Authors: Kitty Ferguson
Hipparcos cannot directly measure the distance to these globular clusters. They are in the Milky Way’s halo, outside the Galaxy’s main disc, too far away for parallax measurements even with Hipparcos. Instead, researchers used Hipparcos to measure the distance and brightness of other stars and compared them with stars of similar composition in globular clusters. Catherine Turon of the Paris-Meudon Observatory, who along with others has calculated 12.8 to 15.2 billion years for the age of a globular cluster known as M92, has admitted there are difficulties with such measurement: the stars used for comparison are often dim stars that have no metals or other heavy elements. Getting models adapted to such extreme objects with low metallicity is problematical. Processes such as fast rotation or metals sunk out of view into the star could skew the conclusions. Michael Perryman, a project scientist for Hipparcos, is even more sceptical. Hipparcos’s own data have shown that some of the stellar models are spectacularly wrong. These new calculations of the age of stars did not put the age-of-the-universe paradox to rest.
However, also in the autumn of 1997, astronomers using the Hubble telescope to watch the collision of two galaxies called the Antennae observed at least 1,000 clusters of newborn stars forming from giant hydrogen clouds in the centre of the merging galaxies, indicating that globular clusters are not all so ancient. Some at least are emerging out of more recent galactic collisions.
Perhaps old assumptions about the history and ages of stars are not unshakeable after all. But Freedman points out that though not all globular clusters are as old as previously thought, a great many in our Galaxy
are
. Also, continuing studies of faint galaxies in the Hubble Deep Field argue for an older universe. They show that some elliptical galaxies were already well advanced in years at a red shift of 1.2. It doesn’t
help
that
some
things are younger than previously thought.
Sandage, Freedman and their associates were treading on the frontiers of science – very recent and ongoing science. The Cepheids in the Virgo cluster that stirred up the controversy couldn’t be seen at all before the Hubble Space Telescope, in fact not before its faulty optics were corrected in December of 1993, less than a year before Freedman’s team’s discovery. Some of the measuring techniques being used were barely past the experimental stage, yielding data whose implications no one fully understood. It would have been the stuff of tabloids to declare anything settled. Those who are uncomfortable when science yields paradoxes rather than certainties would have to go on being uncomfortable for a while.
But suppose the Hubble constant does end up indicating that the universe is younger than some of its stars. It may seem relatively easy to think of the rapidity with which the universe is expanding as being the result only of a two-way tug-of-war between gravity (which is working to make the universe contract) and the expansion energy resulting from the Big Bang (which is working to make it expand). There may, or may not, be another player involved, and that player is Einstein’s old ‘mistake’.
Popular science books and articles habitually describe the cosmological constant as a repulsive force that might counter the effect of gravity. It would be well to acquire a little more sophistication and realize that the cosmological constant can actually work either way.
If the cosmological constant is a positive number, then it will indeed counter gravity, joining in the struggle on the side of expansion. However, if it is a negative number, then the effect will be to weigh in on the side of gravity. If it is zero, then it will do neither. Think of it then as a theoretical property of the vacuum of space that, if it exists and isn’t zero,
might
act to stretch space and thus counteract gravity’s contracting power, or
might
do the opposite. To put that another way, imagine
yourself
facing a dial. If the arrow is pointing to zero, then the cosmological constant is having no effect at all. If you move it to the minus side of the zero, the further you turn it the more it will contribute to the contraction of the universe. If you move it to the plus side of the zero, the further you turn it the more it will contribute to the expansion of the universe.
Even this slightly more sophisticated view of the cosmological constant does not begin to do justice to the complications involved when physicists and astrophysicists play with its value. The cosmological constant can seem to be working both ways at the same time, allowing us to have our cake and eat it too. However, what seems a contradiction is not, because of the way the cosmological constant fits into the equation for omega.
Theories of quantum mechanics – the study of the very small (atoms, molecules and particles) – have it that everywhere in the universe particles are spontaneously popping into and out of existence. Their life spans are unimaginably short. Nevertheless, ‘empty space’ seethes with this energy, and ‘empty space’ does not mean only what is out there dark and remote between the stars. This quantum energy fills the enormous amount of empty space within the atoms that make up chairs, tables, human bodies and all other things familiar and unfamiliar. ‘Emptiness’ is full of energy. Theory suggests that the energy of the cosmological constant might be this energy of virtual particles which wink in and out of existence at all times and everywhere in the universe.
For Einstein, the cosmological constant was only a mathematical device, and not long after he put it into his equations in order to avoid the implication that the universe must be either expanding or contracting, he decided it had been a mistake – for of course the universe
is
expanding. After visiting Hubble at Mount Wilson in 1931, Einstein rejected the whole idea, calling it ‘theoretically unsatisfactory anyway’. But the cosmological constant didn’t go away. Lemaître in particular enjoyed fooling around with it and adjusting its value, discovering that by
fiddling
with this theoretical dial he could construct universes that started out very slowly and then sped up, or universes that started out fast and then slowed down, or universes that began expanding, stopped and then expanded again. Something like that stop-and-start version was evoked in the 1940s as a possible remedy when new discoveries indicated that the universe was younger than the solar system. When it then turned out that the Hubble constant had been overestimated, the cosmological constant wasn’t required after all and was packed away once again.
