Interview with Dennis Sciama

0:00 Is it working? I think so. Yes, it is. It's the 16th of October, 20 minutes past 11, and I'm talking with Professor Dennis Shannon. So, you were beginning to tell me. Well, I suppose there are two different but related problems about which there have been debates. I'm not sure how far back you want to go, but mentioning Bondi and Durrani suggests you want to go far back. And one debate was, as I'm sure you, of course, know, do gravitational waves exist or not as a matter of principle or as a property of geometrality? And the second question is, what's the right formula for the gravitational radiation reaction? and I remember still the debate about the first point and it goes back to early now, not early compared to relativity, papers in particular Infeld, have you Infeld wrote papers, Infeld has been working on the equations of motion in GR and arising out of that work he wrote a paper or papers saying as I remember, he didn't think I haven't looked this up, by the way, in preparation for the talk, because I've been so busy. It's said that gravitational waves don't exist. And the very, very simplistic rationale for that was the fact that there isn't a localizable energy momentum tensor with gravitational radiation or gravitational events. only some kind of non-local thing that you can integrate to get something that has decent invariance properties and so on. And partly for that reason, but partly I think arising out of detailed study of the motions of particles of systems, which if there were waves would then be affected. He said he didn't think waves existed. And so there began, I think that probably began the debate at that time. I don't remember the year. It must have been in the 50s. I suppose I was then fairly immature as a research student. I can't remember when

2:30 I got my own PhD or not. I felt vaguely doubtful about that because my approach to things is very very physical one, rather than a mathematical one on physical grounds. It seemed unlikely that they didn't exist even if in order to define them you had to pay attention to these non-local questions. Then as I remember the crucial discovery that settled that part of the discussion but not yet at the level of the radiation damping formula was Bondi, Piranha and Robinson wrote a paper producing an exact solution with plane gravitational waves and it was obviously it wasn't necessarily a wave that was wavy depending on maybe some profile you assume because they sometimes talk about sandwich waves in those days, that is you had flat space proceeding on a long surface, and then a thickness with some Riemann tensor, as it were. And then behind the slab, empty space again. And you could choose something about that profile in between. And it was played in the sense that in a certain embedding sense the waveforms were playing. And it was an exact solution. It was a wave in that sense. It wasn't a wave in the wave. So maybe you could make the profile oscillate a bit if you like, but I mean the main point was the physicist would say okay, that's a way, even if it's idealized, because it's playing, and therefore couldn't have a sort of a band of distances. But still, that exact solution, and it had certain interesting algebraic properties. Those were the days when people started studying the Bianchi types of solutions, and solutions with null arrays that had certain properties. Oh, darling, would you meet Mr. Kidd. Hello. beginning to study things like empty space times with null rays that had certain properties they did or didn't curl all that kind of thing and their algebraic is special a whole lot of algebraic work was done at that time and this solution fitted in a bit to that because it was type N I guess in the

5:00 Penrose classification or whatever you call it properties were of interest rather to this as a mathematical construct, but from the physical point of view it helped people to feel that Infeld must be wrong, and that gravitational waves did exist, even if you couldn't talk about their energy density as a local property. and when it was felt that Infel's arguments weren't very convincing so I think I was saying that I can't remember whether it was deciding what went wrong with Infel's view partly it had to do with the fact that the equations of motion business, you know, where you reduce the equation of motion directly from the field equations. At that time, in particular, it had a very elaborate technology, and therefore, maybe one could go wrong, because the whole discussion was rather intricate, and approximations had to be made, because if you treated the particles with motion you were interested in as point particles, then you had a singularity, you had to cure So it was all messy. And then in that mess, one might mistakenly say, there's no waves of, let's say, either acting on the particle or the emission by the particle affecting this motion, or those would occur in some high-order approximation, sort of to how you did it and so on. It was just not very reliable, and so my memory is that once you had an exact solution, at least it showed that weight as a matter of principle could exist, even if that weight had no source. Someone like me who was very much a physicist, but felt that GR, while it had a complicated mathematical structure, was still describing a physical phenomenon. the one-on-linear, et cetera, was still vaguely similar to electromagnetic. It was unlikely that the physics would be totally different, but when you weigh in something, it would be elevated in these ways. So, when you mentioned earlier the physical argument, your initial physical feeling that the gravitational wave must exist, this was largely by analogy with the electromagnetic case.

