based on the isolation of compounds from nature. If terran life had provided silicon-based molecules, then our knowledge of silicon-based chemistry would now be advanced.

The natural tendency toward terracentricity requires that we make an effort to broaden our ideas of where life is possible and what forms it might take. Furthermore, basic principles of chemistry warn us against terracentricity. It is easy to conceive of chemical reactions that might support life involving noncarbon compounds, occurring in solvents other than water, or involving oxidation-reduction reactions without dioxygen. Furthermore, there are reactions that are not redox. For example, life could get energy from NaOH + HCl; the reaction goes fast abiotically, but an organism could send tendrils into the acid and the base and live off the gradient. An organism could get energy from supersaturated solution. It could get relative humidity from evaporating water. It is easy to conceive of alien life in environments quite different from the surface of a rocky planet. The public has become aware of those ideas through science fiction and nonfiction, such as Peter Ward’s Life as We Do Not Know It.2

The public and the scientific community have become interested in authoritative perspectives on the possibility of life in environments in the solar system very much different from the ones that support life on Earth and life supported by “weird” chemistry in exotic solvents and exploiting exotic metabolisms. To NASA those ideas would help to guide missions throughout the solar system and permit them to recognize alien life if it is encountered, however it is structured. Given the inevitability of human missions to Mars and other locales potentially inhabited by alien life, an understanding of the scope of life will improve researchers’ chance to study such life before a human presence contaminates it or, through ignorance or inaction, destroys it.

In broadest outline, this report shows that the committee found no compelling reason for life being limited to water as a solvent, even if it is constrained to use carbon as the scaffolding element for most of its biomolecules. In water, varied molecular structures are conceivable that could (in principle) support life, but it would be sufficiently different from life on Earth that it would be overlooked by unsophisticated life-detection tools. Evidence suggests that Darwinian processes require water, or a solvent like water, if they are supported by organic biopolymers (such as DNA). Furthermore, although macromolecules using silicon are known, there are few suggestions as to how they might have emerged spontaneously to support a biosphere.


For generations the definition of life has eluded scientists and philosophers. (Many have come to recognize that the concept of “definition” itself is difficult to define.3) We can, however, list characteristics of the one example of life that we know—life on Earth:

  • It is chemical in essence; terran living systems contain molecular species that undergo chemical transformations (metabolism) under the direction of molecules (enzyme catalysts) whose structures are inherited, and heritable information is itself carried by molecules.

  • To have directed chemical transformations, terran living systems exploit a thermodynamic disequilibrium.

  • The biomolecules that terran life uses to support metabolism, build structures, manage energy, and transfer information take advantage of the covalent bonding properties of carbon, hydrogen, nitrogen, oxygen, phosphorus, and sulfur and the ability of heteroatoms, primarily oxygen and nitrogen, to modulate the reactivity of hydrocarbons.

  • Terran biomolecules interact with water to be soluble (or not) or to react (or not) in a way that confers fitness on a host organism. The biomolecules found in terran life appear to have molecular structures that create properties specifically suited to the demands imposed by water.

  • Living systems that have emerged on Earth have done so by a process of random variation in the structure of inherited biomolecules, on which was superimposed natural selection to achieve fitness. These are the central elements of the Darwinian paradigm.

Various published definitions of life understandably incorporate those features, given that we are the life form defining it. Indeed, because the chemical structures of terran biomolecular systems all appear to have arisen through Darwinian processes, it is hardly surprising that some of the more thoughtful definitions of life hold that it is a “chemical system capable of Darwinian evolution.”4

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement