 Furthermore, making an argument that I have to change my motive activity, or I had to change my motive activity around 1971, most scientific work is what is called normal science by the historian of science Thomas Kuhn. Normal science is adding tiny bits of knowledge on top of the edifice of the science that exists and the science that is accepted by almost everyone. Occasionally, the basic approach that science is being done on comes under attack and something really new is done. This is called revolutionary science. In the 1970s, I had read Kuhn and could recognize that my past was revolutionary, but I was likely to have only normal science in my future. So be it. But within this limitation, I could do worthwhile things. I could try to exemplify a standard of good work. I could lead my colleagues into new areas of research. I could offer worthwhile commentary on the intellectual issues of the world around me and I could offer careful criticism of the scientific work about me. So new directions. For decades, now we're in 1975 or so, for decades, physics and chemistry departments have been studying atomic and subatomic phenomena and mostly ignoring the observable world of wind, water, waves, and sand, but partially because the existence of computers made possible new models of investigation. There were major opportunities in the scientific study of things that happened in the world about us. My colleagues, Albert Liebscher Bay and Sidney Nagel at the University of Chicago introduced the study of these things into that university and through Chicago and to physics departments all around. For example, I did a theoretical study focused on the shape of this little neck that connects a drop coming off of a pipette to a larger mass of fluid. This shape has a mathematical behavior described as a singularity and is quite universal in its behavior. At the same time, Degenne in France and Sam Edwards here in England were introducing these kinds of new subjects, I repeat, wind, water, waves, and sand into physics departments in their respective locations. Backwards, why do I need to be backwards? Along the same lines, I began a series of calculations to understand what would happen in real collisions amongst small pieces of material like grains, misspelled, of sand or sugar. In Newton's mechanics, when two particles collide head on, they come off with unchanged relative speed. This happens because they don't lose any energy in the process of their collision. In contrast, when two pieces of real material collide, because of friction-like effects, they move away from one another with a smaller relative speed than they had before the collision. I point out that this leads to strange effects, like as in the behavior of a bunch of bouncing balls moving around in a narrow pipe. Here are the balls, here's a sketched picture. What happens is, if you let them bounce for a while, the balls towards the end come almost to a stop, and one ball moves up and back to the two ends with large speed. This is a striking result, and there are other striking results like that. Perhaps in response to this theoretical work by experimental colleagues, Nagel and Heinrich Jaeger, they did a series of investigations of real granular materials. One of the most amusing is seen in the next slide. Here's a movie of a sand shooting out, at least from where I stand, it's hard to see what the movie is, but I will try to make up for that. We have a jet of sand shooting out, and it hits a little target here. As it hits the target, it splashes around the target and forms a cone around the target. The remarkable thing about the formation of the cone is that the sand sticks together. The beam of sand doesn't fall apart. That's a big surprise, and not something that we expected. Another threat of work is what is called convective turbulence. It sounds formidable, but is exemplified by a most common situation, heating water to boil an egg. You'll heat the water from below and cool it from above. There is hotter water towards the bottom of the pot and cooler water towards the top. The heated water tends to expand, so the water at the bottom is less dense than the water at the top. Thus the water at the bottom cries to rise. You'll excuse me for giving volition to the water. It's not supposed to be a scientific thing to do, but all scientists do it. The water at the bottom tries to rise. The resulting motion was being investigated in the laboratory of my colleague Albert Liebscher Bay. I drew a picture, here's the picture, of what I thought was happening in what is called the convective cell, the box in his laboratory. What I saw in the box, literally with my eyes, but it takes some interpretation, was plumes. Plumes are little pieces of hot water going up and cold water going down, and a sort of complicated machine. I'll come back to this machine. The machine involves several different working parts. There's a jet of water form going up this way, and a jet of cold water form this way, and plumes being formed at the bottom. Now first I should tell you what a plume is. A plume, hot plume, is a region of heated, less dense water in a higher density environment. It is trying to rise. In order to describe what happens, I want you to imagine that I am a bag of hot fluid. Well, I am sort of. The thing that's different is that here I see a situation in which the fluid around the hot fluid is more dense than the hot fluid, and in the case of the air, it's less dense. But imagine that I have a more dense environment around me, and therefore I am trying to rise. I try to go up that way, but the fluid above me pushes down. So instead of going up, all I could do is go out. As I go out, I'm pushing against the fluid there, and it curls under like that, and you get something that looks like this. Here's a small plume of the source that forms in the laboratory. Here is a big open air plume, much more unfortunate for our world. This is a thermonuclear explosion, and the hot fluid rises there and forms a cap. The hot fluid rises here and forms a cap. Here's a picture of a lab cell with fewer plumes in it than existed in a Lebsche Bay cell. But you can see hot fluid going up that way, cold fluid coming down that way, in the form of plumes, the thing in the center is a thermometer to measure the temperature. Here's a movie, a movie of a computer simulation of that behavior. We have fluid swirling around. You can see the plumes, the caps aren't fully formed because we have a relatively small cell, but you can see the plumes if you want to look for them. I'm going to let this go through twice. We've just gone through the first time. What I want to look for is