In 1948, researchers detected the effects, on atoms, of the vacuum energy decreed by quantum mechanics, but no one went on to study its possible influence on the universe as a whole until 19 years later when Zel’dovich, the Soviet theorist, realized that this vacuum energy would enter into Einstein’s equations in just the same way that the old cosmological constant did. Before long it became evident that if Einstein had been right about mass causing spacetime to curve, and if all this vacuum energy really does exist, then the vacuum energy ought long ago to have curled the universe up into a tiny ball or something even smaller, or else driven the expansion so that even atoms – much less galaxies – could never have formed. Even by making the cosmological constant extremely small, Zel’dovich couldn’t show how the universe could have turned out to be the way it is. So it seemed the value must be zero, and that is what most theorists since have been assuming. That zero does not, by the way, mean that there is no vacuum energy, only that by some truly remarkable coincidence, all the positives and negatives in that vacuum energy cancel out exactly.
As we’ve seen, the cosmological constant is still with us, hovering like a ghost in the equation for omega. If its value is zero we could write it off and forget it, but the symbol for it would still be sitting there. I am reminded of a recipe one of my more eccentric friends gave me for broccoli soup. The recipe had been passed from cook to cook, many times. Always the list
of
ingredients included a can of won ton soup, with parenthetical instructions: ‘Don’t put this in.’
Even before the recent discoveries that the universe may be expanding much faster than previously supposed, late-20th-century astrophysicists had been feeling once again an itch to reach for the cosmological constant dial. There was a possibility it might offer solutions to some intractable problems such as the still-missing dark matter. Nearly everyone was approaching the idea super-cautiously, having been burned twice before. John Noble Wilford commented in a
New York Times
article that one thing that makes physicists particularly uneasy about assigning the cosmological constant a value other than zero is that this reminds them too much of the way medieval astronomers designed increasingly complicated celestial mechanisms to explain the planets’ motions, in order to preserve their beloved earth-centred Ptolemaic universe. As was true with those mechanisms, there was nothing to indicate that a cosmological constant value other than zero was wrong. But there was also nothing to indicate it was right. The best argument for it was that it allowed physicists to cling to theories in which they had a vested interest! Being able to turn the cosmological constant dial to a number (negative or positive) of their choice that ended up supporting the currently favoured version of the Big Bang theory was just a little too easy and allowed for too much leeway. With a friend like the cosmological constant, did a theory need enemies?
In 1990, Michael Turner of the University of Chicago and Fermi National Laboratory proposed a recipe to add up to critical density: 5 per cent ordinary matter; 25 per cent cold dark matter (including both invisible and ‘exotic’ types); 70 per cent the cosmological constant or something like it. According to Turner, the energy of the cosmological constant could compensate for some of the missing mass and serve as an additional brake on cosmic expansion, balancing things out in such a way that the universe would neither eventually collapse
nor
expand into an ever darker, thinner, colder infinity, but instead perch for all eternity on that highly desirable knife edge between the two. The lower density of matter that such a cosmological constant value would allow might be an added boon to theorists, making it easier to explain how matter congealed into such enormous structures as the Great Wall of galaxy clusters.
After Freedman’s team’s discoveries in late 1994, physicists began to consider much more seriously these suggestions that the cosmological constant might not be zero. Adjust the dial, and the cosmological constant’s energy, over time, could change the rate at which the universe expanded. If expansion was slower when the universe was young, that would have given more time for stars and large structures to develop. Later, the energy of the cosmological constant could have influenced the expansion to speed up. The current measurements of the rate of expansion, by Freedman and others, would be only measurements of the
present
rate of expansion, and unreliable indicators of the age of the universe.
However, though a non-zero cosmological constant was looking more and more tempting in terms of explanatory power (in other words, useful to explain what was going on), there was still the major hitch that no one had yet been able to find any direct observational evidence that the value was anything other than zero. The first hint that this might change came in 1996, not, itself, from direct observation.
The age-old method of testing alternative ideas and making up for insufficient evidence by using mathematical simulations came into its heyday with the advent of supercomputers. In the 1980s and 1990s, as never before, it was possible to feed in some observed and assumed conditions in the early universe and find out what this might lead to after billions of years. In 1996, an international team headed by Carlos Frenk of Durham University in England, using Cray supercomputers at Munich and Edinburgh, ran simulations to find out whether
temperature
fluctuations observed in the early universe could have led from a Big Bang fireball, where everything was almost uniform, to today’s universe of galaxies, clusters and voids.
The starting point for the simulations was the universe as it is thought to have existed 300,000 years after the Big Bang. That was when the cosmic microwave background radiation originated (the radiation that Penzias and Wilson detected in 1964). In 1992 George Smoot and his colleagues had been able to discern wrinkles, tiny energy fluctuations predicted by Big Bang theory, in this otherwise smooth cosmic fabric. Frenk’s team simulated the growth of those initial wrinkles. The results supported the possibility that the cosmological constant should indeed be summoned from limbo.
Frenk and Simon White at Munich (assisted by Adrian Jenkins, Frazer Pearce and Joerg Colberg) ran four different simulations. One of the ways they differed from one another was in the estimates of the mass density of the universe. A second difference was in whether or not Frenk and his colleagues allowed the cosmological constant to be other than zero. Of the four (
see Figure 8.2
), the one capable of producing the universe as we know it today was the model in which the mass density was only 30 per cent of what experts think would be needed to produce omega-equals-one, or critical density,
and
in which Frenk also factored in a non-zero cosmological constant.
In this simulation, the expansion rate changed over time and was slower in the early universe than it is today. The computers demonstrated how the wrinkles might have attracted surrounding matter. Lumps of matter, according to the simulation, collapsed onto themselves and grew larger by merging with other lumps, eventually forming a complex filamentary network of large, twisting ridges surrounding vast empty regions. Gas and dark matter flowed along these filaments. Where the filaments intersected, galaxies and galaxy clusters formed. In the simulation, the last few billion years don’t show much gross
alteration
, for the universe is expanding fast and the mass density is too low for the large structures to change very much.