Barry G. Clark, Interviewed by Kenneth I. Kellermann on 18 November 2024

Kellermann:  00:01

Okay. This is Ken Kellermann. I'm with Barry Clark in Socorro. It's November 17^th [2024]. Is that right? And we're going to discuss Barry's--

Clark:  00:16

November 18^th [2024].

Kellermann:  00:18

18, November 18 [2024]. We're going to discuss Barry's career at Caltech and then at NRAO. So to start with Barry, tell me more about the years before Caltech. I know you grew up on a farm and came to Caltech knowing how to drive a tractor. What else?

Clark:  00:40

That's about it.

Kellermann:  00:44

Keep this near you [the microphone].

Clark:  00:45

I lived on a farm near Happy, Texas, until I was 7. Then we moved to Canyon because it had a better school. And there were two of my older siblings still in school at that time. I'm the youngest of the family. And so the school I went to was associated with what was then called West Texas State College, newly renamed from West Texas State Teachers College. We were to be practiced upon. They unified with the public schools and they stopped supporting the school at the time I entered middle school, and thereafter, it was just the normal public schools. In high school, I had a high school physics teacher who I count as one of the great mentors of my youth. And, well, the thing that started it was I entered what was then called the Westinghouse Science Talent Search --

Kellermann:  02:28

Oh, yes. I'm familiar with it, right.

Clark:  02:33

--more or less the equivalent of today's science fairs.

Kellermann:  02:40

What was your project?

Clark:  02:43

It was extremely stupid. It was orbit calculations for interplanetary travel, and as I said, they're very stupid. I knew no calculus, so I could only work with things that Kepler knew about, and this project submission was accompanied by a test, and I suspect that I did very well indeed on the test because, as I said, the project was stupid.  [Post-interview comment added by Clark:  My first year or two at Caltech, the national show of the Westinghouse Science Talent Search was in LA, and I went.  Exhibits there were a league beyond my project.]

Kellermann:  03:26

Well, only with the hindsight of your future.

Clark:  03:30

Yes. Well, anyhow, I had an honorable mention in the regional show . I'm not sure what the region was, but anyhow, my physics teacher said that that's good enough that I should try applying to prestigious universities. So I did that. Caltech and MIT, of course, were familiar names. And I also applied to University of Chicago.

Kellermann:  04:12

Did you know then that you wanted to do science, I guess?

Clark:  04:15

Yes. Yes, I have known since maybe fourth or fifth grade that I wanted to be a scientist.

Kellermann:  04:24

Did you know what a scientist was?

Clark:  04:26

Sort of. Yes. My first love was geology. Looking at rocks, finding interesting rocks, dinosaurs, what have you. But my older brother, who was nine years older than I, was also very interested in science. He graduated from high school shortly after I started school. And so he had a chemistry lab in the basement and that sort of thing. And we made a balance for determining specific gravities of rocks out of tinker toys and that sort of thing.

Kellermann:  05:31

So I presume you were accepted at MIT?

Clark:  05:33

Yes, I was.

Kellermann:  05:34

So what motivated your choice?

Clark:  05:38

I was sort of scared by the size of MIT. Caltech, if anything, is ridiculously small.

Kellermann:  05:54

Yeah, right.

Clark:  05:56

My high school graduating class was 65 people. And that was the year the Conant Report on Secondary Education came out and said you can't possibly get a decent education with less than the graduating class of less than 100. But the only obvious problem with the size of the school was that they didn't offer a trigonometry class.

Kellermann:  06:34

Oh.

Clark:  06:35

And I had to take that by correspondence.

Kellermann:  06:39

So that means you didn't have calculus or even advanced algebra?

Clark:  06:45

Well, we had reasonable algebra. At least semi-reasonable algebra. We talked a lot about determinants, which are interesting theoretically, but they're very impractical.

Kellermann:  07:06

So that means you only took, what, two years of math in high school or?

Clark:  07:11

Three years.

Kellermann:  07:12

Three. Okay.

Clark:  07:13

Yes. Introductory algebra and simultaneous equations.

Kellermann:  07:22

Because, I mean, even at that time, the norm to go to Caltech or MIT was four years of math.

Clark:  07:31

Plane geometry in high school was interesting because it took me maybe two weeks into the course to realize that it wasn't really about geometry, but it was about logic.

Kellermann:  07:52

Yeah. What area of science were you thinking about at the time?

Clark:  08:01

Early in high school, I got interested in astronomy. I had an uncle who was just retired and had taken up amateur astronomy as a thing. So he showed up with his 7 by 50 binoculars, and we looked at a few things, which I found interesting.

Kellermann:  08:32

So you were thinking of astronomy right from the beginning of your undergraduate work?

Clark:  08:37

Oh, yes.

Kellermann:  08:38

I see. Okay. And then what year were you when you started to work with John Bolton?

