One On One, Part II; No. 8; Using Bob Newhart Comedy to Teach Physics, with Dr. William Bennett

This program is made possible by a grant from Sigma Xi, the Scientific Research Society. I'm Faith Middleton, and this is one-on-one. My guest is Laser Specialist Bill Bennett, the CB Sawyer Professor of Physics at Yale University in Connecticut. He has some candid things to say about government secrecy on campuses, and about how academic discoveries can make some people very rich. Dr. Bennett is best known for making the subjects he teaches interesting to just about everybody. One day, as you'll hear, he decided to use comedy to teach his students about the laws of probability. Back in 1972 or three, I was asked by the Chairman of my department to put together a new, exciting computer course, which no holds were to be barred, and we could talk about any exciting topic that came up under the sun. One of the things I wanted to do was to seduce the students into a discussion of information

theory, which is inherently a dull, boring subject if you do it the standard way. The thought I had was to sneak up on it on the basis of the well-known concept of the monkeys at the typewriter, for example, there had been a popular nightclub act by Bob Newhart, which he talked about this concept in popular terms, and I started the lecture off on that particular subject by playing the Newhart recording, which actually he gave me permission to use. In fact, maybe you'd like to hear that. We can start right there. Okay. Let's listen. In a number of mathematics books, they made reference to something that either proves infinity or the law of probability. They claim that if you take an infinite number of monkeys and an infinite number of typewriters, and you set them down, and they just type away that eventually, given enough time, they would type all of great books. Yes, now they're going to type a lot of gibberish, but eventually they will type all

of great books. If they ever tried this, they would have to hire guys to check the monkeys to see if they were turning out anything to respond. So, I would like to present a day in the life of one of these monomers. I'm going to check post-14, see if he's got anything. Oh, really. I'm going to get to a devil's ever going to write anything in one way. Harry, hold on, post-15 here, or something. I think this is famous or something to be or not to be. That is the gazornin' plan. Anyway, that illusion to the monkeys of the typewriters is certainly a very familiar concept. Why is all this important in a computer course?

Well, probably it isn't. The main importance it has is getting people interested. So, this is a come-on, and that's a come-on, right? The real deadly boring stuff is not far below the surface if you want to follow it up. But it was strictly a come-on, and on the other hand, it's a very successful come-on. It's amazing how many people have picked it up in the computer world in the intervening few years, but the notion that I then went on to was a more contemporary illustration of the concept that was actually due to Kurt Vonnegut. He had published a short story in Collier's magazine around 1950 called Epicac, which he depicts a poor, unfortunate computer programmer working late at night trying to write love letters to his girlfriend and getting nowhere so that, in fact, all he could come up with are things that sounded like they belonged in the Journal of the American Physical Society.

So, sitting there, just worrying about what to do, and finally, he types in a childish numbers for letters code into the computer, A equals 1, B equals 2, C equals 3, E equals 5, A equals 8, N equals 14, so on, T equals 20, goes through the whole alphabet. And presumably, what he has his hero do is write some kind of a program that isn't really discussed in great detail, which makes use of a random number generator in the computer, that then starts coming out with pros. In fact, it starts writing love poetry, apparently very good stuff, and he sends it to the girl, and the girl likes it a lot, and everything is going great until it turns out that the computer really wants to marry the girl and then has a nervous breakdown and burns up and blows excuses. Going on from there, what I did was illustrate what happens if you type in a childish numbers for letters code into your everyday home, B, C, or other computers that are available. And indeed, although there's a certain finite probability that you might get something as

exciting as the new heart result, even taking the first few words to be or not to be, that is the, say, up to the word gazornian plots and stopping before that. If you were to estimate how long it would take a good monkey to come up with this result in real life, assuming it was typing at something like 10 characters a second, it would be something the order of 10 to the 36 years, number far in excess of the age of the universe. Anyway, it was quite remarkable that Bob Newhart got that incredible result so quickly. Well, why have people been fascinated with this question throughout time, it seems? Well, I think fascination shows up in all different levels and students, the notion that there might be a way of knocking off term papers without a lot of fussing the order of stuff that's sort of intriguing. I think the other extreme, there are people who really wonder if maybe the human mind somehow works this way and some rudimentary form.

