About James Hillier
James Hillier is best known for his work on the development of the electron microscope. He began his research as a graduate student at the University of Toronto in the late 1930s. He and Albert Privas developed the magnetic transmission electron microscope in 1938. Hillier was also involved in developing techniques for specimen preparation which allowed broader applications of the electron microscope. Hillier spent most of his career with RCA, where he was also involved in video disk research, and became director of research.
The interview covers Hillier's early work in the development of the electron microscope as a graduate student and with RCA, which he joined in 1940. In 1941 Hillier developed the first scanning electron microscope in the United States. Much of Hillier's work in electron microscopy during the years 1940-1953 focused on expanding the ways to make specimens usable in electron microscopes. He also developed the procedure that allowed magnetic lenses to be perfectly symmetrical. Hillier discusses the importance of relating technology to the economy at large, as well as technology's role in addressing broad social problems. He is particularly sensitive to the often conflicting, yet complementary relationship between research goals and business goals. He also discusses his management philosophy, defined in part during his later career in administrative research at RCA.
About the Interview
JAMES HILLIER: An Interview Conducted by Mark Heyer and Al Pinsky, IEEE History Center, 16 July 1975
Interview # 029 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc.
This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.
Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, IEEE History Center at Stevens Institute of Technology, Castle Point on Hudson, Hoboken, NJ 07030 USA or email@example.com. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.
It is recommended that this oral history be cited as follows:
James Hillier, an oral history conducted in 1975 by Mark Heyer and Al Pinsky, IEEE History Center, Hoboken, NJ, USA.
Interview: James Hillier
Interviewer: Mark Heyer and Al Pinsky
Place: Princeton, New Jersey
Date: July 16, 1975
Family and Education
Heyer and Pinsky:
Why don't you start off by telling us something about where you went to school? How you got into engineering?
Alright. I was born in Bradford, Ontario on August 22, 1915. Bradford is a little industrial town in Southern Ontario. I went through elementary and high school in Bradford and at that time I was very much interested in becoming a commercial artist. My family had come from artists, writers and musicians, but my father was an engineer in spite of that. That was by necessity rather than choice. I think he was really an artist. I had started out to be a commercial artist, but a geography teacher I had in high school convinced me I ought to go the university and think about science. He got me interested in amateur radio. I had an amateur radio license in high school, and finally he persuaded me to apply for a fellowship at the University of Toronto in science.
Lo and behold, I got one, so I went to the University of Toronto. There, I was in the mathematics and physics course, which in Canada was called an honors course. You had to have a special level of grades in order to had gain entry to it. It was a normal four years of undergraduate work, though I must say that in my senior year, even in those days they had a flexible program, so that we were allowed to try all different sorts of research, more as technician assistants to graduate students, but we did get outside of the normal course work and laboratory work. Then came graduation and the question arose as to what I was going to do after that. I really had two choices: to go on to graduate school or to become a high school teacher. Becoming a high school teacher at the time didn't enthuse me, so I explored the possibilities of graduate work.
Perhaps I should explain that, at that time, Toronto was well known for its work in low-temperature physics. They had a helium machine there at that time, one of the earliest in the world. They were also experts in the early work in nuclear physics and spectroscopy. Somehow or other, though I wasn't particularly enthused with any of those at the time, partly because Dr. [John C.] McClennan, who had been the director of the physics department in Toronto for many years, had retired a few years before and the place was coasting at that point.
There was a professor there by the name of [Henry J. C.] Ireton, who talked to me to one day and said, "What are you going to do as a graduate student?" He just assumed that I was going to be one. He started talking about the electron microscope. I had never heard of such an animal, so it intrigued me. I guess within hours I decided to look at the electron microscopes as a career. I should explain, at that time, the concept of the electron microscope was really quite old. The origins went way back into the last century, and even the basic principles in which it could conceivably be built had it been thought through and some work in Germany had been going on in a very primitive way to try and implement these thoughts. All of this greatly intrigued me.
So, in a sense, my life was a matter of chance, of getting to where I was at that point. As a result, right away I became very interested in the electron microscope work. I started reading everything I could find on it and by the fall I was joined by another graduate student by the name of Albert Prebus, who came from the University of Alberta. The two of us started out as neophytes to develop electron microscopes or find out about them and see what we could do to make them work.
I should also mention that we had been preceded by Cecil Hall, who was a graduate student who had started on a similar project a couple years ahead of us. He had built two very primitive, preliminary models of electron microscopes, which he had in large part copied from German work. Those were standing there not working at the time. So when Privas came and we went to work, we really took Hall's two instruments and put them into working condition to see what they would do. These were what they call an "emission-type" electron microscope. In other words they just looked at the electron emission coming from a cathode. Not a very exciting subject unless you were interested in studying electron cathodes at the time. The magnification was very low, 1,000 or so; the resolution was even lower than you would get with a light microscope. So they didn't hold much promise.
High-Voltage and Magnetic Transmission Electron Microscopes
- Audio File
- MP3 Audio
Play (029 - hillier - clip 1.mp3)
When we got to the point where we could produce pictures at high resolution and with fair reliability, we made this public. We were bombarded — this is the best way to put it — by requests from all the medical people and all the colloidal physics people — everybody that had things too small to be seen in a light microscope. We quickly realized that two graduate students working in a makeshift lab with a makeshift instrument couldn't satisfy the needs for this tool. So we started exploring the possibilities of getting it into production. We even considered the possibility of going into business for ourselves. We were entrepreneurs of the Route 128 type even in those days.
Thank God, we gave that idea up, real quickly because the amount of the research that had to be done before this would become a truly commercial instrument was enormous, the amount of the investment it would take was enormous, and the payoff time was a long ways in the future. Very logically, Prebus and I started looking for a big company that would be financially able to support this type of research. It was a pretty speculative type of undertaking. We had knew that Dr. [Vladimir] Zworykin had been to Europe and had actually hired a man, , to set up a research organization on electron microscopy at the RCA [Radio Corporation of America] plant in Camden. We knew that GE [General Electric Company] had some interest in electron microscopy.
So we decided to send letters to see if they were interested in hiring us to do this. We were just graduate students looking for a job at the time, but we had a special skill that we thought might give us an edge. We did not look much in Canada. The Canadians get very annoyed at me for having left Canada. I'm part of the brain drain, I guess. But our logic was very simple, namely that the companies in Canada simply didn't have the kind of money to spend on this type of project. Moreover, it didn't really matter who did it, as long as it was done somewhere in the world, because the value of it would be attributed to anybody who wanted to use it. So, the result of all this was we immediately got replies from GE and RCA.
