[Cover of Sullivan's book 2009, Cosmic Noise]
Sullivan's Cosmic Noise, Cambridge University Press, 2009


NATIONAL RADIO ASTRONOMY OBSERVATORY ARCHIVES

Papers of Woodruff T. Sullivan III: Tapes Series

Interview with Dennis W. Sciama
At University of Washington
12 January 1978
Interview time: 35 Minutes
Originally transcribed by Pamela M. Jernegan (1979), retyped to digitize by Candice Waller (2017)

Note: The interview listed below was originally transcribed as part of Sullivan's research for his book, Cosmic Noise: A History of Early Radio Astronomy (Cambridge University Press, 2009). The original transcription was retyped to digitize in 2016, then reviewed, edited/corrected, and posted to the Web in 2017 by Ellen N. Bouton. Places where we are uncertain about what was said are indicated with parentheses and question mark (?).

We are grateful for the 2011 Herbert C. Pollock Award from Dudley Observatory which funded digitization of the original cassette tapes, and for a 2012 grant from American Institute of Physics, Center for the History of Physics, which funded the work of posting these interviews to the Web. 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.

Click start to listen to the audio for tape 84B of the 1978 interview.

Begin Tape 84B

Sullivan

This is talking with Dennis Sciama on 25th, Ė 24th January 1978 at the University of Washington. So could you tell me just what your undergraduate and graduate training was in and when you first heard about radio astronomy?

Sciama

I read Maths and Physics as an undergraduate, which was towards the end of the War, and then I went into the Army. But during the later part of my Army period, I worked at a government research establishment, called TRE.

Sullivan

Ah, yes.

Sciama

Where much of the radar work was done during the War. And I worked on the lead sulfide class of protodetectors which were being studied because of the possibility of detecting enemy aircraft from the heat of -

Sullivan

Yes.

Sciama

And I worked on the quantum mechanics and group theory of the energy level distribution. And as a result of various memoranda I wrote then, I was accepted as a research student at Cambridge in 1949.

Sullivan

So this is a different TRE lab than the one that Ryle worked at.

Sciama

I canít remember if he was actually at TRE Ė he said that he worked on radar during the War, but there was more than one place that did radar. Thereís only one TRE.

Sullivan

I think he was.

Sciama

Yes, but that was in an earlier period.

Sullivan

So then you went to Cambridge as an undergraduate and graduate student also?

Sciama

Yes, I went back to Cambridge as a research student, and I started working actually in statistical mechanics, but in the middle of that I got interested in relativity and I got to know Herman Bond and Tommy Gold through that. I got my doctorate in 1953.

Sullivan

I see, and youíve already mentioned to me that you were at this conference in 1951 held in University College in London.

Sciama

It was, I believe, yes.

Sullivan

Do you have any recollections as to what this field was at that time?

Sciama

Oh yes! I remember quite clearly because at that time there were already a number of discrete radio sources known, of course, but they hadnít been identified. And there was much argument as to just where they were located, particularly whether they were nearby in our own galaxy, or were extragalactic. And there was an amusing argument where Martin Ryle took the view that on the basis of whatever details were known about their distribution and intensity and so on at that time, that they had to be in the galaxy. And that there were so many of them that their integrated radiation could be held responsible for the diffuse background. This was before the days of the synchrotron. Whereas I recall Fred Hoyle and Tommy Gold suggesting since the sources were distributed isotropically around us they could either be very close, and thereby not revealing the shape of the galaxy, or they could just as well be very distant and be extragalactic. And that there were some reasons for thinking that it was fruitful to consider them extragalactic. And so there was a conflict on that point.

Sullivan

Now were Hoyle and Gold already at that stage interested in the possibility of using this as a cosmological test?

Sciama

No, I donít think the idea of counting radio sources and using the log-n/log-s was around at that time. I suppose too few sources were known. The first published discussion was the Ryle-Scheuer paper in 1955. That was a bit later.

Sullivan

So when did you personally get started in this whole bruha-ha that developed?

