Interview with Gart Westerhout on 21 July 1993
Description
Creator
Rights
Type
Interviewee
Location
Original Format of Digital Item
Interview Date
Notes
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event.
Series
Transcription
Begin Tape 5 Side A
Dick
Last time we talked about some of the issues in organization and reorganizations, and we got into some of the budget matters. I thought we'd start a different tack today. We mentioned last time a little bit about computers. It seems to me that there's two major trends in the last thirty years that have affected the observatory to some degree or another. One is the Space Age and one is the Computer Age. I'm not sure how much the Space Age has affected us, but probably the Computer Age has affected us more.
You mentioned last time that the Nautical Almanac office is always the place where the computers were centralized, and during your time there has been quite a big change in that, from the mainframe to the distributed computing. I wanted to get the story behind that, and whether or not that was a straightforward or contentious change in the computer environment at the observatory.
Westerhout
Let's start at the beginning. Of course there was the mainframe computer.
Dick
When you came, it would have been the IBM 360.
Westerhout
That was still the 360.
Dick
The long history of the IBM at the USNO, going back to '57 when the 650 was used.
Westerhout
That's right. And then we had a 1410. That's when PL-1 was introduced and so on. Of course, before the 360, before the 650, we were pioneers in using computers for large-scale calculations, because during World War II, the whole lab, with all the Hollerith machines and so on, was already there. Eckert and so on were here.
Dick
Eckert was here for four or five years during the war.
Westerhout
Right. That's when a lot of computing was done. Of course, that you will get from the historians in the Almanac Office. We were pioneers in the government in using computers for all sorts of things, for publishing technical data and so on.
Dick
But those weren't really electronic computers in those early days.
Westerhout
They weren't electric computers yet, but I'm sure we had the very first electronic computers that were available. But they may not have been electronic computers; they were computers all the same, because we, in Leiden, we had the first real computer as such, which wasn't electronic, it was an IBM calculating punch--602A. You programmed that on one of those wire wrap boards, and you could calculate tables with that. There was no question. You could do sines and cosines. Always relays, because you didn't come in a sine; you put in the approximate formula for sine. So much times X, plus so much times B, and so on, all these AX plus BY plus CZ plus FZ2, and so on.
Dick
That's an interesting question, actually, if we can divert to Leiden for just a minute. Do you remember how the computerization developed there during your time? You left there by '62, wasn't it?
Westerhout
Yes. As I say, that 602A was the first machine.
Dick
When did that come?
Westerhout
That came in - Hügbom was there, so let me think when that could have been. '54, '55.
Dick
Was that an IBM machine?
Westerhout
That was an IBM 602A calculating punch. Hügbom was a very enthusiastic character.
Dick
Was he the one responsible for getting it?
Westerhout
No, but he sat down, and one night he programmed it to do a Fourier transform, a 32 point transform. You fed in one card and it did the Fourier transform off the data on that card, and it spewed out a second card which gave the Fourier transform data. The board had wires hanging out all the way to the ground, because you could, in fact, inside those wires then put splitters or switches that automatically switched, so you could sort of build on to that computer. It took four minutes. The IBM representative didn't believe it, so he called his boss, and a whole group of IBM people came up to see that wonder. "How could you keep that machine calculating for four minutes?" It worked. That was, of course, just an intellectual exercise, but we did a whole lot of tabulations regarding radio astronomy.
Then Shell had the first IBM in The Hague. Royal Dutch Shell, remember? Shell is a Dutch company. I didn't know whether you knew that. Shell--the official name is Royal Dutch Shell. They had the first IBM 650, and several of our people traveled to The Hague to do radio astronomical calculations on that machine. That's how we slowly got into it. But the Leiden observatory never had a mainframe computer, not until I left.
Dick
Not before you left.
Westerhout
Not before I left. There were small things, but there was nothing that resembled anything like that. We did that sort of work elsewhere. I think in my story about Leiden, I talked about digitizing a lot of stuff, and that went by means of an electric typewriter--one of the first electronic typewriters. On the back of the carriage, I equipped it with a row of contacts, and as the carriage moved, new contacts were made. I think it was a triple row of contacts. You could make three contacts at the same time. That fed into itself, so you would type the name of a star in there, and when you came to the end of the name of the star, it was on the next contact, and that contact would feed to the telescope, put in the necessary filter, and having done that, would send a signal back which started the next letter, which made a jump. That started the observation. The observation was done. When the observation was done, after so many seconds it would move the typewriter again, and the typewriter would then type out, read through another two of those contacts, the output of the digital voltmeter that read the integrated energy of the star, and so on. By the time you were finished with one observation, there was a line of 80 characters which had all the information that you wanted, and you could start with your next star.
Dick
Frank Drake tells the story about you doing something, digitizing at Project Ozma in Green Bank in '61.
Westerhout
That was not quite digitizing. There was a Hewlett Packard printer, and you read your data. We didn't have punch cards yet or stuff like that; it was a Hewlett Packard printer. You read your data, you printed the data. Every ten seconds there would be a printout. The operator was forever walking there, making notes on that printout. I said, "That's completely unnecessary." So I attached a wire to the thing that spaced it. In other words, you got a space. That wire went over the ceilings, through two loops, and hung in front of the operator, and the operator simply had to pull that wire every time a minute had passed, or every five minutes or so on, and that, of course, then got a space on the thing. Only once an hour did he have to write something on it, instead of every minute.
Dick
Was this during Project Ozma, then?
Westerhout
No, this was my survey of 3 cm, which didn't succeed.
Dick
But it was about that time?
Westerhout
It was the same time, just before Project Ozma.
Dick
This is off the subject, but do you have any recollections of what was going on there during the Ozma time?
Westerhout
No, because that happened later. They started to think about it. When I was there, they had an Iron Maiden. I forget what name. I think it was called the Iron Maiden. It was an unbelievably complex receiver system, backend system, to measure the radiation of the planets at 3 cm wavelength. It was not your ordinary computer, but three of them in great big boxes to be kept at constant temperature. The whole room was full with that stuff. Drake had invented that. I think he used it once.
Dick
Looking at Jupiter or something.
Westerhout
Yes. I think he used it a few times, and it was never appropriately used.
Dick
But then for Ozma, they brought in the parametric amplifier, so that was, I think, the first time that was maybe ever used in radio astronomy.
Westerhout
Was it the parametric amplifier for Ozma?
Dick
Yes.
Westerhout
No, I think it was the traveling wave tube.
Dick
No, it's parametric amplifier.
Westerhout
Okay. Because what I had was a traveling wave tube amplifier. That was one of the first ones. I mean, what I had on there which was for the 3.8 cm that Drake had installed to do the planets.
Dick
This is the 85-foot?
Westerhout
That's the 85-foot. I'm trying to think what frequencies Ozma used.
Dick
Twenty-one cm, I think.
Westerhout
Probably 21 cm. So they must have had the parametric amplifier already at that point, because Field was the one observer before me, other than the locals. Field was the first outside observer - George Field. I was the second outside observer in Green Bank. George Field was there. The telescope was commissioned in the end of March, George Field was there in April, I came in May.
Dick
And they didn't have much in the way of computing equipment there?
Westerhout
None. There was a 1421 desk computer which wasn't really a desk computer, it was about the size of this thing, in which you could program some things. But that was completely unattachable to anything, so you had to feed everything in by hand. That auto correlation receiver was a marvel of computer technology inside. It really did everything digitally, and out came the auto correlation function of the spectral line. So he had the 200-byte-per-inch stepping recorder, and it recorded that on the 200-byte-per-inch stepping recorder. That's where it stopped, because there was nothing that could read the stepping recorder.
So in the first experiments of the area, we would take the tape, or somebody would drive the tape to Maryland, where my computer programmer had programmed an IBM 1410-something-or-another to read the tape and do the Fourier transform. That was only a very short period, because they were in the process of installing an IBM 704A computing system at NRAO Green Bank. But as a result, a lot of the initial programs were programs that we had programmed in Maryland. But we were so keen to figure out whether everything worked, that Art Shalloway had a little vial with iron filings which we sprinkled on the tape, and with a magnifying glass we read bytes and started writing the bytes, and we soon found out, yep, what we put on there digitally is actually caught there.
Those were the days. Those were the good old days. Art Shalloway then developed a little gadget whereby he could read that tape recorder and input it into the desk calculator. A few weeks later, that's what they were able to do. They only had one tape recorder, so after he recorded the tape and then carreid the whole tape recorder to the lab where the computer was, put the tape recorder beside the computer, plugged all the stuff in, and then read the tape. This was all during the experimental work, of course, before we were on the air.
Dick
Is Shalloway spelled the way it sounds?
Westerhout
Yes.
Dick
Then when you went on to Maryland, did you have to worry about computers there?
Westerhout
This was in Maryland. This was in '63. I was in Maryland in '62.
Dick
So the astronomy department had its own -
Westerhout
No, the computer science center was always well ahead of things. They had good computers all the time. We were well equipped. For a long time we were considerably better equipped than NRAO. So that helped a lot in this very computer-intensive 21 cm line work that I was doing. I got line profiles every ten seconds. Initially those were 100-point line profiles. Later on they became 1,000-point line profiles. There's a lot of data in there, as you can imagine.
Dick
Now you have 10 million channels, multispectrum.
Westerhout
But we also have fast computers. So that doesn't worry me. But back to the Naval Observatory. Yes, I think the observatory had these big computers right from the start, and they went from one IBM to the next. I don't think they were very much behind for the size computers. We never went to really big super-type computers, although they didn't even exist in the 1970s. But I think the 360 - and then we had a 370; we upgraded it to a 370 - was at that time the state of the art, although IBM, as everybody else was working on much bigger ones, and indeed the one that was then replaced, the 4380 - no, no, there was one in between. I think the 4341. Then 81A, which is an upgrade, basically.
Dick
That was the last.
Westerhout
That was the last big one we had. Now the question that you had was how did that then develop into the business that we have now.
Dick
Because there must have been some considerable discussion.
Westerhout
There was.
Dick
Controversy, or not?
Westerhout
Well, in the first place, of course, already in the late sixties, the astrometry department acquired an IBM 1800, and so did Time Service. In fact, when the automatic transit circle moved to Flagstaff, an 1800 was installed there also. So there were three 1800s in the observatory.
Dick
The 1800 for the transit circle was there when I came in '79.
Westerhout
That's right. Sure. The Starscan measuring machine had a Hewlett Packard F Series, or even smaller than the F Series, attached to it. Starscan arrived in 1972 or so. So the various departments that were doing automated work did have computers, but they were all, of course, on-line computers.
In about 1972, and therefore each of the departments that had on-line computing had indeed that available relatively soon after they needed it. Of course, then in Time Service, all the clock programs slowly but surely got developed on that, but it was still very much the on-line work. In other words, reducing data as you go in real time, the same as the transit circle computer. In both cases, they then started developing programs where the computer could also be used to reduce data afterwards, still using partly the same programs, but going on from there.
I think that went until about the late seventies, when the first Hewlett Packards appeared. Those were Hewlett Packard instrumentation computers. Again, Hewlett Packard pioneered in having instruments attached to a computer, for a computer to control instruments.
In the transit circle department, it went slowly. They transferred one function after the other. First it was off line, then a few functions. Finally, the Hewlett Packard would talk to the 1800, which controlled the transit circle, and eventually then they had the Hewlett Packards control the transit circle itself. At that point the IBM was turned off. At Flagstaff, the IBM was turned off in 1985 or 1988 or so, very late in the game.
The first PCs started getting here again in the scientific departments, and that must have been around '84. They were strictly for scientific work. We had word processors - NBI word processors - in a number of the departments that were especially for word processing. As the PCs became more popular, it became clear that the PCs did at least as well as the NBIs for word processing. So then PCs - and that must have been then around '86 or '87, PCs started appearing on secretaries' desks. Then the Zenith contract was signed by the DoD, whereby Zenith had a big contract to be the sole supplier of the IBM PC, the IBM-clone PC, and that suddenly opened the floodgate, so everybody had the PC.
At that point, several people started worrying about proliferation, particularly proliferation of software. Everybody is going to reinvent the wheel. Everybody will invent their own software. So we set up a Computer Committee, and the Computer Committee, which originally had only looked at what should be programmed and what are the requirements for the IBM mainframe, now started looking into those sort of things. That was during Captain Anawalt's time, I think, or shortly before that, even. This was quite a while ago.
Dick
Mid-eighties.
Westerhout
Mid-eighties. Networks were still not heard of. We did not really get connected to the outside networks, which was before we were connected to the inside inventory, until about 1990, I would think, even though a lot of people were already using the networks to communicate with each other. We were a little bit late. Our first connection came when we got the VLBI correlator and that, of course, started sprouting. Everything kind of developed through the departments budgeting for it, somehow or another, nobody actually controlling it, because everybody defended the computer he needed on his desk, equipment needed for this special project and nothing else.
The first workstations came in about 1990. At that point we suddenly started screaming, saying, "We get the workstations, we must unify. There is no question that we should have workstations by four different companies. Everything should be the same." So Seidelmann, who wanted the Suns, said they should all be Suns. Winkler and Gauss, who wanted an HP, said they shall all be HPs. Nutile & Co., who liked DEC, said they were all DECs.
Oh, yes, wait a second. As part of the Space Telescope stuff, we bought a VAX. We bought a VAX in parallel with the IBM machine, and that VAX, which Rich Schmidt spent a long time on developing from scratch an imaging processing program, that was a big mistake. We should have imported those things. He did a lot of very beautiful work from scratch. Then the micro VAXs came, and they suddenly started proliferating in the observatories everywhere around the country. So I think we again, through the Space Telescope, got a few micro VAXs. But we were at that point already so wedded to HP that there was no question that the micro VAXs would take over as they did in some universities. Micro VAX never took over. We had two micro VAXs - one micro VAX first and the second one later, and that's the way it stayed. They're, of course, still in use [unclear].
So we had both the VAX and the IBM. The VAX we shut off first. The IBM, we had a number of long discussions about whether we indeed were going to go to distributed processing. Captain Hagen was a very strong proponent for distributed processing. So was I, but Captain Hagen was a lot more vocal than I was, because Captain Hagen simply ordered that there shall be distributed processing. But the idea was still the concept of everybody connected to everybody else, and using each other's workstations. We wanted two workstations and lots of terminals. Of course, that didn't work either, because the workstations became cheaper and cheaper and cheaper as these plans developed. But it was under Captain Hagen that the whole LAN network - the backbone LAN and the individual departmental LANs - were developed. In some departments, that went faster than in others, but I would say by early '92 or late '91, we pretty well were connected with a LAN, not perhaps through the whole observatory, but certainly everybody had his Local Area Network in the departments and the backbone LAN connected all those together.