the change in direction of the motion. The motion is now going this way, but after a while it will turn around and go the other way, and that change in direction, come on, okay, back it up and I'll show you the change in direction. So back up one, go. So it goes along this way for a long time. You can see the plumes. What's going to happen is a plume is going to rise from here and the motion, there it happened, it happened. The motion started going in the opposite direction. My colleagues were interested in understanding how that change in direction occurred and a lot of other things and they studied the movie carefully in order to see what was going on. We can all agree that what we've just seen in the heated cell is a pretty complex motion. Complexity is a fraught word in U.S. science and culture. That's because there's a large body of people in the United States who refuse to believe that humankind is a result of an evolutionary process, but instead argue that we are the product of an intelligent design by a creator or more likely a creator with a capital C. At this point, as you can gather, I'm describing a segue from a situation in which I'm trying to help my colleagues move into new fields of science to a situation in which I'm helping them and perhaps others think about what the science implies in a social context. There is this view of people being constructed by a creator was, for example, put forward by the Reverend William Paley. In present day, Michael Beehe, who is a chemist, and Bill Dempsky, a mathematician and philosopher and ex-student of mine, are all intelligent design theorists. In my view, one of the responsibilities of a scientist or other intellectual is to use his or her expertise when relevant to add to the discussions of the day. My contribution to the intelligent design argument is to describe how complexity naturally arises in fluid systems. Beehe's argument starts out with a biological cell. There's a picture of a biological cell. Beehe points out that biological cells are amazingly complicated. He points out that they have lots and lots of different pieces and that the pieces all work together. He cannot imagine how they could have risen naturally through any process other than construction by a creator. I try to reply by saying that fluids are basically very simple. They could be described as little balls bouncing around and occasionally bump into one another. This then is a picture of what a fluid looks like. However, they can do things like this and now I'm going to describe for you in a little bit what this is. So we have on the bottom here a thin layer of fluid called a boundary layer. There are waves in this thin layer of fluid and the waves throw up spray. The spray forms into these mushroom-shaped objects which are plumes. The plumes then come up and move through the center of the box or move in a jet along the wall. When the plumes come up, they hit the top and they go splash. The splash produces waves that move across the top of the box. That produces cold plumes, inverted. Those cold plumes move across form a jet over here, move into the center, come down and go splash and this continues over and over and over again as you saw in the moving, as you saw in the computer simulation. That behavior is a behavior like a machine with many complicated parts. Nowhere as near as complicated as that, but many complicated parts. These include this jet here, the boundary layer there, this layer of moving fluid, the plumes and so forth. This machine forms naturally with doing nothing more than turning on your stove in the morning. Pot of water heated from below, cooled from above, that's what it is. It is conceivable to me at least that a biological cell is formed by a process of making itself more and more complicated, starting out with the natural ability of nature to be complex. We don't have much knowledge about how complexity arises in the present biological world or especially for the earliest biological forms. The intelligent design theorist, Stemsky and Beehe, do us some good by pointing out our ignorance, but neither they nor the intelligent design people more generally seem to have much positive to add to this discussion. A better view than theirs in my opinion is that of St. Augustine of Hippo who said, at least said according to the translation of a modern Catholic scholar Howard J. Van Til, I must admit that my Latin is not up to doing translations. The universe was brought into being in a less than fully formed state, but was gifted with the capacity to transform itself from unformed matter to a truly marvelous array of structure and life forms. I'll drink to that. In the last 1600 years, we've learned some specifics to help flesh out Augustine's good overview. We should not let the particular views of the universe provided by intelligent designs followers replace science in the school curricula. Instead, we should notice that evolutionary biology, paleontology, cosmology are not speculations. They're root parts of science and human knowledge. They belong in schools as part of the basis of our curriculum and of our understanding of the natural world. Other subjects unrooted in experiment should be avoided in school curricula. I have one final example that I'm going to give you all in concluding this talk. I'm going to be talking about large-scale computer models and I'm going to be asking implicitly how reliable are the largest-scale computer simulations in the physical sciences. This is another set of turbulence calculations. You can imagine a cold fluid sitting on top of a hot fluid because the hot fluid is less dense than the cold fluid. The hot fluid tends to rise and here is a picture, before we come to the picture, I should point out its practical importance. This instability, called the Rayleigh-Taylor instability, the instability of this fluid trying to rise, is very important for the U.S. nuclear weapons labs. They saw variability of the early studies of this instability. They sponsored more than 20 major studies of this instability and the turbulence that it causes. First I should say something good about the nuclear weapons labs. I'm not going to say altogether good things, but they have the capacity to study over long periods of time problems that they think are important. That is rarely true. Governments have in general a very short attention span and industry is getting a shorter and shorter attention span every year. This fluid flow is very much like the convective turbulence flow mentioned earlier except that here there's no friction or rather the friction of the liquid is so small that there's nothing to stabilize the motion. So