Tag: Edge.org

Scientific Concepts We All Ought To Know

John Brockman's online scientific roundtable Edge.org does something fantastic every year: It asks all of its contributors (hundreds of them) to answer one meaningful question. Questions like What Have You Changed Your Mind About? and What is Your Dangerous Idea?

This year's was particularly awesome for our purposesWhat Scientific Term or Concept Ought To Be More Known?

The answers give us a window into over 200 brilliant minds, with the simple filtering mechanism that there's something they know that we should probably know, too. We wanted to highlight a few of our favorites for you.

***

From Steven Pinker, a very interesting thought on The Second Law of Thermodynamics (Entropy). This reminded me of the central thesis of The Origin of Wealth by Eric Beinhocker. (Which we'll cover in more depth in the future: We referenced his work in the past.)


The Second Law of Thermodynamics states that in an isolated system (one that is not taking in energy), entropy never decreases. (The First Law is that energy is conserved; the Third, that a temperature of absolute zero is unreachable.) Closed systems inexorably become less structured, less organized, less able to accomplish interesting and useful outcomes, until they slide into an equilibrium of gray, tepid, homogeneous monotony and stay there.

In its original formulation the Second Law referred to the process in which usable energy in the form of a difference in temperature between two bodies is dissipated as heat flows from the warmer to the cooler body. Once it was appreciated that heat is not an invisible fluid but the motion of molecules, a more general, statistical version of the Second Law took shape. Now order could be characterized in terms of the set of all microscopically distinct states of a system: Of all these states, the ones that we find useful make up a tiny sliver of the possibilities, while the disorderly or useless states make up the vast majority. It follows that any perturbation of the system, whether it is a random jiggling of its parts or a whack from the outside, will, by the laws of probability, nudge the system toward disorder or uselessness. If you walk away from a sand castle, it won’t be there tomorrow, because as the wind, waves, seagulls, and small children push the grains of sand around, they’re more likely to arrange them into one of the vast number of configurations that don’t look like a castle than into the tiny few that do.

The Second Law of Thermodynamics is acknowledged in everyday life, in sayings such as “Ashes to ashes,” “Things fall apart,” “Rust never sleeps,” “Shit happens,” You can’t unscramble an egg,” “What can go wrong will go wrong,” and (from the Texas lawmaker Sam Rayburn), “Any jackass can kick down a barn, but it takes a carpenter to build one.”

Scientists appreciate that the Second Law is far more than an explanation for everyday nuisances; it is a foundation of our understanding of the universe and our place in it. In 1915 the physicist Arthur Eddington wrote:

[…]

Why the awe for the Second Law? The Second Law defines the ultimate purpose of life, mind, and human striving: to deploy energy and information to fight back the tide of entropy and carve out refuges of beneficial order. An underappreciation of the inherent tendency toward disorder, and a failure to appreciate the precious niches of order we carve out, are a major source of human folly.

To start with, the Second Law implies that misfortune may be no one’s fault. The biggest breakthrough of the scientific revolution was to nullify the intuition that the universe is saturated with purpose: that everything happens for a reason. In this primitive understanding, when bad things happen—accidents, disease, famine—someone or something must have wanted them to happen. This in turn impels people to find a defendant, demon, scapegoat, or witch to punish. Galileo and Newton replaced this cosmic morality play with a clockwork universe in which events are caused by conditions in the present, not goals for the future. The Second Law deepens that discovery: Not only does the universe not care about our desires, but in the natural course of events it will appear to thwart them, because there are so many more ways for things to go wrong than to go right. Houses burn down, ships sink, battles are lost for the want of a horseshoe nail.

Poverty, too, needs no explanation. In a world governed by entropy and evolution, it is the default state of humankind. Matter does not just arrange itself into shelter or clothing, and living things do everything they can not to become our food. What needs to be explained is wealth. Yet most discussions of poverty consist of arguments about whom to blame for it.

More generally, an underappreciation of the Second Law lures people into seeing every unsolved social problem as a sign that their country is being driven off a cliff. It’s in the very nature of the universe that life has problems. But it’s better to figure out how to solve them—to apply information and energy to expand our refuge of beneficial order—than to start a conflagration and hope for the best.