Could it be that great creative works of literature and art, music, and so on are strictly the work of some sort of random process and, of course, we all believe that human beings are somehow really needed for that enterprise, but nevertheless, I think that's one of the basic underlying questions of popularity. What is all this probability? Yeah, it's a kind of conditional probability theory, which is the underlying basis of information theory, and therefore makes a very good background tightly of problem to introduce an unfamiliar subject. If we're not scientists, we might be, but if we're not scientists, how do you explain what information theory is and why do you care about it so much? Well, the main underlying questions involved are the problems of transmitting information really, and the subject deals with the efficiency in which information can be transmitted.

If one analyzes the underlying statistical properties of it, it turns out that, for example, there are ways of increasing the efficiency of transmission of language. One could imagine, for example, this in practical terms, if you thought of it in terms of the phone company wanting to sell long distance telephone transmission to its subscribers, you could ask, you know, what sort of bandwidth you really need to transmit the human voice and typical messages, or what sort of bandwidth in a simpler problem do you need to transmit typewriter messages or teletype signals? And one of the things that comes out of this subject is that you find that there is a certain redundancy in a typical language text which can be taken advantage of in a statistical way to improve the efficiency of sending messages. Is this what you spend most of your time studying, applied science? Yeah, lots of different areas of applied science.

My real feel is laser physics, particularly gas lasers and the relationship to atomic molecular physics. So, do you consider laser work your main line? Well, it's certainly the line that I'm best known for in the research world. Is it the thing you like the best? I like lots of different things. It's hard to answer the question, what do you like the best? It's like, who's your favorite composer? I can't answer that because there's so many of them. But I mean, I certainly have very fond of laser physics, but on the other hand, I've been a little disheartened in the direction that laser research has been taking in the last few years. Why? Increasing involvement of the Defense Department directing this research in particular, the big pork barrel in the sky that's currently being opened up, the Star Wars program, and so on.

There are dreadful mistakes that are being made, and I think there's a terrible effect that this emphasis will have on basic research, particularly on the university level. As more and more emphasis is put on the military applications. The secrecy that is involved in one side, the absurdity of it, the other side, the notion that the entire gross national product will probably be going into this field if President Reagan has his way, and so on. What do you mean about the secrecy and science in the university? There is an increasing tendency due to more and more domination by Pentagon-oriented people in various aspects of applied physics and research to insist, for example, that they see articles and approve them before they're published. They have a tendency to want to cancel talks that are being presented at meetings, even though the talks are on subjects that aren't classified per se.

It's not really happening. It's happened enough times in the past few years to make one worried. It has happened, for example, in the computer science field and subjects involving cryptography and things that might be of interest in the national security agency and so on. I'm told that it happens very frequently in the laser field these days, particularly people who are working in the national laboratories that are peripherally, at least connected with research on lasers for military purposes. It's getting very, very difficult for some of those laboratories, I think, to attract outstanding people because of the fact that these people are having trouble publishing and particularly at the junior level in the field, postdocs and the like. Why is the Star Wars plan, in your opinion, a big mistake? There are many reasons, and I think it's a mistake. I think it's going to get in the way of future negotiations with the Russians.

I think the only sane approach in the long run to this subject is increased negotiations and treaties. On the other hand, I don't think it could conceivably work the way these people are hoping it will. When you start asking what does this one need in the way of a defensive screen against ballistic missiles, it seems pretty clear that if only a half a dozen or so missiles were to get through a defense system and blow up half a dozen American cities or half a dozen Russian cities, that this would be an intolerable, hopeless, tragic situation for civilization. You look and ask the question, how many missiles or how many warheads are involved in each side is something like 30,000. Here comes the applied science. Right. All you have to do is do an elementary calculation of percentages and the odds of getting anything that is good to 99.99% reliability are very low in this field, and in the history of all military problems in the past, nothing has ever worked with a sort of reliability

that would be needed to ensure this type of strategic defense plan that the president wants to encourage. At the same time, you can see that the expenses involved in pursuing that program are horrendous. I mean, there's three billion that people are wanting to pour into it right now is just to drop in the bucket. And if you start imagining what will happen over the next decade or so, when that kind of program takes hold, it's clearly going to start approaching a trillion dollars. I mean, so much money that the average person wouldn't be able to visualize it. That, in fact, trillion dollars would be about a hundred dollars each for every person that ever lived in the surface of the earth and the history of the earth, and I don't think it makes economic sense, and I'm particularly concerned that small are not so small, big, special interest groups will get a foothold in this area that will make it very hard

to stop the program. Once it gets going, once you invest three billion dollars or four billion dollars or still more in this program, it won't be possible to stop it. You know, it'll have an inertia, which will keep going and it will just get in the way for any peaceful, of any peaceful, sensible solution along the line of negotiations. By special interest, do you mean laboratories laser company? Yes. Small laser companies, small laboratories, and big laboratories too. That's the military industrial complex that President Eisenhower warned about in his parting address, but taking place with the vengeance. One sees, for example, I mean, I know one recent case, even at Yale, where a junior faculty member was offered a salary that was just astronomical compared to normal university appointments to go work in one of these Star Wars-related industries, and good factor of four or five

has involved. Once those people become committed to this area, of course, you're very hard to change the funding in any way that would be meaningful. I mean, if we get to a point where a substantial fraction of the U.S. population is making his living out of this process, it will be very hard to stop it no matter how you approach the problem. If you've just joined us, we're talking with Bill Bennett, who is a professor of applied physics, among other things at Yale University. What would be, if there, if someone were dangling at a check in front of you, what would be the interesting thing about laser application that you'd like to see advanced? Well, in terms of laser applications, I think there are all kinds of very interesting things that are being exploited at the moment, and unfortunately a lot of them are not being developed by us, but by the Japanese and the Europeans, for example, one of the things that really depresss me greatly is to realize that the Japanese have been the ones who have used lasers

to make compact digital recordings available in the average home, or to make laser video disks available for television, for transferring information, for peaceful purposes, and so on. Did you ever imagine when you were working on lasers at the start that laser disks would come along? Not at the start. I did think of that actually about 10 or 12 years ago, and suggested it to a small laser company and Bethel Kineticett, who told me that I should go back to my ivory tower and that they were busy building surveillance systems for the Air Force. Is that right? You never talked to anybody else about it? Just not that I recall particularly, but it was a kind of foreseeable thing. The magnitude of the development problem was, of course, something that most companies wouldn't have been able to handle, and it's a credit to Sony and Phillips that they were able to carry that out, but the concept is straightforward enough, and once anyone realized that you could reliably focus a beam of light to a wavelength in diameter with great intensity

and turn it on and off at millions or billions of times a second, it became clear that transmission and storage of information that way was relatively straightforward. We have, on the one hand, the development invention of a laser in this country, we've had to development of digital recording techniques, pulse code modulation was something that was also developed at the Bell Laboratories in the 1930s, 1940s, and the whole computer field has been something that was largely an American development. Why is it that all these items have only recently been so successfully put together and exploited as they have been in the digital recording field and television field by people outside of this country? And I think one of the reasons is that everyone here is worried about defense research so much, whereas you go to a country like Japan where they don't do defense research as a result of the treaties that we imposed upon them at the end of World War II. What they do is spend their time and resources developing things that are used to the general

public and to civilization, and that aspect of things rather depressed is me. I think we have ourselves to blame. Well, let's do the flip side. What would be your dream project to work on? Many of these things get taken over by large corporations that have the resources to go through like a snowplow and dominate the field, and one of the problems that one faces while being in a relatively small university department is to try to nibble at an area where the rest of the herd hasn't been grazing recently, and you come up with some exciting new development, and then you have to almost step out of the way as a large number of people can stamp eating towards you. That happened, for example, with the Argonne Islands laser, that was something that we had developed ourselves in my laboratory with some students here at Yale. What is that?

It was the first blue laser that was developed in the visible region. It was a gas laser that ultimately turned out to have very high power capabilities, and in fact is still very much in use, particularly medical applications and laser cutting and welding applications. Well, that's so much welding, but narrow focusing problems where you need the subtlety available from such a well-controlled laser, but in any event on that occasion we were actually doing this with support in the Air Force, but over their modest support at the time. We got this laser to work and sent off a letter to one of the journals, which unfortunately got lost in the males internally at the location of the journal for a few days, and then discovered that two other groups were working on the same problem, one at Hughes and France, all of whom published somewhat the same results at about the same time, but once this discovery had been made, which was for years, the only lasers for a couple of years anyway, the

only laser that had been available in the visible spectrum on a CW basis, capable of this sort of precision where the helium neon lasers, they've read and sometimes in the yellow orange part of the spectrum, but this was the first blue laser of that type, and suddenly it's attracted the interest of the entire laser scientist population, and all the major laboratories began working on it, and the next thing you knew there were hundreds of papers being submitted on the same subject, but what I'm saying is that one can make a small discovery in one area, this field, and then immediately hundreds and hundreds of people start working on it, and you have to have a certain philosophic approach to the problem to understand what might happen. So similarly, the things I'm working on now, I'm sort of nibbling at, and I don't think too many other people are working on them right now, and I don't want to arm my competitors

