H. William Koch – Session III

Aaserud:

After the National Bureau of Standards period, you went to the American Institute of Physics to become its director. How did you learn of the position?

Koch:

There was a gentleman by the name of Ralph Sawyer who was chairman of the governing board of the American Institute of Physics. He was also chairman of the National Academy of Science Advisory Committee to the Bureau of Standards, and there were then many such advisory committees. I think that whole concept of using committees designated by the National Academy of Science to advise the Bureau of Standards was originated by Ralph Sawyer. I don't really know that history, but he, at the same time, was also vice president for research at the University of Michigan. At the time the University of Michigan was getting more research money from the federal government than any other university in the country. So he had quite significant responsibilities in science and in the academic community. He had observed the developments at our laboratory in Washington, and then also out in Gaithersburg. When the third director of the American Institute of Physics, Van Zandt Williams, died suddenly of a heart attack after having only been in office for one year — he died in London while he was on AIP business — Ralph Sawyer, officially was designated I believe, acting director of the American Institute of Physics. He then looked for somebody, because it was partly in his own interest to find a replacement for himself, as acting director.

Aaserud:

That was the second time he was acting director.

Koch:

Right. I don't really recall any more of the details, but I think in August or September 1966, on one of his advisory committee trips to the Bureau of Standards, he asked to see me, and asked me if I would be interested in being considered for the job. He said that they had a list as large as maybe 80 people that they had identified as potential candidates, some of whom didn't know that they were potential candidates. Some committee had just put down names. And they were trying to screen those names, and I don't recall if I'm repeating something that was said in the previous interviews, but at the time I was what I regard as still young, 46. My wife and I had these five children who were about to go to college, and I was concerned about how to finance their education. The Bureau of Standards politics looked like it was getting worse and worse because of what was obviously going to be tougher and tougher times to justify the Bureau's role and get funding for the various research programs. So I said, sure, I wouldn't mind being considered. The next thing I knew, I was asked to come to New York for interviews. And then the next thing I learned was that I was invited to come to the governing board meeting in New York, and it was at this governing board meeting that I mentioned earlier Condon who was a member of the board was present. When I was introduced to the board, he said that he had hired me at the Bureau of Standards and he was very pleased and proud to participate in hiring me at the American Institute of Physics. So that's how the whole question of hearing about the job at the American Institute of Physics came about. It was all through Sawyer.

Aaserud:

Were you already active then in looking for other opportunities?

Koch:

No.

Aaserud:

Did you apply at any other place?

Koch:

Oh no.

Aaserud:

No, it was just the AIP.

Koch:

Well, you know, it all happened so fast, and it seemed like a wonderful opportunity. I knew really very little about the American Institute of Physics; as with most physicists, they don't know much about the American Institute of Physics. I knew only about my own professional Society — the American Physical Society.

Aaserud:

So the state of the NBS made your decision easier but it didn't lead to it, so to speak.

Koch:

No. When I learned about the AIP, I really got quite excited about the things that sounded like a real challenge, and it seemed appropriate after my 17 years at the Bureau of Standards. I started there September 19, 1949, and then left on December 26 of 1966. I'm always amused about the fact that I'm no job hopper.

Aaserud:

What was the specific reason for your enthusiasm about the AIP as you came to learn more about it?

Koch:

Well, it looked like a place that could make an important contribution at the national level, from the point of view of administration and management, similar to the important contribution that I had always thought the Bureau of Standards could make at a research level. So it looked promising. I didn't realize at the time the complications that exist in an institution like the American Institute of Physics. It looked so simple to say that there were, at that time, seven member societies of AIP, and it sounds great to say that AIP supplies services to those seven societies — collects their dues and fulfillment, publishes the journals, helps in running meetings and finding jobs. Those were all the things that seemed obvious, and it wasn't so obvious that there would be difficulties in trying to pull all this together and try to keep it together. As Seitz said in at least one editorial in Physics Today, there is a continuous tendency in the physics community to fragment. It's almost the very nature of science and physics, where ideas come up and get started and get to be larger and larger, and then get more and more applied, and then become engineering, so you have to chuck them off. So in that regard, things originally in physics then tend to fragment. And this fragmentation problem, and it is a very real problem, has its counterpart in AIP. The research aspects and the scientific aspects have a counterpart in the management and the administration. There is a tendency to continuously fragment. We can talk more about that.

Aaserud:

So you saw that already then as a natural continuation of your career, so to speak.

Koch:

Oh yes.

Aaserud:

It was from the administration of actual research at one laboratory at NBS to the administration of physics and physicists at AIP.

Koch:

Yes.

Aaserud:

And that was part of your reasoning for choosing.

Koch:

Right. In fact, during the first five or ten years of my being here, people kept on asking me whether I didn't find it really quite different from being at NBS. I think I've remarked already that I didn't really find it too much different, partly because in the kind of research that we did at NBS, you have to build models and you have to try to think about how particles and fields interact, whereas when you're dealing with people, you also have to do the same thing — develop organizational models, and try to see what things work and what things don't work. The fascinating thing at AIP is this question of how physicists communicate. How do you really pass information back and forth? How does an idea get generated and where does it go? How do other people pick up on that idea? How do you make sure that things are reviewed by peers, and that you don't have favoritism? Of course you do have some of that, but by and large the scientific process in physics works extremely well. I think it's the most successful of any field, and that's a fascinating thing. My involvement in AIP has gone from the time that we were heavily involved in information that's in print, and as of now we're going into electronic information. Everything is being re-evaluated in terms of electronic means, and how physicists communicate electronically, rather than by printed journals and printed books. This is a whole new era that's starting now. I'm so pleased that I have had an opportunity to be involved during this whole time period up until now. It's really been an exciting period. So I would say my feeling about and projections for AIP, when I first considered coming here, have been confirmed.

Aaserud:

But of course, there must have been some physicists, then and now, who see physics as the proper thing to do and administration as somewhat —

Koch:

— low grade.

Aaserud:

Yes. In connection with that, did you get any flak from other physicists at the time for choosing the position?

Koch:

No, but one always senses that, though. You know, there is a hierarchy in any field, and in fact there's a hierarchy among the ten member societies of AIP. The feeling is, and I think it's justified, that the American Physical Society is the place where the fundamental work is being done — where the world class physicists are to be found, where those world class physicists become presidents of the Physical Society and so on; whereas the other societies infrequently have people, some would say, of that same caliber. Which isn't true because there can be just as outstanding work done of an applied nature — in optics or acoustics or whatever — as in fundamental physics. But there is a feeling among people generally that the American Physical Society is the place to be. You have to be a member of the Physical Society or you're not a physicist, some would say. And similarly between the research workers and the administrators — administrators have tended to be support people. That in fact has always been my role — supporting things. AIP is a service organization providing service to the ten member societies. And again, we're not quite as high in a pecking order as people doing outstanding research work. I think I mentioned earlier the differentiation between research people and educators, people who teach. Research people are at a higher level, in some people's estimation.

Aaserud:

Yes, you mentioned that.

Koch:

With time, the teachers and the administrators and the managers will get increasing recognition. However, they have had to play a little bit in the background always, and should.

Aaserud:

That's something you had to battle with at various times during the development of the AIP, I'm sure. While we are at the origins of it, was there competition among physicists, do you think, for the directorship of the AIP?

Koch:

Oh yes. It was amusing to me. I didn't know anything about the competition, but it was amusing to me later to be told that so and so really had been competing for my job, and was very disappointed that I was picked.

Aaserud:

We needn't mention names in that connection. But Van Zandt Williams was not a physicist, correct?

Koch:

Yes, he was.

Aaserud:

He was? I'm sorry.

Koch:

He was a physicist who had worked at the Perkin-Elmer Company, and he had been a president of the Optical Society, and he was an opticist.

Aaserud:

Yes, not a member of the APS but a physicist nonetheless.

Koch:

That's right. He was a member of the Optical Society. And I think he was highly regarded.

Aaserud:

At the time that you accepted the position, how specific were your goals for AIP? Did you have any specific ideas on what you wanted to do?

Koch:

Yes, it was to provide physicists and physics societies with improved support services. The first thing that was done during the first year I was here was to start the NSF grant program. And that was in the area of science information that was one of those lower on the pecking order of fields — famous physicists don't worry about how they communicate. As an example, I remember one time I was in Yugoslavia at an international conference. I had the opportunity of taking a walk in the woods with Maier-Leibnitz — a very famous physicist from Europe. He and I started a conversation about our involvements. I was very much involved in science information, trying to understand how physicists communicate, and I said, "How do you do your research? How do you work with your students?" And of course he was a European Herr Professor. He said, "I don't have any problem of keeping informed." This information explosion — he didn't understand it at all. He said what he did was to take the most prominent journals of the world, and assign them — this to this student, this to this student, this to this student. Those students read them and he held a weekly staff meeting with them, and he said, "Well, does anybody have any new ideas, anything that you've been reading that's new and exciting?" And this student comes up with this and says, "You should be sure to read this, look here, and you should be sure to read this, and look at this calculation — we could do experiments based on that calculation." He said, "I don't have a problem keeping up with the world." Well, not everybody's a Herr Professor, and not everybody has the same resources. And soon after we were doing our research at AIP. We shouldn't really refer to it as research, because I've been told over and over again that AIP is an administrative organization, and we don't do research here. But we felt we were doing research in the same way we thought we were doing research at the Bureau of Standards. A good example of this is an investigator at MIT, Tom Allen, who tried to understand how engineers communicate said it was really an important phenomenon to him that in an industrial company with a large number of engineers there was always one person that read all the journals, read all the newspapers, and was always fully up to date. He, then, was the prodder. He would go around and see what research others were doing, and he would heckle. I thought the same thing was true at the Bureau of Standards. We had a Ugo Fano and then later Mike Danos who were just that kind of person. Now, this engineer at MIT called them gate-keepers. The gate-keeper phenomenon is a very important one. Certainly in physics. I think it continues to be an important one. And if one can somehow take care of the mechanical aspects of this gate-keeping, that would be a very important role for AIP to play; to work with different levels of people, and see how to get information efficiently and effectively to those people. At the time that we started our work here, most of the outstanding research work in physics was being done in the United States. With time, other countries have been putting bigger and bigger investments into physics research, doing more and more outstanding work. It is now obvious that if US physicists are not apprised of what's going on outside, they will soon be left in the dust, not able to keep up with the leaders in world class physics. In our research with this NSF money, it became evident that the AIP had gotten to do more and more work. AIP had started with NSF funding, in 1954 to 1957, translating Soviet literature, when it became obvious that we had to translate other important literatures. Now the important literature that we're focusing on is Japanese, and so few physicists understand what's going on in Japan. We think we know ways of making more of that body of literature available here, and trying to provide cross-fertilization and to provide our literature to the Japanese more effectively. That has been our approach, so in those few words you can see the importance of science information, and the whole subject of communication among scientists. In those words you can also appreciate why, when I go to talk to NSF on the 28th of October, I'm going to make an important pitch for the fact that the research divisions of the National Science Foundation must increasingly concern themselves with supporting research and development that make it possible for US physicists to keep up with the rest of the world literature in physics.

Aaserud:

But this has been just as much a problem for the AIP convincing the physicists as vice versa?

Koch:

That's right.

Aaserud:

And that brings me to one possible distinction between two activities of the AIP, which is, providing services, and doing independent research as you might call it. The services are what it was established for, I suppose.

Koch:

That's right.

Aaserud:

It was intended as kind of an umbrella organization to deal with publications and things like that. And then as the AIP developed and became its own institution, then it also developed more independent tasks.

Koch:

Yes.

Aaserud:

Do you think it makes sense to distinguish between those two aspects?

Koch:

Yes. Very much so, because an interaction with another government agency, the Internal Revenue Service, has demonstrated that it was important that AIP develop a role that's separate and distinct from the role of service to its societies. But before I jump into that, let me just backtrack a little bit, and be sure to emphasize some earlier aspects of this distinction. The AIP started publishing Physics Today in about 1947.

Aaserud:

1948.

Koch:

1948. And it was an AIP vehicle that was clearly identified with AIP and that was available to the community of scientists who were members of our member societies. In the late fifties this same concept of trying to work as an independent organization to serve the community at large was promoted by Elmer Hutchison, who had just become the second director of AIP. Elmer, having been a dean at Case Institute of Technology, now Case Western Reserve, was very academically oriented. He saw the potential of a variety of programs that he started — in education and manpower as a division and another in history. These were some of the programs that he initiated. He was accused of not having given enough attention to the bread and butter jobs of AIP; one such job was a disastrous experience with handling dues and subscription billing during his tenure here.

Aaserud:

Accused by the governing board or by others?

Koch:

By the governing board and various committees, and I've never probed very much into that, but my interpretation of what happened was that when you provide the societies with a good excuse — and it was a good excuse, it was a disaster and it should not have been allowed to happen — then they'll give you a hard time. But I don't know whether Elmer could have done anything to have prevented it. It may have been one of those things that was bound to happen. But I think he was told that, "Look, you've been spending too much time on these things that you feel very comfortable with — education and manpower and history and so on, these academic things. You get back to your knitting." And he was not able to do much in the way of correcting the problems, and I believe he was encouraged to leave AIP. Well, you can look at the analogy of my experiences. During the first five years, the NSF supported AIP to the tune of a million dollars a year, and this was in the area of science information. This is this lower level subject — not as important as science or the publishing of science. Here, the director of AIP was spending a lot of time, in fact, responding to the wishes and desires of the government agency that is supplying all this money. And we also had dues and subscription fulfillment problems. And there were some real task masters in these societies. When I first came here, Shirley Quimby was the treasurer of the American Physical Society, and he was followed by Joe Burton who was very similar to him. They were very hard task masters, who, when there was a legitimate excuse, would give you a hard time. These things should not be allowed to happen, but many times they do. Fortunately for me, we did develop solutions. I promised solutions and we got the solutions, and there was one experience that I'll tell you about in some detail, maybe...

Koch:

...of too closely following the analogue of Elmer Hutchinson — of being diverted, in my case, to science information activity, when the bread and butter handling of dues and subscriptions and publishing of journals was less than perfect. For example, there is the question of keeping journals on schedule. I think I mentioned to you, in an earlier session, that Pasternak — once a very dedicated editor of The American Physical Society — caused a major problem for AIP, in that he would come with briefcases loaded with manuscripts. He had been working on them for weeks, and he would finish working on a batch over a weekend. He would come in Monday morning, put them on the desk of our publication division, and say, "Get these done by Friday." We couldn't possibly get them done. They also were completely focused on authors, and they would hold up a journal issue for one author's manuscript because they had promised that author he would be in a particular issue of Physical Review. You couldn't possibly get the thing done. So I arranged a meeting in our Board room with Bill Havens of APS, who as far as I'm concerned was always very fair about this. We'd bring the Physical Review editors in. I said, "Who do you think these journals are being produced for?" And one of the editors said, "For the authors." I said, "No. They're being produced for the readers. And we have to maintain schedules. If you load us with manuscripts, if you hold up an issue for one author, how can we maintain schedules at AIP?" That wasn't their problem, that was our problem, they responded. But when the journals were late, that was everybody's problem. And the solution had to involve cooperation from the editors. So I'm saying, that's an example of problems that are bound to happen, if you continue to work with a given system. What you've got to do is figure out how to change that system in order to avoid problems before they occur. As Bill Havens has told me recently, "A prominent characteristic of a director of the American Institute of Physics is to have a thick skin," and he said I had a thick skin. I must therefore have had a thicker skin than Elmer Hutchison had, in order to be able to get through my periods of problems. For we did look for solutions, and we found them. As a result, we now, I believe, have an outstanding dues and subscription handling system, and we're going to make a major improvement in the next six months or so, going to the next generation of systems. In the journals, we're within two days of every journal being on schedule. And that's a major accomplishment, in view of the fact that we've had problems, when we first went to Woodbury, in our new facility. But those things are bound to happen in any assemblage of people. Something gets out of kilter and you have to then figure out how to correct it. Well, we were able to find out how to correct it.

Aaserud:

But that was not the result of the AIP not seeing through its service, though it might have been perceived as that.

Koch:

Well, it was perceived as my spending too much time on the NSF program, and not spending enough time trying to figure out how to keep all these things on track.

Aaserud:

To what extent were you made aware of that, during the early period? To what extent did the governing board really try to pin you down on this?

Koch:

Well, at one time a committee of society treasurers and secretaries had a special meeting, and it was when we converted from the Univac 9300. The Univac 9300 was a tape oriented machine, and the technology using discs had come in during the time we had that machine. Univac had a Univac 9300 that was a disc oriented machine, and we converted the programs from the tape to the disc. It must have been in August 1975 or 1976 when we were ready to do billing. I think I may have mentioned this earlier but let me state it again. The programmers assured me that although they had just a little bit more work to do, they could get it done easily. Therefore we relied on their statement, and decided to bill with the new system and the new machine. Well, it turned out that by October-November, when millions of dollars were flowing in here, we had no system for keeping track of those millions of dollars! So a major crash effort was made to help the programmers finish their job, and they did finally get it done in November or so, when literally we had millions of dollars stacked up that we could not allocate to which journal, which society, and so on. Then by lots of good work — lots of sweat, people worked into December and January — by February the problem was really very well licked. But by February, when you leave the old subscription year and you go to the new year, some of the members were complaining. They hadn't received their journals and they'd paid their bills. "Where are my journals?" They complained to their societies who then complained to AIP, just in time for the March Governing Board meeting, and of course I was questioned, "What was wrong?" I said, "We have had problems, but they're all solved." Do you think the board would believe that? No. And of course, they wanted to be sure that it was now solved. So they told a group of society secretaries and treasurers to meet; even at that time this was not a formal committee. It was an identifiable group, obviously, and they told them to meet with the management of AIP, and see whether this thing was really solved. I remember very distinctly having that committee in executive session; the officers of AIP were not part of that discussion. They then called me in, and didn't ask me to sit down. Joe Burton, treasurer of The American Physical Society, was leading the group. Joe stood just a few feet away from me. He said "We've had our discussions with the committee; we want you to know that we regard you as personally responsible for this disaster, and we want to have your personal assurance that the thing is going to be solved and solved quickly, and we want to have you prove to us that it's solved." I said, "Yes, I think we can prove it to you. In fact, I will institute regular reports," and every two weeks from then on I sent them reports. I think I mentioned earlier that after trying to figure out what could possibly be done to prove to this committee and to the board that things were right, I somehow had the idea to date-stamp every piece of paper that came into this building — every bill, every complaint letter. I would then put that date-stamp — if it was an order for a new journal or if it was a renewal of an old subscription — onto the computer tape of the new 9300 system. And we then put the second date of when the money had been allocated to the particular journal, when it all was completely processed. All they needed was to produce mailing labels. We took the time difference between when it entered the building and when it left the building, and then I would plot frequency distributions of time differences as a function of weeks. I'd show these nice plots, and then I'd show how the base of that got smaller and smaller with time. They could then believe that the system was right. Well, we had the solution. But you see, in that case, they didn't believe me, and they really had not too much reason for believing me. They were very tough and very specific: either I got that solved or I was to leave.

Aaserud:

Was that the worst instance?

Koch:

That was the worst experience, although when Shirley Quimby was treasurer of The American Physical Society maybe up to 1971 or 1972 when I was first hired here — he was a complainer about AIP costs. Our costs, just because of the large volume of work, have really been very advantageous, very low, compared to the way journals are produced elsewhere, except for the question of timeliness. And we were late because we had not figured out how to deal with this author driven system, and not worrying enough about the eventual consumer, the user. So that in some cases we were late. Then Quimby would always try to demonstrate that we were very expensive. In publishing there is a cost associated with what are called "author's alterations" that you get when you go to a commercial supplier. They will give you a price for composing a page, and then they will tack on 20 percent for "author's alterations." In the case of some commercial suppliers, in order to get your business, they will forget about the author's alterations initially. They promise you everything until they get you in the store. But Shirley would then say, "Look, your author's alterations are unreasonably large," and he called them "editor's errors." He said, "They're not author's alterations, they're editor's errors." And he would plot what he called editor's errors as a function of time. He would demonstrate, he thought, that we were improperly staffed — didn't have high enough quality staff, and so on. That was during the time period that we had the NSF grants. So I was trying on the one hand to do a good job of the NSF grants, on the other hand trying to respond to the problem of journals being delayed and being presumably expensive. It was a difficult period.

Aaserud:

Yes. Was that something you were told right when you came in here — that you had to take your service role seriously, that you shouldn't try to have too much independence?

Koch:

Oh no.

Aaserud:

It wasn't stated as a condition from the outset? Koch: Not at all. You picked that up quickly. For example, one of the things I always remember is the following story. We had a man by the name of Wallace Waterfall whom I mentioned earlier. He was, I guess, the secretary of AIP for 40 years. When I first came here, I said: "Wallace, how much vacation does the director of AIP get?" He said, "Well, here's what we give the rest of the staff, but you can take as much as you want. If you want to take the summer off, you can take the summer off. You can take the summer off, if you get the work done. If everything is going smoothly, then you can take the summer off." It sounded good to say you could take the summer off, but when you realized the details and realized that you're serving at the pleasure of the board and you'd better get the work done — I had very little vacation.

Aaserud:

How many summers did you get off?

Koch:

I never got a summer off. Those are things you pick up rapidly.

Aaserud:

These problems I'm sure had different expressions at different times, so maybe we should get back to these things chronologically. When you came in, there had been a long time without one director here. The continuity had been broken two or three years back when Hutchison left. Then it was Sawyer, then it was Van Zandt Williams for a short period and then it was Sawyer again. Then it was you. Did that affect your beginning work in any way?

Koch:

No, just in that maybe the honeymoon period for me was maybe longer than one would have expected it, because there had been problems in filling the office, and they were so pleased at having gotten someone. Maybe they would have been satisfied with anybody, and were perhaps more willing to be tolerant for a longer time period. Maybe that explains why we were able to do five years’ work of NSF, which was quite a perturbation on the staffing and the administration of AIP generally.

Aaserud:

Was the NSF work your suggestion?

Koch:

Yes and no. You see, science information at the National Science Foundation was organized separately from the research divisions, partly because the research divisions were unwilling to make investments in something that was not strictly research. They didn't want to make investments in education. They didn't want to make investments in libraries, communication techniques, or anything. So I believe it was 1959 that the Office of Science Information Service, OSIS, was formed, with Burton Atkinson as the first head. Now, I was picked at that October meeting, and you have to realize I was in Washington. So for the two months before I came to New York in December, I took whatever time I could to go and visit the federal agencies. I visited Atkinson. I visited the National Library of Medicine, the Library of Congress, and talked to many people. I learned that Atkinson had a concept of various national information systems in the major science disciplines, and that he had heavily supported developments at the Chemical Abstract Service. In fact, they invested in the chemical classification system. It was a way of labeling chemicals by computer. And we now have some four million compounds, and that system was developed I think largely with NSF money — about 25 million dollars, quite an investment. And Atkinson looked for how to invest in engineering. He came to the engineering societies, and the engineering societies were very fragmented, very political, and they feel they're competing one with the other. I don't understand that phenomenon either, but they had that feeling, and the engineers had tried various federations. In an earlier age one such federation was called an Engineers' Joint Council. That Council was really very ineffective. The most recent reincarnation of this concept in engineering is the American Association of Engineering Societies, which also is not working very well. But Atkinson tried to work with them. He went to biology. Biology is terribly fragmented. There are something like 200 biomedical societies. And in fact, last Friday, the president of the American Institute of Biological Sciences and the Chief Executive Officer visited me here in New York at Woodbury. They were trying to determine AIP's characteristics and how they can emulate us. It was the same thing back in Atkinson's time. He tried to see how it was in other disciplines. When I came along, and when there had been this sequence of directors, he was just so pleased to see me, because he wanted to try his concepts in physics and he found a real reception. I was excited about it. And that's why I was so enthusiastic just before I came, because I'd learned of this concept. And Atkinson said, would I be interested in making a proposal to NSF? So I worked hard before and after I came to AIP. For six months after I came here, I was commuting weekends to Washington. My family was still in Washington. So during the week I had lots of time to think. AIP had an apartment that we rented, and that's where I lived, a block away. I didn't have any loss of commuting time or whatever, and so I made up proposals and sent them to NSF and got them to react. I finally convinced the board that we should take the money.

