Epigenetics Can Help Fix Nutrition Science

Aeon recently posted a great essay critiquing the popular 'selfish gene' analogy, exploring new layers of complexity biologists have since uncovered within the process life uses to propagate itself. One of the hottest new concepts has been 'gene expression' or epigenetics. It's basically the insight that not all genes affect their organism in the same way; some heritable changes can occur that aren't caused by changes in DNA sequence. This promising new field muddles the nature/nurture question in a most intellectually-rewarding way.

One of the greatest potential applications of epigenetic thinking is in the field of nutrition science. Explaining exactly why someone's body fat percentage is what it is remains a murky and contentious area, ironic given the global obesity epidemic. Much of the confusion stems from the difficulty in parsing out heritable effects from lifestyle and cultural factors. As an example of the theoretical clarity epigenetics can provide, consider this question that I've long wondered about (and never found a satisfactory answer to).

Most people agree body fat percentage is in part driven by heritability: kids tend to share the body type of their parents, and we all know people who can eat whatever they want and remain skinny. This heritable component likely contributes to the difficulty people experience when trying to change their basic body type in a sustained way.

So here's my question: say you have a heterosexual couple, both with a propensity to fatten, and they've lived much of their lives overweight. If they had a kid, odds are the kid would share this propensity. Now consider what might happen if the couple decided to buckle down, lose weight, and get in shape. What would happen to the kid? The parents' genes certainly haven't changed, so one might expect the kid to still be cursed with a propensity towards accumulating body fat. But it seems strange to imagine, say, parents who've been thin and fit for 10 years having a kid with a propensity to fatten.*

Epigenetic effects can help explain why this scenario seems unlikely: after 10 years of thinness, the parents have probably undergone major biological changes that are somehow 'locked in' and get passed on to their children.

Hopefully researchers will someday be able to understand the exact dynamics of interactions such as this. If we knew definitively, for example, how long parents must sustain weight loss in order to gift a thinness propensity to their children, it would be a tremendous benefit to society. One can easily imagine compassionate parents being strongly incentivized to lose weight and get healthy themselves in order to enable their children to live better lives with greater opportunity.

*admittedly there's some confounding effects here: parents who are able to make big lifestyle shifts like sustained weight loss probably are insane fitness nuts who are likely to influence the behavior of their children

Calories Are Not Just Calories

"Another way to view the connection to the Black Swan idea is as follows. Classical thermodynamics produces Gaussian variations, while informational variations are from Extremistan. Let me explain. If you consider your diet and exercise as simple energy deficits and excesses, with a straight calorie-in, calorie-burned equation, you will fall into the trap of misspecifying the system into simple causal and mechanical links. Your food intake becomes the equivalent of filling up the tank of your new BMW. If, on the other hand, you look at food and exercise as activating metabolic signals, with potential metabolic cascades and nonlinearities from network effects, with recursive links, then welcome to complexity, hence Extremistan. Both food and workouts provide your body with information about stressors in the environment. As I have been saying throughout, informational randomness is from Extremistan. Medicine fell into the trap of using simple thermodynamics, with the same physics envy, and the same mentality, and the same tools as economists did when they looked at the economy as a web of simple links. And both humans and societies are complex systems."

That's philosopher Nassim Nicholas Taleb, from the new section, 'On Robustness and Fragility' in the updated edition of The Black Swan. It's a bit jargon-heavy, but goofy neologisms are a signature trait of Taleb's, and probably constrains original thinking less compared to other more stylistically-traditional authors. It's just a taste of what's to come in his forthcoming book 'Antifragility,' which extends the ideas in Fooled By Randomness and The Black Swan into positive territory, rather than just negative (e.g. discussing what we can and should do, rather than just what we can't and shouldn't). Expect fascinating sections on economics, philosophy of biology, epistemology, medicine, and nutrition.

For a further preview, check out this great EconTalk interview with Taleb. I'm especially excited to get Taleb's full take on the potential health benefits of fasting, especially after reading this great cover story in Harper's.

