Geoffrey Burbidge is Professor of Physics at the University of California, San Diego. He served for six years as Director of the Kitt Peak National Observatory. His latest major award, jointly with Margaret Burbidge, is the Gold Medal of the Royal Astronomical Society. Jayant Narlikar served as
Founder Director of the Inter-University Centre for Astronomy and Astrophysics in Pune, India, until his retirement in 2003. He is now Emeritus Professor at IUCAA. Among his current interests is exobiology.
Both of us were asked to describe our views of the ways we first approached this topic. We have decided to combine our contributions but present them separately because we came to the basic ideas from different directions. Geoffrey Burbidge became interested in the CMBR from his early association with the fundamental problem of the origin of the chemical elements. Jayant Narlikar had been interested in alternative cosmologies and was, therefore, concerned with the problem of how the CMBR could be produced without a hot big bang. Each of us has given a "first person" account. As we had the benefit of close interaction with Fred Hoyle we have folded in his views also wherever necessary.
The Approach Taken by Geoffrey Burbidge. My first interest in this area came during the period 1955 to 1957 when Margaret Burbidge, Fred Hoyle, Willy Fowler, and I were solving in detail the problems of the origin of the elements (Burbidge et al. 1957).
I realized that the large abundance of helium in stars (MHe/MBaryon = Y = 0.24) meant that there must be a very special place, or an era, when there had been a great deal of hydrogen burning. At that time, the value of the Hubble constant was thought to be 180kmsec-1 Mpc-1 (Humason, Mayall and Sandage 1956), so that the Hubble time was H-1 ~ 6 x 109 years. Taking the luminosity of the Milky Way to be about 1044ergsec-1, this meant that over 6 x 109 years the total mass of helium that was produced by hydrogen burning would be far less than 24% of the total mass of helium ~ 2.5 x 1010 solar masses.
I did not realize at the time that my argument was very similar to that which had been made by Alpher, Bethe and Gamow (1948) a decade earlier. At the time of the first calculation by Gamow, Alpher and Herman, Hubble and Humason (1931) had given a value of H0 = 550 km sec-1 Mpc-1, so that H-1 ~ 2 x 109 years and the discrepancy between the observed abundance of helium and the amount which could be attributed to hydrogen burning in stars was even larger. However, in contrast to me, Gamow and his colleagues had discussed the basic physics of the big bang and concluded that helium could only have been made in the early universe. Up until then it had been assumed that in Friedmann models, in the beginning the rest mass energy is much greater that the radiation energy. The immediate effect of the change to a radiation-dominated universe was to require that the scale factor of the universe a(t) is proportional to t1/2. Omitting electron-positron pairs, the radiation temperature T is inversely proportional to a. Thus the radiation temperature T is proportional to t-1/2. With radiation alone and no neutrinos Tg = 15.2 xt-1/2 where Tg is measured in units of 109 K and t in seconds. However, the numerical coefficient 15.2 is modified by the presence of electron-positron pairs and by neutrinos. For temperatures high enough for the electrons and positrons to be relativistic, and for two mass-less neutrino types, the numerical coefficient is changed from 15.2 to 10.4. So long as the energy in the early universe is dominated by radiation the equation above holds.
But the next step in the discussion was completely ad hoc. The mass density of stable nonrelativistic particles, explicitly neutrons and protons, decreases with the expansion of the universe at a rate proportional to a-3, i.e. as t-3/2. Calling this density pb, Alpher and Herman (1948) took pb = 1.70 x 10-2t-3/2 gcm-3 with the coefficient 1.70 x 10 2 being the ad hoc step. There is nothing in the theory which fixes this value. It is a free choice, chosen to make things right, in this context to obtain the calculated value of the helium abundance Y to agree with observation. Thus, while the big bang theory can explain the microwave background, it tells us nothing about the helium abundance unless we choose a numerical value which enables us to do this.
This is fine if you come to the problem of the helium with a belief in the big bang. And this is what most contributors to this book have done. But I came to the problem with no cosmological beliefs.
