An infinitely good read

Via CP at the Knight Science Journalism Tracker (a highly recommended site for general science news) comes the suggestion for this excellent article by Science News editor Tom Siegfried: Success in coping with infinity could strengthen case for multiple universes.

Despite the strangeness and fanciful-seemingness of some of the ideas suggested in the title and the article itself, it's a pretty good summary of some current thinking about life, the universe, and everything. It's even understandable on one level – as long as you don't insist on knowing the mathematical details of things like string theory and cosmic inflation.

In outline, some reputable physicists, including some of the originators of inflationary cosmology, are arguing that they may have mathematical "proof" that there must exist multiple universes. The argument is based on the idea that without an infinite number of existing universes, similar in some respects to ours, yet possibly different in radical ways, the probability is nil that all the characteristics of our universe could be so precisely tuned as to allow the existence of sentient life.

While some may question whether sentient life does in fact exist in our universe, just take that as an assumption for the present. The idea just described is sometimes known as the Anthropic Principle. There are several forms of this principle, and all are rather controversial, in some degree or another, in the minds of people (such as physicists and philosophers) who think about such things.

Tom Siegfried is reporting (among other things) that physicists like Alan Guth and Andre Linde think they may have found a way to prove mathematically that an infinity of multiple universes must exist in order to explain the highly unlikely existence of sentient life in the one universe we know about.

Using the word "unlikely" here means that what's involved in a rigorous argument has to use the mathematical theory of probability. And to employ that theory, it is necessary to define what's called a "measure", which is a way of assigning a specific number to the relative size of a subset of a larger set. What has to be done, to support a mathematical argument for some validity in an application of the Anthropic Principle for deducing the necessity of multiple universes, is to find a suitable measure that makes it exceedingly unlikely that the universe we are aware of, with its particular forms of life as we know them, could exist if there were only one universe (or a finite number of them).

There are certain unobvious problems that have to be dealt with, for example the problem of "Boltzmann brains". That refers to something exactly like a human brain that could arise purely by chance in a universe that's infinitely large and infinitely old.

There are physicists who object violently to the idea of the Anthropic Principle, in any form, as an explanation for why the universe we can perceive seems to have the properties which allow the existence of sentient life. An alternative that many of these physicists prefer is the existence of mathematical principles that uniquely determine the properties of our universe – rather than have it all be a matter of chance, which leads to an infinite number of universes, each with very different characteristics and physical laws. Einstein, too, was a believer in the existence of deterministic principles, long before inflation was even thought of, but also before it was recognized just how finely-tuned a universe has to be to support life.

Unfortunately, a huge problem remains with this deterministic view, that our universe, if it's one of at most a finite number of other possible universes, can have the life-friendly properties we (think we) observe. Namely, why should mathematical principles dictate exactly this kind – and only this kind – of universe? Yet if such principles could allow more than a finite number of other universes, we'd be back in the multiple universe scenario, whether or not it's the scenario string theory seems to call for.

Further reading

Boltzmann brains and the scale-factor cutoff measure of the multiverse – August 2008 arXiv paper by Alan H. Guth, Andrei Linde, Alexander Vilenkin, and others

Life, the Universe, and Everything: A Conference Looks to Ultimate Origins – Sky and Telescope report on the conference