Watching physicist Adrian Bejan as he demonstrates how snowflakes form tells you two things. One, he's not just a scientist. The confidence and balance he projects as he draws are a give-away to his training as an artist and as an athlete. And two, that there is a certain predictability to each snowflake.
I've quoted Adrian Bejan numerous times in books and articles about evolution, about
academic mafias and peer review, but somehow we never got around to having a full conversation. So I called him recently at Duke University, where he is now J.A. Jones Distinguished Professor of Mechanical Engineering, to chat about both his constructal law of design in nature -- which he considers one of the few laws of physics -- as well as his formative years in the 50s and 60s in communist Romania. Bejan was born in Galati, Romania on the Danube River.
Despite growing up in a country where hunger was an ongoing fact of life, Bejan thrived, indeed triumphed. He became both a math champion and star athlete, years before the West knew about Romanian star athletes -- like "perfect" Olympic gymnast Nadia Comaneci.
In 1969, Bejan was allowed to leave Romania to study in the US. Over the next several years Adrian Bejan would receive his BS (1971), MS (1972) and PhD (1975) in engineering from MIT.
In the interview that follows Bejan defines life as "the flow system that morphs and persists." He says the origin of life on Earth is 100% physics and that it coincides with the origin of Earth itself. He also questions linking of the term "growth" with "evolution" in science. Bejan thinks D'Arcy Thompson's book On Growth and Form was an inspiration for the "mistake."
Adrian Bejan's interest in art first evidenced itself during his childhood. He went from drawing high speed motorcycle racing to the human anatomy -- no doubt inspired by his father's work as a veterinarian. Bejan now regards illustration essential to his science, and he attributes the success of his books in part to his drawings.
Adrian Bejan is the author or co-author of 28 books, best known is Design in Nature (Doubleday, 2012), and of 600 scientific papers. The Bejan numbers (Be) -- in thermodynamics and in convection -- were named for him 30 years ago. Adrian Bejan has received 18 honorary PhDs from universities in 11 countries.
Our interview follows.
Suzan Mazur: Your last name Bejan, what is its derivation? Are your family roots in Romania for many generations?
Adrian Bejan: Yes, I am of Romanian origin. My family comes from two-centuries-old villages on the lower Danube. The name Bejan is Old Slavonic, it means "a person on the move." It's found everywhere from Poland to the southern tip of Greece.
During the Dark Ages, the whole of central and eastern Europe was invaded from the northeast by waves of Slavic land-seeking peasants. They brought a lot of stuff with them, to some places they brought language -- like to Czechoslovakia, Yugoslavia, Poland and Bulgaria. To other places, names.
In all these places where people at the time spoke either Greek or Latin, local residents would ask the newcomers: "Who are you?" And they would say: "We are bejan."
Romanian is basically Latin, so I am able to speak the Latin languages. Alongside my native Romanian, I speak Italian, French, Spanish and Portuguese.
Suzan Mazur: I understand you were trained as an artist as well as a scientist. Would you talk about your interest in art?
Adrian Bejan: It was because of the very nice parents I had, Marioara and Anghel. The story, which I remember very well, is that my father took me to watch a motorcycle race one day in the city. I was a child, less than 10 years old. When I came home I began to draw for everyone the motorcycles at high speed. My parents decided that evening that I had talent in the graphic arts, so they sent me to art school in the afternoons after my regular classes. It was a good decision on their part because for several years I was taught how to draw using perspective, light, shadow, etc. I looked at the famous paintings and drew them my way. I did studies of the human anatomy -- of the hands, the face. . . .
My training in art has paid dividends in engineering, which requires drawings and blueprints. It has also enabled me to think more effectively, because thinking is done through images.
I'm not an artist now, I'm a physicist, one who is at ease with images and the fact that science is images. Geometry is images or figures or relations between figures. Mechanics requires pictures that move.
My art now is for the purpose of illustrating ideas in physics. Graphs, drawings, little sketches that have developed legs -- which other scientists also find useful. My illustrations have been a selling point for my books as well.
I don't see any difference between art and science. They're both about images in motion. The inner pleasure is the same in making a fantastic piece of art that inspires the viewer as coming up with a scientific idea that triggers explosions of images in the mind of the same viewer. Scientists and artists are specimens of the same species.
Suzan Mazur: Who are a couple of your favorite artists?
Adrian Bejan: Alexander Calder with his mobiles is an idol. He was a mechanical engineer, by the way. I like Matisse a lot. Most recently I saw the Broadway show about Edgar Degas, Little Dancer. His life is very moving, including his comments about critics.