In the 1950s a debate was going on between the majority of cosmologists, who believed in a beginning, and a few, particularly Hoyle, Bondi, and Gold, who had developed an alternative, the steady state cosmology (Bondi and Gold 1948; Hoyle 1948). By the late 1950s, standing on the sidelines in Cambridge, I realized how unpopular the steady state theory was, since at the time there was a very unpleasant dispute going on between Ryle and his group on one side, and Fred Hoyle. In the early 1960s, Hoyle and Narlikar (1961) gave an alternative interpretation of the radio source counts to show them as consistent with the steady state theory, whereas Ryle insisted these provided strong evidence against the steady state.
Returning to my own work on the origin of helium, I made a calculation assuming that all of the baryonic matter of the universe with a density pb = 3 x 10-31 gcm-3 had the same helium abundance. I then showed that if it were produced by hydrogen burning the energy density must amount to ^4.5 x 10-13 erg cm-3 (Burbidge 1958; see also Bondi, Gold and Hoyle 1955).
In my paper I offered several possible scenarios for the production of helium. It could have been produced in the early universe if there was one;
it could be due to higher luminous phases in galaxies for periods during their lifetimes; or I speculated it was possible that we were overestimating the real cosmic abundance of helium because the ratio of helium to hydrogen was much smaller in the low-mass stars which make up a large part of the total mass, than it is in the hot stars and nebulae in which the abundances can be determined spectroscopically.
The key point that I missed, as did Bondi, Gold and Hoyle (1955), who had made a similar calculation in 1955, arguing that the energy must have come from red giants (in 1958 I had missed the Bondi, Gold and Hoyle paper), was that the energy density corresponding to the production by hydrogen burning when the energy was degraded to blackbody form would give a blackbody temperature of 2.75 K!
If these results had been publicized, they might have been seen as predictions based on observed quantities of what the temperature of the blackbody radiation would turn out to be, if it were detected. But of course this never happened.
As he told me many times later, Fred Hoyle had realized all along that the hydrogen burning in stars was a possible source of the helium and that it would lead to a powerful background radiation field. Much later he and I took very seriously the fact that the CMBR energy density is so close to what the prediction from the hydrogen burning origin would give, and concluded that all of the light isotopes D, 3He, 4He, and 7Li also have a stellar origin. In other words all of the isotopes in the periodic table are due to stars. Our paper on this topic was rejected by Physics Review Letters, obviously because very convinced big bang advocates refereed it. However, it was finally published in 1998 in the Astrophysical Journal Letters (Burbidge and Hoyle 1998).
A key point that most physicists were unaware of throughout the 1950s, 1960s, and 1970s, and in particular the large number of those who believe in the standard model still appear to be unaware of it, is that in 1941 A. McKellar at the Dominion Astrophysical Observatory in Victoria made an estimate of the radiation field in which the interstellar molecules CN and CN+ are bathed, and stated that if this was blackbody the radiation temperature is 1.8K< T <3.4K. The exact quote from his paper (McKellar 1941) is as follows:
Dr. Adams has kindly communicated to the writer his estimate of the relative intensity, in the spectrum of Z Ophiuchi, of the A3874.62, R(0) interstellar line of the A3883 CN band and the A3874.00, R(1) line, as 5 to 1. B0J"(J" + 1) + ... has the values 0 and 3.78 cm-1 for the 0 and 1 rotational states and for the two lines
R(0) and R(1) the value of the intensity factor i are, respectively 2 and 4. Thus from (3) we find, for the region of space where the CN absorption takes place, the "rotational" temperature,
If the estimate of the intensity of R(0)/R(1) were off by 100 percent, this value of the "rotational" temperature would not be changed greatly, R(0)/R(1) = 2.5 giving T = 3°4K and R(0)/R(1) = 10 giving T = 1°8K.
Had this been generally known in the 1950s, and been put together with the result quoted earlier, the history of what most people want to believe about the CMBR and its origin might be different.
At the time in the early 1960s when Fred Hoyle and George Gamow were debating cosmology, Fred was aware of this result, and used it when Gamow would argue that the temperature was likely to be much higher. I first learned of this result from Fred in that period.