RCA was a little slow in replying, I remember, but the net result was that I was interviewed by both the GE people and RCA people. Finally, I came down to this country for an interview and, interestingly enough, the reason I came to RCA was because Zworykin asked me a question that none of the others did. He said, "How soon can you make one of those things?" He did not talk much about anything else. A lot of the other companies, including GE, told me about their retirement plans and their beautiful new laboratories and not much about electron microscopy, but Zworykin showed me an old abandoned factory building. It was really a decrepit place. He asked me how soon can I build a microscope.
So, I came to RCA. The whole objective at that point was to create instruments not that could be mass-produced, but produced in enough volume so that they could be sold to other people and were simple enough to operate that they could be operated by a good technician. As a result, I did come to RCA in February 1940, where I was put together with a mechanical designer and an electrical designer, given a small shop, and told to go ahead.
By July 4th of 1940 we had our first prototype together and operating. It turned out that it cost the company $10,000 for us to do that. It also turned out that Zworykin didn't have the budget for it. He had counted on nobody catching up with him until after we had done the job, which is an interesting sideline on the whole situation. The eventual way he balanced his budget, finally, was that he got permission to sell the prototype to a company, got paid for it, and got his money back. So he came out whole.
It's interesting to note that the same project today would probably cost two million dollars to do. It would be done under much more constrained conditions because of things like safety and all the things that we have to put a lot of attention on today. They were considered in those days — we didn't do anything that was unsafe — but we didn't put all the great elaborate bureaucracy and organization into doing it, so that things were less expensive, relatively speaking, in those days anyway.
So, that got us started. We were certainly the very first to have an electron microscope on the market in this country. Meanwhile, in Germany the Siemens company had hired Ernst Ruska, who was one of the pioneers over there, and they designed a commercial instrument. We were never really sure who delivered the first so-called "commercial" instrument. We recognized that the German effort was slowed down to a certain extent because by now they were well into the war. Also by then the communications were very definitely a little more sparse.
Having got the electron microscope to the point to where we were now getting ready to put it into production, I devoted myself to some of the other aspects. There were several of them: one was to look at higher-voltage electron microscopes. We were working with 50,000 volts on these early instruments. It looked as though 200- to 300-kilovolt electron microscopes might have certain advantages. It also turned out that they had a lot of extra difficulties too, so they were a little bit slower in coming.
Scanning Electron Microscope Mode
There also appeared to be the possibility of using an electron microscope in a different mode, which we called a "scanning electron microscope mode" at that time. While I think very few people have recognized it, a group that I had — Dick [Richard F.] Snyder, Les [Leslie] Flory, and myself — put together the first scanning electron microscope ever made in this country, if not in the world. I've never confirmed the world part, but I know it was the first one here. It was in early 1941, which was very early. Most of the pictures you see in the papers nowadays taken with an electron microscope are scanning electron microscope pictures at relatively low magnification because they always have a nice three-dimensional look that makes them attractive.
At the same time, I got back to worrying about the specimen part of the electron microscope. We had a scheme then which we would use for many years of attracting people from universities, medical schools, and other companies, bringing them in, showing them what the electron microscope could do, usually encouraging them to bring a specimen with them, and then struggling to find out how you had to prepare their specimen in the right way so it would give them the best picture — or give them a picture at all. Through that I began to find that it was much easier, at least for me, to learn some of the language and some of the principles of their science rather than try to teach them all the intricacies and the oddball things of an electron microscope. I found — particularly with biologists — they simply couldn't comprehend this mysterious pile of equipment. I found that very interesting because it enabled me to become quite familiar with other sciences and pull these things all together.
Wendell M. Stanley
One exciting event in about 1941  was when Wendell [M.] Stanley, who was at that time working at the Rockefeller Institute, which was located in Princeton here at the time, came to Camden, where we were located, with a bottle of tobacco mosaic virus. Here was a "filterable" virus, as it was called, completely invisible to the optical microscope. But by many techniques, such as centrifuging and X-ray diffraction, he had been able to get a fairly accurate estimate of the size of this virus. He was a little bit surprised to have it always come out that it had to be a rather thin rod-shaped affair.
I remember the excitement when he walked in with this vial of virus. We were reasonably familiar with this sort of specimen preparation, so we prepared a specimen and put it in the electron microscope and showed him a picture within an hour. He was able to put a ruler to it and said, "My gosh, that's exactly the same size I said it would be." He later got the Nobel Prize for this work. These were some of the thrills that we got. At that time, there were places the electron microscope didn't seem as though it could be used. While we built the scanning microscope it still wasn't a very practical instrument at that time.
One place where the transmission microscope didn't seem to be useful was for metallurgy. But as usual, necessity is the mother of invention. People said, "Collodion films are nice and transparent to the electron microscope, so why don't we make a replica in the collodion film of the metallic surface?" It was found possible to put various plastics on a properly polished and etched metal surface and take it off after it dried or hardened. You had it very, very precise, right down to the molecular dimensions, a replica of the surface of the metal.
Then somebody else had a bright idea and said, "Metals tend to be opaque to electrons, so why don't we evaporate metal from an angle on this replicated surface?" You got the effect of sunlight illuminating a mountain range from an angle, and a three-dimensional picture snapped right out. These were things that just came along one right after another, to solve the problems.
Viewing Animal and Plant Material
The one that really gave us the most trouble was the hope of looking into regular animal or plant tissue because the normal microtome sections that people have been making for years for the light microscope were a micron or more in thickness. This was much too thick for our electron microscope: they would get burned up. When we did our calculations, we found out we had to make the sections a tenth of that, or even a fiftieth of that, in thickness, but nobody knew how to make a knife, a microtome, or anything else that would cut sections that thin. We also found out as we got into it, that even if we had been able to cut sections that thin, it wouldn't have been any good because the techniques of preparation had all preserved the structure only down to the limit of the light microscope. Below that it was badly distorted. So for several years we went through a lot of developments that finally got us to the point where we could literally make a hundredth-of-a-micron sections of any kind of tissue, to the point where we could take a bacterium and make about 20 or 30 sections for one germ.