Suzan Mazur: Are you an inventor, do you have patents?
Adrian Bejan: Very few. Tinkering is something I don't spend time doing.
Suzan Mazur: What is the significance of the Bejan numbers?
Adrian Bejan: There are two Bejan numbers. Two independent groups in the 1980s proposed the Bejan number be used for two different phenomena. One of the Bejan numbers relates to fluid flow. We call it convection, with heat flowing along with the fluid. There are two sources of entropy generation, which means irreversibility or dissipation, two sources of dissipation. One of these is due to heat transfer, which means heat flow. The other is due to fluid friction or fluid flow. The ratio between the two effects is the Bejan number from thermodynamics. This ratio appears first in my 1982 book, Entropy Generation Through Heat and Fluid Flow.
Another group in the 1980s came up with a proposal around dimensionless temperature differences. Heat is driven by temperature difference. Temperature difference in a dimensionless form in heat transfer is known as the Rayleigh number. But as I said earlier, when you have convection you also have fluid flow because without fluid flow there is no conveyor belt. The fluid flow is driven by a pressure difference. This corresponding dimensionless pressure difference number in convection has also been named Bejan number.
Suzan Mazur: How influential have the Bejan numbers been in science?
Adrian Bejan: These two numbers because they are dimensionless constitute a shorthand for a science that's mature. The shorthand is useful in abbreviating the text, in simplifying the presentation. We have lots of these numbers. The Reynolds number (Re) is another one of them. Very famous, very much used. Prandtl number (Pm) is another. There are more numbers in Earth dynamics, etc. In the field of thermodynamics and also convection, the Bejan numbers have been used regularly since the 1980s by colleagues, for two generations. It's a nice story of how science evolves, how scientific text evolves toward simpler while not losing its content.
Suzan Mazur. Thank you. I'd like to discuss your life in Romania a bit more. You were a member of the Romanian National Basketball select. Can you share some of that experience? Did you travel internationally? Did you ever think about defecting?
Adrian Bejan: Everybody under communism dreams of defecting. I grew up in the 50s and 60s. And under communism, especially in the place where I grew up, which was a provincial city -- there were two games in town. One was the Mathematics Olympiad and the other was performance in sports.
Being an athlete with Olympic aspirations, I was not only attracted by both of these challenges but thought I was equipped to do well. And I did well in both. I was a Mathematics Olympiad performer and champion, and also a champion in sports. . . . I practiced all sports -- basketball, tennis, handball, soccer, etc., but focused on basketball from an early age. . . .
When I left to become a student at MIT, I'd been a player in the top league for three and a half years, but I was also a member of the national select and I benefited very briefly from the opportunity to travel abroad as a player. It was not only a huge badge of honor but a ticket to culture for a young person from a country where traveling abroad was illegal.
Suzan Mazur: What contributed to your good health and performance as an athlete? Wasn't there a food scarcity in Romania? What foods did you grow up on?
Adrian Bejan: We did not have special food. My family was not privileged, we were the opposite of privileged. Both of my parents were imprisoned. My father lost his job for opposing the regime. Yes, there was hunger in Romania, even during the time I was growing up. Roots were a staple in the diet. We had meat twice a week, on Thursday and Friday -- the meat was fish heads discarded from the fish-packing plant. The rest of the week it was beans, stews, thick soups, and a lot of bread.
This situation went on for a long time. The deprivation and repression destroyed my family, my parents. What was a good idea for a few people who benefited from the communist regime was a recipe for destruction for the population en masse, which is why in 1989 the regime disappeared overnight. Most of it went peacefully, effortlessly, because it was so unnatural.
Suzan Mazur: But your parents were inspirational in terms of your career. They were both professionals. Your mother was a pharmacist. Your father, a veterinarian.
Adrian Bejan: That's correct. My parents were my heroes, role models. Self-made. They came from the village. They were the first generation to make it out of the village to become somebody in the city. They were exactly the kind of examples that should have been followed by people growing up. Instead, as intellectuals my mother and father were targeted as enemies of the state, even though neither was doing any political organizing.
Suzan Mazur: Are you still playing basketball.
Adrian Bejan: I exercise every day, but I'm not a maniac. I'm aware of the fact that just as in the constructal design of anything from animal to vehicle, the flow has to be maintained. Life is movement. Everything needs to be kept in tip-top shape, from walking to physical dexterity to reflex. And yes, I still play a little basketball.
Suzan Mazur: How's the basketball at Duke?
Adrian Bejan: Not as good as my theories about sports evolution!
