Field of Science

How to thrive as a fox in a world full of hedgehogs

This is my fourth monthly column for 3 Quarks Daily.

The Nobel Prize winning animal behaviorist Konrad Lorenz once said about philosophers and scientists, “Philosophers are people who know less and less about more and more until they know nothing about everything. Scientists are people who know more and more about less and less until they know everything about nothing.” Lorenz had good reason to say this since he worked in both science and philosophy. Along with two others, he remains the only zoologist to win the Nobel Prize for Physiology or Medicine. His major work was in investigating aggression in animals, work that was found to be strikingly applicable to human behavior. But Lorenz’s quote can also said to be an indictment of both philosophy and science. Philosophers are the ultimate generalists, scientists are the ultimate specialists.

Specialization in science has been a logical outgrowth of its great progress over the last five centuries. At the beginning, most people who called themselves natural philosophers – the word scientist was only coined in the 19th century – were generalists and amateurs. The Royal Society which was established in 1660 was a bastion of generalist amateurs. It gathered together a motley crew of brilliant tinkerers like Robert Boyle, Christopher Wren, Henry Cavendish and Isaac Newton. These men would not recognize the hyperspecialized scientists of today; between them they were lawyers, architects, writers and philosophers. Today we would call them polymaths.

These polymaths helped lay the foundations of modern science. Their discoveries in mathematics, physics, chemistry, botany and physiology were unmatched. They cracked open the structure of cells, figured out the constitution of air and discovered the universal laws governing motion. Many of them were supported by substantial hereditary wealth, and most of them did all this on the side, while they were still working their day jobs and spending time with their families. The reasons these gentlemen (sadly, there were no ladies then) of the Royal Society could achieve significant scientific feats were many fold. Firstly, the fundamental laws of science still lay undiscovered, so the so-called “low hanging fruit” of science was ripe and plenty. Secondly, doing science was cheap then; all Newton needed to figure out the composition of light was a prism.

But thirdly and most importantly, these men saw science as a seamless whole. They did not distinguish much between physics, chemistry and biology, and even when they did they did so for the sake of convenience. In fact their generalist view of the world was so widespread that they didn’t even have a problem reconciling science and religion. For Newton, the universe was a great puzzle built by God, to be deciphered by the hand of man, and the rest of them held similar views.

Fast forward to the twentieth century, and scientific specialization was rife. You could not imagine Werner Heisenberg discovering genetic transmission in fruit flies, or Thomas Hunt Morgan discovering the uncertainty principle. Today science has become even more closeted into its own little boxes. There are particle astrophysicists and neutrino particle astrophysicists, cancer cell biologists, organometallic chemists and geomicrobiologists. The good gentlemen of the Royal Society would have been both fascinated and flummoxed by this hyperspecialization.

There is a reason why specialization became the order of the day from the seventeenth century onwards. Science simply became too vast, its tendrils reaching deep into specific topics and sub-topics. You simply could not flit from topic to topic if you were to understand something truly well and make important discoveries in the field. If you were a protein crystallographer, for instance, you simply had to spend all your time learning about instrumentation, protein production and software. If you were a string theorist, you simply had to learn pretty much all of modern physics and a good deal of modern mathematics. Studying any topic in such detail takes time and effort and leaves no time to investigate other fields. The rewards from such single-minded pursuit are usually substantial; satisfaction from the deep immersion that comes from expertise, the enthusiastic adulation of your peers, and potential honors like the Nobel Prize. There is little doubt that specialization has provided great dividends for its practitioners, both personal and scientific.

And yet there were always holdouts, men and women who carried on the tradition of their illustrious predecessors and left the door ajar to being generalists. Enrico Fermi and Hans Bethe were true generalists in physics, and Fermi went a step further by becoming the only scientist of the century who truly excelled in both theory and experiment; he would have made his fellow countryman Galileo proud. Then there was Linus Pauling who mastered and made seminal contributions to quantum chemistry, organic chemistry, biochemistry and medicine. John von Neumann was probably the ultimate polymath in the tradition of old natural philosophers, contributing massively to every field from pure mathematics and economics to computing and biology.

