The best of all possible worlds: the anthropic approach to cosmology asks, what makes the universe compatible with intelligent life? Was it just the luck of the draw?

Cogito ergo sum," wrote Rene Descartes in 1641: "I think, therefore I am." Anyone who has a thought, in other words, must actually exist--quite a relief for those who had their doubts. Three and a half centuries later scientists and philosophers are investigating an intriguing, related question that Descartes passed over: Sum ergo ...? What conclusions, if any, can be drawn from the fact that we do exist?

Attempts to answer that question have given increased support to a line of thought that scientists and philosophers have termed the "anthropic principle," the idea that the simple fact of our existence offers fundamental insights and information about the entire universe. First named in 1973 by the cosmologist Brandon Carter, now at the Paris Observatory in Meudon, France, the anthropic principle can be better described as the "anthropic approach" to the cosmos, examining the universe for what can be deduced from the fact that we are here. (To emphasize that the "we" in question could be any form of intelligent life, some prefer the term "biophilic [life-loving] principle" to the terms "anthropic principle" or "anthropic approach.")

The anthropic principle rests on the assumption--maintained by almost all physicists and cosmologists--that a single set of physical laws holds true throughout the universe. All of modern cosmology, including the theory of the expanding universe, rests on that assumption. The principle itself has at least two forms. One, called the weak anthropic principle, notes that for us to exist and thus to observe the universe, the values of the parameters that describe the cosmos must be consistent with the fact that we have evolved within the universe.

The second version, called the strong anthropic principle, requires that the basic parameters describing the cosmos--the strengths of fundamental forces, for instance--cannot have random values, but instead must have values that enable life to develop at some stage in cosmic history. In other words, no universe could exist that did not allow the possibility of life.

What can either or both of these versions add to science? Some scientists find them no more than a tautology: We're here because we're here. Certainly from a biological viewpoint, the fact that so many possible sites for life exist in the cosmos, and the fact that life as we know it thrives in a range of environments, suggest at most a weak connection between cosmic conditions and the evolution of life on Earth. The story of life on Earth is the story of how life has evolved to be fit for our planet. The biologist Paul P. Ehrlich of Stanford University makes the same point for the universe in general: "To say that 'the universe is fit for life," he writes, "has things exactly backward: We are fit for the universe, not the other way around. Had things been different, we would have evolved differently."

Yet some scientists--though certainly not all--are convinced that the anthropic principle has intrinsic scientific worth. Certain facts, or clusters of facts, about the universe as a whole once again, assuming that the same laws of physics hold sway throughout its extent--suggest that the physical parameters of our universe do not have random values.

Until recently, the "cosmic coincidences" most frequently cited as being favorable to life included the following three items:

* The development of structure in the universe during its first billion years gave rise to galaxy clusters and to individual galaxies made up of hundreds of billions of stars. That process depended critically on minute deviations from a totally smooth distribution of matter at the time clumps of matter could first form, about 380,000 years after the big bang. If those deviations had been a trifle smaller, galaxies would never have formed as the universe expanded; if they had been a bit larger, almost all the matter in the universe would have ended up in super-massive black holes.

* The strong nuclear force holds protons and neutrons together. If that force were more than 2 percent stronger than its actual value, protons would bind themselves in pairs to form "di-proton nuclei"--a state of matter that does not occur in the universe we inhabit. In that case, nuclear fusion during the first few minutes after the big bang would have left nearly all the protons in the universe bound up in di-proton nuclei. Those nuclei would have quickly become deuterium nuclei (each made up of one proton and one neutron) as one of the two fused protons turned into a neutron. The deuterium nuclei would then have rapidly paired off and fused to form helium nuclei, each made up of two protons and two neutrons. As a result, most of the ordinary matter in the universe would have become helium, not hydrogen. In our universe, about a quarter of the ordinary matter did end up in helium, but almost all the rest remained in the form of protons (hydrogen nuclei). A helium-dominated universe could never have given birth to stars that generate energy by fusing hydrogen into helium. Nor could it have produced water, the solvent that many scientists think is essential to life.