Sciama

Well, when I switched over to relativity, I got personally friendly with Bondi and Gold and because they helped to develop the steady-state theory, I got interested in those questions. Although I hadnít worked on that theory; I think it came into my thesis in one place, but it wasnít my main concern. I was working on Machís principle for my main thesis. But through knowing them, I got interested in all the astronomical aspects of cosmology. So already by 1951, which was before I finished as a student, I was interested enough to go to this conference. And then I started working on the steady state theory just after my thesis Ė I discussed, for instance, galaxy formation in the mid-fifties. So when hostile evidence began to emerge, I was interested in that. And by that time, Iíd become attached to the theory for its esthetic virtues as I took them to be at the time. So when the hostile evidence began to come out, I tried to see whether that evidence could be refuted. So I got a bit involved in the discussions from then on.

Sullivan

So you became one of the camp of steady-state factors at that time. Is that fair to say?

Sciama

Thatís certainly true.

Sullivan

All right. And so what was the main argument that you had against, letís be careful about when, when the 2C results first came out as well as their analysis by Ryle and Scheuer, did you have arguments at that time about this interpretation?

Sciama

I donít think I was involved, myself, until a bit later. Although, I followed what was going on. And it wasnít many years after that that there was the (?) by Mills, which disagreed in detail in the region of overlap of the two surveys, they were quite different and it became evident that the Cambridge survey was confusion limited, and that their results were unreliable in the interesting region where there were many sources where the slope was in excess of what it ought to have been. And the Millsí survey, as I recall, gave, in fact, essentially the Euclidian slope to within the -

Sullivan

1.84 is what they came up with.

Sciama

Oh, thatís right. It was not 1.5, it was already 1.8 in 1958. But they said, I think, itís 1.8 but not incompatible with 1.5, given the uncertainties. That was their position at that time. But we developed various ways of trying to deal with this hostile evidence. One was clearly to exploit the fact that there was this conflict between different sets of observers, and one view we took was simply, ďWell, the observers canít agree. We just keep quiet until they do agree, and then see what opposition is.Ē But it also began to be possible to make the following point: the nature of the alleged anomaly was that there were too many faint sources than could be accounted for by this very steady-state theory. But, of course, it could be accounted for by evolution in a Big Bang theory. But an alternative way of interpreting the same data would be to say that there were too few bright sources because it depends where you normalize the curve, so that makes a difference, because too few bright sources is something that one could accommodate by statistical fluctuation, or attempt to.

Sullivan

That was the argument as to whether that made it too special a location, wasnít it?

Sciama

So it was a matter of detail how big, particularly if we were in a local hole as Hoyle and Narlikar discussed for an extragalactic distribution of sources, you try and make the parameters such that the hole is not too big, nor too isotropic Ė that was another problem because we shouldnít be at the center of the hole. And there was much argument about how convincing that explanation was. Then, in the early sixties, I proposed ironically enough, a bit the model that Ryle had favored in 1951, namely, that there was a class, by then, of course, some of the sources had been identified and we knew they were extragalactic, Cygnus A and so forth. But many sources, even today, in what then became the 3C catalogue Ė which was superceding 2C Ė even today, many of those sources are not identified.

Sullivan

Sure.

Sciama

And, of course, even more were not identified then, and this was before quasars were discovered.

Sullivan

Oh, there must have been only 10% identified in the 3C survey in the beginning, at most.

Sciama

Thatís right. So I investigated the possibility that there were two classes of radio source, this was before quasars, so I had a certain kind of prescience, although Iím not really claiming that seriously. So there were two classes of radio source Ė the extragalactic ones were kind of already known by then like Cygnus A, and another class invented ad hoc, I completely admit, to save the steady state theory, but they were galactic. The point being that then the size of the local hole, it was the galactic ones of course that gave the enormous slope and the idea was that we were in a local hole, so there was a shortage of bright sources for that reason. But the hole didnít have to be all that large according to the parameters I could pick. And the distribution of those sources was rather similar to the distribution that Ryle had thought to exist in the galaxy in 1951.

Sullivan

(?) radio stars?

Sciama

Yes, it was integrated radiation made a lot of the background, and whose individual properties were similar to the one Ryle (?). But as I said, the size of the local hole was then, as far as I remember, something like 15 parsecs whereas Hoyleís local hole was so many megaparsecs that it was already not small compared to cosmological distances Ė at least for the galaxy distribution.