These polymaths not only kept the flame of the generalist alive, but they also anticipated science ironically coming full circle. The march of science from the seventeenth to the twentieth century might have been one toward increasing specialization, but in the last few years we have seen generalist science again blossoming. Why is this? Simply because the most important and fascinating scientific questions we face today require the meld of ideas from different fields. For instance: What is consciousness? What is life? How do you combat climate change? What is dark energy? These questions don’t just benefit from an interdisciplinary approach but they require it. Now, the way modern science approaches these questions is to bring together experts from various fields rather than relying on a single person who is an expert in all the fields. The Internet and global communication have made this kind of intellectual cross-pollination easier. 

And yet I would contend that there is a loss of insight when people keep excelling in their chosen fields and simply funnel the output of their efforts to other scientists without really understanding in what way it’s used. In my own field of drug discovery for instance, I have found that people who at least have a conceptual understanding of other areas are far more likely to contribute useful insights compared to those who simply do their job well and shove the product on to the next step of the pipeline.

I thus believe there is again a need for the kind of generalist who dotted the landscape of scientific research two hundred years ago. Both the poet Archilochus as well as the philosopher Isaiah Berlin have fortunately given us the right vocabulary to describe generalists and specialists. The fox, wrote Archilochus, knows many things while the hedgehog knows one big thing. Generalists are foxes; specialists are hedgehogs.

The history of science demonstrates that both foxes and hedgehogs are necessary for its progress. But history also shows that foxes and hedgehogs can alternate. In addition there are fields like chemistry which have always benefited more from foxes than hedgehogs. Generally speaking, foxes are more important when science is theory-rich and data-poor, while hedgehogs are more important when science is theory-poor and data-rich. The twentieth century was largely the century of hedgehogs while the twenty-first is likely to be the century of foxes.

Being a fox is not very easy though. Both personal and institutional forces in science have been built to support hedgehogs. You can mainly blame human resources personnel for contriving to make the playing field more suitable for these creatures. Consider the job descriptions in organizations. We want an “In vivo pharmacologist” or “Soft condensed matter physicist”, the job listing will say; attached would be a very precise list of requirements – tiny boxes within the big box. This makes it easier for human resources to check all the boxes and reject or accept candidates efficiently. But it makes it much harder for foxes who may not fit precise labels but who may have valuable insights to contribute to make it past those rigid labels. Organizations thus end up losing fine, practical minds who pay the price for their eclectic tastes. Academic training is also geared toward producing hedgehogs rather than foxes, and funding pressures on professors to do very specific kinds of research do not make the matter any easier. In general, these institutions create an environment in which being a fox is actively discouraged and in which hedgehogs and their intellectual children and grandchildren flourish.

As noted above, however, this is a real problem at a time when many of the most important problems in science are essentially interdisciplinary and would greatly benefit from the presence of foxes. But since institutional strictures don’t encourage foxes to ply their trade, they also by definition do not teach the skills necessary to be a fox. Thus the cycle perpetuates; institutions discourage foxlike behavior so much that the hedgehogs don’t even know how to be productive foxes even if they want to, and they in turn further perpetuate hedgehogian principles.

Fortunately, foxes in the past and present have provided us with a blueprint of their behavior. The essence of foxes is generalist behavior, and there are some commonsense steps one can take to inculcate these habits. Based on both historical facts about generalists as well as, well, general principles, one can come up with a kind of checklist on being a productive fox in an urban forest full of hedgehogs. This checklist draws on the habits of successful foxes as well as recent findings from both the sciences and the humanities that allow for flexible and universal thinking that can be applied not just in different fields but especially across their boundaries. Here are a few lessons that I have learnt or read about over the years. Because the lessons are general, they would not be confined to scientific fields.