too much. I tell you exactly what I'm doing. It's not secrecy, but a question of strategy and how one pursues these matters. Well, the times, how long ago did you discover the blue laser thing? That was an early spring of 1964, if I was around March, I think of 1964. Do you find that your thinking has changed or your sense of, I don't know whether I should call it a naivete about this, but does your mind now jump to thinking about patents and money and becoming rich, or having your university become rich? I know there are departments in various universities around the country now that are set up almost like businesses to copy right and patent things, to bring in money for the university. Is that happening more and more? First of all, I'm not really that concerned about money, as long as I have a comfortable stipend to live on. I want to go to a site that I didn't want to even try to become a

millionaire. First of all, I don't have much aptitude at that kind of thing. I tend to buy stock that goes down in value and take a loss and so on, but as long as we can continue to exist comfortably, that isn't a concern. The problem, of course, is that the minute any one of these inventions or new discoveries gets well known, there will immediately be large numbers of other people working on them. In fact, many people will be have been working on them already. There's a tendency for new things to be discovered simultaneously and independently by several different areas. The large industrial research laboratories have huge patent departments. Hundreds and hundreds of lawyers will be put to work on this problem. You don't have a chance against them when you're working on a legal facility is available at a place like, yeah. In the case of the Argonne Analyzer, I actually went around and tried to talk to several people about this. The first reaction I was given

was that, oh, it's an obvious thing. It's too late to patent. Still, another gas lazer, there had already been several lasers starting with the Helium Neon Lazer, which actually I was co-author of the patent. The first advice I got was it was ridiculous to even try to patent it because it wouldn't be patentable because it was just an obvious extension that would be well understood by experts in the field. Well, that turned out not to be the case. Hughes aircraft did indeed go ahead and patent it long after we had published the thing. Of course, they had published their own work on it. Bill Bridges did some excellent work on the problem of Hughes. Indeed, got the first patent on that laser. Let me stab you. That's twice now that you have gone to people and said, hey, let's patent this. I think laser discs might be interesting here. That person said, forget it, go back to the ivory tower. Now you've said to another group, let's patent this and they said, no, we can't handle that and somebody else did. What does that start to do?

Well, one could become paranoid, I suppose. There have been a number of things I suggested that didn't turn out to be very good. I think it's unfairness. Here comes applied science again. One of the things you have to realize is the economic involvement that goes into taking out a patent. I remember once being told by the patent lawyers at Bell Labs that on the average, they spent something like $300,000 for every patent that they got granted because the minute something interesting comes up, there are immediately lots of other people who file suits and litigation gets very involved and testimony goes on for decades in some cases. Well, if you don't want to be rich, what's the point of getting the patent anyway? That's an interesting question. I think scientifically they don't have very much value or a steam that'd be better to have a good publication if you're just thinking of it in terms of brownie points. One of the things that I found was kind of irritating, though, is that if you don't patent that someone else does and you might end up getting in a situation

ultimately where you had to pay royalties when someone else's patent for a device that you discovered, that literally happened at one point with me in the Argonne Island. I was actually consulting with a small company in Bethel, Connecticut that I mentioned before that didn't seem to be interested in laser disc recording, but they were interested in Argonne Island lasers, and indeed they built some, and after they'd been in production for a while and I helped them design some fairly nice Argonne Island lasers for medical purposes in fact, and they immediately started getting bills from his research laboratories inviting them to pay royalties on the patent that had been taken out on something that I thought I had discovered, which I found a little ironic, and I think indeed they did end up paying these royalties, and I guess that's one of the ironies of business, and it's that kind of consideration which makes me feel glad to be in a university and not worrying about this type of economic reality that goes on in a small business world.

Bill Bennett, physics professor, and master of Yale University's Siliman College. One-on-one is a production of Connecticut Public Radio, the series is made possible by a grant from Sigma Zai, the Scientific Research Society. For a cassette copy of the conversation you just heard, call 203-527-0905, or this member station of the Public Radio network. The engineer of one-on-one is J. McDermott, Michelle Press and I co-produce the show. I'm Faith Middleton, thanks for listening.