My view of the subsequent history (as I saw it) is as follows. In the early 1960s Robert Dicke and J. Peebles reworked the ideas of Gamow, Alpher, and Herman. Since Dicke was a superb experimentalist, he proposed that an attempt be made to detect the radiation. This is what he and David Wilkinson set out to do. But, of course, before they achieved any result there was the serendipitous discovery by Penzias and Wilson (1965a).
But throughout the 1960s the ideas emanating from Princeton and also from Moscow from Zel'dovich's group led almost everyone to believe that the radiation could only be a remnant of a big bang and would be of blackbody form.15 It would be proof that the steady state theory was wrong. With the Penzias and Wilson discovery, while there was still no proof that it was blackbody, it was thought that the verdict was in.
Even Fred Hoyle began to doubt the correctness of the steady state cosmology, and in his address to the British Association in September 1965 he came as close as he ever did to concluding that the steady state would not work. Starting at that time, he began to discuss a modification of the steady state which in the 1990s, with J. V. Narlikar and me, was turned into the QSSC - an oscillating model still over the long term a steady state universe (Hoyle, Burbidge and Narlikar 1993).
15 It was in this period that my view that cosmological ideas are driven as much by the views of leading scientists as by actual observations was strengthened. I was present at meetings where early rocket observations were reported which did not confirm the blackbody idea. Those were immediately severely criticized by leading theorists who did not understand the experimental details but were absolutely convinced that the blackbody nature must be correct. They eventually turned out to be right, but their prejudice was obvious.
Jayant V. Narlikar's View. I recall that one day in 1964, Fred Hoyle walked into his office in the DAMTP in a rather disturbed mood. He confided: "I believe, I have found the strongest proof for the big bang." With his previous encounters with Martin Ryle and his colleagues in the Cavendish, I wondered if there was some new evidence from radio astronomy that had unsettled Fred. "No," he added, "my own calculations suggest that helium was mostly made not in stars but in a high temperature epoch in the past. I find that if the density-temperature relationship is properly adjusted one can get almost 25% helium."
For someone who had worked long and hard on stellar nucleosynthesis to demonstrate that most of the chemical elements were made in stars, this finding had come as a shock, even though it was he himself who had done the calculation. His work with Roger Tayler was subsequently published in Nature (Hoyle and Tayler 1964) and quickly became a much-cited paper . . . probably it was the only paper Fred wrote with conclusions close to favoring the big bang scenario. Nevertheless, he left an alternative possibility open, namely the existence of supermassive objects that allow stellar nucleosynthesis to generate adequate helium. This possibility is also discussed briefly in the classic paper on nucleosynthesis by Wagoner, Fowler and Hoyle (1967).
Even so, Fred did not relate the 1964 finding with the possible existence of relic radiation. The result struck him as very important only in 1965 after the discovery of the radiation by Penzias and Wilson (1965a). Although the blackbody nature of the radiation had not been established in 1965, its finding together with helium abundance apparently had the effect of convincing him of the existence of a high temperature phase early in the universe.
It was against this background that he delivered his oft-quoted speech to the British Association for the Advancement of Science (Hoyle 1965) in which he came close to supporting the big bang cosmology at the expense of his own steady state theory. One popular magazine in the USA likened this reaction to the problematic situation of Lyndon Johnson abandoning his membership of the Democratic Party to join the Republicans!
I had worked with Fred on many aspects of the steady state theory, and felt that Fred had "given in" too soon. Dennis Sciama, another strong adherent of the steady state idea, also felt the same, although within a couple of years he changed over to the big bang point of view. In the meantime, Fred had second thoughts on the matter. Both he and I, along with Chandra Wick-ramasinghe, felt that alternative explanations of the radiation background should be looked for. The reasons were mainly as follows:
(i) There are radiation backgrounds at various other wavebands and these are mostly traced to astrophysical sources. Can the microwave background be shown to originate from astrophysical sources radiating mainly in infrared and microwaves?
(ii) Following a more general line of argument, there are galactic and extragalactic astrophysical processes with energy densities comparable to the newly discovered microwave background (CMBR), for example cosmic rays, magnetic fields, and galactic starlight. So to ascribe a relic interpretation to the CMBR gives an unexplained coincidence of energy density.
(iii) The fact that if all helium in the universe were made in stars the resulting energy density would be comparable to that of the microwave background which has already been highlighted in this paper suggested a nonrelic interpretation.
I will discuss these possibilities briefly from a modern standpoint.
It was shown by Wolfe and Burbidge (1969) that the multiple source hypothesis would generate a microwave background that was too inhomoge-neous for agreement with the preliminary limits on anisotropy. The only way to escape from this conclusion was that the sources were far more numerous than galaxies and typically weaker than galaxies. Such a population was considered rather unlikely and has not been found.
The search for an astrophysical process to generate the CMBR in the Milky Way Galaxy or in clusters of galaxies led Hoyle and Wickramas-inghe to various scenarios involving interstellar dust: dust that could convert starlight or other energy into a thermalized form with the energy density found in the CMBR. Narlikar, Edmunds and Wickramasinghe (1976) wrote a paper suggesting how this could happen using dust grains in the form of whiskers. The scenario was plausible but it was not clear that it would meet the various observational constraints that were being placed on the properties of the CMBR.
The idea of Narlikar, Edmunds and Wickramasinghe (1976) could be applied to a situation in which it was assumed that there had been a lot more starlight initially because of greater stellar activity, which led to most of it being thermalized by whiskers. This idea, however, ran into problems with the original formulation of the steady state theory, which would not allow any epoch-dependent process. Nevertheless, Hoyle and Wickramasinghe persisted with the efforts to study the thermalization process in detail.
Eventually the process was shown to work, not in the original steady state cosmology but in its variant, the Quasi-steady state cosmology. This cosmology was proposed by Hoyle, Burbidge and Narlikar (1993) and it envisages a long-term steady state universe with short-term oscillations. The e-folding time of the long-term steady state is around 1000 Gyr, whereas the period of a typical oscillation is around 50 Gyr. We refer the reader to the details given in Hoyle, Burbidge and Narlikar (2000) and to later references (Narlikar et al. 2003). So far this alternative is able to achieve the following:
(i) Explain the CMBR as a relic of stars burnt out in the previous oscillations with the present temperature of 2.7 K related to stellar activity at present observed in the universe. See Hoyle, Burbidge and Narlikar (1994) for details.
(ii) A Planckian spectrum at all wavelengths except possibly at wavelengths longer than 20 cm. (There the galactic noise anyway dwarfs the cosmological effect.)
(iii) An angular power spectrum that explains the main peak at around l = 200, as arising from typical clusters at the last minimum scale epoch (Narlikar et al. 2003).
(iv) The dust density required for thermalization being consistent with that needed for dimming distant supernovae.
(v) A weak polarization on the scale of clusters arising from magnetic alignment of whiskers scattering the radiation.
(vi) Independent evidence for the existence of whisker dust from various astrophysical scenarios.
Fred Hoyle firmly believed that an alternative interpretation of the CMBR along the above lines would turn out to be closer to reality than the standard interpretation. What were the attitudes of the other two coauthors of the steady state theory? I never had the chance to discuss the CMBR with Tommy Gold. By 1965 he had already moved away from cosmology and I do not think he worried too much about the issue. Hermann Bondi had likewise developed other interests. However, I had met him on several occasions. Once in an interview on the All India Radio, Pune, during the 1990s I had asked him what he felt about the steady state theory in the light of the observations of the CMBR, especially by COBE. He replied that to him the steady state theory had been attractive from the Popperian point of view: it made definite statements which could be checked against observations. That the CMBR spectrum had turned out to be so close to the Planckian was, in his opinion, a very difficult observation for the steady state theory to explain. So he had felt that the theory was no longer viable. Like most cosmologists he had been unaware of the above work on alternative cosmology, but seemed pleased that perhaps such an explanation of the origin of the CMBR might succeed.
Going back to 1965, one can say today that while the big bang scenario has been taken a good bit forward in the last four decades, the alternative explanation has also made considerable progress and deserves to be critically examined side by side with the standard explanation.
Was this article helpful?