Suzan Mazur: You've said that "For a finite-size system to persist in time (to live), it must evolve in such a way that it provides easier access to the imposed (global) currents that flow through it." You term this the constructal law. Why do you call this a law and not a theory?
Adrian Bejan: The distinction is due to the definitions of the two terms, law and theory. A law of physics is a very brief statement, a summary about phenomena of the same kind, occurring in the billions. There are the first and second laws of thermodynamics. There are Newton's laws of motion. Another law in physics, evolutionary design in nature, is the constructal law, which I discovered in 1996. The constructal law brings together such phenomena as self-organization, complexity, networks, etc.
Laws of physics are few. However, there are many theories. For example, on my desk here I have a theory about predicting city sizes, one of river sizes, and a third theory about the shapes of Egyptian pyramids. Each theory is self-standing, each possible because of the constructal law.
Suzan Mazur: What do you think of the idea that there is no entailing law for the becoming of the biosphere?Adrian Bejan:
Yes, I just read that and I think it is, how shall I say -- amusing. The law is known already. "Order for free" is just another observation of the natural tendency recognized in physics as the constructal law. For example, the climate is free organization (circulation), also known as free convection.
Suzan Mazur: Are interdependencies that manifest themselves along the way secondary to your constructal law? I'm talking about plants encountering fungi and finding a way to negotiate a mutually-beneficial partnership, their leaves becoming bitter to defend against beetle predators, the resistance to antibiotics by microbes, etc. Do you consider these secondary to your constructal law?
Adrian Bejan: All these things you are listing are observations of phenomena and they are not at all secondary. Each of these observations is a documented and studied manifestation of the constructal law in action.
The reason why you have two species collaborating, a symbiosis, is because moving stuff physically on the landscape is easier when done together than individually. For the same reason two rivulets come together to form one larger stream. It is also the mother of social organization in the human realm.
It isn't that getting together and making one big thing out of a huge number of small things is the best arrangement. There's a balance to be reached between the few large and the many small. That balance is also predictable. Big is not necessarily the answer. The answer is that it is easiest to moving stuff on the landscape in the social, animate realm by a special combination of few large and many small. The combination is predictable based on the constructal law in every domain we've looked at. This is how flow proceeds easiest in the available space.
Suzan Mazur: But where do you cut off what where one system ends and another begins? What is the finite system you describe? Isn't it all interrelated? Noam Chomsky has said, for instance: "We see the world in terms of trees and dogs and rivers, etc. But often the question is, what are those concepts?"
What are your further thoughts?
Adrian Bejan: Let's consider the architecture of the human lung. It has a few large airways and many small airways and even more numerous alveoli. The alveoli are where the ramifications appear to end or cut off. The alveoli are where Benoit Mandelbrot with his supposed geometry of nature out of thin air decided to cut off his algorithm of growing things that look like the branching of a tree.
The constructal design of the whole volume is quite different. It's about facilitating the flow of oxygen from one point, in this case from the mouth throughout the entire finite volume, which means to an infinite number of points, to the entire tissue that inhabits the volume. How is this done? It is done when the time for the oxygen to flow along the airways matches the time the same oxygen needs to diffuse across the tissue that separates the air in the air sac and the blood in the capillaries that vascularize the tissue of the same sac.
Suzan Mazur: But it's not finite, it's all connected because otherwise we wouldn't be alive.
Adrian Bejan: True, however, "finite" refers to the entire flow system. The volume of the thorax is finite, macroscopic and constrained because of the power needed by the human vehicle in order to carry the thorax.
It's one flow that covers the entire area. One flow fills the entire volume. In the case of the lung, it is a flow from one point through the entire volume. And in order for this flow of oxygen to connect completely to every piece of tissue inside the thorax, the flow must travel through a hand-in-glove of airways and vascularized tissue. So they're perfectly mated or mating in the tiny surfaces of these alveoli. That is where the mating is occurring. But it's one flow. The length scale where the mating happens is a constructal design feature, and it is predicted. It is not a "cut-off" picked arbitrarily, out of thin air, as in the descriptive practice called fractal geometry.
Suzan Mazur: Some would say the comparisons you've made to the patterns of lungs, river basins etc. are merely analogies.
Adrian Bejan: I actually don't make analogies and comparisons. I predict every single drawing. I come up with drawings from a purely mental viewing of things. I have published drawings of river basins. My drawings of river basins, not river basins I've observed. I've published drawings of the human lung, my constructal lung. It is not a lung you'll find in a mouse. It's a lung that you find in my book. I predict the drawings of nature, and yes, they are truthful these drawings of nature. They have airways and alveoli. They have branches and leaves. They have nothing to do with molecular construction.