Aaserud:

Your connection with Atkinson was actually from before you got the job offer?

Koch:

No, right after.

Aaserud:

Yes, exactly, so it was always in relation to AIP.

Koch:

I then went around and knocked on doors and said, "What do you know about physics and physics information?" So I learned a lot quickly.

Aaserud:

Good timing there.

Koch:

It was.

Aaserud:

When you entered your position here, the economic tide of physics was beginning to come down already, I guess.

Koch:

Yes.

Aaserud:

What you said earlier about the worst times at the NBS was part of that.

Koch:

Yes.

Aaserud:

How did that affect your position here at that early time?

Koch:

Well, the importance of the manpower statistics operation here was obviously increased, and the fact that physics had done the best job of collecting education and career data of any discipline is related to that.

Aaserud:

Due to AIP?

Koch:

At AIP, that's right. We still do. We do a better job of collecting that than in any other discipline. It's been a tradition here. And so in 1968 we started to see Ph.D. degrees leveling off. There were 1500 Ph.D.s in 1968-69, and that then started to go down. We kept on tracking it, and in 1972, the Bromley Report was written.

Aaserud:

"Physics in Perspective," yes.

Koch:

That's right, a super job. And I've always been good friends with Alan Bromley, and so he asked me to write up the chapter on manpower. I had this wealth of information here from AIP files that had been collected for years by Suzanne Ellis and years later by Beverly Porter and her colleagues, and I wrote up the chapter. I calculated all kinds of ratios. I took the number of freshmen in college, the number of juniors, seniors, first year graduate students, Ph.D.s, post-docs, and later careers. I had tables of these provided primarily by Suzanne, and they were all in that report. Physicists on the committee, in 1972, were not yet convinced that we were going into a recession. I was warning that the demand for physicists was going down. There was an oversupply of physicists at the time, and I warned that we'd better be prepared for that, and in fact train people to be broader so that they could go into other fields. The Bromley Committee then assigned an individual from IBM, Dale Teaney, to work with us here at AIP to try to understand what was being said about manpower and the projections. I think the end result was that it was pretty much the way we had originally written it. The first draft of that chapter was one that I wrote, and the conclusions were pretty much what we had originally said. The work on the Bromley report is an example of an impact on AIP of this recession. That was the question that you asked. I think it certainly put into greater prominence the importance of good manpower statistics at AIP. There was also the question of the way the journals were going to suffer. We started to see losses of subscriptions of 2, 3, 4 percent a year, all during the seventies, and still do. There are just a few journals, such as Applied Physics Letters that are still on the increase in subscriptions. But there's been a concern about what's going on. Shouldn't AIP study this more? And the climate in the mid-1970s was more and more in agreement with what I was proposing — that we should have a heavy emphasis on planning at AIP. Let's talk about planning for a minute, because it's such an important part of AIP during my stay.

Koch:

The late sixties and the early seventies saw the start of a recession in physics that I think continued until just the last few years, and we're now starting to come out of it. Now we've bottomed out and the number of Ph.D.s produced per year is starting to go up again, even among US nationals. During such a recession and the period of recovery was a good time for AIP to do some planning. Not only were our number of member societies increasing, our volume of work got bigger, and we were crowding out of this headquarters building. Therefore, I think I told you earlier that Fred Seitz agreed to be chairman of a long range planning committee in 1974. Did I tell you that?

Aaserud:

No, I don't think you talked about that.

Koch:

In 1974 he had agreed to do that and he had various subcommittees. After a year and a half of work he just could not get the subcommittees to agree on what direction AIP should take and so on, and he finally decided that he knew what to say in the report. So he wrote it up, and he submitted it to the governing board, and the governing board was very critical of it. They lambasted it. They tore it apart.

Aaserud:

Were you part of that committee work?

Koch:

Yes, I would listen in, but they were in principle advising me and advising the board. At that time the concept was one of an advisory committee advising me as director. In the late seventies, the concept was changed so all the committees were more and more advising the board. That was the result of a committee on committees. Because Seitz couldn't get an agreement, he got very disillusioned with long range planning, and was not involved formally after that. [But I've always gone and talked to him at Rockefeller University. I have found him a great friend of AIP, and have always found that very helpful advice was obtainable from him.] In about 1977 or so I was bold enough to propose that we should institute long range planning again at AIP, but with a different concept. The thing that I had realized was wrong was that there was not agreement from the societies on what job AIP was supposed to do. So I proposed to them that we first start out with an enumeration of the jobs that we had to do. We started with seven functions which, with the addition later of information services, became eight functions. We then instituted something called function planning, and if you look at the annual reports now, you'll notice that they have been subdivided into the eight functional areas — publishing services is one, and information services is another.

Aaserud:

1979 is a case in point —

Koch:

Yes, I think you'll find it in there.

Aaserud:

Yes, the seven basic functions. I picked the right one.

Koch:

Great. The board thought that it was not an unreasonable approach to get them to agree on the jobs we were supposed to do. Then we would analyze everything in terms of those eight functions — look at the money investments, the dollar investments, make some projections — so it all looked quite reasonable. And it has worked quite successfully as an administrative tool to keep the staff informed and to keep the members of the board and society officers informed. I remember specifically that Jim Gerhart — secretary for many years of the American Association of Physics Teachers — told me that as a result of the function planning it was the first time that he really understood something about what was going on at AIP. He had been involved for many years with AIP. So the function planning was continued in pretty much the same way, and you've seen these books that are color coded? I have a stack of them around in my office. They have one color for each year. And there's one thin document that's called the Function Plan, and then there are other subsidiary documents that are development plans. For example, we had a document one year, in 1982 I think, called the Publishing Development Plan. That's a typical five year plan, where we indicate that we're going to start a big book publishing program. You're going to get a new generation of composition equipment, we're going to use staff in the following ways, we need more building space, we need more money, and here is how we're going to pay for it all. So that's what was all part of function planning. And as I said, I think it was very successful. But then I got a critique of our function planning by some professional planners. I said, "Tell me what's wrong."

Aaserud:

How did they get their hands on it?