Dick
That is still all centralized in the AA department.
Westerhout
That is still centralized in the AA department, and it should be, because there must be a central organization that takes care of the centralized stuff.
Dick
I guess the other possibility would have been to have another department.
Westerhout
And that's what somebody sometimes talks about, because the AA department director protests vigorously every now and again that that takes away all his resources. Of course, my nasty solution to that is to say, all right, we will take five people from the AA department and maybe one from Time Service and one from AD, and set up a new department, because AA was what was left from the Nautical Almanac Office after the R&D part was taken away, and the service part was left. If the service part now doesn't want to do servicing, all right, we'll make a separate department. That's my nasty solution.
So a very conscious decision was made about the distributed system, and sometimes people think that that conscious decision was made to get away from the rule that computing in the Navy was going to be centralized and that, in fact, we were going to report to the Navy Yard by '93 or '94. All our computing was going to be done at the Navy Yard, and we would be connected by lines and so on. Indeed, in '92, some of that was tried out with some of the programs.
With some of the programs we indeed were operating a computer at the Navy Yard through our terminals. However, that was an extremely slow connection and did not really satisfy anybody. Besides, we did not make much of an effort, because we had already decided the IBM machine was going to go. Now, some people think that that was our subterfuge of getting away from going to the Navy Yard. It was definitely not. It was an additional item that came in there, but it was definitely not a subterfuge. I want to make that point very strongly. It was clear we were going to distributed computing. Every modern workstation is already better than the IBM machine we had.
Even four years ago, five years ago, maybe, we didn't think that. Five years ago, we were still thinking that for a number of things, including ephemerides and so on, we would need a much more powerful computer, and we talked about a small super computer. In fact, we put funds for that in the OPN budget, requested funds, and got it for FY-'93. FY-'93 was the year we were going to get a small super computer. That money, $2 million or so, was there when we decided to not have any computer at all anymore. Therefore, it was distributed among more workstations and partly used for other purposes, because there's no way you can spend $2 million worth of money on workstations and so on. So a lot of that simply disappeared out of the computer domain and went into other things. We reprogrammed some of that OPN money.
Dick
So that's a real turning point in the observatory when you go from centralization.
Westerhout
Absolutely.
Dick
As far as people's work habits, the way they do their daily work.
Westerhout
Yes, absolutely. However, I would say the development of the mainframe computer, whether it was the IBM which had terminals all over the place or the AD Hewlett Packard or the Time Service Hewlett Packard, they all already had terminals, so people were always used to not working with a mainframe computer as a thing where you brought in a batch of cards, but programming it yourself and everybody collecting somewhere a reel of tape or putting on a new reel of tape. That was already there for quite a few years, because, as you know, the IBM terminals, the terminals to the mainframe computer, were all over the place. In fact, that's when we started having terminals, we started the bulletin board system, which, when the IBM disappeared, was put on the micro VAX, which nobody can access, and therefore the bulletin board system is still there, but it's basically disappeared because you have to have all sorts of security codes for Mr. Nutile. He wants you to change your password every six months and all that baloney, so forget about the VAX.
Dick
Is there anything else we want to say about computers? We're almost at the end of this side, maybe one minute.
Westerhout
At the moment, I don't think I want to say. There's a lot more to say when you start discussing it, but for the moment, for this interview, let's leave it at that.
End Tape 5 Side A
Begin Tape 5 Side B
Dick
Another item I wanted to cover was the mission of the observatory. Would you say that the mission has changed during your tenure? Or maybe, first of all, you should state what the mission is, just for the record.
Westerhout
No, I'm not going to state what the mission is for the record, because that's written up, but I will state that the mission changed during the days of Captain Roberts. Somewhere halfway through his tenure, we changed the mission and got it approved all the way up to SECNAV. The way we changed the mission is that we added to it some of the words recording the fact that we were working not only for the Navy, but for the DoD and for the general public, and we added the words that we could do research. The important aspect there was providing astronomical and timing data and giving the reason why we did that. That was a change in the wording. It was what we were doing, but we felt strongly that should be in the mission statement, and that the important thing was to get the word "research" in, to conduct relevant research, which is very unusual in the mission statement of a second-echelon operating command, to conduct relevant research. We got that through. As I say, that was in the middle of Captain Roberts' reign.
We also changed the function statement rather thoroughly. You know we have this statement of function, and that you must simply look at and compare it with the previous one, which I don't remember anymore.
Dick
That was 1984.
Westerhout
Right. You may remember also that in the previous mission statement, there was the words "celestial objects, both natural and artificial." We removed the "artificial" again.
Dick
Why was that?