Power Players: Arthur C. Clarke

Amal Siriwardena of the Sri Lanka Guardian has a great profile:

While researching for this article I came across a searing indictment by Clarke on the American capitalist system. After observing that the structure of American society may be unfitted for the effort that the conquest of space demands he continued, "No nation can afford to divert its ablest men into essentially non-creative and occasionally parasitic occupations such as law, insurance and banking". He also referred to a photograph in Life Magazine showing 7,000 engineers massed behind a new model car they had produced as ‘a horrifying social document’. He was appalled by the squandering of technical manpower it represented. All this indeed makes one wonder whether he really was a closet socialist.

For more, check out this interesting clip from 1964 of Clarke making some predictions about the future.

A Different Kind of π-Day

March 14th, Pi Day. On this most important day math geeks and students-with-no-choice alike celebrate the number π: the mathematical constant equal to the ratio of a circle's circumference to its diameter, usually by baking or buying a delicious pie to share with friends.

This year, I want to bring to attention an interesting idea: that the number π is stupid. A circle is defined by its center and its radius, and by defining π using diameter (two radii), we complicate things unnecessarily. Instead, mathematicians could use Tau, defined by the ratio of a circle's circumference to its radius.

Now, obviously, the constant π is enshrined in mathematics (could we ever really change the beautiful Euler's Identity, e^(iπ)-1=0 ?), but it's certainly fun to imagine alternative structures based around Γ. For more information, read The Tau Manifesto by Michael Hartl. Happy π-Day!

The Horror of Science

This morning I read an article about a fungus that turns ants into suicidal zombie-ants, and the whole day I couldn't stop thinking about it in the context of zombie horror films. Granted the ant brain is simpler than the human brain, but it still gives me pause just thinking about these ants. I just love it when science fiction becomes real science.

Neologism Watch: 'Anthropocene'

Anybody who witnessed the strange furor over Pluto's demotion from official planet to dwarf planet knows that scientific categories matter. More than factual data, categories shape our perception and influence our values. That's why the recent push to establish a new geologic epoch defined by human impact, called the 'Anthropocene,' is cause for celebration and support.

To be sure, geologic categories should be determined by geologists through rigorous inquiry and public reasoning. Yet the fact that human-driven processes are leaving permanent imprints on the Earth with magnitudes equivalent to other past epoch-defining events should make it a one-sided debate.

An official Anthropocene category could generate a fantastic branding campaign to combat the corrosive belief in a static, historically-fixed Earth. Many environmentalists, by focusing solely on avoiding future environmental damage, neglect a powerful message concerning the changes already wrought by human technology. Bill McKibben's Eaarth is perhaps the best manifesto of this sort, methodically listing current environmental changes caused by global heating and ecological destruction. Fostering environmentalist preferences among a global population still reeling from economic catastrophe has been difficult, not least due to the supposedly vague, far-off nature of environmental costs. But the fact is humanity has made the world its own, whether we like it or not. It's time we formalized this harsh reality.

Exotopia

There's been a lot of buzz lately about the startling discovery by NASA's Kepler orbital observatory of six new extrasolar planets (or exoplanets) orbiting a single star. Further tentative results suggests evidence for over a thousand more exoplanets. Although this type of discovery inevitably spurs talk of aliens and the search for extraterrestrial life, exoplanets are most valuable for what they can tell us about Earth and our own biological origins. Up until now, all of our science (perhaps excluding physics and chemistry) has been based in observational evidence from the only location we've ever had access to: the Earth. The emerging field of comparative planetology unleashes the power of comparative analysis on our most vexing scientific questions: where did we come from? How common is life? How unique is our own geologic and chemical situation? Access to many diverse exoplanets allows us to control for variables in a way that was previously impossible, opening up new avenues for scientific inquiry.

The future of exoplanet science will be a long, gradual process, but investments in deep-space science should be a global priority for several reasons. The long-term survival of the human race crucially depends on migrating into space in order to establish resilience against catastrophic risk; inhabiting only one spaceship (the Earth) puts all our eggs in one basket, so to speak. Secondly, comparative planetology and deep-space inquiry will have spill-over effects in other areas of science (such as geology and biology) that will generate concrete benefits here on Earth. Thirdly, the sci-fi nature of exo-science will surely awe a new generation of youngsters, creating interest in science and re-establishing NASA's unique position as a government entity with vast cultural influence. Inspiration matters. Finally, scientifically demonstrating the insignificance of Earth and human existence will aid in a re-alignment of policy priorities on a global level: establishing a species-wide identity will assist in better risk-management and decisionmaking in issues varying from global climate change and ecological destruction to catastrophic risk mitigation.

Carl Sagan once said, "the universe seems neither benign nor hostile, merely indifferent." Perhaps so, but the knowledge and promise contained within the vastness of space is undoubtably a positive resource. The fact is this: if the human race has a future, it lies in space. Let's get started.

What's The Opposite Of Catastrophe?

Discover Magazine is out with a fascinating article about the current state of nuclear fusion energy research. Generating energy via nuclear fusion is seen as a holy grail for the human race--it produces no radioactivity or carbon emissions, and is fueled by hydrogen extracted from seawater. Unfortunately, the technology is about 20 years away, and has been for about 50 years.

I recently finished Catastrophe: Risk and Response by Richard Posner, a book about high-impact, low-probability events and their treatment within governing institutions. While Posner focuses only on bad, "catastrophic" events like bioterrorism and asteroid collisions, many high-impact, low-probability events are beneficial. An unexpected research breakthrough, say, in the field of nuclear fusion, could cause untold global benefits. For lack of a better word, we can call these beneficial, improbable events positive black swans.

Just as our governing institutions are ill-suited for dealing with catastrophic events, so too are they incapable of properly assessing the rewards of positive black swans. Our political and policy structures abhor uncertainty, and greatly discount future costs and benefits. This results in over-exposure to catastrophic risks and under-exposure to positive black swans. Policy and funding priorities more influenced by risk assessment, cost-benefit analysis, and expected utility calculations could supply a corrective.

Question: What is a better name for beneficial high-impact, low-probability occurrences? I can only think of 'miracle,' which is a somewhat religiously loaded term. This macro-concept is simply begging for a cleverly-titled non-fiction book to be written about it. Somebody get me Malcolm Gladwell's phone number...

Science and Law

Governments worldwide are not doing enough to prevent the occurrence of low-probability, high-impact events, also known as catastrophic risks. This is the primary message in the book Catastrophe: Risk and Response by Richard Posner, a federal judge and University of Chicago professor. But aside from the horrors of asteroid collisions and bioterrorism, Posner sketches a fascinating critique of our current legal system.

Among the many deficiencies in our catastrophic risk management system, the law stands out as an institution particularly ill-suited towards dealing with issues of science and risk. As the wealth of human scientific and technological knowledge has grown, so too has its complexity. Understanding many scientific topics requires more and more specialization, something anathema to the law in its current structure.  Although some lawyers specialize in scientifically rigorous fields, judges and juries by design do not. This can result in tragically misinformed legal outcomes, such as the 1980 Diamond v. Chakrabarty which established that living organisms can be patented just like new chemicals.

A second problem concerns the culture of the law. With its emphasis on verbal logic, legal professions overwhelmingly attract students who are either bad at, or uninterested in, math and science disciplines. In fact, the absence of math and science in the law is a big factor that attracts many students from humanities disciplines. Science illiteracy is a big problem everywhere, but the law's unique importance in our policy process makes it especially damaging.

A third concern relates to the law's obsession with personal testimony and human cognitive bias. It's best explained here, go to 1:50 in.

Lastly, a more philosophical point. In all areas of science, but especially in quantum physics and biology, the idea of isolating simple cause-effect relationships is breaking down. It turns out that reality is a lot more complex that we thought. Instead of simple, linear, causal relationships we have interacting feedback loops of overlapping conditional and causal networks. Chaos theory and fractal geometry have shown us that randomness is a massively powerful and pervasive force that affects everything. In light of these new truths, the law's singular structure of "one party versus one party; who's to blame" seems almost quaint. Take environmental law, for example. Does this company's pollution make it guilty of contributing to this community's health problems? This type of question is terribly suited for the law's "guilty/innocent" decision structure. Maybe the low socioeconomic status of the community reduced its members' overall health and made them especially vulnerable to the effects of the pollution, and the same pollution had no effect on the richer, healthier, more resilient communities in the area. In this case, assigning blame on the basis of one direct causal relationship is naive in light of other causal factors. 