7:30 In effect, yes. Oh, I don't know, not only, I mean, what sort of scale, I mean, other fields that one felt ununderstood, and one didn't feel that long linearity has totally destroyed the phenomena, that they make it more intricate in India. But my views aren't relevant. On the other hand, my reaction is probably similar to that one. But if you ask what in the development of the subject that's now readable in the literature led to people taking it from you, I suppose that paper, followed by further studies of that kind, making more complicated exact solutions where I guess it was at first a plane wave and then was a thing called a plane-fronted wave and so on. So there was five-figure exercises done on that, which was very valuable, of course, but it didn't add a lot to the question of whether it was a matter of principle that existed or not. So then a new phase of the subject really began at that point. Then later, I suppose, in fact relatively recently, people started saying, we don't believe the quadruple for Bill Einstein, which of course goes back a long way. of 1918 or something, where you just linearize and do the obvious thing. And it's completely like electromagnetism. But because it's a tense theory, well, you know, it lies in everything a bit more complicated. It's similar. And so then people started saying, ah, well, but then all these subtleties, and we've got to be careful, and the approximation procedure is tricky, and maybe the Einstein formula is wrong. other people wrote papers, working a bit harder. I don't know, you would notice better than I do, like D'Amour, I suppose, and so on, wrote papers, going more carefully into the whole procedure. And I guess it all went in the direction that apart from fine fonts, the quadruple formula was not a bad approximation. And then the binary pulsar came around and fitting and experimenting with that. And I suppose that more or less clearly does. But then there was this further question, was it Christodoulou produced this memory thing which Kicker works on. And again, you will know what I think, perhaps your paper will be more. So I'm not going to tell you what else you know about. Though again, I felt

10:00 Okay, this may be an interesting effect, but it can't change the broad features of it. Of course, being simplistic can be dangerous, and sometimes it's correct. And sometimes the subtleties actually dominate. And you can't be sure very well. I mean, I have vision of that. Well, I mean, but since you probably know all that, I don't know what was really helpful to you. But anyway, that's my memory of those steps. Sure. No, it's interesting to hear that. One thing I was interested in asking, since you touched upon the binary pulse earlier just at the end, and obviously at that point the radiation reaction problem, per se, became of direct astrophysical consequence, because here was a system that actually people were observing some effect at this time. To astrophysicists, or from a physical point of view, had any of the earlier stages of the problem been of interest, to relativists. You mentioned, for instance, that some of the spaces involving the description of the Ways were interesting from a very mathematical point of view. Was it a problem that astrophysicists gave any thought to, say, in response to Info's idea that perhaps the Ways didn't use this, John? I think it's fair to say, though, that many exceptions, of course. Well, let me start again, it's a cultural problem. I mean, one group of physicists, physics is so elaborate, etc. One group of physicists, even if they need a result from another part of physics, will tend to take some feature of this other part of physics, which they need to use, but will not care about all the details and technicalities of the other. Often mistakenly, in my opinion, because again, don't say my style, because it may influence everything I say to you, And I think it's a mistake not to pay attention to the physical discussion of the other part of the subject. Because it may be dangerous. Anyway, the fact is I would suspect that any, well, sort of, more than one of the best physicist probably didn't even know what would be the case. Because the literature is so enormous, etc. So they probably barely knew that there was some doubt about what the right answer is. I would suspect. Because astrophysicists are just a different breed in relativists.

12:30 And even astrophysicists who care about things in astrophysics that have to do with relativity would still very much know engineering relativity, as you might as well. And slightly hyperglutant, as they would think, debates about the exact correctness of certain conditions. My suspicion is they barely at all knew that such debates occurred. Now, there will be exceptionally individuals who think that's not true, but broadly speaking, there is this culture. In fact, I have a little joke there, that the literature, this is all true in observational astronomy has an interesting topology because you get sets of ideas or papers that form a disconnected set in the space of the literature. So something like, say, what quasars do at far ultraviolet energies, there might be 200 papers about that. Someone who ought to know about that, but actually is educated in a different part of the subject, will not be aware that those 200 papers exist because they form a disconnect set in this frivolous topological space. But you see what I get in there. So therefore it's quite easy, although you're supposed to know your book work kind of things, it's just so much material, everyone's so busy, that it's quite easy to be unaware of a whole development of ideas that's just in a part of the literature you've never been connected to. And you perhaps don't go to seminars on that kind of thing, or don't talk to people and do that sort of thing. So you remain virtually in English. And that's sort of the sociological explanation. Plus, maybe not be very sympathetic with the label regards by the anti-mathematical problems. So I would guess, just to go back to your question, that astrophysicists would barely be aware that these events were serious. Yes, that's interesting. Certainly I'm interested in that aspect of it Because, although, as you mentioned to begin with, this is somewhat a study of quite a technical problem, I think that some of the sociological implications and so on are interesting in themselves. That is, the different groupings involved in this. It's interesting, as you say, that this, in any case, was certainly a problem, but it was really only an interest to Roger this.

15:00 Well, it didn't even last a long time as I remember. highfalutin doubts, you might say, but maybe genuinely doubt about, about the damping formula or voice. In only a year or two, there were many papers written saying, no, when you do it a bit more carefully, you get the, basically you get the end of the entry result at a certain level of approximation. So it's not that it was a problem for 10 years or something, which perhaps would then have come to the notice of developing. If it was all set in rather rapidly, then that would be another reason, I guess, why it would have much impact outside the circle of the medium interested profession. One interesting point that a number of people have mentioned to me on one avenue of reading that I'm trying to cover is that of conferences. A number of people suggested certain conferences that either were arranged with certain, something of this problem in mind, for instance, Aymar who's organized a summer school at Brenham with a particular discussion of the quadruple from the radiation reaction problem. I vaguely remember that. I remember that he was involved in more precise, because he has a very precise mind. He was one of those who wrote slightly more precise papers than he'd done before, and I vaguely remember he organized the meeting, although of course there were so many meetings, it's difficult to keep track. What year was that meeting? That was 1972. You see, in any case, I ought to be interviewing you because you don't know about this. I vaguely remember that there was a meeting to address this question. some meetings at which there was very vigorous discussion between certain people who were really concerned. In your memory, being someone who wasn't one of these who was so maybe vitally interested in the problem, do you have any memories of conferences at which this was a big topic? No, I thought it was mentioned at conferences I was at. It's not a big topic. Of course, don't forget, when I say I was not vitally interested, hadn't it gone the other way, I would have quite too interesting, because that would have said that for subtle reasons, GR is quite a different sort of field theory for the electromagnetic, and that would have been important, but if