Clark:  08:50

After my sophomore year, I started looking for a summer job and went and asked Jesse Greenstein, and he said, "I don't have anything, but radio astronomy might." And so I went and talked to John Bolton.

Kellermann:  09:12

Had you heard of; had you ever heard of radio astronomy, or…?

Clark:  09:16

I had heard of radio astronomy, yes. I don't think I had heard of John Bolton. But then that first year, as I said, was mainly driving a tractor, clearing brush, digging post holes for a low-frequency interferometer.

Kellermann:  09:47

Oh, right. So I think I read in your interview with Woody. I hadn't realized that that array was not built for the Crab occultation.

Clark:  10:03

No.

Kellermann:  10:04

Yes.

Clark:  10:07

It was sensitive enough to see exactly four radio sources. So we milked them for whatever they were worth.

Kellermann:  10:18

And so you worked on that after your-- I mean, the summer after your sophomore year, and again, after your junior year, when was the occultation? It's every year.

Clark:  10:33

Yes. The one that was published, I think, was probably a year and a half after it was built.

Kellermann:  10:45

So I read the paper. I was surprised to see that the elements were full-wave dipoles. Do you have any idea-- I mean, it seems to me the natural thing would be half-wave dipoles. A full-wave dipole is non-trivial to feed.

Clark:  11:04

Yes. He wanted the narrower element beam to be able to separate the Crab from M87.

Kellermann:  11:17

Okay. During that time, radio astronomy was involved the 32-foot dish at Palomar. And do you have any idea why they built it at Palomar? I mean--

Clark:  11:37

They built it at Palomar because Caltech had a presence in Palomar, and that was before they had Owens Valley.

Kellermann:  11:44

Ah, okay. All right.

Clark:  11:48

And again, they just wanted to get a foot in the door.

Kellermann:  11:51

Okay. I didn't realize that it did not have the Owens Valley then because, as we know, tops of mountains are not good places for radio telescopes.

Clark:  12:03

But sometime in there, I guess in probably late '57, we had an expedition to Palomar led by Dan Harris, which disassembled that telescope, got it ready, and to put on a truck and drive to Owens Valley.

Kellermann:  12:35

Okay. So you were involved in that?

Clark:  12:37

Yes. And the surface, which had been a wide mesh, was to be a fiberglass surface with aluminum foil glued to the surface, which sounds very adhoc nowadays, but that's what it was. So I helped spreading fiberglass resin on the form.

Kellermann:  13:12

Well, I spent part of one summer taking off that aluminum foil.

Clark:  13:20

I don't think anything ever came of that telescope.

Kellermann:  13:25

But there was talk of using it as a portable antenna with a radio link just running up and down the valley. I've always been disappointed. I think I even pushed that at the time, thinking one could even use amateur radio to provide the link. But it never happened. I think Dick Read was interested. That would have been very useful, I think.

Clark:  14:00

Yeah. We certainly got a lot of information out of the Owens Valley experience.

Kellermann:  14:11

Well, I didn't mean so much the technical experience, which the case of NRAO is valuable for the VLA, of course. No, but I meant astronomically.

Clark:  14:22

Yes. I’m trying to remember when I saw Al Moffet’s paper on the discovery that everything was double.

Kellermann:  14:37

That was either during our first or second year. It would be '60, '61. And John reported on that at the URSI meeting, I think it was '60, summer of '60.

Clark:  14:56

Yeah. It really took that before it became obvious that long baselines were helpful.

Kellermann:  15:05

Yes. No, I was thinking about it during my later years when we knew all that. All right. Here's a question, Barry. OVRO had two movable antennas years before Cambridge. Did we miss out on doing proper aperture synthesis and why? I think we've discussed this, and I certainly have discussed it with Al.

Clark:  15:36

Yes. We used the equations of aperture synthesis. We couldn't do very good rotation synthesis because of the 60-degree hour angle limits. But I remember doing the calculation of where is this baseline in the UV plane to decide whether we needed to move an antenna to get there.

Kellermann:  16:19

I wasn't talking so much about the complete full beam synthesis they did at Cambridge. But the individual sources, where you don't necessarily need to have the full hour angle tracking. I remember it must have been our first year taking, --- no, it was the radio astronomy course.

Clark:  16:52

From Westfold?

Kellermann:  16:54

No, no, no. Did John ever teach the class?

Clark:  17:04

I don't know.

Kellermann:  17:05

I don't think so. So Maarten Schmidt was the official instructor, but in practice, Al Moffet did most of it. And he was just finishing up his thesis. And I remember he had us do a one-dimensional synthesis, including amplitude and phase data. But after the first year, there was a north-south interferometer, and there was a little bit Moffet and Maltby had some observations, but it was still all model fitting that they did.

Clark:  17:50

With two antennas, model fitting is about what you can do.

Kellermann:  18:00

Well, Cambridge had three antennas.

Clark:  18:04

Yes. Makes a big difference.