Richard Nisbett (a social psychologist) has a great one — a concept we've hit on before but is totally underappreciated by most people: The Fundamental Attribution Error.

Modern scientific psychology insists that explanation of the behavior of humans always requires reference to the situation the person is in. The failure to do so sufficiently is known as the Fundamental Attribution Error. In Milgram’s famous obedience experiment, two-thirds of his subjects proved willing to deliver a great deal of electric shock to a pleasant-faced middle-aged man, well beyond the point where he became silent after begging them to stop on account of his heart condition. When I teach about this experiment to undergraduates, I’m quite sure I‘ve never convinced a single one that their best friend might have delivered that amount of shock to the kindly gentleman, let alone that they themselves might have done so. They are protected by their armor of virtue from such wicked behavior. No amount of explanation about the power of the unique situation into which Milgram’s subject was placed is sufficient to convince them that their armor could have been breached.

My students, and everyone else in Western society, are confident that people behave honestly because they have the virtue of honesty, conscientiously because they have the virtue of conscientiousness. (In general, non-Westerners are less susceptible to the fundamental attribution error, lacking as they do sufficient knowledge of Aristotle!) People are believed to behave in an open and friendly way because they have the trait of extroversion, in an aggressive way because they have the trait of hostility. When they observe a single instance of honest or extroverted behavior they are confident that, in a different situation, the person would behave in a similarly honest or extroverted way.

In actual fact, when large numbers of people are observed in a wide range of situations, the correlation for trait-related behavior runs about .20 or less. People think the correlation is around .80. In reality, seeing Carlos behave more honestly than Bill in a given situation increases the likelihood that he will behave more honestly in another situation from the chance level of 50 percent to the vicinity of 55-57. People think that if Carlos behaves more honestly than Bill in one situation the likelihood that he will behave more honestly than Bill in another situation is 80 percent!

How could we be so hopelessly miscalibrated? There are many reasons, but one of the most important is that we don’t normally get trait-related information in a form that facilitates comparison and calculation. I observe Carlos in one situation when he might display honesty or the lack of it, and then not in another for perhaps a few weeks or months. I observe Bill in a different situation tapping honesty and then not another for many months.

This implies that if people received behavioral data in such a form that many people are observed over the same time course in a given fixed situation, our calibration might be better. And indeed it is. People are quite well calibrated for abilities of various kinds, especially sports. The likelihood that Bill will score more points than Carlos in one basketball game given that he did in another is about 67 percent—and people think it’s about 67 percent.

Our susceptibility to the fundamental attribution error—overestimating the role of traits and underestimating the importance of situations—has implications for everything from how to select employees to how to teach moral behavior.

Cesar Hidalgo, author of what looks like an awesome book, Why Information Grows, wrote about Criticality, which is a very important and central concept to understanding complex systems:

In physics we say a system is in a critical state when it is ripe for a phase transition. Consider water turning into ice, or a cloud that is pregnant with rain. Both of these are examples of physical systems in a critical state.

The dynamics of criticality, however, are not very intuitive. Consider the abruptness of freezing water. For an outside observer, there is no difference between cold water and water that is just about to freeze. This is because water that is just about to freeze is still liquid. Yet, microscopically, cold water and water that is about to freeze are not the same.

When close to freezing, water is populated by gazillions of tiny ice crystals, crystals that are so small that water remains liquid. But this is water in a critical state, a state in which any additional freezing will result in these crystals touching each other, generating the solid mesh we know as ice. Yet, the ice crystals that formed during the transition are infinitesimal. They are just the last straw. So, freezing cannot be considered the result of these last crystals. They only represent the instability needed to trigger the transition; the real cause of the transition is the criticality of the state.

But why should anyone outside statistical physics care about criticality?

The reason is that history is full of individual narratives that maybe should be interpreted in terms of critical phenomena.

Did Rosa Parks start the civil rights movement? Or was the movement already running in the minds of those who had been promised equality and were instead handed discrimination? Was the collapse of Lehman Brothers an essential trigger for the Great Recession? Or was the financial system so critical that any disturbance could have made the trick?

As humans, we love individual narratives. We evolved to learn from stories and communicate almost exclusively in terms of them. But as Richard Feynman said repeatedly: The imagination of nature is often larger than that of man. So, maybe our obsession with individual narratives is nothing but a reflection of our limited imagination. Going forward we need to remember that systems often make individuals irrelevant. Just like none of your cells can claim to control your body, society also works in systemic ways.

