Our lack of data is reflected in our attempts to define life. Koshland (2002) lists seven features of life: (1) program (DNA), (2) improvisation (response to environment), (3) compartmentalization, (4) energy, (5) regeneration, (6) adaptability, and (7) seclusion (chemical control and selectivity). A simpler definition is that life is a material system that undergoes reproduction, mutation, and natural selection (McKay 1991). Cleland and Chyba (2002) have suggested that life might be like water, hard to define phenomenologically, but easy to define at the fundamental level. But life is like fire, not water—it is a process, not a pure substance. Such definitions are grist for philosophical discussion, but they neither inform biological research nor provide a basis for the search for life on other worlds.

The simplest, but not the only, proof of life is to find something that is alive. There are only two properties that can determine if an object is alive: metabolism and motion. (Metabolism is used here to include an organism's life functions, biomass increase, and reproduction.) All living things require some level of metabolism to remain viable against entropy. Movement (either microscopic or macroscopic) in response to stimuli or in the presence of food can be a convincing indicator of a living thing. But both metabolism (fire) and motion (wind) occur in nature in the absence of biology.

The practical approach to the search for life is to determine what life needs. The simplest list is probably: energy, carbon, liquid water, and a few other elements such as nitrogen, sulfur, and phosphorus (McKay 1991). Life requires energy to maintain itself against entropy, as does any self-organizing open system. In the memorable words of Erwin Schrödinger (1945), “It feeds on negative entropy.” On Earth, the vast majority of life forms ultimately derive their energy from sunlight. The only other source of primary productivity known is chemical energy, and there are only two ecosystems known, both methanogen-based (Stevens and McKinley 1995; Chapelle et al. 2002), that rely exclusively on chemical energy (that is, they do not use sunlight or its product, oxygen). Photosynthetic organisms can use sunlight at levels below the level of sunlight at the orbit of Pluto (Ravens et al. 2000); therefore, energy is not the limitation for life. Carbon, nitrogen, sulfur, and phosphorus are the elements of life, and they are abundant in the Solar System. Indeed, the Sun and the outer Solar System have more than 10,000 times the carbon content of the bulk of Earth (McKay 1991). When we scan the other worlds of our Solar System, the missing ecological ingredient for life is liquid water. It makes sense, then, that the search for liquid water is currently the first step in the search for life on other worlds. The presence of liquid water is a powerful indication that the ecological prerequisites for life are satisfied.

Orbital images, such as the canyon in Figure 1, show clear evidence of the stable and repeated, if not persistent, flow of a low-viscosity fluid on Mars at certain times in its past history. The fluid was probably water, but the images could also suggest wind, ice, lava, even carbon dioxide or sulfur dioxide. Recently, results from the Mars Exploration Rover missions have shown that this liquid carried salts and precipitated hematite in concretions. The case for water, we could say, is tight.

Life on Mars is also likely to be the same at the top of the table: at the ecological level. Primary production in a Martian ecosystem is likely to be phototrophic, using carbon dioxide and water. Heterotrophs are likely to be present to consume the phototrophs and in turn to be consumed by predators. Darwinian evolution would result in many of the same patterns we see in ecosystems on Earth. While it may be similar at the top (ecological) and bottom (chemical) levels, life on Mars could be quite alien in the middle, in the realm of biochemistry. Pace (2001) has argued that alien biochemistry will turn out to be the same as biochemistry on Earth, because there is one best way to do things and that natural selection will ensure that life everywhere discovers that way. Only observation will tell if there is one possible biochemistry, or many.

The question is open and possibly urgent. On space missions already being planned, we may have the opportunity to analyze the remains of alien organics on the surface of Europa or frozen below ground on Mars. The instruments that will make this analysis must be designed in the next couple of years.

One approach appears promising. I call it the “Lego Principle.” It is based on the patterns of the molecules of life. Biological processes, in contrast to abiotic mechanisms, do not make use of the range of possible organic molecules. Instead, biology is built from a selected set. Thus, organic molecules that are chemically very similar to each other may have widely different concentrations in a sample of biological organics. An example of this on Earth is the 20 amino acids used in proteins and the use of the left-handed version of these amino acids. The selectivity of biological processes is shown schematically in Figure 2 by the distribution of spikes in contrast to a smooth, nonbiological distribution. General arguments of thermodynamic efficiency and specificity of enzymatic reactions suggest that this selectivity is required for biological function and is a general result of natural selection. Different life forms are likely to have different patterns, and at the very least we might find the mirror symmetry of life on Earth, with d- instead of l-amino acids.

This approach has immediate practical benefit in the search for biochemistry in the Solar System. Samples of organic material collected from Mars and Europa can be easily tested for the prevalence of one chirality of amino acid over the other. More generally, a complete analysis of the relative concentration of different types of organic molecules might reveal a pattern that is biological even if that pattern does not involve any of the familiar biomolecules. Interestingly, if a sample of organics from Mars or Europa shows a preponderance of D-amino acids, this would be evidence of life, and at the same time would show that this life was distinct from Earth life. This same conclusion would apply to any clearly biological pattern that was distinct from that of Earth life.

Organic material of biological origin will eventually lose its distinctive pattern when exposed to heat and other types of radiation, (examples of this include the thermal racemization of amino acids), but at the low temperatures in the Martian permafrost, calculations suggest that there has been no thermal alteration (Kanavarioti and Mancinelli 1990). An interesting question, as yet unanswered, is how long organic material frozen into the surface ice of Europa would retain a biological signature in the strong radiation environment.