So how can the law improve its relationship with science? Increasing scientific specialization among lawyers, judges, and juries in the form of a "science court" could go a long way towards improving the quality of scientifically complex legal decisions. Massive investments in science and math education could also help, as well as mandating higher levels of science and math coursework in basic education guidelines. That many of today's students will never take an introductory statistics course in their lifetime is quite a chilling thought indeed.

Power Players: Benoit Mandelbrot

Last thursday Benoit Mandelbrot died in Cambridge, Mass.  Mandelbrot was a hugely important figure in mathematics: he refined the concept and study of fractals--an object that looks the same the closer you zoom in on it.  The discovery of fractals showed the world that mathematics could be applied to realms previously off-limits to rigorous inquiry.  Just as calculus unlocked the mysteries of change, fractal geometry revealed the inner structure of nature's jagged contours.  A nice obituary is here.

In honor of Mandelbrot, I'm re-posting this spectacular documentary about fractal geometry done by Nova.  It includes many great clips of Mandelbrot and of the crazy computer-generated imagery we've all seen but never really understood.

What Would Happen If...

..you put your hand in front of the beam of the Large Hadron Collider?  Let's ask the experts:

Why Sci Fi?

I love science fiction, maybe because I'm a huge nerd, but also because sci fi deals with interesting concepts (social arrangements, trends, technology, etc.) in an extremely accessible way.  Sci fi isn't shackled by the requirement of believability, so concepts and ideas can be dealt with explicitly and without subtlety.  A lot of sci fi occurs in the future, which makes commentary on contemporary concepts and issues much easier: the reader doesn't have to analyze and speculate on their own about about the impact of current trends or technologies.

Here's a great list of some major works in science fiction, although a few big ones like Dune and Foundation are conspicuously absent.

Note: Unfortunately sci fi isn't known for it's cover art.

Quantum Poker in Star Wars

Hardcore Star Wars aficionados may be familiar with Sabacc, the fictional poker-like card game popular with lowly spice smugglers and interplanetary tycoons alike.  Without delving too much into the nerdy arcana, players receive cards grouped by value (1-15 plus 16 face cards) and suit (Coins, Flasks, Sabres, Staves).  There are several rounds of betting, the pot going to the player with the strongest hand based on a specific hierarchy.  The main feature of the game is the shifting nature of the cards: while in a player's hand, every card may randomly change its value and suit based on a predetermined probability distribution.  During each round of betting, players have the option to place cards face down on the table.  This action locks in the card's value and suit, preventing further shifts.

So why is this relevant?  Sabacc provides a perfect real-world analogy for the concept of quantum superposition, or the condition of an object existing in several mutually-exclusive states simultaneously.  Emerging from modern particle physics, quantum superposition is perhaps best known by the famous thought experiment Schrödinger's Cat.  A key feature of quantum superposition is its collapse: whenever the superpositioned object is measured, its strange quantum condition ends and a single observable state emerges.  Returning to the Sabacc analogy, we can view a player as holding cards that exist in a probabilistic superposition.  Due to the random shifting, each card possesses not one value, but rather a probability distribution of every possible value.  Placing a card on the table collapses this distribution into a single value and suit.

The concept of quantum superposition, once relegated to the netherworld of quantum mechanics, is seeing an explosion of applications.  Most interesting is the fledgling field of quantum game theory, which allows for the quantum superposition of strategies by players, in addition to regular pure and mixed strategies.  For more information about quantum superposition and its exciting new applications, check out episode three of the Math For Primates podcast.

Culture And Social Science

Michael Keating says it all:

The social sciences face four enduring problems in understanding and explaining behavior.  First is how to account for both continuities and change over time within societies.  Second is to explain the connection between micro-level changes and the larger, macro level.  Third, and related, is how to explain the connection between individual decisions and the aggregate behavior of a society as a whole.  Fourth is the relationship between the hard facts of the social world and the way in which these are interpreted by people.

Cultural explanations of social phenomena go directly to the collective level, they are essentially social and in many respects (but not quite all) they represent a challenge to methodological individualism.  They also seek to bridge external explanation, by reference to a social world, and internalist explanations, which rely on individual interpretation and decision.  Yet if culture allows us to identify and explain differences in behavior among groups -- be these nations, classes, genders or localities -- it is an extremely elusive and slippery idea, prone to all manner of abuse

[Culture] can help to understand and explain social and political institutions and behavior, but only if it is understood in a sophisticated way.  Although there are numerous difficulties in operationalizing and measuring it, these are not totally insurmountable. . . . insight into the complexity of culture can be gained through triangulation and combining different methods.  Thus surveys can tell us a lot about popular attitudes, while ethnographic work may be needed to explore their meaning.

Happy Pi Day Everybody

I almost forgot, it's 3/14 today: π-day.  If you're confused, check out this short podcast.

Thoughts On The Current Policy Quagmire

It's become the conventional wisdom that our government is incapable of solving big problems.  Health care, global heating, peak oil, federal deficits, structural reforms, and most environmental problems all typically make the list.  So why can government tackle some issues, but not these?  Perhaps each issue's unique characteristics are to blame?  While each problem obviously contains unique challenges, certain common characteristics exist that allow us to generalize a bit.  Most of these problems are massive and complex, requiring major policy solutions.  These solutions require simple, fixed, upfront and well-understood costs to achieve benefits that are vague, amorphous, far-off and complicated.  So what about these big, complex issues makes them so difficult for government to solve?  There exist many explanations. 

The most basic answers focus on individuals.  Cognitive biases in human psychology make certain issues troublesome to deal with.  People have a difficult time conceptualizing the benefits of solving environmental issues, for example, yet they understand the costs of higher taxes very well.  People discount the future too heavily, so they favor and reward short-term promises over long-term solutions.  People tend to think that the status-quo is safer than it actually is.  Some explanations centered around individuals focus on leadership: Obama isn't making the right decisions.  Leadership and inspiration is difficult with big, complex issues.

Other explanations emphasize social groups and collective decisionmaking.  Democratic and Republican lawmakers don't have casual friendships anymore, so rejecting compromise is dispassionate and easy.  On the larger scale, commentators claim the public cannot make up its mind.  Party identification has been falling for decades, and now we have a massive group of impatient independents in the ideological middle who simply ping-pong back and forth, preventing either party from amassing enough power or popularity to solve all but the smallest issues.

Still other explanations do away with people altogether, instead focusing on institutions and big, structural factors.  In this category, the senate filibuster is to blame.  Legislative committees and executive agencies, jealously guarding their turf, don't coordinate.  Solutions to big, complex issues lose cohesion.  Legislative politics incentivizes short-term strategic action and hard-nosed opposition.  Gerrymandered districts pack The House of Representatives with extremists, who go on to win Senate seats, infecting that body.  Our constitution grants too much power to states and minority factions.

All these explanations have merit, and are probably all true to some degree.  Yet most of them have trouble explaining the recent change.  Government has solved big, complex issues before.  We've always had the filibuster, and the senate managed to pass big, transformational legislation.  Human psychology hasn't changed too much over the past 40 years, I'm guessing.  What is it about now that makes these issues especially impossible to solve?  I don't know the answer (or even if there is one; it could just be random factors), but I do have a thought.

Over the course of history, and especially in the last 40 years with the information revolution, technology and scientific knowledge has increased rapidly, leading to new levels of understanding about how the world works.  This pattern is apparent in all aspects of life: the world is becoming increasingly technocratic.  Better statistical techniques identify more and more correlations.  Experts know more and more about less and less.  Modern social science is giving us drastically better information about individuals and populations.  Our understanding of the world is becoming better and more complex.  The implication for government is clear: while politicians may not be more calculating than they were in the past, they are certainly better informed and more accurate.  The 2000 presidential election was a revealing flash-point.  That incident showed everybody that nothing matters except winning.  Bush became president, end of story.  The ends do justify the means, because voters only remember the ends.  If the minority party can hold out in opposition long enough to deny the majority any achievements, the minority will soon find itself in the majority, pursuing its ideal policies.  We may not like the harsh, misanthropic findings of social science, but their explanatory power is impossible to ignore.  Increasing quality and analytical rigor of political decisionmaking means it's easy to be "rational" in the formal sense.  An unfortunate and unintended result of this medium-term consequentialism is that long-term policy solutions are increasingly politically irrational.

Toxicology

Nicholas Kristof has a great column today about possible environmental causes of autism spectrum disorders.  For more information about the crazy conspiracy theories and misunderstandings surrounding autism and vaccines, check out Denialism by Michael Specter.