So, the next time the house of cards collapses, remember to focus on why we were building a house of cards in the first place, instead of focusing on whether the last card was the queen of diamonds or a two of clubs.

The psychologist Adam Alter has another good one on a concept we all naturally miss from time to time, due to the structure of our mind. The Law of Small Numbers.

In 1832, a Prussian military analyst named Carl von Clausewitz explained that “three quarters of the factors on which action in war is based are wrapped in a fog of . . . uncertainty.” The best military commanders seemed to see through this “fog of war,” predicting how their opponents would behave on the basis of limited information. Sometimes, though, even the wisest generals made mistakes, divining a signal through the fog when no such signal existed. Often, their mistake was endorsing the law of small numbers—too readily concluding that the patterns they saw in a small sample of information would also hold for a much larger sample.

Both the Allies and Axis powers fell prey to the law of small numbers during World War II. In June 1944, Germany flew several raids on London. War experts plotted the position of each bomb as it fell, and noticed one cluster near Regent’s Park, and another along the banks of the Thames. This clustering concerned them, because it implied that the German military had designed a new bomb that was more accurate than any existing bomb. In fact, the Luftwaffe was dropping bombs randomly, aiming generally at the heart of London but not at any particular location over others. What the experts had seen were clusters that occur naturally through random processes—misleading noise masquerading as a useful signal.

That same month, German commanders made a similar mistake. Anticipating the raid later known as D-Day, they assumed the Allies would attack—but they weren’t sure precisely when. Combing old military records, a weather expert named Karl Sonntag noticed that the Allies had never launched a major attack when there was even a small chance of bad weather. Late May and much of June were forecast to be cloudy and rainy, which “acted like a tranquilizer all along the chain of German command,” according to Irish journalist Cornelius Ryan. “The various headquarters were quite confident that there would be no attack in the immediate future. . . . In each case conditions had varied, but meteorologists had noted that the Allies had never attempted a landing unless the prospects of favorable weather were almost certain.” The German command was mistaken, and on Tuesday, June 6, the Allied forces launched a devastating attack amidst strong winds and rain.

The British and German forces erred because they had taken a small sample of data too seriously: The British forces had mistaken the natural clustering that comes from relatively small samples of random data for a useful signal, while the German forces had mistaken an illusory pattern from a limited set of data for evidence of an ongoing, stable military policy. To illustrate their error, imagine a fair coin tossed three times. You’ll have a one-in-four chance of turning up a string of three heads or tails, which, if you make too much of that small sample, might lead you to conclude that the coin is biased to reveal one particular outcome all or almost all of the time. If you continue to toss the fair coin, say, a thousand times, you’re far more likely to turn up a distribution that approaches five hundred heads and five hundred tails. As the sample grows, your chance of turning up an unbroken string shrinks rapidly (to roughly one-in-sixteen after five tosses; one-in-five-hundred after ten tosses; and one-in-five-hundred-thousand after twenty tosses). A string is far better evidence of bias after twenty tosses than it is after three tosses—but if you succumb to the law of small numbers, you might draw sweeping conclusions from even tiny samples of data, just as the British and Germans did about their opponents’ tactics in World War II.

Of course, the law of small numbers applies to more than military tactics. It explains the rise of stereotypes (concluding that all people with a particular trait behave the same way); the dangers of relying on a single interview when deciding among job or college applicants (concluding that interview performance is a reliable guide to job or college performance at large); and the tendency to see short-term patterns in financial stock charts when in fact short-term stock movements almost never follow predictable patterns. The solution is to pay attention not just to the pattern of data, but also to how much data you have. Small samples aren’t just limited in value; they can be counterproductive because the stories they tell are often misleading.

There are many, many more worth reading. Here's a great chance to build your multidisciplinary skill-set.

Our Genes and Our Behavior

“But now we are starting to show genetic influence on individual differences using DNA. DNA is a game changer; it's a lot harder to argue with DNA than it is with a twin study or an adoption study.”
— Robert Plomin

***

It's not controversial to say that our genetics help explain our physical traits. Tall parents will, on average, have tall children. Overweight parents will, on average, have overweight children. Irish parents have Irish looking kids. This is true to the point of banality and only a committed ignorant would dispute it.