Koch:

Well, I called them in. I'd been looking to find out how to improve it. There were things that bothered me about the function planning. I couldn't quite put my hands on it. I knew nothing about planning, certainly, before I came here; I learned it on the job. So I went to an organization called the Support Services. They have eight offices in eight major cities around the country. And they are a not-for-profit business management organization that assists not-for-profit agencies. They're a wonderful organization. They give seminars in accounting — computers and business, you know, that kind of thing — and they charge very nominal fees compared to the American Management Association, which charges ten times as much. These people charge $35 half a day, or $60 for a whole day seminar. Very useful. So I went to them and I said, "Who in your organization knows something about planning?" Well, to make a long story short, I ended up with a tall man, Jon Cook in the Washington office — a very bright young attorney — and he wrote me a report. He said, "Your basic problem is that you've got eight different long range plans. They're not coordinated." It sounds obvious, doesn't it? Well, what are the implications of that? The implications are being presented on Monday of this week, but I've gotten approval. For the last two years I've built up to it slowly. We're now going to the next generation of planning at AIP. And I can explain it very quickly and simply. We have a Part A, a Part B, a Part C, and let's think backwards, and look first at Part C. Part C are these development plans — the publishing development plan, educational services development plan — and in the next few weeks, I'm going to work with the division managers to try to get the next edition of the publishing development plan and the education services development plan. We'll project five years ahead, and they're typical five year plans. And those can be generated any time during the year. Whenever it seems desirable to make a building development plan, you project: How much space do you need? How much money is it going to cost? Who's going to fill the building? And so on. So that's Part C. After we had our labor union problems in late 1981, we developed a personnel administration development plan to figure out how to avoid a labor union, and how to improve personnel administration; and it's worked like a charm. That's also part of Part C. Part B is now called operations plan. Part C looks to the future. Part B looks to the past. Part B looks at the operations and decides what things have been done during the past year and the last few years in the major areas that AIP is involved in. And now, instead of listing eight functions, we list three "businesses" — the administrative and financial business, the publishing business, and the educational services business. Very simple. Not as complicated as eight functions. And you'll see the real merit of the three businesses in just a moment. So we had those three businesses, and we reviewed the operations in those three businesses, and Part B has the advisory committee reports of every one of the advisory committees that gave us advice in the three business areas. Then Part A is called policy planning, and has to precede all the rest, because Part A has every policy, procedure, rule, guideline, contract, agreement and principle in it. We do not supply many details in Part A, just supply the means of accessing, so if you want this contract, you go to that file, if you want that contract agreement, you go to that file. But all the basic policies are enumerated in the policy plan. As I saw it described somewhere, policies are always a pyramid of policies, where the most important ones are at the top of the pyramid, and then as you go down the pyramid you have the rules and procedures and guidelines for how you're going to operate, for example, the Letters to the Editor of Physics Today, what the role of the editor of Physics Today is. Part A contains the basic policies in the three businesses, and the three businesses are really quite different, one from another. Now, I said that there's real merit in this way of breaking down AIP, because now you can look at AIP as a unit — not eight different functional areas, not eight different plans, but one plan, the AIP. In 1977 and 1978 AIP almost lost its tax exemption. It was a strange phenomenon at that time, because the Internal Revenue Service agent who was trying to make a name for himself — an IRS man by the name of Doman — attacked not only AIP but the Chemical Society, the IEEE, and the other engineering societies. He said, "Look, those societies just should not be exempt from paying federal taxes." And he was particularly confused by coming to AIP, because we're not just a simple society. At that time we had nine member societies related to AIP. He couldn't understand who was on first, who was on second, who was on third, and how these were interrelated. The end result was that we finally prevailed. In fact, one source of information in this history of AIP is the legal brief that our attorney in Washington prepared. It's a super analysis of AIP. What it concluded was that AIP is an entity onto itself, with more than 50 percent of its operations being conducted for AIP as a corporation. If the nine member societies were to go away and do their business elsewhere, the self-contained unit AIP would still be able to exist in principle. In practice, this is nonsense. You don't want to separate AIP from the physics community. But in principle, this lawyer argued, and it's a very valid point —

Aaserud:

— for the tax purposes.

Koch:

We can exist, we're a corporation. That corporation generates a net income in its publishing business — in the Soviet translation journals, for example, that it owns, and the journals like the Journal of Applied Physics, Applied Physics Letters, where I appoint the editor; we have six journals like that, and Physics Today. We make money and we apply that money to our educational services programs that are absorbers of money; they can't create money as publishing can. So we generate money in one area, plow it back in for educational purposes, so here you have the legitimacy of AIP. That's why it's so nice to think in terms of administration and finance, publishing, and education as the three major businesses of AIP. That's the simplest form of AIP. And now you can also think of societies coming and saying: "Look, we want some financial services from you." We now provide APS with accounting software. We collect all of their money and whatever money we owe to a society, we pay interest to that society, by the day. We collect $10,000 for one society, hold it for a week. We pay by the day — interest we can earn ourselves or they can earn. In the publishing area, similarly, they come and we do publishing for them; educational services also. So these three "businesses" are very basic for AIP. They also provide a rationale for AIP as an institution. And they show that we not only provide administrative services to our societies, but we also supply some services that require some professionalism. We may jointly produce a journal with one or two of them. We may jointly have a program. Now, to finish with the Internal Revenue Service. This Doman, investigator, auditor, said, "AIP is just supplying administrative services to its societies, and isn't an entity unto itself." And if that were so, then we would be like a hospital cooperative. Now, a hospital cooperative, on behalf of tax exempt hospitals, takes in dirty laundry from each hospital and supplies clean laundry back. So Doman said, "You're like a hospital cooperative, you take in journals and you take in manuscripts and then you supply journals back, it's just like a hospital cooperative." But because we're a self-contained entity, because more than 50 percent is being done for AIP as an institution, and because the publishing business made a profit that has been consumed in our educational business, that floored him. That was the fallacy in his argument, and that prevailed, and that again provides now also the rationale for AIP as an institution and the right way to do planning. That's why I'm so pleased now that after agonizing about how to improve on the planning, we have now hit on this procedure, and I hope it will serve well even after I'm gone from here.

Aaserud:

So it seems that the IRS auditing helped the AIP a great deal in defining its role and its independence in relation to the societies.

Koch:

That's right. Now, we have just had the IRS auditors return, and if you know the way IRS auditors in the United States work, they are told to examine the tax returns of an organization or even an individual for a given year. They were told to look at 1981. And they just came here. And Jerry Gilbert, our Treasurer, who interacts with these auditors, said: "Do you know anything about the past history of AIP?" "Oh no. But we want to have you show us all your books for the last six months and we will carefully spend months of effort looking those over, looking for problems and so on." And Jerry said, "You'd better look at what happened in 1977." Jerry pulled out this legal brief and the history and the fact that we are now a famous case study — we're a classic case. And when they saw that, they then turned around and went away.

Aaserud:

So that was not the reason they came; they just didn't know about it. Picked you out of the hat.

Koch:

That's right. So we feel very comfortable now with the legitimacy of AIP as a membership corporation, its relationship with the societies, and the important supportive role it can play for the physics community. I really must emphasize support and services to the members. We don't regard individual physicists as really being our members, but in principle they are members through the societies, and we've discussed that. But AIP's prime members are our societies and we are here to serve those societies. That's where the strength of AIP has been and will be. We now think we understand that strength better, and that's serving in fact to attract other organizations to join AIP, partly because there increasingly is an obvious role for AIP, but also because in the physical sciences there is no other organization like AIP to which smaller societies can adhere. So it's a very comfortable position, I think, we are in at AIP now. We finally have found our place in life. We really have an organization with a character that is definable now. We are not like the Bureau of Standards who by analogy has been struggling to find its character, its role in life.

Aaserud:

It still has the same problem.

Koch:

That's right. But we now know where we're going at AIP. Now we can plan. Now we can do a better job for physicists.

Aaserud:

I don't know if you want to go back in time again. When you entered, Van Zant Williams had been concentrating on the information side of it, I believe, to a large extent.

Koch:

He had been approached by Atkinson to try to get AIP involved in some larger concepts, and the way Atkinson worked at that time was obviously to try to get the proposals from the organization, not that he write the proposal. That's the way all of NSF works. It tries to prod and stimulate, and I think he had been trying to prod and stimulate to get a proposal from Van Zandt Williams that would be supportable. Now, I don't know really how much had been accomplished by Van during the short time he was here. It just was obvious that when I came that enough preparatory work had been done that it was then possible even for a novice like myself to take the pieces and frame them up and put them into a logical whole and then make a proposal.