There, too, we learned many things. It's hard to say who did what; we were all contributing in a very broad, cooperative sort of way. One fellow found out that, if you broke a piece of glass just the right way, it made a beautiful knife for this. Another one of us found out that by impregnating the tissue with the right kind of plastic you could make it hard, so you could preserve the finest structures right down to molecular dimensions. Another person learned that you can infuse it with osmic acid or something like that, which, put in a very dense metal, would stain the material so you could see with it better with the electron microscope. And so on.
Through that whole period from 1940 to the time when I really stopped electron microscopy in 1953, we were greatly expanding the ways of making these specimens so that they revealed what people were looking for in them. This meant that I worked with the Sloan-Kettering Institute for Cancer Research, biology departments at Princeton and Penn, and up at Woods Hole in the summer. We also worked with a lot of the chemical companies on how to look at pigments and asbestos fibers. A great deal of the things that we started back in those days are now just routine techniques.
At the same time we were continually improving the microscope itself. Dr. [Edward G.] Ramberg — a theoretical man — was working here. He and I were very interested in really improving the images. We had noticed several years earlier that when you went through focus there were fringes and all sorts of peculiar image effects in the very finest of the image details. These were not symmetrical; they tended to be astigmatic. Slowly but surely we recognized the fact that our very accurately made magnetic lenses were astigmatic. We did not know how to make them sufficiently spherical to work properly.
Solution to Astigmatic Problem and Other Improvements
I guess one of my major contributions was made one night when I was half asleep, thinking, "How the heck can you stretch a magnetic field to straighten it up, to make it really circular, instead of slightly elliptical?" The idea of putting some soft iron screws out a ways from the lens gap to do that came to me, and the next day I was here and we got some welding rod, which was the softest iron we could find. We got the machinists to thread some holes through one of the lenses, and within a few hours we had almost quadrupled the resolving power of our microscopes. It was just one of those creativity type of things. We had all the background that we needed to solve the problem, literally for years, but all of a sudden a click came and the whole answer came out. To this day there is not an electron microscope made that doesn't have an "astigmator", as they called it then. They use magnetics, electrostatics, but the idea is always the same. By a certain criterion of what you see on the screen, you can pull the lens so it's perfectly symmetrical rather than slightly squashed.
To give you some idea of the scale of things, a light microscope using visible light will resolve somewhere around a quarter of a micrometer, 2500 angstrom units. In our early work in Toronto, we aimed at 100 angstrom units — in other words, an improvement of a factor of 25 over the light microscope was the goal. When we achieved that, we had champagne and cheered. In 1945, when we did this astigmatic correction of the lens, we went down to 10 angstroms, ten times better than that. We had been kicking down below 100 in the 40 and 50 range and overnight we went down to 10. It was not only that we went down to 10 as a limit. We were able to reproduce 10 on almost every picture. That was a much more important gain. Before that we were frittering around between 40 and 200, and you never knew where you would come out.
Let me give a little bit of philosophy. The electron microscope had many, many technological problems in it. Essentially, they all had exactly the same end result — they blurred the picture. You could look very carefully to see if there were any differences in the blurring and you could see some differences. You would have a theory; you would take out a few defects. But when you were taking out one or two defects out of several hundred, you didn't see much difference. By 1945 we had gotten rid of most of them. We had three or four left, this improvement came along, and all of a sudden we only had a couple left. It was the easiest thing in the world to clean up those final ones because they were isolated and were easy to work with. But when you had 100 problems, taking out one or two didn't make really that much difference, and you didn't know whether you had done anything most of the time. It was an interesting thing, how, over a long time, nothing happened and then all of sudden you get rid of all the problems.
Of course, since then, with better vacuum pumps, better understanding of electron optics, improvement of the lenses, better regulation of the power supplies, and technical developments in a number of other electronic fields which we pulled into the electron microscope as quickly as possible, much progress was made. By the late 1950s, people were getting down into the 2 and 3 angstrom range, which is almost 1,000 times better than the light microscope. By this point you are getting down to, I think, some pretty fundamental limits.
Princeton, Westinghouse, and Back to RCA
I had a lot of fun as you can see, working with a great variety of problems. I suddenly woke up to the fact, in the early 1950s, 1951, 1952, and 1953, what I was doing was in an unofficial way directing a peculiarly dispersed research program. I was really directing it, and I had a choice at that time in my career as to whether or not I would go into a hospital, which was one of the possibilities, and really push the electron microscope into the medical applications, or whether I went "legitimate" and became a research administrator. Both of those alternatives gave me problems because I was still working for RCA. In the meantime we moved to Princeton, but I was sort of a separate entity, a law unto myself. I had a little cubbyhole down here. I was really selling, designing, and doing everything with electron microscopes. But most of RCA at the time was very much interested in television, so I was, in a sense, the odd ball.
I decided I wouldn't go into the medical field because I didn't have a medical degree and the status implications of that would be terrific — they still are. I would have become a technician sooner or later. I saw I was really directing research, but going along that route in RCA didn't really have many doors open. So, I left RCA and went to Westinghouse Air Brake, actually to one of their subsidiaries known as Melpar. Very quickly I was made the director of research for the central research lab for Westinghouse Air Brake.
I won't go into the details of that, but it was interesting because as a central research lab they were completely naive, and everything they did was completely wrong. I was completely naive, and I suspect a lot of what I did was completely wrong. But in any case I got the best education about research management in one year that I could have possibly got anywhere in the world because it was a place where everything exploded four times a day. To put a central research lab for a big diversified corporation in a small sub-department of a specialized subsidiary is not how to run a research lab. Then to put a neophyte like me running it....
I did such a good job of liquidating that lab that they wanted me to go up to Pittsburgh and take over another problem laboratory they had — a much bigger one — but instead I explored the possibility of coming back to RCA. I was called back to RCA by Elmer Engstrom, who was head of the labs at the time, and made an administrative engineer on what we called the Data Corporate Staff. I got very much involved in the long-range programs of RCA, primarily in the computer business, and in a lot of other things.
From there on, my story is really of switching around. I went down as chief engineer of our Industrial Products Division, where we made mobile radios and broadcast equipment. I was chief engineer of that group for just over a year. I was just beginning to pull that thing together when, all of a sudden, I got a call to come back to the laboratories here.