It's slightly more controversial to talk about genes influencing behavior. For a long time, it was denied entirely. For most of the 20th century, the “experts” in human behavior had decided that “nurture” beat “nature” with a score of 100-0. Particularly influential was the child's early life — the way their parents treated them in the womb and throughout early childhood. (Thanks Freud!)

So, where are we at now?

Genes and Behavior

Developmental scientists and behavioral scientists eventually got to work with twin studies and adoption studies, which tended to show that certain traits were almost certainly heritable and not reliant on environment, thanks to the natural controlled experiments of twins separated at birth. (This eventually provided fodder for Judith Rich Harris's wonderful work on development and personality.)

All throughout, the geneticists, starting with Gregor Mendel and his peas, kept on working. As behavioral geneticist Robert Plomin explains, the genetic camp split early on. Some people wanted to understand the gene itself in detail, using very simple traits to figure it out (eye color, long or short wings, etc.) and others wanted to study the effect of genes on complex behavior, generally:

People realized these two views of genetics could come together. Nonetheless, the two worlds split apart because Mendelians became geneticists who were interested in understanding genes. They would take a convenient phenotype, a dependent measure, like eye color in flies, just something that was easy to measure. They weren't interested in the measure, they were interested in how genes work. They wanted a simple way of seeing how genes work.

By contrast, the geneticists studying complex traits—the Galtonians—became quantitative geneticists. They were interested in agricultural traits or human traits, like cardiovascular disease or reading ability, and would use genetics only insofar as it helped them understand that trait. They were behavior centered, while the molecular geneticists were gene centered. The molecular geneticists wanted to know everything about how a gene worked. For almost a century these two worlds of genetics diverged.

Eventually, the two began to converge. One camp (the gene people) figured out that once we could sequence the genome, they might be able to understand more complicated behavior by looking directly at genes in specific people with unique DNA, and contrasting them against one another.

The reason why this whole gene-behavior game is hard is because, as Plomin makes clear, complex traits like intelligence are not like eye color. There's no “smart gene” — it comes from the interaction of thousands of different genes and can occur in a variety of combinations. Basic Mendel-style counting (the sort of dominant/recessive eye color gene thing you learned in high school biology) doesn't work in analyzing the influence of genes on complex traits:

The word gene wasn't invented until 1903. Mendel did his work in the mid-19th century. In the early 1900s, when Mendel was rediscovered, people finally realized the impact of what he did, which was to show the laws of inheritance of a single gene. At that time, these Mendelians went around looking for Mendelian 3:1 segregation ratios, which was the essence of what Mendel showed, that inheritance was discreet. Most of the socially, behaviorally, or agriculturally important traits aren't either/or traits, like a single-gene disorder. Huntington's disease, for example, is a single-gene dominant disorder, which means that if you have that mutant form of the Huntington's gene, you will have Huntington's disease. It's necessary and sufficient. But that's not the way complex traits work.

The importance of genetics is hard to understate, but until the right technology came along, we could only observe it indirectly. A study might have shown that 50% of the variance in cognitive ability was due to genetics, but we had no idea which specific genes, in which combinations, actually produced smarter people.

But the Moore's law style improvement in genetic testing means that we can cheaply and effectively map out entire genomes for a very low cost. And with that, the geneticists have a lot of data to work with, a lot of correlations to begin sussing out. The good thing about finding strong correlations between genes and human traits is that we know which one is causative: The gene! Obviously, your reading ability doesn't cause you to have certain DNA; it must be the other way around. So “Big Data” style screening is extremely useful, once we get a little better at it.

***

The problem is that, so far, the successes have been a bit minimal. There are millions of “ATCG” base pairs to check on.  As Plomin points out, we can only pinpoint about 20% of the specific genetic influence for something simple like height, which we know is about 90% heritable. Complex traits like schizophrenia are going to take a lot of work:

We've got to be able to figure out where the so-called missing heritability is, that is, the gap between the DNA variants that we are able to identify and the estimates we have from twin and adoption studies. For example, height is about 90 percent heritable, meaning, of the differences between people in height, about 90 percent of those differences can be explained by genetic differences. With genome-wide association studies, we can account for 20 percent of the variance of height, or a quarter of the heritability of height. That's still a lot of missing heritability, but 20 percent of the variance is impressive.

