“LIFE IS simply this: an entity that can make a copy of itself from parts all
of which are far simpler than itself,” says Carl Woese, an evolutionary
biologist at the University of Illinois at Urbana-Champaign. Even a robot could
demand the right to be called alive if it was designed to construct and maintain
a copy of itself with no outside help, he claims.
But even as Woese makes that bold assertion, it feels wrong. Something within
him rebels at the thought of calling such robots alive, because human engineers
would have to build the first one. So abruptly, in mid-interview, Woese changes
his mind. “Evolutionary history is going to have to be part of any really
workable definition of life,” he muses. In other words, to be alive an entity
must not only create itself from simpler parts, but must also participate in an
evolutionary epic like the one that created the life forms that prowl today’s
Earth out of the bare bones of an inanimate world.
This is an odd task I’ve set myself: Ringing up biologists and asking them to
tell me what life is, exactly. (Imagine a sportswriter ringing up coaches and
asking them to define the game of football.) But this shockingly basic question
is moving to centre stage as researchers draw near to the moment of truth in
their quest for new forms of life. Spacecraft have already travelled to Mars to
search for signs of life, and in 2003 NASA plans a trip to Jupiter’s moon
Europa—perhaps the most likely home for life elsewhere in our Solar
System. Earlier this year, NASA announced plans for a new astrobiology research
centre at the Ames Research Center near San Francisco. Two years ago, of course,
the infamous Mars meteorite ignited a debate that’s still raging about whether
its cracks and fissures held traces of life. And closer to home, biochemists are
getting tantalisingly close to creating life in a test tube, while some
researchers claim they have already generated artificial life within the
electronic guts of a computer.
Advertisement
You might think that a precise definition of life would be a prerequisite in
all these searches. But as I try to find one, I dredge up one of biology’s dirty
little secrets: the best minds in the world may have no problem separating the
quick from the dead in ordinary experience, but they still can’t agree on what
life is. Living things eat, move, and excrete? So does your gas-guzzling,
exhaust-belching car. Life maintains order in the face of entropy? A flame can
do that. Life is the ability to replicate? Then crystals are alive but not so
mules, old women and many old men.
You begin to see my problem?
One of the difficulties is that we know of only one kind of life: the form
that arose here on Earth. That makes it hard—some say impossible—to
separate the essential features of life in general from the idiosyncrasies of
our single sample, while at the same time excluding objects or systems that gut
feeling tells us aren’t really alive. Still, that doesn’t stop people from
trying.
Many scientists start as Woese did, and claim that the linchpin of life is
metabolism, the biochemical processes that harness energy to maintain order and
stave off chemical equilibrium. To see how important this is, visualise what
happens when someone dies and stops metabolising: they quickly crumble to dust.
Ergo, metabolism equals life.
Of course, a candle flame—which is clearly nonliving—uses energy
to maintain order, too, so some biologists add another requirement to their
definition: “identity”, or some sort of boundary (in practice a membrane) around
the metabolising entity. “Life is a self-bounded system where the boundary is
made by the material in the system. It’s not a thing, it’s a process, and these
processes involve the production and maintenance of identity,” says Lynn
Margulis, a biologist at the University of Massachusetts at Amherst, who is
coauthor of a book titled What is Life? (Simon & Schuster, New
York, 1995).
But there’s that sample-size-of-one problem again. Must life really have
identity and boundaries, or could there be other possibilities? There could,
according to researchers who study the origin of life on Earth. Jeffrey Bada, a
biogeochemist who directs NASA’s exobiology research at the Scripps Institution
of Oceanography in La Jolla, California, pictures life beginning as a loose scum
of replicating molecules with not a boundary in sight. The first membranes, he
says, probably turned up by chance, and gave the new organisms a huge
evolutionary advantage. Because they could keep valuable metabolites to
themselves, they took over the world. Indeed, Bada thinks a simple boundary-less
soup may be exactly the sort of life that NASA finds on Europa. “There’s
probably a bountiful supply of nutrients supplied to the surface of Europa from
cosmic dust. There may be very little evolutionary pressure to change much,” he
says.
If metabolism equals life, at least that gives researchers something definite
to look for. “When you think about how you recognise life, you think about what
life does, not what it is,” says Michael Meyer, who heads NASA’s astrobiology
programme in Washington DC. Living organisms should generate hallmark
molecules—the free oxygen in Earth’s atmosphere, for instance, would not
be there were it not for photosynthesis. If NASA’s as-yet unnamed Europa orbiter
mission finds free oxygen, one interpretation is that Jupiter’s moon harbours
life, although free oxygen could also be created by some unexpected non-living
chemistry.