17:30 it was going to go what became the conventional way, then it was a matter of exactly what the correct approximation is at the such-and-such level of approximation, and then it was a matter for the experts. So once it became evident that it was a false alarm, then people like me, slightly on the sides, would say, well, then I'll leave it to the experts. But had it gone the other way, would it have been a measure of concern for us all? Yes. So the fact that within a reasonable space of time, the Conventional view was supported really was the main reason we feel that people outside of the immediate group weren't attracted to really take action. Sure. I would imagine so, yes. it's reasonable Well, if you can, I think you've already given me a certain amount of detail that you remember, but I am interested in this work that Bondi and Pranay and Robinson and so on did in response to Infel's objections, or skepticism, let's say, about the existence of gravitational radiation. And as I've read, for instance, a letter that Bondi wrote to Nature... I remember that. I was about to say that Bondi, Pirani, and Robinson was not the first publication. In fact, I'm a close friend of all of these people. Well, Bondi's a bit senior to me, because I've been associated with him for a long, long time. Pirani and Robinson, you might say, be more my age. I was close, personal friends with him. For geographical reasons, I don't see much of him recently. although Ivor Robinson was in Monadie in Cesar last year, because he works regularly with Troutman on there, up to now, so he goes to visit Troutman there. And I guess Ivor was probably in Cambridge at that time, as I was, and I remember he was working all these ways, and he was even, perhaps you shouldn't put this in where you write it up, but he was slightly annoying. In fact, you can talk to him, that Mondi published, I think, this letter to nature, where he perhaps already given a talk. In fact, it might have been that Felix Birani and I used to organize a group of people in Cambridge to talk about interesting things in relativity and

20:00 fundamental physics. And we jokingly called him the Eddington group, because, you know, Eddington, in his fundamental theory, had a group which he called E numbers with a capital E, which were rather like a drag gamma algebra for spinners, relating to that sort of thing, but he finds it slightly differently. So there was this group of E numbers, Edmonton numbers, which, I forget how many now, but let us say, I don't remember, there were 20 members of the group. So we tried to get 20 people interested in fundamental physics to get together. We made a little joke that in order to join this group, you had to write a statement saying, believe that the laws of physics could be deduced by pure thought, because this was in Edmonton's philosophical the right into the later life. But we certainly had serious meetings when we would discuss questions of fundamental physics but more centered on relativity than quantum field theory because we came from the relativity side of the divide. And I guess Imer at that time was also working on the equations of motion probably in the interstellar sense if I remember right in his very early work. and they had some weight solution. He talked about it at some meeting. You better check with him, because I don't want to say anything that might cause bad feeling. But I think he felt that he talked about it and that he had an exact solution or something. And then Herman wrote this letter that he obviously looked up with a solution. I forget if it was an exact solution you'd have to remind me whether it was some approximate discussion. But it was along these lines anyway and I felt a bit that he'd be overtaken I guess. Do you remember? Because Herman sometimes did approximations. That's right. The letter as I recall it describes solution very briefly and refers to the fourth coming publication by Robinson, Brownian, Barney. Well, maybe Herman, I don't remember the details, you see, whether Herman had talked about it, and then they got together. I mean, I can't remember that,

22:30 but you can talk, also, I think you can talk to Felix. Yes, I'm planning to talk to him in London, and to go to see Barney in Cambridge, and hopefully it's a point to talk to Arne Robinson. Well, it's much more sensible if you talk to them, because they will remember it even more vividly. I mean, I was friendly with them at the time, and I was still a relative, even if I wasn't working on those things. But my memory was that I was slightly mixed, that Herman had stolen a march on me. Though that's not the sort of thing you want to put when you write up an excuse. It's a very minor matter, actually. Anyway, then they must have got together and produced this nice paper where the exact solution was figured out. No doubt they made it happen that way, because they wrote the John's paper. But it's true, I was about to mention, just before you did, that there was an earlier paper by Dr. Bondi by himself. Dr. Bondi's a very creative and independent and original thinker, and it's the style that G.R. is good at, on the basis of his other books, I'm not trying to cause trouble with many thoughts. I'm a peacemaker. If there's ever friction in physics, I'd like to have a peacemaker. But since you're asking me these things, I just remember that many thoughts. Yes. The other aspect of the letter that Bondi wrote concerned a thought experiment in which he which I think a number of people have mentioned to me as being quite influential in which he described that he may have been influenced by I think the work that Pirani and Nidhi Robinson were doing that's in fact is then discussed in the found Nidhi-Rani Robinson paper yes I think so but it also was specifically that's right it's there in the paper but it's also mentioned in the letter Was this argument just a part, was this argument, do you remember it specifically as being important, or...? Well, I liked it because it appealed to my physical way of... This is how you measure the Lehmann tangent, how you measure the way, because afterwards the particles are moving if they were pressed before, and things like that. So I found it a Lehmann tangent, because that's why I remember it to this day, indeed.