Kellermann:  18:07

Well, they weren't using any phase closure or anything. So three antennas.

Clark:  18:16

No but three times as many baselines.

Kellermann:  18:18

But still I've always thought Caltech missed out on that. Al claimed that it didn't have the computing capability, but I think that's not true.

Clark:  18:35

I don't know what machinery Al used for his fitting, whether it was a calculator or a slide rule.

Kellermann:  18:50

I have vague recollection of you and I going to evening classes with a Burroughs computer. And I quit after a week or two because I couldn't see putting a number, register A, and moving it to B, and then adding C. But you stuck with it.

Clark:  19:22

Yep.

Kellermann:  19:23

But even before I left, there was an IBM something, 709, I guess, or 704, which ran Fortran and I used it.

Clark:  19:37

Yeah. Did we have one on campus, I think not?

Kellermann:  19:42

Yes, yes.

Clark:  19:44

It started out by sending a deck of cards to JPL.

Kellermann:  19:51

No. I remember going over somewhere, I think. Well, even if it was at JPL, that would have been sufficient.

Clark:  20:04

That JPL had a 7090. Wow. Almost as much power as getting a greeting card these days.

Kellermann:  20:20

So your thesis topic is very different from what the rest of us were doing. How did you get onto that?

Clark:  20:31

Well, when John was about to leave for Parkes, he divided the graduate students and gave me to Rad [Radhakrishnan], who, of course, is the third great mentor of my youth. And Rad was interested in the neutral hydrogen. Rad was interested in a remarkable variety of things.

Kellermann:  21:01

Right. And done a remarkable amount. So it was Rad that sort of steered you.

Clark:  21:14

As I said that was given to Rad. I have always gone with the flow.

Kellermann:  21:25

In your interview with Woody, you suggested that the development of the mechanical fringe stopping was developed independently at OVRO and Jodrell. I think that's not right. I think Dave Morris was the link.

Clark:  21:50

I see.

Kellermann:  21:51

And I specifically remember -- there are certain things that embedded in your mind -- of Dave working on something, staying up all night when I went to bed. And the next morning, he had it working. It was levers and--

Clark:  22:16

Yes. Yeah. He was like that. He would get something stuck in his teeth, and it would stay there.

Kellermann:  22:33

Until it was done, right. So both you and I were present at the time of the discovery of quasars in 1963. Do you have any personal recollections of what was going on with Schmidt, Greenstein, and--

Clark:  23:02

Not personally. Mostly got things second or third hand.

Kellermann:  23:12

And Tom Matthews was the other key person. And I have talked to Tom. And I've never known any discovery where the people involved seem to have gone out of their way to say, "I didn't do that. Maybe so-and-so did." As you know, or may know, the occultation position was wrong because they were using the coordinates of Sydney rather than the coordinates of Parkes to time the occultation. So they had the wrong position for the moon. And that incorrect position was the one that John Bolton sent to Maarten Schmidt and the one that he used for the Palomar observations. It wasn't until afterward that they sent him to the right position. 3C 273 was one of the 10brightest sources in the sky. There must have been a-- well, there was a Caltech, an OVRO position. But it's not clear who actually pinpointed the radio position with the stellar object.

Clark:  24:55

I see.

Kellermann:  24:58

Both Schmidt and Matthews don't remember. Okay. And I, too, I was busy with my thesis. And so I hadn't paid any attention at the time. I got very interested in that subject afterward. Now, while you were still a graduate student, you got involved in the Owens Valley Array, the design?

Clark:  25:36

I don't remember getting very much involved with that. After I came to NRAO, I got more involved because I felt it would help improve relations between OVRO and Caltech

Kellermann:  25:55

Ah, I see. What's your understanding of who is the intellectual driver behind the Owens Valley Array?

Clark:  26:13

Probably Al [Moffet]. It was not entirely clear who was doing what, but Al is the guy with sufficient brain power to do it. And it was not clear that anybody else had both the brains and the inclination.

Kellermann:  26:37

Well, I think John Bolton had-- I remember him sketching out four movable antennas that would move out in a cross, which is kind of elementary.

Clark:  26:55

That was years before?

Kellermann:  26:58

Yes. It was before he left. But I always thought of that as the foundation that other people built on. But I agree with you about Al. It wasn't Gordon [Stanley]. You were clearly one of, if not the most, outstanding graduates in astronomy. Were you ever asked to come back?

Clark:  27:45

Not really. As I said, I've always gone with the flow. I had a happy, happy place, and I was happy to stay there. And if somebody did suggest that I might like to go somewhere else. I wasn't very responsive.  Govind Swarup tried a couple of times to recruit me.

Kellermann:  28:09

But you don't recall a Caltech effort?

Clark:  28:10

No.

Kellermann:  28:12

Don't you think that's strange? I mean, and do you have any thoughts about why Caltech had difficulty finding a more qualified director?

Clark:  28:31

Oh, I'm not qualified to be a director.

Kellermann:  28:34

No, no, this is a separate question. They went all those years-- well, until Marshall [Cohen] came. And he wasn't the director either. But until Marshall came, there was a void which-- and even Al was long delayed in getting a faculty appointment.

Clark:  28:55

Yeah, right. Yeah, I don't understand that. But he was very clearly an extremely bright guy. His only disadvantage was a great reluctance to tolerate fools. [laughter]

Kellermann:  29:21

How close did you work with Bolton before he left? Do you have any feeling for why he left right at the peak of--?

Clark:  29:42

No, I have no feeling on that. The science work I did with Bolton was that he had this map of Centaurus A at 960 Megahertz, and I did the data reduction for the repeat at 21 centimeters. Which wasn't worth much because for measuring spectral index because  it was much less sensitive.

Kellermann:  30:23

Yeah, I don't think I've ever seen that. Did it have sufficient resolution to separate the lobe from the galaxy?

Clark:  30:32

Yeah, yeah. In fact, when Per [Maltby] was doing his work on Centaurus A, he came to me and we discussed what percentage of the radiation in the center of the source was the galaxy and what was nucleus and what was lobe. It's mostly nucleus at 20 centimeters.

Kellermann:  31:16

Yeah. So I remember John was still pushing the-- I don't know if he was pushing it, but he certainly pointed out to me that the Centaurus A radio source may just be an extension of the galactic radiation.

Clark:  31:36

That doesn't sound right. But the 960 Megahertz map had the two north-south lobes and then also had an east-southeast extension that was indeed probably galactic radiation not associated with Centaurus A. I remember that being discussed.

Kellermann:  32:15

I'm trying to think about the current maps, whether it shows that extension or not. I remember 960 MHz

Clark:  32:25

Wasn't really visible at 20 cm.

Kellermann:  32:29

Yes. But John liked to think of non-conventional interpretations. As you know, in their original radio galaxy paper, he concluded that M87 and NGC 5128 were galactic. Are you aware of that?  He said there was-- I guess there was no spectroscopy at the time, and he said-- the conclusion of the paper was that there were nebulosities in our own galaxy. And the title of the paper was “Galactic Radio Emission from Discrete Sources at 100 Megahertz,” or whatever.

Clark:  33:29

Yes. Other ancient history.

Kellermann:  33:33

Yes. Yes. Well, that's what I've spent my last years on. Here's another piece of history. What is your recollection or understanding of the discovery of CTA 21 and CTA 102?

Clark:  33:51

CTA 21, I believe, was due to the nature of the encoder readouts on the main console at OVRO. Harris set up for a drift scan with one degree wrong declaration and found this interesting radio source that wasn't in 3C. I don't know about 102.

Kellermann:  34:33

Well, I've talked to Dan about this and his story is slightly different from what you said. And maybe this refers to CTA 102, and he just fell asleep in-- they were all drift scans because everything didn’t work when he moved the telescope. And he fell asleep because he had too many beers or something, and he woke up it showed the other source. Dan told me this decades ago. I have quoted it and written about it multiple times. I was very surprised to see, in Woody Sullivan's interview with Al Moffett, Al claims to have discovered it because, again, he didn't necessarily fall asleep but was resting or didn't pay attention, and the other source came up. I have full faith in the integrity of both Al and Dan. I mean, Dan has issues, but I don't think he'd make up a story like that. And so I suspect it happened independently with both of them. But the real question is why did nobody recognize the value of a short wavelength survey?

Clark:  36:02

Took a while, and yes

Kellermann:  36:08

Because that's what happened, in effect. I mean, survey a small piece of sky and came up with these – actually three sources including CTA 26.

Clark:  36:18

Yes. The thinking was that all the interesting stuff was in the lobes and that the lobes with the steepest spectra tended to be the highest power sources. And so, low frequency was where to go.

Kellermann:  36:53

Well, George Seielstad and I actually did a survey.  I think even after my thesis was all finished, I guess George also.   Actually, we stayed on for -- at least I stayed on for the summer afterwards. Or maybe it was just the spring. But we did a 21-centimeter survey. Didn't find anything exciting. You really had to go to centimeter wavelengths to see the flat spectrum population. Well, you already commented on Al. Do you have any other comments on -- well, you also commented on Rad a little bit. Working with Rad, Al, Gordon, their pros and cons.

Clark:  38:05

Gordon knew the machinery, of course. Al didn't think much of him.

Kellermann:  38:18

That was a problem, yes. I think I sort of know this story. What brought you to NRAO?

Clark:  38:46

The original genesis was measuring the flux from Venus, which I presented at a meeting at JPL.

Kellermann:  38:58

I was there, yeah.

Clark:  39:00

Frank Drake was there and said, "When you graduate, come to NRAO." That planted the idea in my mind. And when I actually came near graduating it came to the fore. Rad and Hein, of course, were great buddies. And so, Rad put in a good word for me at NRAO. And as I said, I go with the flow. It was easy.