Depending on what the orbiter finds, NASA may send a follow-up probe to land
on the moon’s surface. The probe could look for chiral molecules—ones that
exist in two mirror-image forms, like left and right hands. Purely physical
processes churn out roughly equal proportions of the two forms, but the
catalysts that assist biochemical reactions strongly favour one form. So an
imbalance between the two forms in chiral molecules such as amino acids would be
a strong hint that living organisms are lurking on Europa.
Biogeochemists could also search for mineral traces of life on Europa.
“There’s no way to form sulphide deposits at normal temperatures except
biologically,” says Kenneth Nealson, an astrobiologist at the Jet Propulsion
Laboratory in Pasadena, California. “So when you see sulphide, either it’s a
hydrothermal event or it’s biological—and if you look at the isotopic
composition, you can tell which is which.”
Some researchers see reason to hope that extraterrestrial life forms will be
carbon-based, just as we are. “No other element comes close to forming such a
diverse array of bonds,” says Bada. That array, he argues, is essential for the
chemical complexity of living systems. A few scientists go even further, and
predict that all life forms, wherever they are in the Universe, will use at
least some of the same molecules, such as sugars, as we do, and therefore have
some similar biosynthetic pathways. “There are only four different kinds of
one-carbon compounds,” points out Harold Morowitz, a biologist at George Mason
University near Washington DC. This means that if you are going to build them
into larger molecules there aren’t many pathways to choose from. Morowitz even
thinks that the pathways that break down molecules will turn out to be familiar.
“When we get to some other planet and find life, I have no idea what that life
will look like, but it will have the citric acid cycle,” he says, referring to
the metabolic pathway cells use to burn sugar and yield energy.
Others are less enthusiastic about using life on Earth as a universal
blueprint. If you make too many assumptions about what you’re looking for, warns
Christopher Chyba, a planetary scientist at the University of Arizona in Tucson,
“you’re in danger of running up against some unanticipated chemistry that mimics
life according to that definition.” Planetary scientists got a brief but intense
thrill in 1976, for example, when an experiment aboard the Viking Mars lander
found that a “bait” of radioactively labelled organic molecules gradually
disappeared, with the radioactivity turning up in carbon dioxide—just as
if a living creature had eaten the bait and metabolised it for energy. “It was
saying there was life in the Martian soil,” says Chyba. “My God, how they must
have felt.” But the thrill was short-lived. The Martian soil turned out to be
completely devoid of organic molecules, and they concluded that the “metabolism”
was caused by some inorganic process instead. The moral, Chyba told me, is to
look for as many different facets of life as you can.
Which threw me right back into the philosophical murk of my original
question: what is life? I hit the phones again to follow up Woese’s second
hunch, that it all has something to do with evolution.
Streets ahead
“You don’t have a living organism until you have Darwinian selection,”
confirmed Stanley Miller of the University of California at San Diego, whose
famous 1950s experiments showed that a “prebiotic soup” of inorganic compounds
could give rise to amino acids and other building blocks of life. Plenty of
other scientists echo Miller’s sentiment (although not always with his
confidence), making evolution the clear front-runner in the definition-of-life
Derby. The definition most often quoted is that of Gerald Joyce, who is
researching the origins of life at the Scripps Research Institute in San Diego:
“Life is a self-sustained chemical system capable of undergoing Darwinian
𱹴DZܳپDz.”
That life has to be self-sustaining—that is, not dependent on outside
help—highlights the big challenge for researchers trying to understand the
origins of life on Earth by creating life in a test tube. In recent years, there
has been huge progress toward test-tube life
(see “Let there be life”, New Scientist, 6 July 1996, p 22),
but their self-replicating systems still
depend on complex enzymes supplied by the researchers. Scientists will finally
have created life if and when they can get the whole thing to run on its
own.
Still, even something that appears as indisputable as self-sustainability
creates problems if over-zealously applied, Chyba notes. After all, humans,
unlike most plants and microbes, can’t synthesise all their complex molecules
like vitamins or certain amino acids, he says. “There are some we have to get
from food. So we would not be able to self-replicate and evolve for long without
making use of material that other organisms provide.”
But surely evolution—like nutrition and reproduction—is merely
something that life does, not the definition of life itself? Not so, says Mark
Bedau, a philosopher of biology at Reed College in Portland, Oregon. “Evolution
is not just a property of life. It’s the thing which explains why all the other
properties are there. The essence, the root cause.”