In 1957 I ended up as the general manager of the Princeton laboratories, which now put me in the position of really directing the research programs for a central research laboratory for all the diversified activities of RCA. I spent almost eleven years at that job, and I suppose it will take a long history to know whether I did a good job or a bad job, because running a central research lab that had a lot of basic fundamental research in it and quite a lot of applied research, and doing that at a time when the government was pouring money into electronics research, when the economy and color television were coming along — all of that was very exciting and very difficult. The best you could say is that you feel your way through a situation like that.
I had a philosophy of management, which was, "Take good minds, give them broad experience, and give them a certain amount of challenge but not too much direction." In other words give them a challenge as to what you want to happen, but don't tell them how they can achieve it. Give them some challenge on time, but don't be too rigid on it. There was a great deal of balance between the flexibility, the amount of direction, and the authority versus complete freedom. I happen to think that we balanced it out right because out of that came, for instance, the work on the video disc, which has every promise of being the next big consumer product for electronics. So we must have done something right through that period. A lot of things came along in that time, though not always as fast as we hoped. But our line of solid-state circuits, integrated circuits, which has given RCA a leadership role in that field, came out of the laboratory during that period.
Since 1968, I've still held that responsibility, though I delegate primarily to Dr. [William M.] Webster. Since that time my job has mainly been to relate the engineering programs to the research programs, and all of it to the different businesses of RCA. That's an area that is very, very difficult to completely understand. It's all in the details because there is a future-shock problem. Something that has become ingrained in my thinking is that technology is not the be-all and end-all of the game; it's just the door opener. It says it can be done technically, but getting it into manufacturing, getting it into production, selling it, developing the market, or recognizing the need for a market — all of these things are just as important to get technology to flow from the laboratory out into the economy because, if it doesn't do that, ultimately it does nobody any good. You are just wasting money and giving some people a nice intellectual hobby.
Integrating Technology with Social Needs
Heyer and Pinsky:
You seem to have an increasing responsibility also in a social area, in integrating technology successfully with social needs.
I think this is right. I think the idea of technology as being as stimulator of new business, a catalyst of new business, has now reached the point of maturity where people realize, as I indicated earlier, that it's the total approach that counts rather than just technology. However, we have developed a society, an environment where we are able to support a much larger population than this earth could have supported 100 years ago, and that is done primarily because of technology: technology of production, technology of transportation, technology of communications.
I think that if we are going to deal with the number of people who are going to appear on this world during the next thirty or forty years before we can get the population to level out, to become more constant, we are going to need even more technology just to make this world operate. Quite apart from any of the intellectual feelings of what technology does to the culture, just to stay alive is going to take technology.
The Need for Technologists
In recent years one of the things I have been stressing and talking about in the universities is that we are going to need technologists. We are going to need them desperately in the next few years, and we are not exactly facing up to that issue at this point. Particularly, four or five years ago, when the aerospace engineers were being put out of work fast, and at the same time people were getting sensitive to the environment and blaming the technologist for that, there was something about a military-industrial complex that they didn't like. So all these kids in high school who were at the decision-making point were turned off in droves. Because the decision to go into technology is really made in high school even today. You can't change your mind after you have entered a university. You would be unlikely to study physics or engineering unless you have been awfully foresighted and happened to be coming in with the right courses.
Everybody knows what happened to the engineering enrollments. That valley in those enrollments has turned up, fortunately, because I think there is more sense and common sense spreading around. But the fact is that we are only in the next year or so going to get the real impact of that sudden shortage of engineers, at a time when the government is going to be pouring out money for energy research. I'm not against that, but I think we ought to keep it in balance a little bit because what they might find is that they price technology out of the market place even though we need it.
There's a nasty bounce there. There are various estimates that people get, but the figures are 20,000 to 40,000 engineers short per year for our needs. The supply and demand still works very rapidly in this business, so that the cost of engineering could go up. It's largely the cost of the manpower. Then you pile on top of that a lot of new funds from the government for energy research. We could find that research is going to be very expensive. This happened once before, when you spent more on research but you didn't get any more research done because that same number of people were all there were to do it.
Heyer and Pinsky:
The whole nature of research since the 1930s and 1940s has really gone through a change.
I'm not sure about that. Unfortunately, the word "research" is almost exclusive to the ear of the listener or the mind of the listener. I have jokingly said that, "Research is automatically defined as the most advanced technical activity that you find in any organization. Whether it's a university or a company, it may be the fellow in the white coat just mixing up mixtures, e.g. colors for the paint factory, but if he's the most technically advanced guy in that factory, sure as heck you'll have 'Research Laboratory' on the door."
There are other definitions, for example the technical activity that you must do in order to survive in any business. But you have to be careful because research is often associated with electronics. Electronics was a fluke because electronics happened to be involved with communications at exactly the time when we needed communications and exactly the time when the government was sponsoring research, and this was a likely area. So we had electronics growing in a great boom situation, and the fact is that electronics today is almost as mature in many senses as the steel industry's research or the textile industry's research, and if it isn't, then it's getting there. We will have to look for some other boomtown in the research area. That's another thing that you have to watch. Research has a connotation that depends on what period of the history of research you are looking at. It could become a very complicated sort of thing to look at, and definitions are the stumbling blocks in many places.
Heyer and Pinsky:
Let's talk about management of research.
Considering the management of research, we have a rule of thumb around here — which has turned out to be a very good rule of thumb, not only for RCA but for the whole country. That is, if you spend a dollar and get successful research out of it, first the chances are you will have had to spend three dollars to get one dollar's worth of good, successful research. Having got the dollar's worth of successful research, you now have to spend somewhere around ten times as much, ten to twelve dollars, to do all the technical things of designing the product, designing the factory, doing all the engineering development, advanced development and design work and process design work. Then you have to spend 100 dollars to actually put it in the economy. That 100 dollars involves the capital investment, new factories, investment in the training of people in the factories, training and the setting up of new marketing strategies and implementing them. All of these things usually ended up costing 100 dollars.
The important thing to remember is there's not much point in spending the one-dollar and certainly no point spending the ten dollars unless you have the hundred dollars, or know where you can get it. I think that is one of the keys of managing research, right in that set of numbers. You have got to look at how many of those hundred dollars you have because that will tell you how much research you really should do, not necessarily how much you can afford to do, but it's how much you should do in an economical situation. When you look at those hundred dollar packages at the far end you have got to look out and see what your business is and what are the true needs in the market place. They might not be recognized yet by the customer, but you, as an entrepreneur, a research man, have got to see what the true needs are of the customer that you can satisfy with your particular factories and your particular technologies.