Kellermann:  39:37

Well, the story I heard from Hein, I guess, was that he asked Rad-- they were thinking about the VLA and how do we get into interferometry. And Rad said, "Hire Barry Clark."

Clark:  39:53

Yes. My version of that was after the two element interferometer showed first fringes, Rad and came to visit, and Hein asked him, "Okay. What do we need for the next step of the interferometer?" And Rad said, “A third dish and Barry Clark.”

Kellermann:  40:20

Okay. Yes. Yes. That's what I've heard, too, actually. Short baselines, which you couldn't get because of the ditch or something. So you came to Green Bank initially.

Clark:  40:47

Yeah. I was just there for a year.

Kellermann:  40:49

Yeah. So what was the status of the interferometer when you arrived?

Clark:  40:58

It needed help.

Kellermann:  41:00

But there were fringes on the two element interferometer.

Clark:  41:05

Yes. And there was a delay tracking machine, so that you could do long tracks. But the output was still chart recorder. And they had taken the first steps to remedy that and that Art Shalloway was told to make a magnetic stepping recorder to record that voltage. And he got the stepping recorder to work. And at that point, people realized that you needed a little software to actually do something with that recording. So I stepped in and did the software for it.

Kellermann:  42:14

And that was still a mostly analog system, though.

Clark:  42:22

Yes. It was directly an A/D converter on the same output that went to the chart recorder.

Kellermann:  42:36

And cable delays?

Clark:  42:38

Yes.

Kellermann:  42:38

Yes.

Clark:  42:43

Yes. We were talking cable delays right up until we started getting serious about VLA and realized that-- what a hell of a problem they'd be, especially for spectroscopy.

Kellermann:  43:05

So you're only there a year. Right?

Clark:  43:10

In Green Bank. Yes.

Kellermann:  43:11

           So why?

Clark:  43:16

For one thing, Betty didn't like it.

Kellermann:  43:18

Right. Not unusual.

Clark:  43:21

Yes. And the office in Charlottesville was under construction already, so there were a few months when I lived in Charlottesville and commuted to Green Bank. But that wasn't a very long time at all before we got into the building in Charlottesville.

Kellermann:  43:50

Oh, okay. So you moved when a lot of other people did.

Clark:  43:56

Yes.

Kellermann:  43:57

And that was the-- when I had just come that summer. So that's where we overlapped, I guess. So that means you weren't actually living in Green Bank when we overlapped because you moved during the summer. For school, like everybody else. And what about after that? You must have come to Green Bank a lot.

Clark:  44:32

Oh, yes.

Kellermann:  44:32

Because I mean, what you were doing involved--

Clark:  44:37

Actually, pushing buttons. Yes. When we started building the computer control for the interferometer it was a mildly interesting setup in that I, as a stupid scientist, was not permitted to touch the telescope antenna controls, but I could replicate that control on the keyboard of the computer.

Kellermann:  45:19

I was surprised because I had the impression that the interferometer group was more integrated than, say, where I was working at 140-foot, where I certainly wasn't allowed to touch certain things. But I had the impression of the interferometer group, the scientists, engineers were more blended together.

Clark:  45:43

Probably so.

Kellermann:  45:54

So let's switch to VLBI, where our paths crossed. We did some things right, and I think we did some things wrong, or at least not in an optimum way. Why did it take two years to get fringes?

Clark:  46:13

Yes.

Kellermann:  46:14

It should have been straightforward.

Clark:  46:18

Well, we started out at 600 megahertz, which is not a good place to do VLBI even today.

Kellermann:  46:35

That's right.

Clark:  46:37

And we tried the big step of Green Bank to Arecibo as the first experiment, which we quickly learned from not to do.

Kellermann:  46:53

Not too quickly because we did it twice. Twice unsuccessfully.

Clark:  46:57

Yes.

Kellermann:  47:05

Yes. So what should we have done differently?

Clark:  47:20

Yes. Going out step by step is the only thing I can think of.

Kellermann:  47:29

Which we did in the end.

Clark:  47:30

Yes.

Kellermann:  47:39

Oh, also, none of us, or speaking for myself, I think, didn't fully appreciate the power and the value of what we were doing. And we were all doing something else. You were working in the Green Bank interferometer. I was measuring flux densities at 140-foot. And Claude [Bare] did other things, and he would be allowed to work on the VLBI stuff now and then. But if we had devoted full time to it, it might have gotten done faster. Or especially if you had devoted full time to it. Although the software was never an issue, I guess.

Clark:  48:33

Is that the cause or effect?

Kellermann:  48:35

As you say, 600 Megahertz. Well, we were being daring because the Canadians were using 400 Megahertz, and we pushed it only to 600 because of phase stability. I think there were issues with the degenerate paramp, which had to have a coherent -- what do you call it?