Westerhout
Basically we felt we were not in the satellite-tracking business. Apparently the mission statement was changed in 1958 or thereabouts when Sputnik went up, and the word "artificial" was added there. In fact, the little dome on top of the Time Service Building was a satellite-tracking antenna, satellite-tracking telescope. Antenna? No, telescope, probably. I forget what it was. That was before my time. Of course, the dome was completely empty when I got there.
Dick
So do you still feel we're not a satellite-tracking organization.
Westerhout
I still feel we're not a satellite-tracking organization.
Dick
Isn't Peter Kammeyer working on satellite tracking?
Westerhout
No, no. Let me just elucidate that. You may wonder what is Seidelmann then doing in all this astrodynamics.
Dick
Right.
Westerhout
We are the standards organization. We will provide the standards by which people whose mission it is to track satellites can do that. We provide those standards in the form of the best possible orbital theories, harmonics to the nth degree, which is stuff that Kammeyer and several others are working. You see the titles of those papers, your head nearly bursts because you can't even pronounce the words. And giving advice as to how to use these marvelous means that are directly at their disposal to get much better data.
To just give you one simple example, the GEODDS system is run by the Air Force, and it's the classical optical satellite-tracking system. It's supposed to find all the satellites it wants to find and track them and determine their position. The way it does that, it has the forerunner of the CCD in it, an electronic imaging system that does not use CCDs but CIDs, charge-injected devices, or something like that.
Dick
Never heard of them.
Westerhout
Electronic backend, in other words, very fast moving telescopes, nice telescopes. So what they do is they move the telescope to where they think the satellite is, they find the satellite, they track the satellite, they move the telescope at the speed of which the satellite goes, reads the coordinates on the telescope, and say, "That's where the satellite is." While they're tracking the satellite, the stars become streaks, and that's always a nuisance. How the hell could you get rid of those stars? That would be really nice. That's a nuisance.
So how then do you do it better? The accuracy of those things is typically 15 seconds of arc or so on. What Kammeyer did was prove what using the CCD in the scan mode will buy you...have the telescope stationary, and as the satellite walks through...this telescope, in other words, follows the stars. As the satellite walks through the field, you move the charges such that the satellite remains a point source. If you move those charges, the stars, of course, become streaks because of the charges of the stars. So you have a dot-shaped satellite and streaks of stars, but now you stop that and you simply move with the stars. Now the stars become points, so you get a little streak of star and then a point.
So the way you do it is you get a little streak of star, you get a point for the star for a while, now comes your satellite. Now you follow the satellite. It's a point. It was a streak first, because you were following the star. Now it becomes a point. Now the stars become streaks. Now the satellite becomes a streak, and the stars become points again. Say there's two stars. You have four points for those stars and one point for the satellite. Of course, suddenly you can determine that with an accuracy depending on the accuracy of the catalog, but with the primitive means that we used here to much better than half a second of arc.
Dick
So the problem of tracking two things that are moving.
Westerhout
On the same frame.
Dick
It's like the Markowitz moon camera. They had the same problem of tracking the moon for ephemeris time.
Westerhout
That's right, except you can now do it entirely electronically by simply transferring the charge. That camera, I wrote a paper or two about. We tried that out. We got a tiny little bit of money, and we had the CCD camera already, so that was a very nice practical use of the CCD camera. With that we have, since for the last two years, been trying to get a million and a half, to prove that as a prototype, to build one prototype tracking camera that could do both high altitude and low altitude devices, but there is a factor of 20 to 100 improvement in accuracy.
Dick
Why can't you get the money if it's such a great new idea?
Westerhout
Because the people who operate those GEODD systems look in their book and they say, "Here are the requirements. The requirements say 15 second of arc." That's what we're getting. Those are requirements written 25 years ago when you couldn't do any better. So that's when the system was built. It's the R&D people who say, "Boy, this is great!" But they need R&D money for their own purposes, so it's a vicious circle. I'm sure eventually we'll get there, but that is how we are involved in satellite tracking, showing people, because now see the consequence of this.
In order to do that with faint satellites you can look at objects down to 10 cm size and determine their orbit with this fantastic accuracy in one go, because you get this position of the satellite and, of course we're moving, so you take two more with respect to a different star field over the next five minutes, and that's more than sufficient to get you an orbit. That is far more accurate than anything that anybody's ever tried with radar and so on, and you can do that for lower earth orbit debris down to 10 cm. That's pretty good going. Therefore, we think that also for NASA, this system has all sorts of future. But what do you need? Star background. For these things that are 15th or 16th magnitude, which is easy to do with the CCD, if it takes the thing a number of seconds to walk across, that's already good enough with a decent-size telescope. You need a star background with that accuracy down to 15, 16, 17, 18th magnitude.
Who is supposed to provide that star catalog? So by educating the satellite-tracking crowd, we now hope that that satellite-tracking crowd comes around and says, "Wowee! These guys had better get on it and start writing requirements for a major star catalog at 15th to 18th magnitude." And that's the thing we're after. We aren't sitting still; we are going to produce that catalog anyway. Indeed, we've already talked the Air Force into realizing that this star catalog is necessary. The Air Force is providing $200,000 a year to Flagstaff to operate the big measuring machine for the Palomar Sky Survey plates as part of an R&D effort at this point, but for this specific purpose.
So we're making headway. In other words, we're selling our product in a sense, because we know that at some point it's needed, but we have to make it clear to the people that need the product as to how they should handle that and what they can reach. I noted, when I talked to some of these people, that...because when you talk about a fraction of an arc second in lower earth orbit, you talk about 2 meters, and I told these guys, "You know, in this system you can see the satellite maneuvering." This was a number of years ago when the Russians had things. They said, "What?" I said, "Yeah, you get 2 meters' accuracy. If this guy suddenly jumps by 2 meters, you see that right there." "What? Ooooh!" You plant thoughts in people. I don't know whether they're interested in maneuvering satellites or not, but it was clear that that word rang a bell to people. That's the way.
Commander Bollwork, of course, has now been appointed to do that full time, and he does that with absolute verve and enthusiasm. I've been wanting a requirements officer like that for years and years and years. We finally got one. He's only here for a year, but, hey, he's going to do a lot of good things.
Dick
Anyway, this program is an example sort of a symbiotic relationship between the pure and the more practical aspects of what we do in a lot of areas, wouldn't you say?
Westerhout
Absolutely. Absolutely. Of course, in the time area it's the same sort of thing. We are telling people why they need precise time, and that they should not ignore the fact that they need the precise time. They need that connection. We give examples. A typical example that happened shortly after I came here was when our clock team - remember in the old days they traveled around with portable clocks - went to Rota, Spain, and found the clock in Rota was off by milliseconds, not just by a few microseconds. It was way off. They said, "Yeah, this clock is sort of behaving badly, but we reset it every now and then."
"How do you reset it?"
"Well, we ask one of the submarines to surface, and we set our clock by that." And this was the station that was supposed to calibrate the submarines! That caused a lot of stir around, because we made sure - Winkler in particular made sure that that message got around to people, that that's definitely not the thing to do.
Dick
So artificial satellites are not in the mission statement, but we are interested in them.
Westerhout
No, we're interested in them as users. Trident missiles are not in our mission statement either. We're very interested. The missions to other planets are not in our mission statement, but we're very interested in NASA as a user of our product. And so on and so forth. But our mission says what we do, and we do it for the general public and NASA and other government organizations.
Dick
One change I noticed was that the old statement used to say "contribute to the advancement of astronomy."
Westerhout
Right.
Dick
That was dropped. Is that significant?
Westerhout
That is somewhat significant, because we felt that contributing to the advancement of astronomy made this observatory perhaps a little bit too suspect. I bet you caught that one. In the sense that we did not think that a statement like that in the current climate you could expect to get CNO-type money easily, that people would start looking at that and say, "What do you do for the advancement of astronomy?"
Dick
We'll cut.
Westerhout
Exactly. That doesn't mean we don't do it; it's a spinoff. Of course, the relevant research is basically the advancement of astronomy in a sense. The word "relevant" also has to be looked at with a grain of salt. In addition, in order to keep research astronomers, all research astronomers have 20 percent of their time to do whatever they damn well please, provided they're doing astronomy and not go look at chickens. There are a lot of people in the observatory who have simply far too much to do to provide that time, either that or do not really have the initiative to use that time on their own accord. Then there are also the lucky ones who are doing an observatory project which becomes the Ph.D. thesis. We have that thing, too. But people forget a little bit that that opportunity is always there, and that people should do it much more, and they don't. And that's unfortunate.
At Flagstaff, of course, it's just the other way around. They do it with verve, because that's how we have been able to attract prominent scientists on the staff, by offering the possibility of doing quite a bit of their own research as well, but never losing track of the fact that they are supposed to put 80 percent of a day into what is ruled by the observatory as being important work.
Dick
So why is there that dichotomy between Flagstaff and Washington here? Wouldn't you say there was a dichotomy?
Westerhout
There is a dichotomy because in Flagstaff we have telescopes. In Washington, we don't have telescopes. These people all come to study objects at 18th magnitude or to do spectroscopy and so on. We don't have any of that equipment here, nor has there been any initiative to use it. We had this beautiful Boller & Chivens spectrograph here for 25 years. We finally gave it away. We gave it to Northern Arizona University.
Dick
What about the 24-inch telescope?
Westerhout
Nobody made use of the 24-inch. That was brought in. All these instruments, the 24-inch, the automatic transit circle, big PZT, Starscan, were all bought in the Vietnam [War] era, when there was lots of money all the way around. Flagstaff was built up in the Sputnik era. The Sputnik era came and went, and suddenly things went way down, and then came the Vietnam era, and suddenly the budget went up again.
Dick
How do you justify spectroscopy for the Navy? Would the Navy have wanted people to use spectrographs?