25:00 Of course, now, I mean, all these things are very obvious that we do things like that, but in a way, just to make a broad remark, in a way, it's interesting how many things in GR, in my opinion, which now, to the modern, young, totally professional person, can keep the obvious, took a long time to understand. And it's partly because in the early days there were very few workers on AP. I think it's slightly more sophisticated than this. I mean, how long it took before people realized that it was important, both true and then important, that you couldn't necessarily cover a manifold with a single coordinate system in a continuous, I mean, in a regular way. And in other words, the understanding of the structure of solution, if you like, that's an interesting question. That goes very slow. People said the most amazing wrong things because they couldn't understand that. Well that and many other things, it seems to me in retrospect people learn that awfully slowly. But in particle physics, there have been so many people, there were a thousand people working in UGR for 1920, and all these questions would have been cleared up at the speed of light, and conceptual development would have been faster. So, what we're now discussing is part of that slow, relatively slow, development of things that, in retrospect, are not that difficult. Although, I mean, I dare say writing down an absolutely correct approximation of radiation and forming that may be a little bit tricky. But generally, some of these conceptual questions are now quite so important, but they're a bit slow in the relative. In fact, as I say, it's part of the largest, you know, lots of features of GR, and I mean properties of the cosmological models and so on, and the way horizons work. People were very slow to understand, put it this way. that was necessary to think about and correct. I don't mean unnecessary sophistication, which I don't like, but correct level of sophistication. So in that sense, this topic that you're looking at is part of the broader feature of GR, where things are a bit slow to be understood. Yes, I think that's an interesting aspect of it. Regarding what you were saying about

27:30 as it were the correct level of sophistication another person who took a small interest in this debate at that time was Feynman seems to have had some fairly trenchant things to say about what he regarded as I suppose the over sophisticated attitude of some of the people working on the problem how much and a number of people that I've talked to have alluded to a difference in approach between more mathematical physicists and more theoretical physicists how much was that a feature of the field at that time in your experience? well that's certainly true and I met Feynman a number of times and I heard him speak at a couple of conferences, and it is perfectly true that he was a bit derisive about the mathematical approach of development. I can remember one lecture where he did produce a bit of mathematics, and he told the audience, and said, I can do that too! And he wrote, of course, this is his 60-page or so, never published a set of lectures on GR If you haven't seen it, you might be interested in it. I've looked at it. I think it was with John Pressley working to publish it now. Yes. Of course, finally, he did things about trying to quantize GR. He contributed the work in understanding how to quantize the gate theories, which had this complicated problem that there are parameters in the theory that you need to keep track of the gauging barriers, but they're not too observable, and so on. There's a lot of formulas that you need, and then you have this ballet of pop-off ghosts, and so on. Well, Feynman contributed to all that. He also discovered something which I had discovered, which slightly he has gravitational waves in the background. at this point, but if you start trying to write a video theory, you find that the theory isn't consistent, you've got to write a correction term, which effectively tells you that gravitation contributes to gravitation. And then you find you end up with an infinite series, a power