Clark:  49:08

Pump.

Kellermann:  49:08

Pump, thank you. And there were issues with that. Again, because Claude, I mean, he could have built a phase lock system, and we bought some lousy commercial system that I think didn't work very well.

Clark:  49:25

I was never really intimate with that receiver.

Kellermann:  49:30

Yeah. Just because Claude had other responsibilities, I think. That's my memory. Let's go to the Russian experiments. And we talked last night about that visit from the Defense Intelligence Agency. You remember that?

Clark:  49:56

Yeah. In my case, the guy didn't say much. He just asked a few questions and left.

Kellermann:  50:09

Well, remember we--

Clark:  50:11

Oh, there were two incidents. Yeah. It was the discussion before and the debrief stuff.

Kellermann:  50:21

Yeah. That's separate. Yeah. The debriefing.

Clark:  50:25

I'm speaking of the latter.

Kellermann:  50:27

Yeah. So if they come to you beforehand and ask you to look into AB, and C that makes you a spy. But if they just let you go and they come to you afterward, ask what you saw, that's okay. Anyway, I was talking about that meeting. It was you and I and a number of people, at least two, if not three, from the DIA in Green Bank. I remember that. In fact, I think we were sitting in Heeschen's office. He wasn't there. It was that big-- with a big table. So what's your memory of that discussion?

Clark:  51:09

I recall that we explained why we needed long baselines and were asked in particular why Russia.

Kellermann:  51:37

Well, I remember them asking if the Russians were going to come to Green Bank to participate in the data reduction. And we said yes. And then they asked us if we could-- they were interested in the geodetic information. Or they were interested in the Russians not getting the geodetic information. And they asked us if we could modulate the data or the phases or something to keep our astronomical data and destroy the geodedic data. And I remember your answer. You remember?

Clark:  52:15

No.

Kellermann:  52:16

You said, "Yes, we could do that, but we won't. We're going to do this above the board, or we just won't do it," and they accepted that. And then toward the end, they finally confessed that they get most of this information from satellites anyway. And they explained that the whole secret of intelligence is you can't keep it secret but to make it hard for the other guy to find out and expensive.

[Post interview comment added by Clark: Yes, I remember telling the DIA that we wouldn’t obscure the geodetic information inherent in the VLBI observation.  But I rather think I wasn’t as stalwart as Ken portrays me.]

Clark:  52:51

I heard that as well.

Kellermann:  52:53

That's my memory. But I also had the memory-- this is the opposite from you, I guess. They knew exactly what we were doing and understood it completely; understood the limitations that we were using in that first experiment, rubidium standards, and not hydrogen masers. I had the impression that they were doing this themselves. You didn't--?

Clark:  53:19

No.

[Post interview comments added by Clark:  Main thing I remember from the DIA debrief was that they were interested in the power structure of Soviet astronomy.  Who was in charge of what, who seemed to speak with more authority than their position warranted.]

Kellermann:  53:20

Okay. So let's go back to Owens Valley Array. You were obviously, the liaison with Caltech.

Clark:  53:31

Right.

Kellermann:  53:40

But you were involved in the almost earliest VLA planning.

Clark:  53:46

Pretty much, yes.

Kellermann:  53:49

Yes, when you came to NRAO.

Clark:  53:50

It started a few months before I got there.

Kellermann:  53:53

Cam [Wade] sort of worked out all the equations, and I think we accept that as the official start. So what did you do?

Clark:  54:05

Mainly talked about the software and served as a cheerleader for various other things. The concept when I arrived varied very different from the VLA. It was Leonard Chow who showed that three arms were better than a tee, and suddenly we realized that we couldn't do cable delays. And I think Hein himself was the one who said we needed 20 kilometers. We had earlier been talking of four or five.

Kellermann:  55:12

Oh, I see. I guess one second of arc at 20 centimeters or something was the goal just because that was what the optical people were doing. There's nothing magic about one second.

Clark:  55:32

Right. But, yeah, we had earlier talked about doing the one second at 13 centimeters.

Kellermann:  55:50

Right.

Clark:  55:51

The arms were shorter.

Kellermann:  55:57

Yes, that's right. The original array was S and X brand.

Clark:  56:00

Yes.

Kellermann:  56:03

I went back and looked at that 1965 report. It wasn't quite write a proposal, but the draft. And there was no scientific justification in the sense that we have to do a scientific justification today. It was just to get better pictures of radio galaxies. So you were mostly involved in the software?

Clark:  56:24

Yes.

Kellermann:  56:28

I think you became and still are or still is the only person that has a complete understanding of the VLA from beginning to end. And I was reminded of this just a week or two ago when there was a flood of emails about the phase noise in the outer antennas. You were the only one who remembered that they have higher power lasers.

Clark:  57:06

I lost the end-to-end understanding of the VLA, though, when the WIDAR correlator was installed. I know some of what goes on in that, but by no means everything.