Sullivan

And that paper was published when?

Sciama

I published two or three papers of that kind around 1962, 1963. But all the time, of course, the radio astronomers at Cambridge were getting better data and the isotropy was being improved. So there was a bit of a controversy between us as to whether the data could admit such an interpretation. And then the quasars were discovered, and that produced a curious twist because that could then have been my second population, except that they have large red shifts. But to complicate the matter, several people including Hoyle himself, took the view, and to some extent perhaps still does today, that the red shift might have had some other explanation than the Hubble effect. And they therefore might be in the galaxy but if so with gravitational red shifts. But I was a bit, I thought I was getting too ad hoc to pile that on to my own model, so I didnít take the view that the quasar red shifts were (?), I mean, were not cosmological. I tried having two populations of objects Ė quasars with known red shifts which were extragalactic and another family of quasars that were in the galaxy. But, of course, it was getting rather less convincing by then, but when major questions were at stakes, you try a bit hard, and of course, you reach a point where your models are too artificial, but you donít mind a little bit of artificiality at first while youíre playing around, but you donít want to give in too easily. You donít want to be obstinate and never give in, but you canít give in the minute a piece of hostile evidence comes because it can always have some meaning you didnít realize.

Sullivan

Or just be plain wrong.

Sciama

Or be plain wrong, so one therefore played with lots of these models and so on and it wasnít until a little later that it became evident that the model couldnít be sustained. But, as a matter of fact, the actual observation that led me to abandon the steady state theory was the following thing which I did in collaboration with Martin Reese who was then a student of mine. In fact, we were the first people to point out that Martin Schmidt did a much better job later, but if you just take the quasars of known red shifts, which you couldnít do earlier because there wasnít the data, and plot something equivalent to log-n, log-s, namely, the number red shift relation of the quasars in 3C. There were too many quasars of large red shifts for the steady state theory to be valid and we published that. We were, in fact, the first to publish that discussion. And that, since I didnít want to have the quasars not cosmological, I really gave in at that point Ė even in a psychological sense inside myself, and then two or three years later Schmidt did a much more detailed job.

Sullivan

Did you have the v over v max -

Sciama

No, I was going to say, because we didnít pay careful attention to all the selection effects, radio and optical and so on, which Schmidt did. And Schmidt had more data, his own data in fact, when he did his discussion. So his discussion certainly superceded ours. And heís now always the one whoís quoted, which I donít mind, but in fact, we were the first to publish the statement that -

Sullivan

So it sounds like that this was the, to you, the final nail in the coffin of the steady state, rather than the microwave background.

Sciama

Because at that time, it had only recently been discovered Ė Iím talking, I think it was 1966, when we did the n-z relation, and the microwave background was only a year old then, and indeed, I also made a model of that attributing it to the integrating effect of the discrete distribution of sources throughout the Universe. Which in 1966 one could just about do.

Sullivan

Two or three points of -

Sciama

Thatís right. You wouldnít want to simulate a whole class of black body, but the bit of ν squared, you could perhaps relate. And later though there was another argument against the high isotropy of the background you werenít seeing (?) n-fluctuations and therefore there had to be a very large number of sources, but then again, the data didnít exist in 1966. So I just, in the same spirit, I made a model of the background which I abandoned of course, a few years later when the data became much clearer. For that reason, therefore, it was the quasar distribution that actually made me recant.

Sullivan

I see. Now going back to the point where you said you could take the attitude when the Mills and Ryle survey existed side-by-side and so flatly disagreed with each other of just waiting and seeing Ė let them straighten it out. I would think it would be very frustrating to the cosmologist who had a decade or more developing to find this great disagreement. But is that all you could do Ė was just wait?

Sciama

Well, first of all, that always happens in science. Often thereís a conflict of data from different observers in any branch of experimental physics or observational astronomy, surely that is so. Thereís nothing special -

Sullivan

Well, this is particularly striking example though, I think.

Sciama

Well, Iím sure if I dredged my memory, I could think of lots of cases of conflicts of evidence, which sometimes is settled rather rapidly Ė it didnít take long in this case, actually, so that was not unusual.

Sullivan

It took almost a decade in this case, really. From the mid-fifties -

Sciama

By 1962, wasnít it, there was the Scott and Ryle paper which -

Sullivan

Yes, thatís right.

Sciama

Which got 1.8 Ė so by then there was more essential agreement, I would say.

Sullivan

I guess there was still some feuding going on, but youíre right, certainly things were coming closer.

Sciama

So that wasnít more than three or four years. As regards being frustrated Ė well, again, this is quite common in physics. If thereís conflicting evidence, unless the theorist himself understands very deeply the experiments and their technique. And he might privately favor one observer rather than another or something, but you know, as long as they are both reputable observers, thereís nothing much you can do except wait. You can go on theorizing, but otherwise, you have to wait until the data settles down.

Sullivan

And essentially, this (?) problem.

Sciama

Yes.

Sullivan

So now you mentioned that, I think the timing you gave me was 1958 or something like that, that it became clearer that you were confusion limited, but was it really that soon that it was so obvious that this was the problem with the 2C survey or is that a bit of hindsight coming in?

Sciama

Iím afraid I canít now remember offhand when it became evident.

Sullivan

Scheuer had the P of D analysis, of course which, in a sense, could be interpreted as admitting they were confusion limited, but look, we can still do something with all our old bumps.

Sciama

Thatís true, and of course, the results they got from that were rather similar to the modern results. I would have thought that was enough. But I think it became evident when the Ryle and Mills surveys were compared that the confusion was occurring because one group couldnít see sources the other group could, and so on. And I think it became very soon evident that that was the reason. But I donít, without looking in the literature, I donít offhand remember the exact year in which the Cambridge people were quite clear that that was going on.

Sullivan

Okay, well. What about the P of D analysis Ė did this have much (?) with you?

Sciama

Oh yes. Because I was making models, you see, accepting a steep slope, I mean my models of the local hole meant that I accepted the steep slope but whether the steepness was one value or another was getting more into the details. There was an anomaly to be explained was what required the local hole business, so the P of D just emphasized that that was the situation, it didnít really change it fundamentally.

Sullivan

And then when the Mills survey came along, I supposed the thing to say that, well then you wouldnít have to work so hard on these models with a steep slope and you could -

Sciama

Well, the less steep the slope, of course, the smaller the local hole or something, the less alarming the model would be.

Sullivan

Thatís right. Thatís what I mean. So then you can sit back and say, ďAh-ha, well this is going more towards what we would have expected before if we had to predict something.Ē

Sciama

Thatís right. And as Mills said, 1.8 with his errors was compatible to 1.5. But of course, even 1.5 is anomalous donít forget.

Sullivan

Yes, right.

Sciama

Because weíve got to make the red shift corrections, and you should really have even less than 1.5, so itís not as though 1.5 resolved the problem.

Sullivan

But itís not too much less than 1.5 on -

Sciama

It depends on the details, of course, depends on how far out youíre probing and so on, because the corrections are small for nearby objects Ė large for distant objects.

Sullivan

But in terms of what the sort of (?) people were putting in, I believe, the numbers were 1.3 and 1.4 Ė is this -

Sciama

Well, itís not strictly speaking, a fixed slope, you see. Itís a slope thatís varying with the density, you see. Because at high intensity, you actually (?) at 1.5 because the red shift corrections are negligible on anyoneís theory, and for very distant objects, the corrections are important. But in any case, one intensity youíre seeing near-by and distant objects because of the spread -

Sullivan

The intrinsic luminosity is that youíre scaling your luminosity function with as to where you are on this continually changing slope.

Sciama

Thatís right.

Sullivan

Okay. Well, weíve concentrated on this particular issue. Before were there any other developments during the 1950s that you as a cosmologist saw in radio astronomy that related to cosmology?

Sciama

Well, of course, the discovery of quasars and the large red shifts.

Sullivan

Thatís a little bit later. In the fifties, Iím thinking.

Sciama

Oh, Iím so sorry, you said the fifties. Excuse me.

Sullivan

For instance, the identification of Cygnus A Ė did you see this as being of import?

Sciama

Well, of course, that occurred before the Ryle-Scheuer accounts Ė that was in 1954.