Suzan Mazur: Could you briefly touch on why the branching stops in organisms, why humans have only five fingers on each hand, five toes on each foot. Why does the branching stop there?
Adrian Bejan: The architecture of fingers, human fingers, and toes is a design (organization) that has happened naturally because of the flow of stresses, small mechanical forces between the whole human body and the body that's being lifted or pushed. The interface the hand touches is finite and on it these stresses have to be distributed as uniformly as possible and as low-level as possible so that the human hand doesn't get hurt. That design is a "coupling." The hand-machine coupling is an amazing design that we see being reinvented every moment in robotics.
Suzan Mazur: Do you know cell biologist Stuart Newman's work on this? Newman gets into this on limb buds. He says morphogens come into play.
Here are some of Newman's comments to me:
"We've taken this self-organizational idea and put it into context of the geometry of the limb. And we've said that at the tip of the limb there's something suppressing it from happening. Cells have to escape from this suppression to organize into spots or rods.
The geometry changes subtly as the limb grows in length. Under some conditions you'll get one skeletal element. Under other conditions you'll get two. Under still other conditions you'll get three or, as in the human hand, five."
Adrian Bejan: This is actually an important thing that you bring up.
Growth is not evolution. [emphasis added]
Suzan Mazur: Can you give me a concise definition of "growth"?
Adrian Bejan: The mistake in linking growth with evolution was committed by D'Arcy Thompson in his book On Growth and Form. A huge following developed and furthered the misuse of language. Growth became synonymous with evolution, which is not even close.
The growth phenomenon is exhibited by anything that spreads in an available space. It could be the water spreading on the ground when you pour it, or the river delta growing, or the spreading of populations -- from bacteria to people -- diseases, new technologies, knowledge. Anything that spreads exhibits a growth phenomenon.
Adrian Bejan: Growth is increase in a volume that's occupied by a spreading flow. Biology looks at the growth of the human body. The change in form over time is evolution. But it is a change in form on a time scale. Evolution is change that has been happening in the architecture over ages. It's long-term.
Growth is the very short time scale of any spreading flow. It's predictable. It's called the S-curve phenomenon. Every animal body grows in S-curve fashion, every mining shaft digs dirt out of the ground in S-curve fashion, every product in business is spreading on the market in S-curve manner. It's called logistics S-curve. The sigmoid function. This is a very old field. The S-curve phenomenon was finally predicted by the constructal law.
Suzan Mazur: There continues to be some debate about which came first in origin of life -- metabolism or genetics. Some scientists, like Ricardo Flores, think a viroid-like entity may have played a role as the first replicon.
Adrian Bejan: The beginning of your question is not in the direction of my thinking. Because with the constructal law life is defined as a phenomenon of physics. Therefore life originated on Earth before the biosphere ever existed. Life was from the beginning with all these freely morphing, flow configurations. Evolution from winds to thunder to river basins and turbulent flow from the beginning.
The biosphere came later with its myriad of designs moving and morphing into more designs as an add-on, in a way in which to allow the Earth's crust to flow more easily and make itself faster and faster, with a more and more complex this, and a more and more efficient that all the way to technology evolution today.
Suzan Mazur: You've said origin of life is 100% physics. Do you see a difference between origin of life and evolution?
Adrian Bejan: In my work, life is defined as a phenomenon of physics. So is evolution. Evolution in physics means design change, meaning changes in a drawing that happens in a predictable, discernible direction in time. So life is the flow system that morphs freely and persists. Evolution is, literally, from the dictionary design change over time in a discernible direction, which is the opposite of random. Discernible is the opposite of chaotic.
I know the river basin will evolve in order to vehicle its way more and more easily. That's river basin evolution. Because of the constructal law I know that everything including my own daily activity or that of the United States of America is to evolve over time through highly complicated superimposed flow designs in order to vehicle the respective bodies more and more easily on the landscape.
This is physics, this has nothing to do with a WHAT. It has solely to do with a HOW. And the HOW is evolution as physics. Life is shared by all things that move and morph freely on the surface of a planet, or for that matter anywhere in space where you see movement. If you see something is flowing or something was flowing -- okay that's physics.
In your book, The Origin of Life Circus: A How To Make Life Extravaganza, you interview Rockefeller University physicist Albert Libchaber, who said life is geometry because geometry is physical. I wrote down this line because this is correct. And geometry is macroscopic. Geometry you see it and you pick up a pencil and you draw it. It has nothing to do with chemistry or genes. Zero.
Suzan Mazur: Libchaber thinks life is both algorithmic and nonalgorithmic, that it mixes discrete mathematics with continuous mathematics. He thinks the synthesizing of life is a problem of chemistry.
Adrian Bejan: His comments about geometry make sense to me.
Suzan Mazur: Here's the Libchaber quote:
"Biology is a mixture of two mathematics. It's partly algorithmic, not totally algorithmic. When the cell divides there's a lot of geometry, a lot of chemistry. And chemistry is not algorithmic, it's continuous."
Adrian Bejan: He made the comment about geometry, "because life is physical," which means palpable.
Suzan Mazur: Libchaber said further:
"It's a mixture of algorithm and geometry. It's the only place where both come together. Algorithmic means that you need a code. You need coded polymers. Otherwise there's no life. And then as soon as you have a code, you have self-reproduction as a possibility. Because once you have a code you have a memory. And once you have a memory, you can reproduce. And geometry because life is physical. It's a combination of both that makes life."
Adrian Bejan: Physics is also a word used without definition. Physics means everything that happens or exists by itself naturally. . . .
Suzan Mazur: Elbert Branscomb, a physicist at the University of Illinois, Urbana-Champaign has told me that catalysis is simply not enough to have kick-started life on Earth, that "all a catalyst can do is accelerate the rate of the reaction."
"Several of the chemical activities needed for the most rudimentary metabolism, the most rudimentary biochemistries are ones that can not happen on their own. They have to be powered, driven against their thermodynamic will. Forced."
Does that mean anything to you?
Adrian Bejan: I agree with Branscomb to the extent that for anything to move, anything to flow, it must be driven. This is the reason why in my statement of the constructal law I used the word "imposed." That word imposed means driven, dictated, pushed, pulled.
So the flow entails power that is being spent on making the flow happen. The power, again with the constructal law, I've shown that it comes from the design called climate. The Earth is a heat engine that produces all the power that is then dissipated in order to drive all the flows from the water in the river basin to the food from my hand to my mouth.
Suzan Mazur: You've made the statement:
"The tendency of living systems to become optimized in every building block and to develop optimal associations of such building blocks has not been explained: it has been abandoned to the notion that it is imprinted in the genetic code of the organism. If this is so, then what genetic code might be responsible for the development of equivalent structure in such nonliving systems as rivers and lightning," etc.
Adrian Bejan: Let me explain, because you just quoted me from 1996, which is the moment when I composed the constructal law. At that time, in 1996, I was very busy in thermodynamics developing a method called the entropy generation minimization. In 1996, I was doing engineering design of machines to make them more and more efficient. That's called thermodynamic optimization or entropy generation minimization or efficiency increase. It's a very old intuitive activity, as old as the steam engine. That was my language at the time. That is why in the passages that you quoted, the word optimal or optimize appears.
For many years already I'm not using that word because optimal in street language means the chosen, meaning the ultimate. The best. The constructal law is, in fact, a statement that negates the existence of such a thing, meaning there is no end design. There is no destiny. If you read the constructal law, it's a statement about the direction of evolutionary design. It's about a time arrow of design change. The direction in which the movie tape is running. This movie tape will continue to run.
Sure, along the way one or more design changes prevail in order for the flow system to persist over time, but they are not chosen forever. Because given freedom, there will always be a better flow configuration for the flow system.
Today I would delete the word optimal from what you quoted because it gives people the wrong idea. And the wrong idea is end design or destiny, maximum, minimum, best, least, and so on. If end design were true, it would be a very boring world.
Suzan Mazur: So does the Modern Synthesis need to be replaced and what should replace it?
Adrian Bejan: Science is an evolutionary design -- a story -- that evolves because of the urges of very many, not because of what one of us says what "should be."
Everywhere I look, I see phenomena that confirm the constructal law, phenomena that tell me and my students that life is physics. That evolution is also a physics phenomenon. Evolution has direction, therefore it is not random. Again, as an educator, I can also tell people what knowledge is and what information is not.
I'm not aware of all the debates. I come from a background of thermodynamics. To me the picture is clear and a lot simpler and a lot more brief. The constructal law is a law of physics, a summary of a tendency in nature. Something more universal than a law of physics does not exist. This is why the book of science is populated by extremely few laws.
Suzan Mazur: Do you think a consensus of scientists is anathema to good science? Science and technology historian David F. Noble once advised, "[W]henever you have a consensus of scientists that should set off alarms."
Adrian Bejan: There's a great line from Galileo, which is: "In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual."