Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.
During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.
Please contact [email protected] with any feedback.
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
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. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
In footnotes or endnotes please cite AIP interviews like this:
Interview of Brian Schmidt by Ursula Pavlish on 2007 July 24,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
For multiple citations, "AIP" is the preferred abbreviation for the location.
Brian Schmidt studied as an undergraduate at the University of Arizona, where he worked on discovering supernovae with the CCD Transit Instrument under John McGraw. He continued his graduate studies in Astronomy at Harvard University, with Robert Kirshner as his thesis advisor, from 1989-1993. He stayed on at Harvard as a postdoctoral fellow before moving to The Australian National University in 1995. When he attended a summer school in Les Houches, France, in 1990, on Supernova, he met many of the supernova greats and marks this as his induction into supernova astronomy. Schmidt started the High-z Supernova Search Team in 1994 at the age of 27. He wrote the supernova search software, much of the simulation software, as well as one of several cosmological fitting software used by the team and led them to their 1998 discovery of the accelerating universe. In this series of interviews, Schmidt discusses the spaces of scientific work, supernovae as scientific objects, and scientific visualization. Schmidt’s outstanding good humor is infectious, and he is an astronomer and an observer highly respected within the profession.
It is July 24, 2007. We are on the campus of Caltech University. I am Ursula Pavlish here to interview Professor Brian Schmidt, the leader of the High-z Supernova Search Team. He is attending a conference on Nuclear Astrophysics here at Caltech. Before we begin, Professor Schmidt, I would like to congratulate you on receiving the Gruber Cosmology Prize with your team.
I hear that the award will formally be given this September.
Yes, we will go as a team in Cambridge. That will be good.
Before I bombard you with questions, I would like to ask you to tell the story of the discovery of the accelerating universe from your perspective.
From my perspective, the story starts in 1990 when I was a first year graduate student. I was working on exploding stars with Bob Kirshner. I was doing massive stars, stars tens or fifty times the mass of our sun. I was trying to measure distances with them in the nearby universe to measure the Hubble constant, which is the current expansion rate of the universe. It was my first year as a graduate student. There was a big meeting in Les Houches, France. It is where Dirac used to run a school for French Physicists. That is what this school was. The school was on Supernovae. It was in honor of Supernova 1987A. It is there that I came into contact with all the greats of the field including many who are here at this conference. For example, I see Stan Woosley running. It is probably ultimately why I was invited to this conference, because of all of that contact there with influential people. That contact was important because it put me in contact with the people who were working on Type 1a Supernovae. We call them both Supernovae, but they are really physically different objects. Consequently they are not necessarily as close as you might think in terms of what you would do to understand them.
The Supernovae 1a are not as close to other types of Supernovae?
A Type 1A Supernova is a little star that explodes by completely different means than these big massive stars. They really are apples and oranges. They have the same name. We observe them the same way. However, the actual physical mechanisms behind them are very different.
Your Ph.D. dissertation was on massive stars?
Then you attended the conference.
This was in 1990, so it was just at the beginning. I was immediately put in contact with all the people working in this area that I was not studying, which are type 1as.
You gave a talk there?
This was a six-week course. I gave a little talk as did everyone, but we also listened to several lectures by many of the greats. There, I met for example, Ken Nomoto, Stan Woosley. Those are the people who were the theorists. At the same time there was Gustav Tammann, who was one of the people who was trying to use those as distance indicators back in the 1980s. There I met Mario Hamuy. Mario Hamuy was from Chile and he told us there about his plans to go out and find Type 1A Supernovae in the local universe. He was young. He was not even doing a Ph.D.; he was there as a research person who did a lot more.
Was he already there as part of the Cerro Tololo group?
This was right before the Calan/Tololo search started. He told us about that.
That it was starting?
Yes, that it was going to start in the next six to nine months. This was important to me because they were interested in looking at Type Ia Supernovae as perfect standard candles, where everyone is identical to every other one. I was quite good friends with him there, and that eventually set up a collaboration where I went down to Chile in 1991 to analyze their Type II Supernovae which they were not using. That was the beginning point.
When did you begin your dissertation work?
In the last part of 1989.
Your collaboration with them was part of your dissertation?
That is correct. I was using the Supernovae that they were not using.
You were really the contact with them?
It is a mix. Robert Kirshner, my supervisor, set the thing up. They knew who they were getting. I was very keen to do it. It set that up and it set that up in the mind of Bob Kirshner as well. I would say that Bob Kirshner set that up. But, then I went down there and became very good friends with all of them in 1991.
You were there for one year?
No, I was there for about 6 weeks. That friendship endured with that whole area. At the same time, Bob Kirshner’s student, Chris Smith finished up and he also went down there to work. There were lots of connections with Chile. I finished my Ph.D. in 1993. I got a job at Harvard Smithsonian [The Harvard Smithsonian Center for Astrophysics]. Sometimes, I think it is largely because I used to bake pastries. The committee that decided who would get the job did not know who to give it to and they knew that I cooked well. They decided, “So, we will take him.”
[laughs] What kinds of pastries?
Oh, anything. Croissants. I like cooking. I like eating too. Anyway, we will not go there. [Colloquial for let us change topics] I was trying to think of things to do. What I wanted to work on, is something I am talking about on Thursday, which is how to figure out the physics of these massive exploding stars. How many elements do they synthesize as a function of the size of the star?
Do you regard yourself primarily as an observer or as a theorist?
I am definitely an observer. I have always been, I guess, good at dealing with hard data problems. That is, lots of data or data that has been hard. I have worked with theorists and I appreciate theory. I even ran, as part of my thesis, a code of Ron Eastman’s that did very heavy duty calculations. But he was the brain behind that. I really am an observer.
Would you label yourself as an astronomer or an astrophysicist or a cosmologist?
I am any of those three, but I call myself an Astronomer. I call myself an Astronomer because it is the least pretentious. Somehow Astrophysicist sounds better than Astronomy and better than physics. [Pavlish laughs] I am an Astronomer, but I am also a Physicist and I am also a Cosmologist. So, as I was looking for things to do, I was exploring things and the work I wanted to do on stars was interesting to people who studied Supernovae but it was very clear that people were interested in cosmology. In 1991, I forgot to say this, when I visited Mario Hamuy he showed me the first batch of Supernovae he found and he was very depressed. The reason he was depressed was that it was very clear the Supernovae, which he was counting on all being the same, were not identical. There were ones that were rising and falling quicker and there were ones that were clearly fainter than they should be.
The light curves were expected to be identical?
That was the paradigm, yes. The paradigm that people were working with in 1990 was that all type 1a Supernovae were essentially the same with a very small dispersion. I also met Hiryuu (Yoyo) Nomoto who is Kenichi (Ken) Nomoto’s wife. She does Japanese books of astronomy.
She is walking by right now. Too bad we do not have a film crew here. [She stops by, speaks with Brian Schmidt; the man sitting next to us asks where and at what time Brian Schmidt is giving his talk on Thursday.]
Anyway, so he was very depressed. Of course, that was the big breakthrough that they made at Calan Tololo, that Supernovae were not all the same but that you could use them because the rate they rose and fell was correlated to their brightness very accurately, it turns out, to about seven or eight percent.
That is when the peak of the light curve and the slope of the fall are correlated?
Yes, the shape of the light curve, as we say now. It was the slope of the fall back then, but we have broadened that. That was the beginnings of that.
They were depressed, then they published, and they were okay with it?
That was two years later, so they realized what was going in. The Phillips relation was published in 1993. This was in 1991. Mario was depressed, yes, because he could see things not working in the way he had expected.
Was this his thesis work?
No, it was not his thesis. He was someone who had an undergraduate degree, and I think even a Masters degree, but he did not do his PhD until later. In an odd turn of events, he ended up working on a thesis related to what my thesis was, and I ended up working on a postdoc that was related to what his work had been on. In 1994 Mario Hamuy came up to Harvard and gave a talk. We saw his data set for the first time, which showed that Type 1a Supernovae were very good distance indicators. Up to that point, I had had my doubts. I really did. I was not convinced by the Phillips 1993 paper. I saw that there was a relationship but it was not clear to me at all how good it was. Mario’s data, which Mark Phillips was part of, as was Nick Suntzeff, Jose Maza, and lots of people, Mario’s data, you could indirectly measure the distance of a Supernova from its recession velocity because the objects were a long way away. That is a very accurate way to measure relative distance. Whereas the distances that Mark (Phillips) had used were very nearby and so you really did not know what the distances were.
They were looking at Supernovae that were pretty far away?
Yes. They are looking at a redshift of what is nearby now, but it was distant then. It was at a redshift of about 0.02 to 0.09. That was a long ways away back then and it was unique, and then the data that they got was beautiful. In 1994 I realized that Supernovae were good distance indicators thanks to that. At the same time, Saul Perlmutter who was always calling us up… I should not just say Saul, because it was that whole SCP group. It might have been Carl Pennypacker; it might have been whoever else was there. They called Bob Kirshner up and had him observe an object that he had discovered. When, Bob went to it with the MMT, and I believe Adam Riess was out there, as was Peter Challis, he saw something, took a spectrum. I got the fax in the morning. Pete had already looked at it. I played around with it as well. We were all convinced it was a Type 1A Supernova a few days after maximum, at a redshift of 0.42. That was the first distant object. We were trying to get them to put it into the IAU, the International Astronomical Union circulars. I was working very hard with Bob to convince them. They were just refusing.
They did not believe you?
No, no. They believed us. It was just not the way they did things. They were not astronomers.
May I backtrack a bit here. You say that you were convinced that it was a Type 1A. How long did that take? There were three of you?
We had it figured out within seconds. It was obvious. It was a nice spectrum. It was a very good spectrum. It was very obvious, no doubts. We, and everybody else, had been chasing false things for five years. They were physicists. We were not comfortable with the way they did things.
What do you mean, chasing false things?
They would send you objects that you would go and look at and there would be nothing there. You would spend an hour and there would be nothing there. You would go yes, okay, okay. That happened. It was not just us. You can ask anyone in the world who was at a big telescope.
You say that you were actually doing this for years, helping them out?
I did not. It is not like we did lots, but we did a few. Anyway, when you suddenly get one that is so good, it is exciting. We were trying to get the IAU circular out and they were refusing to do it. I remember that Bob and I were talking (I was a postdoc at the time), and Bob and I agreed that dammit, we took the spectrum and we were going to give them an ultimatum, this is several weeks later, two days from now, we are going to put it out. We will mention (say) that they discovered it, but we are going to publish our spectra. Then, at the last minute they published their IAU. It turns out they had found seven of these in 1994 of which we only knew about one. They were very secretive. That was the way they did things.
They had obtained those through other observers?
They had found them. Ours was the only one they really had a good spectrum of, but they had found six other things that looked plausible, one of which they had a spectrum of. A few of which they had redshifts of the galaxies for. In general, I believed that they had seven objects. Those are the first seven objects in their first paper.
You believed it from their paper?
No, I believed it from the stuff they showed me.
No. I am trying to remember how they showed us, convinced us. I think it was with IAU circulars. Other people had followed them up in a way that they seemed to be real. They had been confirmed one way or another. The primary source of interest to this, if you want to look into it further, would be the IAU circular from 1994. That was interesting. As we were trying to work with them, to figure out how to publish this object, we got what I can only describe as frustrated. Very frustrated, knowing that only a month earlier, I had seen Mario and realized finally that these were good [objects to undertake distance measurements with], We (myself, Kirshner, and most of our Supernova contemporaries) had been incredibly critical of this group — We did not like the way they did things. They had not shown that you can actually use these to find distances, and not been particularly good at finding them. Suddenly, within a month, it had completely changed. It was very clear that they could find them, and it was very clear that you could use them to. I had been, as I said, doing various bits of things. I got an idea in about July of 1994 that we should compete with them.
I went down to Chile to observe. When I was there, I was talking to Nick Suntzeff, and we hatched a plan for how we could compete with The Supernova Cosmology Project.
What were you down there to observe?
I was there trying another project out with Joe Mohr on observing clusters. I knew that I was moving to Australia. Australia had this great new instrument that I could use to study the evolution of clusters. We were taking some data.
What is the instrument?
It was the 2DF Spectrograph which was used to do the big redshift survey. That project never went anywhere. I could not get time, and compete against people in Australia. It is just as well. That is why I was down there. I did hatch a plan while I was down there to start this up.
How did you do that? Were you at the telescope together?
No, I was down visiting, I think it was before and after. I discussed the possibilities with them. I basically said, “I cannot do this alone. I will write all of the software but I need help observing and helping the analysis.”
Rather than just straight facts, but for the story factor, how did this work, was it at a bar over drinks, or was it in the office?
No, it was just at the office. Nothing so exciting as a bar and drinks. It was just hanging around, going into offices, talking about it. We put in a proposal that I think went in, in September.
I have a copy from Adam.
Yes. That is the copy I put that on the website. That is the first 1994 proposal, yes.
That is really the formation of the team?
For how many months was it in the making before that? Was it a difficult process to get everyone together?
No. Everyone immediately said, “Yes, I will be a part of that.” Especially since I said that I would do all of the work. Nick helped to write the proposal but everyone else basically sat back and said, “Okay, we will help.” That formation process happened, I would say, in August and September. It was mainly stalled by figuring out who we wanted to ask to be on the team.
Who was the core of that?
It was the Harvard group, and the Cerro Tololo guys.
You were going to work on it full time and the others were going to work on it half time?
We were just going to see how it went. It was very clear that someone would have to write the software to find the Supernovae and that was my job. I just started writing it. I was in the process of having a child and moving to Australia at the same time. My son was born on the first of October. I moved to Australia on the 28th of December. The proposal was submitted on the 28th of September. I believe that Nick put it together at the end. I was reading my email. It was pretty tight. You would have heard from Adam about him getting married when we submitted, right? That seemed to be that way that this groups. Big days in our lives! I do not really remember what happened. I just worked, worked, worked, worked, worked, on trying to make sure we had search software. I had it working pretty well. But, because I had just moved to Australia, I did not feel as thought I could easily go to Chile, and I did not have money to go to Chile from there. So we sent the software over to Chile. Mario Hamuy, Mark Phillips, and Nick Suntzeff, had to make it all run. The first time you run software, in real life, things do not work very well, or not at all. [Laughs] It was an unmitigated disaster.
May I ask how writing software works? You were experienced from doing that in your thesis?
Yes. I had written code.
Do you write it to direct the telescope?
The telescope does that already.
It sifts the data?
You have to take the objects. You have to align the data. You have to match the data, to make sure that the stars have the same shape. You have to scale them to the same brightness, and you have to subtract. Then, you have to search the subtraction. We had to do this for many, many gigabytes of data. Disks did not have many gigabytes of storage. It was a real challenge. I found data. I added things. I really did go through things pretty carefully. It was working, more or less.
Did you have cosmology in mind or was this a technical challenge?
It was very clear to us. Our group’s title was: Measuring Cosmic Deceleration and Global Curvature. Note the word ‘deceleration.’ We knew what we were doing.
So, it was clear.
Yes, it was very clear what we were doing.
This is a conceptual question. When you were writing the software, were you thinking of Supernovae while you were doing this? Were you thinking of them as little dots? Were you thinking of them as processes? Someone told me that Supernovae are not objects; instead, they are phenomena.
I call them objects.
Did you visualize them while you were writing the software? Or, is that asking too much for a melodramatic method of storytelling?
I did not visualize them as I was writing. I was very intrigued to see if I could find them. I was having to add them. They are very artificial to me.
What do you mean, add them?
I did not have any data that actually had Supernovae in them. What I had to do, is take data, which I had two pictures of the same field. I had to add a Supernova to one, do all kinds of stuff to it, and then find it. The process was very artificial, until you actually got to looking for real ones. Then, they became these things you were hunting that became your prey, or something. Then they became real objects. I am afraid that adding objects was very artificial.
The kinds of objects you were adding were based on what had been found?
Yes. The added things were the right brightness. They looked about right.
How much material did you consider in making your artificial Supernovae? Did you take into account the Perlmutter group’s results?
I knew what they had done. I knew how bright they should be, just from cosmology. I knew the basic technique that Perlmutter had used. Yes, I used it. As best I know, they had never published how they did it. I just heard talks, and thought that I needed to do something similar to that.
You mean the simulation aspect?
No. The simulation was just for me to know that my software was working.
They had by this time published their Supernova light curves?
No, they had not. They never published their Supernova light curves. They still have not. It is a point of some contention. No, I had no light curves, nothing. I only had hearsay. I was writing this in a time when I did not know how bright they were. I roughly knew how bright they should be. The Supernovae ended up being a little fainter than we had expected because of the acceleration. They were within a two or three tenths of a magnitude of what had been expected. That is within twenty percent.
You had the Cerro Tololo Supernova data?
I knew how bright they were in the nearby universe, so I could extrapolate a cosmology with the curves. We were trying to measure the effect to twenty percent. We knew well enough how bright they would be.
When you were writing the software, did you work exclusively with your computer? Did you work on paper? Did you go back and forth?
I worked almost entirely on my computer.
Did you think about it outside of work?
I thought about it continually. In the shower, I always do think in the shower. Yes, it was continually on my mind and I was working very long hours. What to do? There are lots of technical problems to solve. I had to write many thousands of lines of software. I was working on that continually.
Then, what happened is that you sent the software to the people in Chile and then they said, “Oops, it does not work.”
The problem was just to get the thing installed, because they were using a different computer operating system. Nightmares!
For accuracy’s sake, what operating systems were you using?
It would have been different versions of the Sun operating system. I believe, that is what they were on. It was written on something called, Ultrix, which was a DEC system. I had to move it over to Suns in Australia, which is what we used there. In Chile, they had a different version of the Sun operating system. When you have complicated systems, no computer is like any other. Especially back then. The Internet connection between Australia and Chile was so abominably bad, as to be hard to describe. It was on typically, about one character per second. A…B…C…We had to talk on the phone, but it was extraordinarily difficult to fix problems. It was very clear that things were not working well. They could not send me data, because it would take three weeks just to get one file. I had them send it on a tape. They had the equivalent of Federal Express. It was not Federal Express, but it was something like them. Then, the tape got lost. It never got to me ever.
That was telescope data?
Yes. We were waiting for the tape. It never showed up.
This is around when?
This would have been in February of 1995. Next, we had four observing runs: two in February, and two in March. The first data we had was in February. They really struggled just to use the software just to get the February data even to the point where you could look at it and the software so that it would run at all. Then, when we ran it on it we had all these problems in the 2nd run in February. I was writing tiny little pieces of code, which I would email to them, and they would insert in where I told them to. Then they would run it, and it would make tiny little postage stamps, 16 pixels by 16 pixels across, which then would be emailed to me automatically. They would take an hour to get to me. Then, I would look at these things, to try to figure out what went wrong in these 16 pixel by 16 pixel images.
The postage stamps were a part of the code?
It was something I had to write, in February, to try to understand what was going on. I was continually sending over little pieces of images so that I could guess what was the problem.
Images from the telescope?
Yes, of the data we were taking. They were describing to me, what they saw. I was getting these tiny little postage stamps. Then, I would change the software, try to figure out what was wrong, send the change, and they would run it again and we would see how things changed. We did that for three weeks, while we were waiting for this data to show up. The data never showed up. Then, we had our third run. At the time of the third run, we had to submit our renewal proposal. You proposed every six months and it was March now. We had not found anything yet. We took the data. We had got the software working better, and three hours before the proposal was due, we found something that looked maybe plausible. We found an asteroid. We found something that was plausible. We stuck that into the proposal. We said, “Well, here is something that might be a Supernova. We have not confirmed it yet, because we just found it. Here is something that is clearly an asteroid but it is what a Supernova would look like if it moved. Of course, Supernovae do not move.” We put that in. We were going through the data from the 4th run, looking for something new. Then, suddenly stamp C14 came back (that is what it was called, stamp C14) and I looked at it because they were asleep at the time. I saw it, and I said to them, “Look at this, because this really does look like a Supernova.” The guys in Chile looked at it, and we agreed that it did really actually look like a Supernova. Mark Phillips took a spectrum of the galaxy and it showed the galaxy was at a redshift of z=0.48 Bruno Leibundgut and Jason Spyromilio from ESO (European Southern Observatory), who were founding members of the group. Jason is the guy who did Supernova physics. Bruno was a part of Bob’s group and then a part of Alexei Filippenko’s group. They took a spectrum at ESO. In the end, it was clear that it was a Supernova at a redshift of 0.48, the same as the galaxy, which was the most distant one ever. That saved our bacon, because we had found the most distant Supernova in our first attempt. That was the only thing we found. At that point, I told my wife, I know it was not planned, but I was going to be going to Chile in the future to observe, because there was no way I was going to try to do it from long distance.
Do you have any of these postage stamps? I have not myself, obviously, done this sort of work and I wonder what the postage stamp image is like. Is it a little patch of sky?
It is a little tiny patch of sky.
You see the stars in it and then you match it up with another image?
What happened in this case is that I would say: I think this is something interesting. Imagine taking a picture. Say, I take a picture of you and then I cut out sixteen little pixels around your eyebrow or something. I said, “Okay, there you go, that looks interesting.” In this case, it was a postage stamp around a star.
The software told them that this was interesting?
The software went through and pre-scanned things. Then it would send each postage stamp to me as emails and for them to look at as well. They were asleep and I was just looking at these postage stamps. I saw one and I said, “Geeze, this looks really interesting.” I do not remember exactly who found it first. I have to be careful. Maybe they pointed it out to me. It became very apparent that there was an interesting object. I thought that I saw it first, but I do not know. I was up for twenty hours a day at a time, so I do not know for sure.
You have to see the spectrum to know, don’t you?
The postage stamp would come in four little spots. It had: image last time, image this time, the last image matched to this time, and then the subtraction. What this showed, was a little galaxy.
With a time delay of?
Three weeks. That is how far apart they were, three weeks. This one showed: galaxy, galaxy with a blob on the side of it, and that did not look like it moved. That was the beginning.
You would see first, a galaxy with no blob, and then a galaxy with a blob.
In this case it is not so clear. It is pretty subtle. You would not have easily seen it without subtracting. Getting the spectrum, confirmed that it was actually a Supernova. We got all this data from everyone we could. It was one of our best light curves ever, actually, that first one.
You get the spectrum with a different telescope?
We got the spectrum with the ESO 3.6 meter telescope. That was by Bruno and Jason, because that is a European telescope. If we did not find that object, that could have been the end of the High-z Team. It is not clear that we would have survived.
How did you advertise it?
We put out an IAU Circular. Then, we got to show it in conferences. We had a press release in Australia on it, which got big press. Front page of the Sydney Morning Herald, I was on the Today Show. I do not know why. It was not the kind of thing that would normally grab. For some reason it grabbed in Australia. I had just come to Australia, so I was just overwhelmed by the whole thing. Someone set it up, someone who I know now quite well. But it was surprising, that discovering an exploding star halfway across the universe would make such big press. Normally, it might make page six. But, front page, and on live television, is not normal. Put it this way, when we discovered the accelerating universe, we did not get that type of press in Australia. No way near! We did get that press in the US but not in Australia.
Once you find a Supernova, it is not a process of verifying whether it is or is not a Supernova? You can be pretty sure of what you have found?
You always want to make sure that it has not moved. It is possible for asteroids to drift away and it is also possible for it to be a Supernova of a type that you are not interested in. You really do need to get your data on it.
You take more data?
You take more images, and take spectra. We got as much data as we could. More often than not, I would say about sixty percent of the time they are what you really want. But, four out of ten times, they are not.
So this was C14?
At this point, were you already using those color filters?
Yes. We had developed filters. Adam showed you the faxes of these things. Those filters were designed to maximize the brightness, or the throughput of the telescope, so that we could get as many photons as we could. Also, we tuned to the right wavelengths. In retrospect, I am not convinced that it was the right idea. I was talking to Adam. My recollection is that that is something I did almost exclusively myself, but I am not convinced it was my best idea. Adam thinks now that he was involved, he may have been. I remember it being an obsession of mine at the time.
Where did that obsession come from? How did you find out about these filters?
I just designed them by myself. It seemed to me that if you do an experiment, you might as well do it right, and spend the money on getting the filters. That was, literally, the only thing we spent any money on. I do not even remember where we got the money to do it. I think that I used part of my limited funds from my fellowship, and I am sure that Kirshner or somebody coughed up the rest.