Westerhout
I don't think it was part of the element. Besides, you could always suggest that in order to research the colors and the constitution of the stars so that you know better which stars to use for navigation and which not, I mean, that's already a statement that usually holds water when you talk to a bean-counter. That's the only people we talk to for those services. We never talk to scientists or oceanographers or anything about that; we talk to bean-counters. Those are the ones who have to believe what you say, and that's a whole different kettle of fish.
Dick
Do you think we differ from other agencies like NRL?
Westerhout
Oh, absolutely, because we're funded by the chief of naval operations, whereas NRL, by its charter, is supposed to do basic research which may be relevant to Navy needs.
Dick
So I want to make that point. We are different in the sense that we're not a Navy laboratory.
Westerhout
Absolutely. We are very different in that we're not a Navy laboratory.
Dick
We're lucky!
Westerhout
In a sense, yes, but in another sense, we have to do the research or our product will stagnate, because there ain't nobody else who does it. NRL does some, but NRL doesn't do all the nitty-gritty type of research that needs to be done.
Dick
But in a sense, we're operating under a different set of rules than naval laboratories like NRL.
Westerhout
We operate under a very different set of rules, and I remind you again of the outcome of the visiting committee in 1970, with Willy Fowler and Bart Bok on it, and a few others, who were so enthusiastic about what the Naval Observatory did, that in their report they recommended that the Naval Observatory should be removed from the CNO and start working for ONR. And every Tom, Dick, and Harry started jumping up and down and screaming, "For God's sake! Don't let that happen!"
Dick
This was in 1970. This is when Strand was -
Westerhout
1970. Because everybody knew that if you do that, there's no way you can keep your long-term programs. You could not run a Time Service for more than five or ten years. You certainly couldn't run a transit circle for 80 years, because ONR expects you to go in a different direction. I see that time and again in my defending what we are doing in the R&D area, because the ultimate aim of everything we do in the R&D area is to improve positions of the stars or the reference frame, to improve the Earth's rotation and to improve the master clock.
Every now and again, particularly from people who know, in hearing that spiel, I try to say it as little as possible, but I have to. You always have to come up with an end user. Every now and again, someone pipes up and says, "When the hell are you going to do something else? You're always doing nothing but improving star positions. Can't you do anything else but star positions?" My answer is, of course, "No, sir. That is specifically what I'm doing in my mission. I'm not even allowed to do anything else." Well, that shuts them up.
But that is the R&D attitude. For God's sake, if you spend $5 million on inventing new paints to paint a hull, the next time you're going to do something else. You may still be investing in paint, but no longer for hulls. You've exhausted that subject. You've got to paint masts or whatever.
Dick
But in a sense, this argument goes back way to the beginning of our history, when we were declared both a naval and a national observatory.
Westerhout
Right.
Dick
Do you think the Naval Observatory, with pretensions to being a national observatory, should go beyond the naval aspect?
Westerhout
With pretensions to be a national observatory? Of course it should. We don't have a pretension to be a national observatory in other than the fields in which we are prominent. We're not prominent in astrophysics, not at all. We have a few people who work in astrophysics. We're somewhat prominent in celestial mechanics, but that's a direct result of the fact that that's in our mission, basically. We're prominent in astrometry, and there's nobody else who's prominent in astrometry in this country. So by force, we are the national observatory in that area.
The question is, will you do the same as the Royal Greenwich Observatory? That, of course, is the end of astrometry as you know it, so that's not what you do. So becoming the national observatory in the days of astrophysics means a whole different kettle of fish, and that would automatically basically negate what we're doing, because, hey, we would do all the work that Fugate and Bender and all these people are doing in Phillips Labs in New Mexico and in Hawaii. They are doing all sorts of beautiful infrared work. They have infrared telescopes coming out of their ears, and they're doing astronomy, and they published astronomy papers. How do the guys at MIT have contacts with these Air Force people? But they're doing astronomy papers, but they're doing it because they're developing equipment to see rockets burn and to see hot objects in the sky. That's where all this comes from. They are building beautiful equipment.
Dick
We do differ from some of the classical national observatories like Paris and Pulkovo, which branched out. They still have some of their original departments, but they branched out to.
Westerhout
They did. That is right.
Dick
Is that a good thing or a bad thing?
Westerhout
They were a national observatory. They were not working for a service. Yes, the Paris Observatory was, because the Superintendent was an admiral. That's right. Is that a good thing or a bad thing? It's a good thing. If we could be as big as the Paris Observatory, hey, I mean, the French government spends a lot of money on astronomy.
Dick
In a sense, things have developed differently in this country, because we were the original national observatory and then 100 years later came KPNO and NRAO, and they became the national observatories. In a sense, those functions were taken on by other national observatories, whereas in Paris Observatory it's different.
Westerhout
So is everything else. The French university observatories all look to Paris, because that's the whole French structure, not only in science, but in everything. It's completely centralized from Paris.
Dick
So part of this is rooted in national history.
Westerhout
And Meudon is not an independent observatory; it reports to Paris. They're going to close it one of these days, I saw. We have a lot of contact with in Grasse, near Nice, where they're doing Earth rotation work. They have lunar laser ranging. That's where Kovalesky sits.