Sullivan

Published in 1954, actually 1952.

Sciama

Was it? I remember the publication. And, of course, thatís what Ryle used then Ė I mean, then he had to accept the extragalactic picture and he realized immediately given that that since Cygnus A was the second brightest radio source in the sky and was already whatever it is, I forget the distance offhand.

Sullivan

It was a red shift of 18,000.

Sciama

Yes. So if an object that distant was the second brightest in the sky, one could clearly see radio galaxies, as they later called, so far away that optically they wouldnít be visible even. And therefore, one could do accounts probing the Universe to much greater depths than the 200 inch telescopes could. So in fact, the identification of Cygnus A was the very development that led Ryles to realize that by counting radio sources, he could probe the Universe to great depths. But I think at that time if I recall correctly, that was the only cosmo-, strictly cosmological consequence of radio astronomy. There were lots of astrophysical consequences, like the processes going on in radio galaxies, but I think that was the only cosmological one. And various things came later, you had to have the quasars to do the Gunn-Peterson absorption business and intergalactic hydrogen, and the fact that quasars of large red shift meant you could look at red shifts of two, all that, as you say, came in 1965 with the first red shift of two.

Sullivan

I think that is right, now that Iím thinking myself. That Ė and you mentioned in the paper in Vistas in Astronomy in 1960 that perhaps the idea of very red shifted hydrogen line observations would be of great service to cosmology.

Sciama

That, as I recall, was on the basis of some -

Sullivan

Some erroneous evidence as it turned out, but I suppose the idea was that this may be a way to get red shifts, just technically, is the better way to get red shifts than optically. Was that really what you were thinking?

Sciama

Well, Iím not sure whether itís better, but by being different, itís likely to be informative. Of course, recently, it looks as though weíve discovered cases of absorption at 21-cm in the spectra of distant quasars with a foreground galaxy containing the neutral hydrogen.

Sullivan

Right.

Sciama

So perhaps thatís beginning to come in after all.

Sullivan

But I think the thrust of what you were saying there was the object itself being redshifted (?) to the object itself and not intervening.

Sciama

Oh, sure, sure. I mean youíre detecting, what I now mean is that you detecting the red shift as the intervening galaxy by -

Sullivan

Thatís true.

Sciama

- 21-cm. Iíve just seen an absorption emission, but that -

Sullivan

Trouble is that you canít see these galaxies yet optically.

Sciama

Thereís one case where you can. Thereís a case, in fact, itís very interesting for another reason is that the case of a quasar which contains an absorption line at 21-cm which has the same redshift as that of a galaxy which is known (?) optical redshift is known which lies nearly in the line of sight, and the line of sight passes within something like a hundred kiloparsecs of this galaxy.

Sullivan

Ah, yes. Youíre talking about the one by Burke and company?

Sciama

But also, they have, Sargent and someone have now found calcium in absorption at the same red shift, which shows that the calcium in this galaxy a hundred kiloparsecs out which has all sorts of implications.

Sullivan

I see. I hadnít heard about the optical part of it. Well, this is a matter of the idea of how much, what is the density of intergalactic matter Ė was this thought of at all in the fifties as possibly radio astronomy might -

Sciama

Oh, thatís true. There were limits on that from looking for, say 21-cm line in emission, not at of course, at the line, but integrated. And George Field, and others, looked at that. They looked for it both in absorption and in emission.

Sullivan

Right.

Sciama

And, as I recall, placed limits comparable with but not quite as good as the density that would be just leave it to close the Universe at that time.

Sullivan

Was -

Sciama

In fact, George Field was working on that, I think, when I shared an office with him in Harvard in 1955.

Sullivan

He was a graduate student, or was he -

Sciama

No, he was then a member of the Society of Fellows, he was a junior fellow Ė heíd been, I think, he got his Ph.D. at Princeton.

Sullivan

Heís on my list, but I havenít gotten to him yet. Speaking of which, this is not directly related to history of radio astronomy, but itís my reading of it that there does not seem to be the concern with the overall mean density of the Universe that is of such great concern now. You know, is the Universe open or closed? In the fifties, is this true?