With schizophrenia, for example, people say they can explain 15 percent of the genetic liability. The jury is still out on how that translates into the real world. What you want to be able to do is get this polygenic score for schizophrenia that would allow you to look at the entire population and predict who's going to become schizophrenic. That's tricky because the studies are case-control studies based on extreme, well-diagnosed schizophrenics, versus clean controls who have no known psychopathology. We'll know soon how this polygenic score translates to predicting who will become schizophrenic or not.

It brings up an interesting question that gets us back to the beginning of the piece: If we know that genetics have an influence on some complex behavioral traits (and we do), and we can with the continuing progress of science and technology, sequence a baby's genome and predict to a certain extent their reading level, facility with math, facility with social interaction, etc., do we do it?

Well, we can't until we get a general recognition that genes do indeed influence behavior and do have predictive power as far as how children perform. So far, the track record on getting educators to see that it's all quite real is pretty bad. Like the Freudians before, there's a resistance to the “nature” aspect of the debate, probably influenced by some strong ideologies:

If you look at the books and the training that teachers get, genetics doesn't get a look-in. Yet if you ask teachers, as I've done, about why they think children are so different in their ability to learn to read, and they know that genetics is important. When it comes to governments and educational policymakers, the knee-jerk reaction is that if kids aren't doing well, you blame the teachers and the schools; if that doesn't work, you blame the parents; if that doesn't work, you blame the kids because they're just not trying hard enough. An important message for genetics is that you've got to recognize that children are different in their ability to learn. We need to respect those differences because they're genetic. Not that we can’t do anything about it.

It's like obesity. The NHS is thinking about charging people to be fat because, like smoking, they say it's your fault. Weight is not as heritable as height, but it's highly heritable. Maybe 60 percent of the differences in weight are heritable. That doesn't mean you can't do anything about it. If you stop eating, you won't gain weight, but given the normal life in a fast-food culture, with our Stone Age brains that want to eat fat and sugar, it's much harder for some people.

We need to respect the fact that genetic differences are important, not just for body mass index and weight, but also for things like reading disability. I know personally how difficult it is for some children to learn to read. Genetics suggests that we need to have more recognition that children differ genetically, and to respect those differences. My grandson, for example, had a great deal of difficulty learning to read. His parents put a lot of energy into helping him learn to read. We also have a granddaughter who taught herself to read. Both of them now are not just learning to read but reading to learn.

Genetic influence is just influence; it's not deterministic like a single gene. At government levels—I've consulted with the Department for Education—I don't think they're as hostile to genetics as I had feared, they're just ignorant of it. Education just doesn't consider genetics, whereas teachers on the ground can't ignore it. I never get static from them because they know that these children are different when they start. Some just go off on very steep trajectories, while others struggle all the way along the line. When the government sees that, they tend to blame the teachers, the schools, or the parents, or the kids. The teachers know. They're not ignoring this one child. If anything, they're putting more energy into that child.

It's frustrating for Plomin because he knows that eventually DNA mapping will get good enough that real, and helpful, predictions will be possible. We'll be able to target kids early enough to make real differences — earlier than problems actually manifest — and hopefully change the course of their lives for the better. But so far, no dice.

Education is the last backwater of anti-genetic thinking. It's not even anti-genetic. It's as if genetics doesn't even exist. I want to get people in education talking about genetics because the evidence for genetic influence is overwhelming. The things that interest them—learning abilities, cognitive abilities, behavior problems in childhood—are the most heritable things in the behavioral domain. Yet it's like Alice in Wonderland. You go to educational conferences and it's as if genetics does not exist.

I'm wondering about where the DNA revolution will take us. If we are explaining 10 percent of the variance of GCSE scores with a DNA chip, it becomes real. People will begin to use it. It's important that we begin to have this conversation. I'm frustrated at having so little success in convincing people in education of the possibility of genetic influence. It is ignorance as much as it is antagonism.

Here's one call for more reality recognition.

***

Still Interested? Check out a book by John Brookman of Edge.org with a curated collection of articles published on genetics.