here's a picture of the Rayleigh-Taylor instability. The fluids of different densities sitting one on top of another and instability causes one fluid to drip into another and in this cartoon that's what results. This is not a cartoon, this is a computer simulation done at the three major weapons laboratories of the U.S. Department of Energy. Now, as I said before, the Department of Energy observed considerable variability in the studies they did of this kind. We at the University of Chicago, we is a funny word for what I'm going to describe to you. I was a spokesperson for the program, but very little of the scientific work, almost none of the scientific work came from me, however I will spoke or I have spoken or something or other. In any case, these guys, particularly Alan Calder, constructed a study which was called a type called a resolution study. As you go from here to here, you have a more and more careful and detailed description of the fluid. Here there are precisely four different boxes as you go across in this direction in which the computer recognizes the temperature in each box. Here there are eight boxes, here there are 512, here there are 2048. Similar scale changes occur in this direction. There are lots more boxes going up here than there are going up there. Now the interesting thing to see is how these studies settled down to result independent of the size of the box. This is what you hope is going to happen in order to result be reliable. How does this thing settle down? The answer is not at all. There's no friction, no friction means there's no way of stabilizing the thing. And what happens is as we go through this whole line of argument, the length of this motion of one region into another changes. It goes up and down. Something qualitatively different happens about here in that you lose the symmetry between the left and the right hand side of the thing. And by the time you get here, that symmetry is all gone. This is an argument that you never see a totally reliable result for this turbulent kind of computer simulation. We were using this kind of situation to calculate the explosion of a star. Many of you saw that explosion as you came in. This was what is called the supernova. A whole big star blows up and produces a flash of light that we can see across the entire universe. My colleagues had a novel picture of how that might occur. Here's the picture. They imagined that a hotspot was produced somewhere off-center in the star. That red dot is my cartoon view of a hotspot. That hotspot rises for the same reason as I would rise if I were a dense fluid. That hotspot rises to the surface of the star, comes up to the top, and produces a shockwave. That shockwave is a burst of pressure coming off of the surface. We had expected initially that when that shockwave was produced that that would produce the explosion that would blow up the star. But no, that's not what happens. We were disappointed to see that the shockwave produced essentially no explosion initially. Here's a picture of the hotspot and of the shockwave, as done by the computer, rather than my sort of primitive drawing. But my colleagues carried on and worked to see what would happen thereafter. They said, I need lots of hands for this, more hands than I've got, but I'll do my best. So here's the star. Here's the thing coming off. You have a burst of pressure here. That burst of pressure sends materials out, and the gravity from the star pulls the materials in, and those materials skip like a stone over the surface of the star and refocus themselves 180 degrees away from their starting point. This refocusing provides a hotspot, and that hotspot provides the starting point for an explosion of the star. This was essentially the first computer simulation which gave an explosion of a supernova. So our sponsors came to visit. We'd previously argued that the basic turbulence simulation was unreliable. We told the computers about our supernova ideas and computer simulation. But our supernova calculation is based upon turbulence and is least as unreliable as the turbulence calculations. How could we justify the support of our sponsors? We argued that our work was an exploratory simulation and was valuable because using calculations of this kind, one could produce a new and interesting scenario for supernova behavior. Scenario is a fraught word, too. It suggests that we're making an argument rather than producing a rigid and reliable scientific result. We could suggest other calculations and observations which might check the scenario. Our scenario was suggestive but not definitive. It is my view that many big simulations in the physical sciences adjust that. They're, in fact, ways of making arguments by waving computers around rather than waving their hands around, but they do make very pretty pictures. Research sponsors, unfortunately, often prize pretty pictures above ideas or scientific understanding. That turns out to be bad for science and bad for public policy formulation. So here's the movie. The movie shows the explosion of a star. There is a hotspot coming. Can we turn down the lights any further? No, I guess not. There's a hotspot coming up. It surfaces at the surface of the star. It then produces stuff which moves across the surface of the star and refocuses back on the other side. Here the process has been stopped and the camera runs around the star or the star runs around the camera. It's a virtual star and a virtual camera, of course. And there we go again. The explosion starts and continues and the camera pans back. And there is a huge explosion which, as I said before, can be seen on the other side of the universe. And the whole thing has, roughly, the same kind of quality as the cartoons that I drew. I thank you. Professor Kadinov, I understand that when you received the American medal for science, you received it from President Clinton. Yes, indeed. Unfortunately, we don't have the Queen today. We don't even have our Prime Minister. But nevertheless, we're delighted that you've agreed to join us to present a fascinating hour for us and to accept from the Institute of Physics the Isaac Newton Medal for Physics. I am honored. Thank you very much indeed. Thank you. I am honored to be even indirectly a part of the great tradition of British and English science. So first of all, the medal. Thank you. And secondly, I'll just read the citation, although you know more about it now than I'm sure that many of us did when we first arrived. The citation for inventing conceptual tools that reveal the deep implications of scaling variance on the behavior of face transitions and dynamical systems. Thank you very much.