Aaserud:

When you did that, you made use of the earlier tradition here? There was a documentation study that began back in the fifties.

Koch:

Yes, that's right.

Aaserud:

So there was some tradition for that.

Koch:

Yes, there was. In fact, some of the people here were more knowledgeable about that than I was and had worked in fact in the earlier groups. It may be that Books Division manager Rita Lerner had worked with the earlier group, and I believe Art Herschmann had been hired by Van Zant Williams to try to put together a proposal for NSF, but I can't really recall that. But obviously the time was ripe.

Aaserud:

Yes, it was in the air. Your first publication coming out of the AIP was "A National Information System for Physics," I believe, published in Physics Today, which presents your thoughts about the information revolution and what you intended to achieve at the AIP with that. So we can say that's in the tradition more or less with what came before it.

Koch:

Yes.

Aaserud:

So it's not a change of direction, in any way.

Koch:

No, except the increment was quite large. The AIP was not a large organization at the time. I guess that we must have had 120 or so people on the staff. On that order. And we ended up with 40 more people. You know, that's quite a step. So the question: "Is the tail wagging the dog at AIP?" was an issue that was raised many times.

Aaserud:

By the physicists?

Koch:

Yes.

Aaserud:

But then you had gotten the NSF grant, correct?

Koch:

Oh yes.

Aaserud:

So that was essentially a description of the plans for how to use that.

Koch:

Now, for example, we talked about the earlier documentation work. One of the things that is so basic to any organization of a body of literature is how you classify it and how you index it. Did we talk about that at all before?

Aaserud:

Not in any detail. You may have mentioned it.

Koch:

But we didn't mention anything about what had gone before. There had been hired a librarian, Pauline Atherton, who worked on the documentation program of the American Institute of Physics.

Aaserud:

That was before your time?

Koch:

Yes. In fact, I think she left just about the time I came. I never really did know her. I had met her. And I think Art Herschmann, who was a former editor of The American Physical Society and a very bright physicist, had worked briefly with her. They had been trying to figure out how to classify documents in physics. And they developed a system of multiple facets, five or six or seven facets. Now, for example it's obvious that in chemistry materials is a facet, so you organize everything according to chemical substances. And what this librarian, and I think Herschmann, had been intrigued by was how accurately you could put something into a file organized by facets, and retrieve it. One person files it, then an individual comes along and says, "I want such and such an article," and if you have enough facets, you can really pin it down, select it out. Well, it happened that we had formed an advisory committee on our information program that had on it a man by the name of Tommy Lauritsen, who was a professor at Caltech who no longer is alive.

Aaserud:

The son of Charles Lauritsen.

Koch:

Yes. And Tommy, very bright guy, came in and said, "What kind of nonsense are you people doing here?" He was not at all sympathetic with the fact this librarian and then more recently Art Herschmann had been working on facets for several years. They were trying to do a conscientious job and they thought facets was the way to go. He said, "Can you imagine a physicist writing a paper and trying to figure out, with your definition of facets, how to classify his or her article? Then how to find it? The rules are so complicated. Can't you make something simpler?" And at the time there was a man by the name of Franz Alt, a computer man, who I hired from the Bureau of Standards. He came up here and joined our program. And he was the one that I remember having told me about the concept of a front of the book and a back of the book way of classifying and indexing. The front of the book is the classification; the back of the book is the indexing. Classifications are just ways of organizing without having the details of the information before you. So the front of the book has a hierarchy in it. The hierarchy would be dependent on atoms, molecules, nuclei; above it, electrons and so on. And these are the standard disciplinary designations, and you get more and more detailed in the branches of the hierarchy. And Franz Alt said, "That's the front of the book. Now, for the back of the book, you need the words in the text. You go through the text and you underline that word, that word, that word, and organize those words alphabetically. Then you say where "conference" is on page 5, 6, 25 and so on. So the back of the book needs the text. The front of the book doesn't need the text." And Tommy Lauritsen said, "Can't we figure out some way of using a hierarchy that physicists are familiar with?" And that's where the PACS scheme came from — the Physics and Astronomy Classification System. In fact, that was one of the criticisms, you see. I think Phil Morse was the chairman of the board at that time, and Tommy Lauritsen, a man for whom Phil Morse had great respect, comes in and says, "Jeez, those guys have been working for years on this classification system, and it's nonsense! Nobody can ever use it. You can't expect physicists when they're writing a paper for Physics Review to tell you what facets they've been working on!" So he was the one that was instrumental, and Phil Morse said, "Gee, you people ought to get away from working off in your corner, get out in the physics world and get more of the prominent people to critique what you're doing."

Aaserud:

Another point of contention with the physicists.

Koch:

That's right. And it's worked out well. We now have a good system. But it took the result of that time period. We worked on many other interesting things. Advance Abstracts, which is now part of PINET, are particularly well organized because so much of the world's physics literature is being produced under the one roof at AIP. Every Friday you can just query all the computer files and pull out all the new abstracts that have come in that week. And it works beautifully.

Aaserud:

That has had a complicated international development as well, right? I'm thinking about the relationship with the IEE in England and Science Abstracts, and the relationship with Germany and all that. I don't know if we should go into that as well?

Koch:

Yes, it might be well, in fact, that all started with the information program, in that physicists have never done much with what are called secondary services. They want the original document. They're not as interested in an abstract of the document. Whereas chemists can work very well with abstracts and with numerical data. And so I gave a talk — it may be in part of the stuff that you've seen there — in which I talked about analogue display journals and digital display journals. Now, analogue display journals are like archival journals, where on a page you will have text that describes something and is an analogue of some device or phenomenon or whatever. You will have a photograph that stands for something, an equation that stands for something else. These are all analogues. And it's a complicated mental process that one goes through, when you look at a page in a scientific journal, because you look at the equation, you look at the shape of the curve, you read the words, and it describes what's in that shape, or you look at the photograph and you read what's in the text, and you say, "Oh yes, that machine there is a betatron, oh yes, that's the depth dose," and so on. So these are all analogue displays. Chemists, on the other hand, can very well do with just a number. They just need a boiling point, or they need the molecular structure — how many carbons, how many hydrogens, how many oxygens — that's all they need, like an engineer. Have you heard the Purcell way of describing the natural sciences?

Aaserud:

You have to remind me. I can't think of it.

Koch:

Well, if I haven't told it to you before you probably haven't, because it's dug deep in the Bromley Report, in chapter 2 written by Purcell. It doesn't have the name Purcell on it, but it was written by Purcell. It's an elegant chapter called "The Nature of Physics." And what Purcell in essence said was that the natural sciences can be subdivided into four groupings. I'm then going to talk about the fifth grouping. But the four groups are two materials sciences — the non-living including chemistry, the living materials including biology, and then you have two systems. You have the man-made systems that include engineering science, and the natural systems that are earth and geoscience, including geophysics and astronomy. So those are the four, and what people who work in those four disciplines do is to observe nature just the way it is — look at the weather, look at the rock, look at whatever. They collect lots of data. And then what the fifth set of disciplines does is to try to make sense out of the other four, and that's where mathematical and physical science comes in. Physics tries to look at these data and provide the basic simplifications for the data collected by the other four. I've always been interested in the fact that in physics, 30 percent of the Ph.D. theses are theoretical in nature, and 70 percent are experimental. In chemistry, a very close discipline, one thinks, 3 percent are theoretical and 97 percent are experimental — a factor of 10 different from physics. Really quite striking. So physics and chemistry are quite different. The people working in the observational, experimental sciences often are quite pleased with just knowing what somebody else's experimental number is. They don't care about the understanding of it, just what that number is. They have measured that particular spectrum and they want the other guy's spectrum. What physicists do is to try to understand it. That's why, I believe, physicists haven't ever used secondary services very much. But with the increasing complication of science, going more and more outside the United States, it becomes more and more important to have a means of keeping an overview. Secondary services are playing more and more of a role to overview the literature, besides just providing the experimental number that's in that abstract. So physicists, I believe, are going to need more and more secondary services, and that's a conclusion that we came to in our information program. That's why we started with Advanced Abstract experiments, and it in fact started us in our negotiations with the British. The problem with the British is that they have never honored our copyright laws. Increasingly, the world society and US society probably only now is regarding information as a business, where commercial interests are very important. The August issue of Business Week has "The Information Business" on the cover. I've used it as part of my talk for the National Science Foundation. Information is becoming more and more of a business and our society generally is going away from the concept of information being a public good, freely available, to information as a saleable commodity. As it becomes more and more a saleable commodity, where you negotiate with other countries — for example, you give me your literature, I'll give you mine, or you sell it or you barter it or whatever — different parts of what is intellectual property are created by an author who wants to be protected. It was just last March, or the March before, that the AIP governing board passed a resolution that said that we are to protect our copyrights, including abstracts, and if anyone else uses our abstracts, we should get a financial return. Now, the British never acknowledged that. To this day they haven't acknowledged it. They don't want to acknowledge it, because just on the physics part of their files, they're making at least a million dollars a year profit, net return. When we started negotiations, the contribution from AIP was 30 percent of the total, quite a sizeable chunk. So in 1972-74 I negotiated with them, trying to sell them a tape of abstracts from AIP. They would never listen to it. Their excuse was that their labor unions were so strong that if they were proposing to buy something that they presently are using their labor for, the labor unions would go on strike and wouldn't allow them to take our tape. Nonsense. So that was the excuse, anyway, and they have enjoyed taking our abstracts manually and without our express permission. But there's a new copyright law now in England coming into force. This past spring we got a copy of the White Paper, and this White Paper claims to be the forerunner of the new copyright law. That new copyright law apparently will place much greater strength in the publisher's copyright, and on the importance of secondary users, by licensing the use of the primary publisher's copyrighted material. So I think things are going to be changing, probably not before I leave here, but it's going in what I think is the right direction. I've never felt that it was necessarily so that every abstract should be paid for. But at least it should be acknowledged, so that if the secondary publisher creates mistakes, you have a way of correcting those mistakes, so that our authors don't have their material re-appearing improperly translated or whatever. We must have those controls, I believe. In 1978, as a result of my being involved in the information business and going over for yearly meetings, primarily to Europe, I would meet with these people. I met with a man by the name of Weidemann from Germany, who had been involved with the German physics abstracting service called Physikalische Berichte. It was a famous abstracting service and very well known. But by 1978, because certainly most of the literature in physics in Germany was going to the English language, most of the journals in physics in Germany were in English, even if they had a German title, and many of them have even changed their German titles to English titles. So they made a national decision that since Physikalische Berichte was really not being used outside of Germany, they wanted to become important internationally. They said, "We will go to the English language." When they did that, I flew to Karlsuhe and I said: "Can we not negotiate an arrangement with you?" And that arrangement was consummated and has been extremely successful. Our abstracts are four to fourteen weeks sooner in the abstracting journal that they publish called Physics Briefs than they are in Physics Abstracts which is published by IEE in London. Similarly, the tape of abstracts is now on STN International. Incidentally, INSPEC and Physics Abstracts are also on STN International, and there you can look from one file to the other by computer searching, and there you can see that our abstracts are much sooner. So we hope that someday the British will see their mistake and that they will work cooperatively with us, before it gets to the stage where we have to sue them. I hope that that will never come, but it might be necessary to at least threaten them.

Aaserud:

You didn't have the same problems with copyright in Germany?

Koch:

No. In fact, they have been most cooperative, and they honor copyright laws. If a publisher tells them that he does not want to have his abstracts appearing in Physics Briefs, then they remove them. If he says, "I want you to pay me a royalty for abstracts," they will pay a royalty for abstracts. The national policy, I believe, is to rigorously protect copyright in Germany.

Aaserud:

Was there a British reaction to the cooperation with Germany?

Koch:

Yes. In fact, they accused the Germans, believe it or not, in a legal brief before the European Community, saying that the German organization is government-subsidized, that the British organization was a private corporation not subsidized, and that the German government was unfairly competing with them. The Germans said, "Nonsense."

Aaserud:

Was that in relation to the talks with AIP?

Koch:

Oh yes.

Aaserud:

That is what triggered it.

Koch:

Yes.

Aaserud:

But they didn't join anyway.

Koch:

Not yet. But we have submitted to the British a proposal exactly like the proposal that we generated with the Germans. They asked us for it, we've now supplied it to them. They have responded that they would like not to take everything that we propose to supply them, and of course they want to get it for a fifth of the price that we're charging the Germans. So we have responded that they can delete whatever they want from the tape that we would supply to them, but they have to pay us the same price. It won't save us any money for them to delete things.

Aaserud:

So for the time being, there are two entirely independent sets of abstracts.

Koch:

That's right. And it's utterly nonsense. Now, I learned last week that North-Holland Publishing Company has now agreed to supply the German organization with a tape of abstracts in physics and mathematics. The Germans have agreed to pay them for this tape, and North-Holland will provide it under an exclusive arrangement. Therefore they will not supply it to the British. So we'll see. You see, as time goes on, if something like five major publishers were to do the same thing that North-Holland and AIP are doing with the Germans, the Germans could get a tape that is much more timely, because they're working with the primary publishers, and it can be much less expensive, because the abstracts are a byproduct of the primary journal production, so as soon as there is 50, 60, 70 percent of the abstract tape in physics that comes in from primary publishers, those primary publishers can very easily out-compete the British tape. I would hope again that the British would see the advantage of trying to work cooperatively.

Aaserud:

Well, you said that at the outset it was a little difficult to convince physicists of the necessity for this kind of information activity. Have you seen some development in that? Are they actually using the abstracts more, or the kind of information that they supply?

Koch:

Well, this now leads naturally to discussion of the electronic information systems that we now have started. On March 3lst we put on line a system called Pi-NET — Physics Information Network — and on August lst we started an electronic mail system called Pi-MAIL. These are very much coupled one with the other, and I can mention some ways that they can be coupled. Pi-NET, being the first one to start, has six different data bases on it. Let me enumerate them very quickly. The first one is job openings, job opportunities. The second one is meeting schedules and meeting programs. Right today, there are five different meetings that have programs in Pi-NET. You can search the file and look for when you as a speaker are going to have to give your talk. The third file is Advance Abstracts that appear two to three months before the journals are printed. The fourth file is an article titles file that takes those articles that have now appeared in print that were originally in the Advance Abstracts file. They now are purged from the Advance Abstracts, and appear for three to six months after as a title. This allows you to look up the volume and page number. The fifth file is announcements and press releases, and the sixth file is a publication catalogue so you can order copies of articles, copies of journals and books, with a credit card. So those are the six files, and you can regard all of them as being secondary services. Most of them are either a title or an abstract, like a job opening is an abstract describing the opening. So these are secondary services, and we are seeing 50 new searches per day occurring. Half of those new searches are new registrants per day, so that the number of people using Pi-NET is increasing I think quite dramatically. Now, the PIMAIL system is presently also on GTE Telenet, the communications system, but it's separated because Pi-NET is presently free. AIP pays the communication charges for the latter. For Pi-MAIL, the user pays for the connect hour use. But Pi-MAIL is also very inexpensive and simple to use, because there is no monthly or annual charge. There's no initiation fee. You simply have to pay, in that case, the connect hour charge that you use, and out of hours — after 9 o'clock at night and before 7 o'clock in the morning — the charge is $5 per connect hour, so that's quite inexpensive. Pi-NET is even less expensive. It doesn't cost you anything, and we think both of them are going to provide quite a service to physicists. We can see, for example, an example of a coupling. We can see someone on the publication catalogue of Pi-NET sixth file, ordering a copy of an article. We can scan that article, put that facsimile into a micro- computer at AIP, send it through a modem through the phone line to the person's mailbox on Pi-MAIL, and allow them then to down-load the scanned article, and of course we'll charge them for that service. Right now Pi-MAIL has Harvard University as an institutional user. Everybody on the Harvard faculty can have the electronic mail charges paid by Harvard. We think other institutions, as soon as they realize that Harvard is doing it, will do it. Then you're going to see them sending messages back and forth between experimenters.