30:00 series, because you keep having to correct. And then the final line is in fact Einstein's equations. So the way a particle physicist might derive from that, you see, is you write down the citrus theory of spin-two theory. And then because of this, roughly speaking, it's because you want the energy of momentum tensor to be converted to conserved, not conserved with an ordinary divergence, you see. And that forces you to this influence theory. Although wrote a nice paper, which I think was the first paper ever published in the GRG journal where by choosing the variables correctly to achieve this, he did the whole thing in one step instead of getting into the series. And that's a big relation to what you're talking about because, of course, as you proceed along with several approximations, you're introducing gravitational waves, in fact. Well, Feynman also discovered, as a particle physicist, that if you proceed in this way, you get the whole gr. And then the gauge theory, the gauge transformations that you get at the first approximation, it's very like the gauge transformation of electromagnetism which is HIJ goes to HIJ plus the HI DXJ plus the HI DXI you see rather like the AI goes to AI plus grad chi and then as you change the gauge transformation appropriately for each level of the approximation as you perceive, this is now the case when you have an infinite number of corrections So Feynman did that kind of thing in his lecture, and talked about it, and I remember, but he was a bit derisive about, it was a very mathematical, as illustrated by Illumar, that I can do it too. Of course you have to be careful about this, this is not really relevant, but just for fun, let me tell you, I was perhaps more like Feynman in my style, I'm not saying in my abilities, but in my style. And I remember finding all this work on exact solutions with curling rays and so on. It seemed so special for the physicists. I remember, you see, I knew Roy Kerr when he was a research student in Cambridge.

32:30 He was working on the equations of motion in the GR. which was simply, I mean, it wasn't recognized at first as a solution for black holes, I mean, taking black holes, it wasn't related to black holes at first, it was just an empty space solution algebraically special with curling rays, you see. And I remember going out and saying, typical of these mathematical relativities, you see, they produce these things with a particular kind of symmetry or something, or some algebraic character, but what use are they in physics, you see, And, of course, it turned out to be not only a rotating black hole, but a unique rotating black hole. So I had to keep my word. Anyway, it is true that Feynman was a bit disrespectful of these very mathematical values, whom he felt didn't understand physics, of course. but of course if there is a genuinely technical problem about the radiation you would be afraid of that it's only mathematics that's not really relevant to the physics of the device of that and that would be perhaps related to what you were saying earlier that there is as it were a correct level of sophistication no but it might be investigation, you're not quite sure what it is, what is the right level, but if there is some doubt about it, then because of certain delicacies in theory, then of course you've got to have a path of magic if they need, I'm sure, to discuss that. Sure. One other person or group that I just wanted to ask you briefly about, I'm not sure what kind of a role was played there at the time. I believe that Felix Pirani was in Dublin at the Dublin Institute for Advanced Studies for a few years. He spent some time there, yes. With, I guess, John Singh's group. Was that group, to your memory, involved in this group at all? Or was it merely that Poor Annie happened to be there? As far as I know, he just happened to be there. Really, he was working with, was he working with, consciously with Robinson and Bondi?

35:00 Or had they all been just taking an interest in something to come together? What should I remember the days? You better ask them. You see, when I remember them starting, they were in Cambridge. You see, Felix had done a PhD with Alfred Schultz in Toronto and then came to Cambridge to do a second PhD with Alfred Schultz. And I'd been an undergraduate and had a PhD student at Cambridge. and I was there as a, I don't know, I don't think I'm a, I'm not even clear whether I did a PhD or not, or whether he, I had started, I think, doing medicine in Oxford, but then changed completely and did mathematics with Cambridge, but then he was doing his GR, but I can't remember his official status because I think he was a PhD, but I do remember Felix, because I think he told me I'm doing a second PhD, And Bondi was in Cambridge at that time. So it was natural that they got together. Then I can't remember precisely at what point Felix went to Dudley. But he was already working on the comic book. I think he must have been, but you could easily find that out. More than loudly. Well, I've already taken up quite a bit of your mind, so maybe I should leave it there. I think thank you Thank you.

37:30 Thank you. Thank you. Thank you. Thank you. Thank you.

40:00 Thank you. Thank you. Thank you. Thank you.

42:30 Thank you. Thank you. Thank you. Thank you. Thank you.

45:00 Thank you. Thank you. Thank you. Thank you.

47:30 Thank you. Thank you. Thank you. Thank you.

50:00 Thank you. Thank you. Thank you.

52:30 Thank you. Thank you. Thank you. Thank you. Thank you.

55:00 Thank you. Thank you. Thank you. Thank you. Thank you.

57:30 Thank you. Thank you. Thank you. Thank you. Thank you.

1:00:00 Thank you. Thank you.