The difficulty of saying what 'life' is brings to mind Justice Potter Stewart's famous description of pornography in 1964: it is hard to define, but we know it when we see it.

Yet do we really? Astrobiologists are haunted by the suspicion of terracentricity: we imagine that life elsewhere will look like life here, and bias our searches accordingly.

Some of this community struggle to shed such prejudice, questioning for example the complacent conviction that life depends on water. Others seek very general signatures that make no assumptions about biochemical specifics.

Now a new kind of fingerprint has been proposed by William Sparks of the Space Telescope Science Institute in Baltimore and his co-workers. They suggest we search for a characteristic signature of life by looking at light scattered from the surfaces of extrasolar planets1.

Vital signs

Proposals to identify a unique signal for life have a long history. In 1965, James Lovelock argued that Mars landers could spot life by detecting a sustained chemical disequilibrium in a planetary environment2. This idea had the virtue that it required surveying only a planet's atmosphere. On Earth, the high proportion of atmospheric oxygen, along with the presence of other trace gases, should be a giveaway, relying as it does on photosynthesis preventing oxygen from getting locked into minerals.

Life, but not as we know it? Credit: NASA / JPL

One of the most ingenious ideas is that life affects a planet's topography, for example by mediating chemical reactions that erode rock, by forming soil and protecting it from erosion, and by dictating climate. Geomorphologists William Dietrich and J. Taylor Perron have argued3 that the types and distributions of landforms on Earth probably betray life's influence, and that a better understanding of how they are formed might lead to a clear distinction between the contours of planets with and without life.

In their paper, published in the Proceedings of the National Academy of Sciences, Sparks and his colleagues argue that the presence of living organisms will be likely to make a planet's light circularly polarized, meaning that the plane of its oscillating electromagnetic fields is not random but has a characteristic twist to it, either to the left or the right.

In a twist

Circular polarization is a feature of light scattered from Earth, owing to the 'handedness' or chirality of the building blocks of biological molecules. All natural proteins, for example, are made from amino acids that have a 'left-handed' molecular shape — cells can't use the mirror-image right-handed amino acids to build proteins.

This molecular-scale twist means that the photosynthetic molecular apparatus of bacteria and plants absorbs light circularly polarized to the left or the right to different degrees, creating a net circular polarization in the scattered light.

It's not obvious that this will show up when the light is seen from afar, however, because the scattering process is complicated. That's why the researchers checked that the signature remains evident in light bouncing off cultures of marine photosynthetic bacteria. It does, as indeed they found also for reflected light from a maple leaf.

Whether this method will work for light from exoplanets is another matter. It depends on what proportion of surface scattering comes from living organisms, and on whether this light can be distinguished from that of the parent star.

But who says life has to be uniquely left- or right-handed? 'Homochirality', say Sparks and colleagues, 'is thought to be generic to all forms of biochemical life as a necessity for self-replication'.

This statement draws on the work of astrobiologist Radu Popa of Portland State University in Oregon4 – but he offers only a plausibility argument based on the idea that homochirality simplifies polymer structure in a way that makes information copying more efficient. This doesn't imply that homochirality is essential, but only that it might help. And we know that life does not always do things in the most efficient way.


However, the new proposal actually goes back a long way. An intimate association between life, chirality and polarized light was made in the nineteenth century, first by the French scientist Jean-Baptiste Biot and then by Louis Pasteur, who sought Biot's advice on his seminal discovery of handedness in organic molecules.

Biot, a pioneer in the study of optics and polarization, coined the term 'optical activity' to describe a substance that rotates the plane of polarized light. It was no coincidence that this in itself suggested the operation of some vital, 'active' agent, rather than lifeless passive matter. Biot came to believe that optical activity was "the sole means in man's possession of confronting the otherwise undefinable limit between life and nonlife on the molecular level"5.

Pasteur became a staunch advocate of this view, to the extent that (contrary to the popular belief) he developed something of an anti-materialist, vitalist stance on what life is: he felt that optical rotation must result from "the play of vital forces".

We now know, partly through Pasteur's own work, that he and Biot were wrong. Sparks and colleagues are on sounder ground, but their idea could be seen to support the suspicion that life is everywhere built in our own image.