The chances are that you will never have as many of those hundred dollar packages as you would like to have or think you have. Then you work back to what you do in research. That is contrary to the way the research man in the laboratory thinks, so one of the jobs of research management is to coalesce these two points of view. He's got to persuade the fellow in the lab that this is a more worthwhile route than maybe the one he's taking. On the other hand, he also has to recognize that new bright ideas are going to come from those people. I always call research management a balancing act, all the way through.
Basically it depends on hiring creative people and giving them a stimulating environment in which they stay creative because that's very important. I am one of the people who believes that a man will stay creative throughout his whole life if you provide the right environment for him. I don't believe a bit about this business about peaking out at the age of 30 or 35; and I can prove it from many, many people who I have had over the years. That's a side issue, but it is a subject where I believe our culture, and the society of our technical community, is designed to kill creativity the minute anybody is creative enough to be successful; it's a self-defeating situation. If you persuade a man to step out of that role and start over again in a new technical area or a new market, or something new, and wash out the background, I'll guarantee that he'll become creative all over again, even if he's fifty or sixty. I have seen it happen often enough to be convinced of that. You have got to take creativity and blend that in and keep the people creative. You have got to support them, to give them the kind of technical support they need. You have to keep the right level of communication going so that they know what the objectives of the organization are. You have to keep a balance between their supporting the objectives of the organization while at the same time developing a career for themselves.
It's a whole range of trade-offs and balances, which I think ultimately end up being tailored as it were almost to the individual research worker. I'm a great believer that people are individuals and they should be treated as such. You tailor the situation, the environment, the communications — you tailor everything. It's not a case of misleading them; it's a case of tailoring it into a form which they understand and appreciate and are inspired by.
There are no rules for research management. There are organizations that struggle with setting up rules, or trying to set a handbook for research management, but the fact is, the number of parameters and the number of variables you are dealing with is so great that no organization even approaches any other. We talk about Bell Labs being like RCA labs somewhat, part of IBM being like RCA, RCA being like parts of Westinghouse or GE. Yes, there are similarities, but the minute you get down to real details where you are talking about running the place, controlling it, and managing it, you find out how different these places are. They are completely different and, just as individuals have individual characteristics, so do laboratories. So you get to the point of requiring a great deal of experience. My principle is, "Don't put a fellow in to run a laboratory from outside of the business and preferably not ever outside of the laboratory." I'm a very strong believer that you need the knowledge of the specific environment to be able to manage the laboratory properly. I think we have been very fortunate at RCA that we have been able to do that so far over the years.
Shifts in RCA Research
Heyer and Pinsky:
Was this an example of that working in your previous comments, about when you were being interviewed for various places, and Dr. Zworykin was the only one who said, "How long will it take?"
He knew how to inspire me, challenge me — exactly! I would say I think we have been rather successful around here. We went through some trauma when RCA decided to get out of the computer business. At the time the laboratories were devoting close to 40% of their effort to the computer business; but when we had to trim [staff], we couldn't keep the laboratory doing that. When we were all finished, we'd only trimmed [staff] by something slightly less than 10%. Considering that we were doing 40% of our work for the computer division, I thought this was a pretty good compromise.
Another thing that happens is that the laboratories evolve; they go through a regular evolutionary development. For instance, when I came in here in 1957, the whole objective was to bring a lot of new sciences and technology into electronics. Electronics had been an electrical engineer's game, it was putting components together: transformers, tubes, and capacitors. The transistor was coming along and we were in a whole new realm of physical chemistry, crystallography, that had to do with the phenomena in solids, which we had never dealt with. So we had to hire in people and really set up a whole new basic type of research. Now, that is going over to being engineering, so those same people go over more to the design of things. As I indicated, there was quite an up-and-down change in the government's support of electronics research of the sort we were doing.
When I came here, we were up to 30% of our effort being government supported, and because of the mix that meant certain parts of the laboratory were essentially 100% supported by the government. That gave you very little flexibility and gave you all sorts of up-and-down problems because the contracts ended and the new ones didn't necessarily start. So we cut that back. Then the government helped us by cutting it back further, but this represented another evolutionary change in the period and through all of this the corporation has been very supportive.
We went through another change because in a very definite sense we were doing research for the electronics industry. In the early days RCA had a very liberal policy of licensing all the electronics manufacturers. For various reasons that I never understood, the government didn't like this practice so we stopped doing that. What it meant was that we became more internal looking in our research. Over the years, instead of being sort of an ivory tower looking out in the very long range with only a certain amount of connection to our specific business over the last eighteen years, we have made a complete turn-around so that our research here is very tightly tied to our electronics businesses.
In some respects that has become shorter-termed, and here is an area of flexibility. The budgeteer says, "We can cut off research because we won't feel the effect of that for five years, so we can save money this year." Of course, that's a very foolish thing to do. There's another way of applying research, which is to help where help is needed, particularly when that help is needed in the form of highly sophisticated technical things, in the factory or the design or whatever. So you have a very flexible operation.
Creativity at RCA
Another role research plays, in talking about managing research — a role that I think pays for itself all by itself — is the fact that we have, for instance, in Princeton here, over 300 professionals. Experienced professionals in the electronics business, experienced in every aspect of electronics. I don't know of any place in the world where anybody in RCA can go for consulting on specific problems that's better than this laboratory. The fact that we have about forty engineers a day from other parts of the corporation coming to these laboratories for whatever purpose is an indication of the communications that are going on. Don't forget all that you have to have is one fellow come from one of our divisions and talk to one of our people and save himself a year of work and you have paid for both of their salaries and maybe two or three more besides in that one day's activities.
So, this is another side. What I end up doing is painting a very confused but nevertheless intelligent, intermixed, interrelated, abstract picture of a research laboratory. I refuse, in a way, to use the analogy that the director of a research laboratory is like a symphony director. I don't believe that because the symphony director is trying to make an interpretation on a very set piece of information that was presented to him by some composer. In this group, you are not only composing the music as you go but you are also composing new instruments to play it on as you go — if you want the total analogy all of which gets into a very intermixed, criss-crossed, confused pattern.