Koch:

Not only will they send messages back and forth, but they will then send the text — the preprints — back and forth. So on Pi-MAIL you can tell the network to put a copy of this preprint in these 500 mailboxes, and you register the mailing list, you see. So here you have a preprint service — very quick, very inexpensive. We think both of these will be big services, and all this description was started by your question of whether we see increasing use of secondary services by physicists. Now we see it. Now I think the 15, almost 20 years of work will now come to fruition. On the cover of Business Week, where it talked about the information business, it said as a subtitle, "When will there be payoff?" Well, for physicists the payoff is going to be when everybody is registered on Pi-MAIL and everybody's using Pi-NET, and the payoff will then be the on-line physics community.

Aaserud:

But this must have been a gradual development, though. From the outset, when physicists were hardly interested in abstracts at all, through the period that you have been director, in which the amount of information, the amount of publications has increased phenomenally.

Koch:

However, just realize something, and that is, that up 'til 1968, everything was expanding, and the United States was top dog. Everything was expanding. And if your invisible college was communicating with you, if you were in the inner circle — and most of these inner circles were in the United States — you didn't care about what was going on in Europe or in Japan or Russia. They only cared about what their buddies were doing, because they knew everybody, and so the question of an information explosion, that didn't really strike people in physics before 1968. But then after 1968, then there is less and less research money, more and more concern that you didn't duplicate somebody else's research, and more and more concern that you could be criticized for not knowing that somebody in Germany just made the same damned measurement that you made. "Didn't you know about that? You didn't cite him in your article." That kind of criticism has been more and more prevalent in the 1970s. And so AIP and the role it can play was really completely in tune with what was going on in the rest of the world.

Aaserud:

So it's not the result of the information explosion as such?

Koch:

Well, it all went hand in glove. AIP, as far as I can see, was at the right place at the right time. I think it will make a major contribution in years hence, and I think that physicists generally, NSF particularly, will realize that not only do you have to worry about producing information, you have to worry about using it and knowing what other people are producing.

Aaserud:

You could even say perhaps that the bad economic time for physics has played into the AIP's hands, or at least, that it has provided the need for that kind of information service.

Koch:

That's right. And you know, we've discussed only the information program here. We have to quickly mention — and we've already mentioned — manpower statistics, the education and career data. We truly are in the information business now at AIP. But not only in publishing research results; we're in the history information business, in the education business, the career business, in the public information business, public relations business, job openings, jobs wanted business. We're in with both feet.

Aaserud:

And there are a lot of things we could talk about, like the expansion of this to the Soviet Union, or to China for that matter.

Koch:

Well, that's going in an exciting way, I think. I mentioned the other day when Bob Beyer of ASA was here. He helped start the translation program at AIP from Russian to English. He told me it was 1954 that this work started. I had always thought it started in 1957. But in those years, the mid-fifties, AIP began translating journals, and the way it began was with an organization here in New York City. I think they were called Consultants Bureau at the time, headed by a man by the name of Earl Coleman. And with NSF. You could not fund Earl Coleman directly with NSF money, so the money came through here and went to Earl Coleman, and that's the way we started translating Soviet journals. We now translate 19 different journals, and at one time I tried to figure out what fraction of the Soviet literature in physics is contained in our journals. I'd roughly estimate about half of their literature — half of the abstracts that were in Physics Abstracts at the time were in our journals. Well, because of the success of our Soviet program, and as a result of a trip I took to China, I thought, we really should try to be aware of Chinese research, and be aware of ours. Chinese physics gets better and better, and it's evident that there are so many very clever, wise Chinese physicists, thousands of them, getting trained in the United States, and very very smart. Those people are going to stimulate Chinese physics experimentalists to do more and more super research work. So we ought to have a good sharing of information, and that program has evolved into the Chinese Physics: Lasers journal. That's one area in which they're doing outstanding work. And we hope it will appear in that Laser Journal of theirs and that we can translate it into English. And we expect that as other areas prosper, and high quality work is done, we'll pull out other journals from our selective article journal, Chinese Physics. In order to follow Japanese physics, we should do a similar thing. We are so ignorant of the work that goes on in Japan. But Japanese is quite different than Chinese or Russian language. And the reason is that in Japan, as a result of their being the defeated nation, they have many English language journals. They're now the leaders, but after the Second World War they were destitute and eager to work more and more with the Americans, and they started a number of physics journals in Japan that are in the English language. In China, almost no journals in physics are in English, so it's quite a different situation. Similarly in Moscow — no journals in English. So in Japan, it's a lot of journals. So next week we're going to present to our board a proposal that we get support from the National Bureau of Standards for a Japanese language program. We market the Institute of Physics journals from London in the United States in an exclusive marketing arrangement that does a great deal of good for the British, and they value our involvement very much because when we have an American Physical Society exhibit and we display our journals, we display their journals along with them, and we put them in our catalogue. We take orders for them. So we do a big business for them and they appreciate it a lot. In Japan, we couldn't do the same thing the way we do it for the British, but we can if we get subsidy from the Bureau of Standards. So I'm proposing that we market a set of English language physics journals from Japan, put them into our catalogue, and exhibit them at our meetings alongside the Institute of Physics journals. Secondly, I propose that we take two journals that are in Japanese, the Vacuum Science Journal and the Surface Science Journal, and translate those cover to cover. Those are bound to be exciting interesting journals. Then we will consider whether we should have a selected article Japanese journal, where we take 15 or 20 journals that are in Japanese and select the best articles from them. Then, as individual journals seem to obviously have a great deal of good physics in them, then we could go to them and translate them cover to cover. So here's an example of worrying about things outside the United States that US physicists just must know about. In Europe it's taking care of itself, because they've gone in most cases to the English language in Europe. But we'll have to add Russian, Chinese, Japanese, and then we can see where do we go from there. We'll probably have to go to Brazil and other places.

Aaserud:

How dependent has the physics community become on this kind of activity do you think?

Koch:

What do you mean dependent?

Aaserud:

On that kind of information retrieval that is accessible now?

Koch:

From the scientific point of view?

Aaserud:

Yes.

Koch:

Well, there are several ways of answering that. One is that in two of our journals, Journal of Applied Physics and Applied Physics Letters, presently 17 percent of their articles are from Japan, and the percentage is increasing. Now, that's a big percentage from one country. That means that they are doing damned good physics that's passing our peer review process and going into those two journals, which they themselves regard as important international journals. It's important prestige for them in their home institutions, to be able to say that they have passed the peer review here. So obviously they're doing important physics, and those articles will increasingly be cited and so on. We've looked at the citations in the English language physics journals in Japan. There aren't too many to the US literature.

Aaserud:

Really?

Koch:

There are some. But there are lots of citations to their own institutions. We further have this new translation of Physics Today into Japanese. We also have the reverse process. We're translating, and I have a pack of things I got from Physics Today Editor, Gloria Lubkin last week — material that has appeared in their magazine called Parity. It's been translated into English, and you can get a sense just from the material there — even a novice like me — for the areas in which they are on top of things and the ones in which they are quite naive. So that's what we need to do in the way of breaking down the language barrier — trying to improve the flow of information in both directions, having them benefit from our experiences and we from theirs.

Aaserud:

This sense among physicists at the beginning of the period that they felt that they knew what had to be known, through their natural or their old way of keeping up — has that feeling changed, do you feel?

Koch:

Oh yes, by the very fact that we're getting governing board support for doing this.

Aaserud:

It might be too strong to say that AIP has led to this change in physics, but it certainly has contributed to it.

Koch:

I think so.