Why would you say that it was not one of your best ideas?
It turns out that the pain dealing with these filters, compared to the gain, was not a worthwhile exercise. I am not sure if other people on the team would say so.
You were one of the people who put up with a lot of the pain of dealing with them?
Yes. They really were my folly more than anyone else’s. So, I only have myself to blame.
That reminds me of another innovation of yours, which is this idea of group work according to “the IT person.” Should we save it for our next conversation, or have a little previous now?
Let us do the “IT.” When I started the High-z Team in 1994, I was 27 years old. So I was the 2nd youngest person on the team (Adam Riess was the youngest). I was its nominal leader. By 1997 we had three Postdocs, who were myself, Peter Garnavich, and Adam Riess. I was very strongly of the view that the senior people in the group were unlikely to work for 80 hours a week for a year on this project, which is what it needed, and they certainly did not need the jobs because they all had jobs. I was very strongly of the opinion that we needed to have someone in charge, to be “IT.” That is, a person who would say, “This is my paper,” and then would dedicate his blood and guts to get the thing done. The reason he would do that is that he would correspondingly get all the glory for the paper as well. At a meeting in Seattle in 1997 I was very distracted, because I was trying to do too many things at once.
You all were at some sort of a Conference?
We had a group meeting to reduce data. I trained everyone for how to use the software that I had written. We had to find the Supernovae, but then I wrote software that would help speed up the reduction process, to convert those images into numbers that we could use. We had to train everyone in how to do that. We did that in this meeting in Seattle.
Do you have notes or anything from that?
No. Lots of people will remember it. Adam would have been there.
Winter of 1997?
Gosh, when would it have been? In March or April, probably. That would be my guess. That was where we decided on who would be “it.” We decided that I would write the first paper, in which we did 1995K. I laid out the experiment: here is the cosmology — here is this, here is that. Here is how we find the Supernovae. Here is our filter system. That is the paper that just took forever to come out.
Was this a risk, at all, to divulge your methods?
I have always believed that science should not be proprietary. That is, I think, a foundation of science. That is how science can move quickly, and it is a big advantage in solving things that we have over industry. Science has gotten more proprietary, and that is going to kill us. I have always been willing to hand my software out, even to the other team. I still believe that.
Did you actually hand it out, or you would have?
For example, (Weidong Li), before joining Alexei Filippenko’s group was in China. In 1996 he emailed me, and said, “Professor Schmidt,” I was a Postdoc at the time, but okay. He said, “I would like to use your Supernova software to find Supernovae here in China.” I just said, “Okay, I will send it to him.” I sent it to him with no instructions. Three weeks later he writes back, and says: “Thank you, Professor Schmidt. I have used your software to find my first Supernova.” I remember thinking, “You have got to be kidding,” because no one could do this.
Was he collaborating with the SCP?
No. He was just some random guy. But I would have been happy to send it to the SCP if they wanted it. A year later, he was applying for a job with Alexei Fillipenko. Alex said to me, “He said I could talk to you regarding the job application.” I explained to Alex what he did, and I said, “He has got to be amazing if he could do that.” I was right. He is amazing. It was not an easy thing to do, what he did, but software was open. I strongly believe in the concept of open source software, still.
Did that help you at all, that concept of open source?
I steal and use software from any number of places. Although I wrote many thousands of lines, I did also borrow code from all sorts of people.
That is interesting; I took a good course on intellectual property and science, and connected to that is the concept of priority in science.
My view is that the stuff I do scientifically, I think should be free, open, always as a matter of course. The U. S. is very good at that. If I get an idea, and I decide to run off and make some product out of it, then fine, but do not fund me with the NSF to do that. That is no longer science. Then you are commercial, with a different set of rules.
Would you even say that work at a research lab like Bell Labs is not science because it is commercial?
A lot of Bell Labs’ work is open.
Okay. I do not know enough about that.
It is not that it is not science. I think that is an important concept of science. I think that the CMB guys have actually been very good at being open.
The famous school at Les Houches, France — L’École de Physique des Houches — was established by Cecile DeWitt-Morette in 1951. Dirac may have lectured there. Dirac died in 1984, before Schmidt attended the school.