Yet, there is a sense to it. Some people do it well and some people don't. I really think that the basic requirement is a knowledge of the people, knowledge of the business, and a broad knowledge of science that goes quite a bit beyond the very specific sciences that you are dealing with. That's why I happen to think I was good. If anybody else agrees with me, I don't know. The fact that I had all this association with all these great variety of scientific disciplines in the electron microscope game gave me the ability to balance these. Coming back to the point on creativity, I am convinced that all the creative ideas are really just a sudden coming together of the right pieces of information in somebody's mind. What you have to put into that mind is all the background that has, within it, those right pieces of information.
If there is one basic characteristic that I found in creative people, it is the breadth of the range of their interest. I have never seen a creative inventor yet who had a very narrow discipline. Once in a while, there are guys who do some things, but when I say creative people, I am talking about the creative people that end up owning a 100 or fifty patents, who are steady contributors over and over again, and come out with novel ideas and keep coming up with them. Those people invariably have a very great breadth of interest and absorb a great breadth of information. That's how they play it off, and suddenly it comes together.
Heyer and Pinsky:
One of the things that has struck me through these interviews was the fact that what the guy studied in school was not necessarily what he did when he began to work.
I think that's right. My studies, for instance, were physics and mathematics. The mathematics that I used quickly became rather rudimentary because I had people like Ramberg who were a lot better at mathematics than I was, so all I had to do was know how to ask me the right question. That was a lot easier than working it out myself.
Similarly, most of the elementary physics I learned is exactly the same courses that they now give as a starter for the engineers. When I go up to Cornell and look at their freshmen and junior years of engineering, where they go on a very broad basis and then specialize earlier, their courses are almost identical to what I got as a physics' course when I was a youngster. That's good background and good thinking, but I use very little of it specifically. Practically everything I used all my life I learned as I went. I had to.
Heyer and Pinsky:
Yes, on almost any particular question, there's always somebody who knows more about it than you.
Designing Information Services
That's right. Besides trying to do what I said earlier about bringing the research and engineering and all the design programs together to ensure they really fit the business objectives of the corporation, I have been working rather specifically with the disk project. What we are trying to do is introduce a new service. It's a new information service — not a player or a videodisc or a record. We are introducing a new information service to the home, which can be used for entertainment, enlightenment, culture, education, or whatever. It will end up being a publishing business. What people are buying is going to be a program, for whatever purposes. Although everybody is marveling at the technology that's in that player, the player is just like a pair of glasses that enables you to read the program.
As a matter of fact, the consumer could not care less what's inside that black box as long as when you put a disk on it and push a button, you get a good picture and none of it costs you an arm and a leg. That's as far as the interest goes. It can have lasers, computers, sapphire styli, or whatever in it. Yet all the people in the technical world are arguing like mad about the various advantages. At this stage of the game that is exactly proper. That'll make it a good player, but in the long run we mustn't forget that you can't go out and sell players unless there are programs to play on them and you can't sell many players if the programs to play on them are not what people want to play on them.
Heyer and Pinsky:
I have been thinking over the last few weeks that the battleground is in syndication of the program material between companies and the experts.
The battleground is going to be on what they are offering and how the culture is going to evolve. There is another thing you have to recognize. I have indicated the labs undergo evolutionary development, so it's a dynamic system. It is a system, an organism that is evolving. The same is true of this disc project, so we offer a new service. People are going to look at it on day one as just a new variant on television, and so the programs they are going to buy in this new variant are going to have some association with existing free television.
What they'll buy five years on could be completely different. They'll find out then how they can use it, how they can get the specific program that they cannot get on free television because the economics deny the possibility of putting it on free television, but they can get this program very cheap. They can own it. They can keep it. They can play it once if they want, or play it 100 times. They can show it to their friends or not. You have in many ways more of the elements of the publishing business. People say, "Who is going to play a video of visual a record more than once?" I say, "Who reads a book more than once?" You still sell books.
Heyer and Pinsky:
You can give it to your friend.
You give it to your friend or put it on the shelf and let people see you've got it. [Chuckles]
But getting a set of old-line organizations to do this is really a challenge, which I consider a research challenge because the technologies that have made this possible have come out of the research area. Until we go all the rest of the way we haven't done anything yet in fact. And we forget. Television was a new information service. It happened to be a home entertainment service, but that's just a specific form of an information service to the home by electronics. What General Sarnoff did for RCA was recognize that he was putting out a new service, that it had a "chicken-and-egg" problem. People won't buy television sets until there are programs there, people won't support programs until there are sets out there. So the programs had to be subsidized. That was the easiest way. You could sell the sets, and when you had enough sets out there other people would support the programs. This is the way you build it.
Sarnoff recognized this. In my view, this was his real contribution, that it was a new total service. He also recognized that you might have to set up a factory to make some very special part of that system on which you would never make any money, but nevertheless you had to have it to make the whole system go. He did all of this, and it was successful. But then an interesting phenomenon happened. People forgot that they were part of a service. The people who made the TV sets became TV set manufacturers. They go to trade associations, where there are other TV set manufacturers. The broadcasters split into two groups: one group that makes equipment for the broadcasters. They go to one meeting and the broadcasters that put on programs go to another meeting. Neither of them ever talk to the set manufacturers. Having got to that, when you want to add a new service do you go to that same group? Heavens no! They are all pigeon-holed, and not one of them has the competence or the resources to do the whole job.
So you have to do it as a corporate job. Keeping it as a corporate job, I guess, has been one of my special jobs for the last five years. Not letting it get loose, thinking that you can stick it here and it'll work, because it won't work. This is still speculative. Any of those existing devices have got their bread-and-butter money to be earned.
Heyer and Pinsky:
A lot of what you have been talking about relates directly to the film industry, which I have been involved in for the last six or eight years. Now you are approaching the film industry here, you are talking about, "What are the exhibitors going to think, the producers going to think?"
They have got a very complicated thing to work through because ultimately there's no theater with a box office as big as a set of homes to which this can play.
Heyer and Pinsky:
At first they will.
Some of them, I already sense, are recognizing that this is a very large box office. Yes, they will hold the priorities over theaters, first in television and then in disk. I'll even make you a small bet that ten years from today, first they will put things on the disk and show them on television after.
Heyer and Pinsky:
It's really the first thing, like radio, telephone, television, movies — the disc is the first thing to come along. However, it won't get you such exposure as the microscope work.