Kellermann:  57:33

Yeah. Fair enough. When did you go to New Mexico? Come to New Mexico?

Clark:  57:40

1975.

Kellermann:  57:43

And you volunteered or were you asked?

Clark:  57:47

I basically said, "Don't leave me behind." Growing up in Texas, I much prefer living in the West. There is a different ambiance.

Kellermann:  58:04

So I was going to ask, actually, you've lived in Texas, Green Bank, Socorro. Do you like that lifestyle?

Clark:  58:13

Yeah. Yes. Green Bank bothered me too. Not as much as it bothered Betty.

Kellermann:  58:24

Well, you had other things to do. Yep.

Clark:  58:26

Yes. I just find it uncomfortable when there is such a large gap between the richest and the poorest in the neighborhood.

Kellermann:  58:48

I had somewhat of a different feeling that the local population had a monetary income that was pretty much below the so-called poverty level, but they owned their own home that had been in the family for generations. They had a farm. They had what I would call a comfortable life. Somebody with that same income in the city, even in Charlottesville, would be a slum or poverty. Well, so you weren't there that long. So we were there 20 years. And also through the children, I think, in going to school.  I mean, Henny [KIK’s wife] did get to know local people and mix with them socially. And I was in some of their homes, so I didn't have this strong feeling of distinction. I thought it took a long time to get integrated to the community, but I think we did. Henny did at least.

Clark:  01:00:17

Our primary interaction was with the family that we occasionally hired as babysitters. And yeah, as you say, they had income at the poverty level, but good people.

Kellermann:  01:00:48

 I remember we asked Michele [Fomalont] to recommend someone because they had a babysitter. I mean, to recommend somebody. We had heard about some girl, and we asked Michele about her. And she said something to the effect, "Well, I don't know what her experience with children is, but she can deliver cows."

So the original VLA did not include any spectroscopy, the plan. And Owens Valley did. Did you have any influence over implementing the spectroscopic system? Obviously, you did implement it, but--

Clark:  01:01:48

Yeah. As soon as we realized we needed to go digital, I realized that there's enough hardware to do reasonable spectroscopy, but we still needed the software.

Kellermann:  01:02:08

Right. Right. Okay. That's an interesting perspective. Okay. Finally, scheduling. So for many decades, you were the key person in this. And you knew every proposal, maybe you still do, that went on the VLA and which ones were successful and which ones weren't. And you knew everything that was published.

Clark:  01:02:38

Yeah, I did enjoy scheduling. And they have since replaced me with a team of three people.

Kellermann:  01:02:56

Well, it's not just that. It's more than three. I mean, there's a whole--

Clark:  01:03:06

Yeah, well, okay. If you start counting the whole lot.

Kellermann:  01:03:10

Yeah. The committees and panels and then committee and meetings. Do you have any comment on-- the goal, of course, is to get the best projects on the telescope, the ones that are going to do the best science. How would you compare the current system with-- well, I mean, my own involvement for, I don't know, for five years or so it was you and me and Fomalont and Miller [Goss]. And you had the last word, generally. And it was in a closed-door, smoke-filled room. Well, we didn't smoke, but it lacked certainly the openness that current things require or become customary. What would you say about the integrity of what was done then and the effectiveness? Effectiveness being defined as getting the best science on the telescope.

Clark:  01:04:14

Yes, yes I have no complaints about integrity, and I think we did a good job of getting the best science on. And mainly, it was more fun.

Kellermann:  01:04:42

Can you remember any proposal that you turned down that with hindsight should have been scheduled? Like, it was done somewhere else or done later.

Clark:  01:05:05

Well, there were numerous examples of people being turned down and coming back with a much better proposal later. Nothing very major that I thought we should have jumped on earlier.  [Post-interview comment added by Clark: There was a proposal from Miller Goss for OH in molecular clouds that we originally turned down, that I sneaked a few hours into holes in the schedule that was very productive.]

Kellermann:  01:05:29

So of all the things you worked on, Caltech, NRAO, over 50, 60, 70 years, whatever, what you consider your greatest success gave you the most satisfaction?

Clark:  01:05:47

That would probably be the CLEAN imaging algorithm. That had a major impact.

Kellermann:  01:06:07

And more difficult, perhaps, what was your greatest failure, or lack of doing something that you've done differently?

Clark:  01:06:23

Well--

Kellermann:  01:06:24

“I wish I had  …”   Complete the sentence!

Clark:  01:06:28

No, I don't have any “I wish I hads.”  There are a lot of, or at least several, projects that I abandoned before finishing that could have been profitably finished.

Kellermann:  01:06:53

So what do you think about the ngVLA? Are you involved at all, or?

Clark:  01:07:05

Almost none. I decided I'm too old for big projects.  The VLA was comprehensible and manageable. ngVLA, there is no way to do it except to have lots of pieces that have to be fit together.

