that it can be difficult to discern what the current state of play might be. 60:15:15:10 is my best guess.

You can see why it is so contentious. If you accept 60:15:15:10, then it has material implications for governmental, familial, labor, education and other policies.

98% of Japan’s 81,000 adoptees are males between the ages of 25 and 30

It started with this tweet.

Japan houses more than half of the world's 1,000 oldest businesses: http://t.co/GK18VfGkA4 pic.twitter.com/LVzPPKOeJw

— Zachary Crockett (@zzcrockett) July 15, 2015

Going to the original article, which is quite interesting, there is a fair amount of discussion about why Japan has so many old businesses. I am always fascinated by the history and culture of Japan. I was stunned by this fact, of which I had been completely unaware. From Why Are so Many of the World's Oldest Businesses in Japan? by Dan Kopf.

Most of Japan’s oldest companies also boast of being “family run” for dozens of generations. But for some of these companies, the reality of this claim is far from the truth: in cases where there is no son to inherit a business, or where a Japanese CEO “desires a better quality son than nature provided,” an heir is legally adopted. For centuries, Japanese business owners have engaged in this practice.

Japanese adult adoption dates back to the Tokugawa era (1603-1867), when merchants sought free labor for field work. Initially considered to be demeaning and emasculating, being adopted as an adult became “more intertwined with family firms and capitalism” in the 20th century, and took on a connotation of prestige and opportunism.

Today, 98% of Japan’s 81,000 adoptees are males between the ages of 25 and 30 — many of whom are businessmen who are legally adopted by the owners of corporations and put in management positions. This way, the “family owned and operated” claim technically remains intact, without the risk of passing on a legacy to a less-than-capable son or nephew. Through a technique called “Mukoyōshi,” a business owner with no male heirs to his company can also legally adopt his daughter’s husband as a “second-birth son.” Even in the rare case that a thoroughly vetted male adoptee fails in his leadership role, he can be disinherited, and another heir can be adopted to take his place.