In looking at my own satisfaction of my career, I think I got a bang out of doing the electron microscope work, and got a lot of recognition and credit for it. On the other hand, nurturing the disc from an impossible technical task through the point where it became technically feasible and then a corporate project I think has been much more satisfying to me, but I get much less recognition for it than for the microscope work. The challenge is great.
You know the one thing that people recognize is that the technical challenges are great. We have a great deal of admiration when someone solves a technical challenge, but the challenge of managing the total business is infinitely greater than the technical challenge, almost invariably. I have developed a great deal of respect for the people who are capable of doing that total job. You need the technical people, and they are a special breed and very competent, but sometimes they short-change the difficulty and the challenge of the total activity.
Heyer and Pinsky:
So it's one of the special attributes of a manager of a research division especially that you are constantly faced with new challenges and you have to keep providing yourself with new challenges?
Yes. I'm picking them up from where they come, which is as much from outside as from inside. It's the reconciliation of the two that is often the most profitable thing.
Heyer and Pinsky:
If you care to prognosticate a little bit beyond the disc, do you see some engineering advances, business changes, changes in the business climate, changes in the educational climate?
You have opened up another several hours of discussion. I'm a little bit concerned at that moment about the business environment, primarily because of the tendency that I see on the part of the government to take money away from corporations, away from capital formations, and give it to what I call a non-productive side of the economy. Our productivity in this country, or its rate of growth, has literally been a disgrace over the last ten years, and yet we are still making it worse. It just happens we are falling behind England instead of leading them, and I'm not sure that's a very good recommendation. That one disturbs me, though I do see a number of people beginning to speak up about the need to get some more capital formation.
Going back to research management and my "1-10-100 bit." You really begin to lose faith in spending that one dollar. It sounds like a small amount, and it is. But an absolute quantity gets pretty large, and you can lose faith very quickly because you are spending that one dollar on the speculation that you can spend the hundred dollars to make a business out of it. If the hundred dollars looks as though it is going to be impossible to get, you'll quickly stop spending the one dollar. As a matter as fact, you won't even get close to spending the ten dollars for the design and development phase.
That is going to set us back in our technical development. It's not so much the reality of whether there's going to be capital out there ten years from now, when most of these long-range things you are doing research on are going to come to fruition. It's not so much whether there is going to be capital out there. It's whether the heads of the corporation and the heads of the research labs today believe there's going to be money out there. If they don't believe there is going to be money out there, you will have all the good research labs turn around into very-short-term problem solvers. That could be a disaster in this country. If you would prognosticate there, I think we ought to do something about the image today, and also do something to make sure that the image today doesn't come true ten years down the road. But the image problem is quite serious.
In the technical world, it's very, very dangerous. In fact, it's easy to prognosticate. What I can't prognosticate is how people are going to entrepreneur, react in the market place. These are the things that really determine our development because we have a great many things on the shelf today that we could do.
We keep talking about electronic mail as an example. Electronic mail could have been done ten years ago and done very economically, but here you have many problems that relate to economics, politics, and other things. For instance, what people forget when you send mail from a point to another point and you start a new system, you only have a few points. So if it's electronic I can't talk to anybody except the other people who have got this until the whole system is installed. It's not economical to anybody until the whole system is installed, but the whole system requires one enormous budget and has to be run in parallel to the other system because it has to keep going. So you double the cost of operation.
That's what people stumble over and don't recognize. It's the same with telex. People in companies started putting in telex machines or terminals overseas because they had enough traffic between those two machines. For example, GE would have one in Germany, or somewhere, and one here, with enough traffic to justify those two terminals they put them in. And then, "By God," they thought, "I can join this to a third terminal in GE." So they put it in and now they have a little three-terminal network within GE.
Meanwhile, IBM was doing the same thing, and it developed a five-terminal network, as did Westinghouse and ESSO. By the time you got up to a few hundred companies, you say, "My Gosh! you got telex terminals in 1,000 offices. Let's link them all together and provide an exchange." Then you have a system, but you have got to get to that point. Electronic mail has exactly that problem, but it has this additional problem that you have got to run the whole system right up to capacity before you can turn off the old system. So the costs get up to almost exactly double, and the economics and politics of that aren't going to work out that easily. When you talk about prognostication in these things, you have to look at the whole picture. As I said, electronic mail can be done very easily. We made a calculation almost five years now that showed you could cut the cost of delivering first-class mail in half by using electronic means.
Heyer and Pinsky:
Essentially it was a facsimile system?
It was a facsimile system using electronic circuitry [Homefax]. It was a perfectly workable system that could cut the cost of delivering the mail in half. But when you looked at the capital cost and these other problems, you couldn't get anybody to listen. I think you'll find, coming back to my telex analogy, they're beginning to develop within companies an awful lot of facsimile systems. I wouldn't be a bit surprised to see ultimately the whole thing developed gradually along that route.
In other words, business will do it itself. It will become economical to use these facsimile units between offices. Then they will get switched into a network. As it get switched into a network, the operation becomes still cheaper. Then as you build, it gets cheaper still. Then all of sudden people say, "Why don't we all send our mail that way?" This the gradual way of introducing it. But those things seem to have to happen both by the joining of technology to specific business needs and by having someone join them altogether into a common need.
I think one of the things that we can safely predict, for instance, is by 1980 you will have 50,000 transistors on a small chip, all wired up and operating. 50,000 transistors represents a pretty high-powered computer. What can you do with a high-powered computer that you can buy for five bucks? I would say, "One hell of a lot." Only the imagination of the entrepreneur is going to stop that one. You will find very sophisticated controls on everything. They will do literally what you want them to do, at the push of a button with no complex learning of set-up buttons. If most women are like my wife, one button is fine, if it says "on" and "off." Two buttons confuse her, and three buttons terrify her. But the time is going to come when you will be able to do very complex things with many buttons.
Sophisticated Improvement of Existing Consumer Technology
Heyer and Pinsky:
One of my best projects has been bicycles. I have been thinking a lot about improvements that can be made on the bicycle, which is an ancient design. It's a good one, but it's old and one of the things you could do is make bicycles a lot smarter. I'm not quite sure if it's a mechanical system or an electronic system that could do it. For example, if you could make a bicycle that you could adjust for your particular input torque, however much strength you have got, it would automatically change the gears up and down to adjust to you and the road condition.