So does that mean that if organisms ever evolved a perfect error-free system
of copying their genes (no mutations, hence no evolution) they’d be dead, even
though they continue to metabolise and reproduce? When asked directly, Bedau,
like any self-respecting philosopher, splits a hair. What’s really alive, he
claims, is the entire system—countless individual organisms of many
species, all interacting, reproducing, and displaying unpredictable, open-ended
evolution. Individual organisms—whose fate is a lot more predictable than
that of their whole evolutionary lineage—are only alive in a secondary
sense, as elements in the grand pageant.
At first, this sounded a bit like mumbo-jumbo, but it made sense when I
thought about it. If you take the long-term view, our hypothetical error-free
genetic system is an evolutionary dead end. And so are mules, nuns and celibate
priests—all those annoying little enigmas that nibble away at the margins
of a definition of life. Bedau’s distinction sweeps away the problem. All these
entities are alive—lower case—because they are the products of
open-ended evolution. But they have ceased to be Alive—upper
case—because they can no longer perpetuate the process.
Still, Bedau’s position has uncomfortable ramifications. If open-ended
evolution is the be-all and end-all of life, you have to admit to the high table
other evolving systems, such as computer programs and national economies. Bedau
can contemplate doing that…just. “This is starting to get sort of
mindboggling,” he says.
One mindboggling possibility is creating artificial life inside a computer.
The best-known effort is called Tierra, the brainchild of Thomas Ray, an
ecologist now at ATR Human Information Processing Research Laboratories in
Kyoto, Japan. Tierra’s “organisms” are packets of computer code designed to make
copies of themselves—with occasional random changes that behave just like
genetic mutations—within the memory banks of computers. Once the packets
are set loose in a computer, they compete for memory space, reproduce, and
ultimately die. Over time, packets with beneficial mutations leave more copies
of themselves, so the population evolves, creating an unpredictable profusion of
different “organisms”—including some that parasitise the code of other
organisms nearby. Ray claims that Tierra goes beyond merely mimicking life. “I’d
say it’s alive, because it’s a pretty rich evolution,” he says.
If Tierra’s creatures consisted of proteins and nucleic acids instead of 1s
and 0s on a silicon chip, no one would argue. But the idea of life that depends
on a computer makes many people distinctly uneasy. “It is as far from life as
dolls are from babies,” says Margulis. “A computer is made from the outside.
There’s nothing about a computer that is internally and intrinsically
ԳٲԱ.”
Such criticism misses the point, says Ray, because the computer plays the
same role for Tierra as the physical Universe does for our flesh-and-blood life.
“I don’t think of the Universe as being alive, and I don’t think of the computer
as being alive, but within each of them is a process which is alive,” he says.
In other words, life is a property of software, and details of the hardware are
irrelevant—an echo of Bedau’s emphasis on evolution, not the organisms, as
the central feature of life.
Just as I’m starting to get used to the idea that computer programs could be
alive, Stuart Kauffman, a complexity theorist at the Santa Fe Institute in New
Mexico, weighs in against the idea that life is totally information-based.
Instead, he says, living beings—autonomous agents in his
terminology—must have one foot in the material world of chemistry and one
foot in the world of information. An organism’s corporeal form and its genome
are different facets of the same being, and either is meaningless without the
other. “An autonomous agent is a new concept. It’s not matter, it’s not energy,
it’s not information, it’s something else,” says Kauffman.
And so, two weeks and a large phone bill later, my search for an
all-encompassing definition of life ends in frustration. But at least I’m in
good company. “I doubt it would be possible to find a single definition of life
that all scientists can accept,” says Ernan McMullin, a historian and
philosopher of science at the University of Notre Dame in South Bend, Indiana.
“It becomes a matter of choice as to which definition is most useful at
organising the field we’re working in.”
Kauffman, ever the visionary, is more upbeat. He is convinced that it will be
possible to find the one single definition of life, or autonomous agency. What’s
more, he says, that definition will mark the doorway to a whole new science.
Someday, Kauffman predicts, when he and his colleagues have pinned life down,
they will be able to create new life tailored to do our bidding.
“We’re going to make autonomous agents that can do honest work, and they’ll
build things for us,” says Kauffman. “My bet is that in 30 or 40 years, this
will become a substantial sector of the global economy.”
- Further reading:
The nature of life by M. Bedau in
The Philosophy of Artificial Life,
ed. Margaret A. Boden (Oxford University Press, 1996) -
The origin of life in the solar system: current issues
by C. Chyba and G. McDonald,
Annual Review of Earth and Planetary Science, vol 23, p 215 (1995)