Sciama

Well, I was very interested in intergalactic medium. And, in fact, I was one of the few people writing papers on that subject until it became popular later. And, indeed, Iím glad to say that, you know, with the Gunn-Peterson thing, not seeing the absorption trough and all that, one attributed that to a very hot gas, and I had in fact proposed for other reasons that the gas would be very hot through essentially things like cosmic ray heating, already a year or two the Gunn-Peterson effect was discovered.

Sullivan

But this lack of concern at this time, to what do you attribute this?

Sciama

For one thing, there was rather a lack of concern in cosmology at all, it was very much a fringe subject, and then there were some people who felt that it was speculative to say there was extra matter. I could never understand that point of view, it was as though God was parsimonious, you see, and would not make more ďstuffĒ than you can actually see at that time, whereas I took the view that God could just as easily make as much matter as wouldnít be detected, because you musnít of course, break the upper limits. And if thatís an interesting form of matter and might lead to future detection, you ought to study its possible properties. So this is, of course, the difference between a rather conservative and a rather open, more adventurous, I donít know what you call it, open-minded attitude which no doubt is a bit similar to why some people like the steady state theory and others didnít. So I did a bit of work on the intergalactic medium in those years, but I was one of the very few people doing this.

Sullivan

Was this simply because of a lack of the observational capability, both optical and radio to probe =

Sciama

I suppose it was mainly that. Whereas Iím attracted to fundamental questions, and you canít always wait to work on a problem only if something detectable emerges in a year from then. Sometimes you have to do rather more academic things.

Sullivan

Well, this is always a question of interest as to which leads which Ė the observations lead the theory or the theory leads the observations, and what youíre saying is that in this case, the observations needed to come first to spark the interest amongst the theorists.

Sciama

Well, among general interest, yes, but my belief is that it seemed rather obvious that itís a two-way process, and one ought toÖ I donít mean you study absolutely any problem however arcane or speculative it is, but since galaxies formed, always existed, itís a natural question to say, ďHow efficient was the galaxy formation process, how much was left over between the galaxies?Ē That would then to discuss the astrophysics of that gas that was left over, and who knows what observational methods would emerge in future years, so you donít wait till theyíve emerged to work on a subject, you prepare for that by working on that anyway. Thatís if youíre in cosmology, but of course, at that time, there were very few cosmologies anyway, and perhaps most of those were more mathematicians interested in the geometry of space time and solutions of Einstein equations. Whereas I started working on the physical properties of the gas and many of the relativists are not in that sense physicists.

Sullivan

You alluded to something which you mentioned to me earlier today, namely, the difference in philosophical bent of one who would believe in steady state and one who might not. Could you, just for the record, say how you see them as being different? In terms of the sort of person whoís attracted to them?

Sciama

Well, Ďbelieveí is too strong an expression; I would say be sympathetic with the steady state theory.

Sullivan

Well, thatís (?) fuzzy.

Sciama

No, I think a better word; no, because why should you believe it to be true if thereís no evidence for or against, which was the early situation Ė there was nothing to make you believe it to be true, the point is rather -

Sullivan

To favor.

Sciama

Yes, but what does favoring mean? I think the hard cash test if favoring means do you work on it? If you like the theory, you actually do calculations in it, or you defend it when hostile evidence begins to come, but you have to then debate it. So it effects how you actually work, the questions of literal belief I donít think are relevant at that stage. Later they would be.

Sullivan

Sorry about that.

Sciama

No, Iím not trying to be pedantic about it; I think itís an important point. Because there wasnít sufficient reason to say it was true, but just that it was a very attractive theory, one tried to help to make it clear that it was in fact true, by working out its consequences and some of the problems that it gave rise to were very interesting theoretical problems anyway, and the link with observation. As to personal types, well, I once made a little joke that A believed in the steady state theory because he wasnít religious and B believed it because he was. Itís an adaptation of the joke of Bertrand Russell. But I, you see, felt that the overall beauty of the model and the fact that life would always be possible somewhere in the steady state universe, was psychologically attractive to me. And that counterbalanced the unattractive feature that you have to change the laws of physics. Which, of course, you donít do lightly. Whereas for other people those attractions werenít so great and looming larger for them was the undesirability of changing the laws of physics. So thatís very much a personal matter.