There's certainly the control of automobiles for maximum efficiency, control of environment, pollution. Also on airplanes or on transport planes. Controlling your TV set to fit the ambient light and the program that's coming in and giving you a lot of other information. Making the TV set into essentially a visual display for a lot more purposes than just TV. You can make a calculator use a TV screen instead of paper tape. Control your cooking, washing machine, servicing the home, monitoring the home.
All of these things will be done and become commonplace because that package of logic or information will become so inexpensive, people will really learn how to use it in many ways to help us. I have heard lots about management information systems and what computers can do for calling up data. As far as an accounting is concerned, I will agree they can do a miraculous thing. But as far as general management is concerned, they haven't even touched the subject yet. I have fifty projects in the air at any one time. They range all the way from critical problems in RCA to incidental problems, social, cultural, protocol-type problems, to almost anything. Every one of those projects is usually stopped primarily for a lack of some piece of information.
What do you do next? What is the deduction for this? You can't work this out because you are missing this piece of information. I think there is going to be a steady development so that the manager can do more of a problem in series, instead of doing fifty in parallel. Figure how much better it would be if he could do each one to completion, then start another one — and how less wasteful that would be. Because quite often now I get back to a problem, a piece of information comes in so I can take the next step on this problem. I have to spend anywhere from ten minutes to half a day getting into the problem to recall all the subtleties of it. Otherwise, it should have been done before it got to me. If you can do the whole thing in series, work it out and get rid of it. But if you can only cut it down to ten in a series instead of fifty, you would be that much better off.
I'm sure that is going to come with these high-density chips. People have been talking about this for years, and the time has been put off continuously. Ultimately, there will be a big building somewhere (not such a big building in Washington) where you will have a terminal and will literally be able to pick it up and get any bit of published or printed literature or any technical information you want. Even if you weren't able to identify it with computer precision, you will probably be able to get it. Another thing, most of the time I get stopped in a problem because I don't have the information. Sometimes I know it exists and I know where to get it, sometimes I can be sure it exists but I don't know where to get it, and sometimes I can't even be sure that it exists. These are all things that you tackle in different ways; if you start analyzing the way you do your job.
Heyer and Pinsky:
If you start looking at the social implications of more efficient management, it will cost society as a whole much less to go from point A to point B.
Then there is a problem we have in that the speed at which we are coming to crises in our culture and in our economy is accelerating, and as a result we have find quicker ways of solving these problems. We are getting more and more superficial in the solution of our economic problems because the time we have available in which to solve them is becoming less. Instead of our becoming more sophisticated, we have got a future-shock problem that's actually squeezing the time we have and making it more superficial and simplistic. I get disturbed about that.
Heyer and Pinsky:
You can't plan for ten years from now because you don't know what's going to happen next week.
Literally. You have only got between now and next week to plan to react to what's going to happen next week. And with the complexity of the problem, you haven't got enough time. Well, that's rambling a little bit.
Thoughts on Education
Heyer and Pinsky:
Do you have any particular thoughts on education and the changes in the role of education?
Well, I have got some superficial thoughts. I think some of our school system is becoming quickly antiquated, partly because of television. I think we underestimate the amount of education that our children have by the time they get to elementary school today. Yet we keep right on going as though they had none of that.
On the other hand, I also think that our educational system has become a peculiar type of bureaucracy. I got what I thought was a very good education in a public school. It was a fairly good sized one, but it was literally run by a principal, a part-time secretary, and a part-time nurse. There was a teacher in each classroom. That school, today, would have a psychologist and fifty other counselors. It's probably nasty, but I think the salaries we pay for those psychologists and counselors is ridiculous, to the point we are not getting our money's worth. The salaries are so low that anybody with any confidence can go out and earn twice as much, therefore only idiots take these jobs, and therefore we are not getting our money's worth. To fill the jobs with incompetents because our salaries are too low is a very famous mistake to make in the management of any business, including our educational business.
Heyer and Pinsky:
Because the educational system is not working properly those people have to be there.
In many ways, they are making it worse. They get into all the political seniority, unions and organizational problems. That's where their security is rather than in their competence.
Heyer and Pinsky:
That's a classic problem.
"Time Constants of Distance"
Coming back to my management bit, I have always insisted that our salaries be competitive. I have good and not-so-good people in the laboratory, and I know that Bell Labs also has good people and not-so-good people. As long as I have a spread of salaries that largely and realistically represents the spread of people and salaries as good as Bell Labs' or IBM's spread of salaries, then I can compete on an even basis and I'll get people to fit.
However, I know the minute my salaries go below the level, I won't get the good people. It's not because scientists particularly care about money, but it's the only measure that they have. The worst thing that can happen to an organization is, for some reason or another, to make it go whack, where you are losing your good people. You fill the slot with other people, you have the same expense, but you don't have the same research lab.
I have seen that happen to many organizations. People don't recognize when this happens, and then suddenly five years later the organization is obviously going downhill very quickly. Then it takes another five years to turn it around again.
The one good reason for a technical guy to be in management, particularly a physicist or an electrical engineer, is because he has a good sense of what I call "time constants of distance." If at any time, you don't match the time constants of an organization or of a system, you are in trouble. You could never move an organization anywhere in less than five years. You can confuse it in less than five years, but you can't move it anywhere in a definite positive direction. You can move it in five years without any trouble, providing that you plan and work at it deliberately for the five years. If you try to do it in three months and decide that it wasn't right and try to do something else for the next three months and the next three months, all you end up with at the end is exactly what you started with. If it's going in a hole, it's just further in a hole with that. The matter of natural time constants, I think, is another thing that gets built into your management sense. It's pretty hard to define and it's pretty hard to measure, but it's still built into your management sense and that's important.
- Ladislaus S. Marton
- This date may be wrong.
- 1 About James Hillier
- 2 About the Interview
- 3 Copyright Statement
- 4 Interview
- 4.1 Family and Education
- 4.2 Electron Microscope
- 4.3 Princeton, Westinghouse, and Back to RCA
- 4.4 Integrating Technology with Social Needs
- 4.5 The Need for Technologists
- 4.6 Research Management
- 4.7 Shifts in RCA Research
- 4.8 Creativity at RCA
- 4.9 Designing Information Services
- 4.10 Career Satisfaction
- 4.11 Predictions
- 4.12 Thoughts on Education
- 4.13 "Time Constants of Distance"
- 5 Notes