Dick
And they're related to Paris Observatory?
Westerhout
It's the Centre des Etudes et Récherches Physiques et Astronomiques. They're not related to the Paris Observatory, but is a center that reports directly to the central astronomical organization in Paris.
Dick
But the Paris Observatory comes under the Ministry of Education.
Westerhout
Yes, so do all these others.
Dick
Okay.
Westerhout
And in the Ministry of Education - and don't ask me the names, because I've forgotten them - but I met all these people when for six years I was on the Council of this Stellar Data Center in Strasbourg. Members of this central French organization were always there, and they were the guys that it was clear you had to satisfy, because those were the people that eventually decided how much money was going to Strasbourg.
There were also a few teaching positions in Strasbourg, but that was a minor part. Most of the researchers were paid from Paris. That's the case here, too, except the country is bigger, so we have NASA, we have NSF, we have the DoD. It's spread more. But look at what happened to the transit circles at all the various universities - Cincinnati. Princeton even had a transit circle. I forget what. There must have been about five observatories in the country that used transit circles. They were all closed, because that is not considered R&D in which you can get a Ph.D. thesis. It doesn't work. Not part of education anymore.
Astrometry in general is not part of education, with a few exceptions. People like Cudworth and Hanson, Lutz, and so on. They're all singular in their university.
Dick
So aside from these isolated instances, now the Naval Observatory upholds the astrometric tradition for the U.S.
Westerhout
That's right. Absolutely. Not only tradition - never mind about tradition - we do the astrometric work.
Dick
You mentioned the RGO a few minutes ago. What's your opinion about where did they go wrong, and where do they stand today?
Westerhout
Well, my opinion is very straightforward, but that's naive. When they left the Navy, that was the end of their work. When they left the Navy, they still did transit work. Tommy Tucker was the last one to operate the transit circle in Herstmonceux, and he was very unhappy because they moved the beautiful transit circle from Greenwich to Herstmonceux, and it never worked the same and so on. He finally retired an old man, and that was the last time a transit circle was operated. Leslie Morrison never operated the transit circle until the Danes built one for him in La Palma. He's become a very well-known transit circle observer, but that's by the grace of the fact that it's an international collaboration with which the Brits collaborate for the sake of international collaboration.
Dick
Some people would say by leaving the admiralty, the RGO freed itself to do more interesting things.
Westerhout
Absolutely. That is quite correct. They are the national observatory. They run all the big telescopes. They're a telescope-operating facility. I hear that more and more from the RGO people, and I heard it very strongly a few years ago when I was in Edinburgh. Edinburgh is in charge of running the Anglo-Australian telescope, and basically three-quarters of the staff is service staff, like at NRAO and NOAO - namely, making sure that all the equipment works, and occasionally they get a little bit of observing time, but they see all the nice projects being done by university astronomers who come from elsewhere. That's what RGO is rapidly going to, also.
Dick
So it the last 50 years of RGO's history the story of the demise of RGO or the progress of RGO?
Westerhout
No, the demise of RGO.
Dick
Why?
Westerhout
The demise of RGO as an observatory that produces its own research. It's an observatory that facilitates the research by others. You may say, hey, Kitt Peak Observatory is a marvelous observatory, and that is indeed the way RGO is going, because I find that the demise of an observatory which was proud of having its own publications and doing its own things with its own staff, like the good old Leiden Observatory, it was never dependent on anybody else, although they are now, of course. They use other people's satellites, and they have collaborative efforts, but they have a telescope at La Palma, they have a telescope on Hawaii, and they send their students up there and so on. When I was a student, of course, we had a telescope in South Africa.
So that was the good old-fashioned observatory, and at quite a number of observatories that still works, whereas these national observatories, they're beautiful places if you're a well-heeled scientist who by now has tenure and is allowed to spend half his time or a quarter of his time on his own research. Or you're a postdoc who is specifically hired by the place to do nothing but research, because they give some postdocs, to keep the rolls going. But in general, the national observatories are not what I call the good old-fashioned observatory like Santa Barbara Street or Cal Tech. Entirely different mentality.
Dick
In a sense, RGO has largely lost its identity, wouldn't you say, as being absorbed by the Cambridge University?
Westerhout
Right. That's what I mean by demise. It doesn't mean that the observatory's gone, but as an institution it - the move to Cambridge was sort of natural. I mean, that was to come sooner or later in the way they were going. They had to be around the astrophysicists. They couldn't be sitting there in splendid isolation in Herstmonceux.
Dick
In a 15th century castle.
Westerhout
Right. There were a lot of modern buildings around it, too. Have you been there?
Dick
Once, just before they moved. You sent me there.
Westerhout
I did? That was nice of me. Why did I send you there?
Dick
To use the RGO archives now in Cambridge.
Westerhout
Has Brenda given you a report on all the discussions she had with the archivists and so on?
End Tape 5 Side B