1. Acknowledge psychological biases.

One of the most striking findings over the last three decades or so, exemplified by the work of Amos Tversky, Daniel Kahneman, Paul Slovic and others, is the tendency of human beings to make the same kinds of mistakes when thinking about the world. Through their pioneering research, psychologists have found a whole list of biases like confirmation bias, anchoring effects and representativeness that dog our thinking. Recognizing these biases doesn’t just help connect ideas across various disciplines but also helps us step back and look at the big picture. And looking at the big picture is what foxes need to do all the time.

2. Learn about statistics.

A related field of inquiry is statistical thinking. In fact, many of the cognitive biases which I just mentioned arise from the fundamental inability of human beings to think statistically. Basic statistical fallacies include: extrapolating from small sample sizes, underestimating or ignoring error bars, putting undue emphasis on rare but dramatic effects (think terrorist attacks), inability to think across long time periods and ignoring baselines. In an age when the news cycle has shrunk from 24 hours to barely 24 seconds of our attention span, it’s very easy to extrapolate from random, momentary exposure to all kinds of facts, especially when the media’s very existence seems to depend on dramatizing or exaggerating them. In such cases, stepping back and asking oneself some basic statistical questions about every new fact can be extremely helpful. You don't have to actually be able to calculate p values and confidence intervals, but you should know what these are.

3. Make back-of-the-envelope calculations.

When the first atomic bomb went off in New Mexico in July, 1945, Enrico Fermi famously threw a few pieces of paper into the air and, based on where the shockwave scattered them, came up with an accurate estimate of the bomb’s yield. Fermi was a master of the approximate calculation, the rough, order of magnitude estimate that would give the right ballpark answer. It’s illuminating how that kind of thinking can help to focus our thinking, no matter what field we may be dealing with. Whenever we encounter a fact that would benefit from estimating a number, it’s worth applying Fermi’s method to find a rough answer. In most cases it’s good enough.

4. Know your strengths and weaknesses.

As the great physicist Hans Bethe once sagely advised, “Always work on problems for which you possess an undue advantage.” We are always told that we should work on our weaknesses, and this is true to some extent. But it’s far more important to match the problems we work on with our particular strength, whether it’s calculation, interdisciplinary thinking or management. Leveraging your strengths to solve a problem is the best way to not get bogged down in one place and being able to nimbly jump across several problems like a fox. Hedgehogs often spend their time not just honing their strengths but working on their weaknesses; this is an admirable trait, but it’s not always the most optimal for working across disciplinary boundaries.

5. Learn to think at the emergent level that’s most useful for every field.

If you have worked in various disciplines long enough, you start realizing that every discipline has its own zeitgeist, its own way of doing things. It’s not just about learning the technical tools and the facts, it’s about knowing how to pitch your knowledge at a level that’s unique and optimal for that field. For instance, a chemist thinks in terms of molecules, a physicist thinks in terms of atoms and equations, an economist thinks in terms of rational individuals and a biologist thinks in terms of genes or cells. That does not mean a chemist cannot think in terms of equations or atoms, but that is not the most useful level of thinking to apply to chemistry. This matching of a particular brand of thinking to a particular field is an example of emergent thinking. The opposite of emergent thinking is reductionist thinking which breaks down everything into its constituent parts. One of the discoveries of science in the last century is the breakdown of strict reductionism, and if one wants to be a productive fox, he or she needs to learn the right level of emergent thinking that applies to a field.

6. Read widely outside your field, but read just enough.

If you want to become a generalist fox, this is an obvious suggestion, but because it’s obvious it needs to be reiterated. Gaining knowledge of multiple fields entails knowing something about those fields, which entails reading about them. But it’s easy to get bogged down in detail and to try to become an expert in every field. This goal is neither practical nor the correct one. The goal instead is to gain enough knowledge to be useful, to be able to distill general principles, to connect ideas from your field to others. Better still, talk to people. Ask experts what they think are the most important facts and ideas, keeping in mind that experts have their own biases and can reach different conclusions.