Kellermann:  01:07:40

It seems to me the VLA was a bigger scientific step than the ngVLA.  So I was critical, or skeptical, I should say, about the VLA being unnecessary. Not being a big enough step over Westerbork, I guess.  14 antennas versus 27?

Clark:  01:08:11

Two-dimensional versus one was the big thing.

Kellermann:  01:08:15

Well, except in a significant part of the sky, you can get to the full resolution [with a one-dimensional array]. And what--

Clark:  01:08:21

It takes a lot of time.

Kellermann:  01:08:23

Well, it's full synthesis. And, I mean, I'm repeating what I said whatever it was 50 years ago. One part of the sky is as good as another part of the sky. You don't need to-- if you can just image half the sky, that's good enough. Anyway, I agree. That was all wrong. But Heeschen’s argument was that the VLA was the first time you could pick an astronomical object and study it in the radio. Anyway, and the ngVLA is not that big a step over-- well, MeerKAT or--?

Clark:  01:09:16

Well--

Kellermann:  01:09:20

It'll be high-frequency, of course.

Clark:  01:09:23

Quite late on the VLA project, the concept of an observation was an eight-hour synthesis; wasn't it?

Kellermann:  01:09:36

Yes.

Clark:  01:09:36

Which I knew was silly, but never said so.

Kellermann:  01:09:42

Oh, I see.

Clark:  01:09:45

And--

Kellermann:  01:09:46

Because you were thinking about CLEAN?

Clark:  01:09:47

Not really, because I was thinking about snapshots and surveys.

Kellermann:  01:09:59

Well, without CLEAN snapshots wouldn’t be very good --?

Clark:  01:10:02

No, they're not all that bad. Yeah, they got the six spikes, but you can still do science even with the raw snapshots, considering the state of science of the day.

Kellermann:  01:10:23

Oh, I don't know. That would have been reproducing Caltech in a way. Beside by mentally ignoring the spikes, you are doing the first iteration of CLEAN. I don't recall snapshots were considered in the proposal or anything. And neither was CLEAN, although we knew about it. Yeah, comment on that. Is it that we didn't trust it or we didn't want to talk about it, or--?

Clark:  01:10:53

Well, 27 antennas in a two-dimensional array, it was not obvious that CLEAN was going to be necessary.

Kellermann:  01:11:04

But that was to reach a goal of a 20 dB dynamic range or, it's 17, really when it went from 36 to 27 antennas. So the VLA is a great example of an instrument that has outweighed and outperformed its design goals. Well, I asked you about your biggest mistake and you said you didn't have any. What about NRAO? You've been here for almost its entire lifetime.

Clark:  01:11:47

Yeah. I have limited imagination, so I don't know what could have been, so I can't say any given thing is a mistake.

Kellermann:  01:12:02

Where do you see radio astronomy going in the next years, decades? Are you sorry you're not 30 years old and looking forward to the next 50 years of exciting discoveries? Or do you think you were there during the golden ages, and--?

Clark:  01:12:27

Yes. I think that I lived in the golden age, and yes, I don't think 30-year-olds today will have as much fun as I had.

Kellermann:  01:12:38

If you were 30 years old today, do you think you could have the impact-- I mean, you had a tremendous impact. Do you think you could have an impact on, say, the ngVLA or other things?

Clark:  01:12:59

Things do tend to be all teams these days, but I think there are certainly individuals I can name now who have great impact, like Greg Hallinan and Sri.

Kellermann:  01:13:39

No, Sri's not 30. Shi Kulkarni? He's not 30 years old. I mean, he's closer to us than-

Clark:  01:13:46

Oh, yes. Of course. He still illustrates this concept of how an individual can have an impact.

Kellermann:  01:13:54

Right. Okay. Fair enough. What do you think about the Hallinan's array?

Clark:  01:14:03

Yeah. It's a very interesting concept. In the early VLA design phase, I actually considered something very similar of 500 or 1,000 dishes with lousy receivers and concluded that for something of VLA size, you really need flexibility, which it didn't have. But yeah, you can put a new receiver on the VLA. It's a major effort. Trying to put a new receiver on a 500-dish array is another step beyond that.

Kellermann:  01:15:07

But you don't need to put a new receiver on the instant it fails if you have 500.

Clark:  01:15:14

No, it's not the failure, but they're limited to bands.

Kellermann:  01:15:18

Ah, right. You think it can--

Clark:  01:15:28

So the DSA 500 is a fantastic instrument for what it is designed to do, but that's all it can do.

Kellermann:  01:15:42

Well, that's all it's intended to do. That's another change in the culture. It's designed for a specific problem. And I think I don't remember 5, 10-year lifetime or something, it's going to do the survey twice or three times or whatever, and that's it. Yeah, I agree. I think that's one of the most exciting things going. Okay. Well, thank you, Barry.

Clark:  01:16:13

Okay.