How old is the universe? It is a question answered for thousands of years with dreams and speculation. Only in the last 50 years have answers based on observation become possible, and even now astronomers disagree.
Such disagreement reached the public's eyes and ears recently when several astronomers announced that the universe was not 15 billion years old, as most astronomers had believed, but only 10 billion years old. Despite the publicity given their then-unpublished results, most astronomers remain unconvinced. To understand why, it is important to know how astronomers measure the age the universe's age.
Asking the age of the universe is a meaningful question because astronomers believe that the universe has not existed forever, but that it began in one unimaginably hot and dense fireball called the big bang. That our universe has a finite age is philosophically intriguing. That we can estimate that age to a fair degree of accuracy is truly impressive.
No single measurement of the time since the big bang gives a specific, unambiguous age. Fortunately, astronomers have at their disposal several methods that together fix the age with surprising precision.
Nothing in the universe can be older than the universe as a whole, so the age of the oldest objects in the universe is the lowest possible age of the universe itself. Stars, for example, live from a few million to many billions of years, depending on how quickly they burn their hydrogen fuel. We know this, not from watching stars age (we don't have the time), but by understanding the star's nuclear reactions and stimulating its history with a computer. Comparing that history with what we observe in space, we find the oldest stars to be 12 to 18 billion years old.
Because stars were not formed until shortly after the big bang, the universe is older than that. How much older? A prestellar cloud requires at least a million years to collapse gravitationally; after a billion years the universe would have diffused to too low a density for stars and galaxies to form from the expanding cosmological gas. (Stars form even today in the much denser gas found inside galaxies.) By this argument, the oldest stars formed no less than a million years and no more than a billion years after the big bang, and the universe began between 12 and 19 billion years go.
The radioactive elements found in some meteorites provide another direct measure of the universe's age. This is much like the carbon 14 dating that archaeologists use to date the remains of living creatures on earth. But carbon 14 has a half-life of only 5,500 years. That is, every 5,500 years one-half of the remaining carbon 14 decays. That is adequate for the few tens of thousands of years needed by archaeologists, but astronomers must measure billions of years. If any carbon 14 had been created that long ago, far too little would remain today to be detected or used for dating.
Yet some elements have sufficiently long half-lives for measuring the age of the universe. rhenium 187, with a half-life of 60 billion years, enters the galaxy when supernovae explode and shower the galaxy with heavy elements that were "cooked" in their nuclear fires. The rhenium immediately begins to decay to osmium. Billions of years later, when the rhenium and osmium are trapped in a newly formed meteorite, a record of the time since rhenium's creation is locked into the meteorite, waiting for the day researchers crack open the rock to probe its secrets.
The latest analyses of meteoritic rhenium show an average age for the rhenium of about 9 billion years. If all rhenium-producing supernovae had exploded at the universe's birth, then 9 billion years would be the approximate age of the universe. But supernovae have exploded recently, as evidenced by the beautiful crab nebula, the remnant of a supernova observed 900 years ago. Rhenium produced in recent supernovae must be balanced by rhenium from older supernovae to obtain the average of 9 billion years. Using this argument, the age of the universe based on nuclear-decay chronology is from 13 to 22 billion years.
The third important method is in some ways the least direct. It has also produced the most controversy, and perhaps should be considered the least reliable of the three.
The first two methods fixed the age of the universe by finding the age of objects in the universe. The third method does not directly measure the age of anything. Instead it relies on the distances to, and velocities of, distant galaxies.
Edwin Hubble discovered in 1929 that the universe is expanding. He found that the galaxies are moving away from one another at speeds proportional to their separations, with galaxies that are farthest relative velocities. If the present velocities of the galaxies are extrapolated back in time, a time eventually is reached when all the galaxies are crushed together in a dense, hot fireball, ready to explode outward. This, by definition, is the beginning of the universe, and this relatively simple picture provides another way to find age.
If the galaxies are hurtling outward at the same velocities now that they have had since the big bang, then the distances they have traveled are simply their velocities multiplied by the time they have been traveling. If their distances from one another are known, then their velocities can be used to determine the time they have traveled to reach their present separations. That time is the time since the big bang.
But the galaxies have not traveled at uniform velocities since the big bang. The mutual gravitational attraction of the galaxies has slowed them down. The age determined in this manner is only an "upper limit" to the true age; the true age is less than this, because the galaxies' average velocities since the big bang are greater than their present velocities.
Measuring the velocities of distant galaxies is straightforward and highly accurate. The Doppler shift (frequency shift) of the light from a distant galaxy gives an unambiguous value for the galaxy's rate of recession. The real uncertainty -- and controversy -- arises from our poor knowledge of the galaxies' distances. Great distances in astronomy can be measured only "bootstrapping" from the fairly well known distances to nearby objects. So many layers of inference are involved that errors made early on propagate all through the analysis.
Hubble's first value for this upper-limit age of the universe was only 2 billion years. That was known to be inaccurate even at the time, because rocks found on the earth were older than 2 billion years. Various improvements and refinements in the distance scale over the past 50 years have produced upper-limit ages from this technique that range from 8 to 25 billion years, with 20 billion years most acceptable.
It is remarkable that three such diverse methods give approximately the same ages. This implies that we are on the right track; we really do know the approximate age of the universe. With that confidence, we can ask how combining the three methods and requiring agreement between each of their assumptions further narrows the age uncertainty.
By adding some rather subtle arguments concerning the universe's density and its abundance of helium and deuterium (heavy hydrogen), astronomers can match the upper-limit age to the other ages through the gravitationally induced slowing of the universe's expansion. Putting these all together allows only ages of 13.5 to 15.5 billion years to explain all the changes we've talked about.
What about our controversy? The recent upper-limit age of 10 billion years comes from placing the galaxies closer together than they are normally thought to be. Moving apart a smaller total distance for a given velocity implies moving for a shorter time -- 10 billion years at the upper limit. But the preponderance of observations of all types still argues against an age of only 10 billion years. There seems little reason to assume we have just seen a change in our understanding of the universe, especially when other astronomers making upper-limit measurements do get distances and ages that agree with the other age determinations.
However, if the galaxies really are closer than previously believed, there is one way to reconcile their "upper limit" age of 10 billion years with a true age of 15 billion years. If some mysterious force has accelerated the galaxies instead of allowing gravity to decelerate them, then the upper limit becomes instead a lower limit to the universe's age.
This mysterious force has been suggested before, and by no less an authority than Albert Einstein. When Einstein introduced his general theory of relativity in 1917, he found a result that seemed nonsensical to him and nearly everyone else. His theory predicted that the universe could not be static, as had been believed for hundreds of years, but must be dynamic, either expanding in spite of gravity's attraction or contracting because of it.
Einstein's solution to this was the introduction of what he called the cosmological constant, a "mysterious force" that precisely balanced the attraction of gravity and allowed the galaxies to remain fixed in space. When Hubble discovered the universe's expansion, Einstein called the constant the greatest scientific mistake of his life.
The cosmological constant, if we are willing to reinvoke it, could reconcile the generally accepted age of 15 billion years with the smaller separations claimed by one group of researchers.But that carries a price few astronomers are yet willing to pay, i.e., postulating something that has no other reason for existing than to tidy up the conclusions of one very uncertain observation. Such invocations seem ad hoc, and for the time being, most of us will still bet that our universe is about 15 billion years old.