A great example of someone who learnt enough about a complementary field to not just be useful but very good at his job was Robert Oppenheimer. Oppenheimer was a dyed-in-the-wool theorist, and at first had little knowledge of experiment. But as one of his colleagues said,

“He began to observe, not manipulate. He learned to see the apparatus and to get a feeling of its experimental limitations. He grasped the underlying physics and had the best memory I know of. He could always see how far any particular experiment would go. When you couldn’t carry it any further, you could count on him to understand and to be thinking about the next thing you might want to try.”

Oppenheimer thus clearly learnt enough about experimental physics to know the strengths and limitations of the field, imparting another valuable piece of advice: know the strengths and limitations of every field at the very least, so you know whether the connections you are forming are within its purview. In other words, know the domain of applicability of every field so that you can form reasonable connections.

7. Learn from your mistakes, and from others.

If you are a fox trying to jump across various disciplinary boundaries, it goes without saying that you might occasionally stumble. Because you lack expertise in many fields you are likely to make mistakes. This is entirely understandable, but what’s most important is to acknowledge those mistakes and learn from them. In fact, making mistakes is often the best shortcut to quick learning (“Fail fast”, as they say in the tech industry). Learning from our mistakes is of course important for all of us, but especially so for foxes who are often intrinsically dealing with incomplete information. Make mistakes, revise your worldview, make new mistakes. Rinse and repeat. That should be your philosophy.

Parallel to learning from your mistakes is to learn from others. During her journey a fox will meet many interesting people from different fields who know different facts and possess different mental models of thinking about the world. Foxlike behavior often entails being able to flexibly use these different mental models to deal with various problems in different fields, so it’s key to keep on being a lifelong learner of these patterns of thought. Fortunately the Internet has opened up a vast new opportunity for networking, but we don’t always take advantage of this opportunity in serious, meaningful ways. Everyone will benefit from such deliberate, meaningful connections, but foxes in particular will reap rewards.

8. “The opposite of a big truth is also a big truth” – Niels Bohr

The world is almost always gray. Foxes must imbibe this fact as deeply as Niels Bohr imbibed quantum mechanics. Especially when you are encountering and trying to integrate disparate ideas from different fields, it’s very likely that some of them may seem contradictory. But often the contradiction is in our minds, and there’s actually a way to reconcile those ideas (as a general rule, only in the Platonic world of mathematics can contradictory ideas not be tolerated at all). The fact is that most ideas from the real world are fuzzy and ill defined, so it’s no surprise that they will occasionally run into each other. Not just ideas but patterns of thinking may seem contradictory; for example, what a biologist sees as the most important feature of a particular system may not be the most important feature for a physicist (emergence again). In most cases the truth lies somewhere in between, but in others it may lie wholly on one side. As they say, being able to hold opposite ideas in your mind at the same time is a mark of intelligence. If you are a fox, prove this.

These are but a few of the potential avenues that you can explore for being a generalist fox. But the most important principle that foxes can benefit from is, as the name indicates, general. When confronted by an idea, a system or a problem, learn to ask the most general questions about it, questions that flow across disciplines. A few of these questions in science are: What’s the throughput? How robust is the system? What are the assumptions behind it? What is the problem that we are trying to solve? What are its strengths and limitations? What kinds of biases are baked into the system and our thinking about it?


Keep on asking these questions, make a note of the answers and you will realize that they can be applied across domains. At the same time, remember that as a fox you will always work in tandem with specialized hedgehogs. Foxes will be needed to explore the uncharted territory of new areas of science and technology, hedgehogs will be needed to probe its corners and reveal hidden jewels. The jewels will further reflect light that will illuminate additional playgrounds for the foxes to frolic in. Together the two creatures will make a difference.

No comments:

Post a Comment

Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="http://www.fieldofscience.com/">FoS</a> = FoS