Sullivan

So it would be a more conservative viewpoint in the fifties you would say, to not believe in the steady state theory.

Sciama

Oh, undoubtedly. Because it was saying the conservation of matter was false, and that was one of the keystones of physical theory.

Sullivan

On the other hand, in those histories that have been told and certainly the elementary textbooks, the two are always put up against each other as being equally likely and the evidence for one and the evidence for the other is not a matter of here comes this radical new theory Ė theyíre sort of given equal weight, it seems to me.

Sciama

Well, this might be a question of purpose of exposition because I think if you counted heads, it was only a minority of people who were really sympathetic. And it certainly wasnít on the same footing that it was a radical innovation saying this law of conservation which has been sacred for so long is wrong, and we must change physics. That canít be on the same footing as not changing physics. Naturally, the, it wasnít a ridiculous suggestion because the actual creation rate was far less than the accuracy of the observations, but nevertheless -

Sullivan

Which was an unappealing thing too, it could never be checked directly.

Sciama

No, but it could be checked by observational tests of steady state which as the proponents of the theory kept harping on, itís a very vulnerable theory and that makes it very scientific. But I think it was more to clarify the exposition in popular books on cosmology that they were given, as it were, equal weight, but I donít think that was a view of most physicists.

Sullivan

And as you said before, thereís also the great problem of the singularity in general relativity in the expanding universe, but that was only at least at one time, whereas youíre violating the laws of physics everywhere, and that was Einsteinís physics which had been around for fifty years and so that was not so bothersome. Would you say thatís fair?

Sciama

I think so, yes. As a matter of fact, I met Einstein when I was at Princeton in 1955 just a week before he died and I had quite a long talk with him, and I mentioned the steady state theory and he immediately said itís not in agreement with the field equations. Which, of course, it wasnít, without tampering with them, because of the conservation thing. So I told him the idea I had at the time about how to tamper with them, but of course, he wasnít very sympathetic.

Sullivan

He tampered with his own, I guess.

Sciama

Yes, but not in quite such a dramatic way.

Sullivan

Okay, so and then you say quasars were the next thing and what was the power of the quasars Ė simply that they allowed you to probe that much further away?

Sciama

Well, we hoped they would. Because of their great brightness, they could be seen to larger red shifts, but then it emerged that when you plotted one thing against another that should have given a nice tight relation that would have perhaps solved the cosmological problem, that they were scatter diagram emerged and that was part of the controversy because Hoyle and Burbidge claimed that that showed that the quasars had to be local and other people said that it only meant there was a wide distribution in intrinsic brightness and so forth.

Sullivan

But even if that were true, however, some of them would allow you to probe very deeply (?) and if you could then come up with some other criterion for luminosity such as spectrum or something else -

Sciama

Well, but theyíre beginning to try to do that only today.

Sullivan

Right, but am I reading you right then that what youíre saying is that as soon as the origin of the redshift was called into question, then the power of the quasars in terms of inciting cosmological questions, at that stage in the early and mid-sixties greatly decreased?

Sciama

No, no, thatís not quite the point I was making. What I was saying was that even granted the redshift is cosmological they werenít useful in probing the Universe because the great spread in intrinsic properties of the quasars that were then implied by that interpretation -

Sullivan

At that time, you could not see any other hope of determining their luminosity Ė no other means of determining luminosity so you decided to give up?

Sciama

Well, I wouldnít put it quite like that (?) one decided to give up as though that was foolish and one shouldnít have given up. No method emerged of getting rid of the scatter in a way that they could be used for probing, so one couldnít do anything.

Sullivan

And as you say, itís taken a decade to -

Sciama

And even now, one is only beginning to perhaps be able to do that, so it wasnít a voluntary decision not to do something which one might have not decided that way Ė one simply had no choice.

Sullivan

Okay, I think thatís all. Thank you very much. That ends the interview with Dennis Sciama on 24th January at University of Washington.

End Tape 84B


Modified on Wednesday, 05-Jul-2017 07:28:59 EDT by Ellen Bouton, Archivist (Questions or feedback)