Column: The Meaning of “Life”

The Meaning of “Life”

Daniel Sprockett (formerly of ‘Just 10% Human’) broadens his scope to look at scientific issues and stories more generally in ‘Shot of Science.’ This week, he discusses our failure to define “life,” and considers whether the failing lies with science of with language.




In three words I can sum up everything I’ve learned about life: it goes on.

- Robert Frost

As a biologist, I probably spend more time thinking about the nature of life than the average person. I’d venture to say this is true of all biologists, and a fair number of biochemists, too. Of course, I shouldn’t exclude biophysicists, bioinformaticians, and natural philosophers, either. With all that brainpower devoted to understanding the nature and structure of life, you would think that we would have at least come up with a coherent definition of what life really is.

Yet, after literally centuries of discussion and debate, a necessary and sufficient definition of “life” remains elusive. Pre-scientific ideas about life dealt with fuzzy, immaterial concepts like souls, vital forces, and essential forms. Aristotle conceived of life in terms of hylomorphism, or congruence of matter (Greek: hulê) and form (morphê). The form of living things was its soul, which came in three distinct flavors. The vegetative and animal soul belonged to “lower” life forms, while the rational soul was reserved solely for humans. Aristotle saw the ability to reason as what distinguished humans from other forms of life, in much the same way others have searched for a single set of characteristics that distinguish life from non-life.

The difficult part about drawing a circle that encompasses all of the things that we deem to be alive, while excluding all things that are not, is that there does not seem to be any single set of traits that universally fulfill this criteria. For example, metabolism, growth, and reproduction are normally deemed to be essential traits of life, but this definition would also mean that a bushfire is “alive” in a literal sense. Charles Darwin gave us another powerful new tool for defining life, in that all life must evolve by natural selection. However, this too is insufficient, since viruses and prions (misfolded proteins like the ones that cause bovine spongiform encephalopathy, A.K.A. mad cow disease) can also evolve by natural selection.

And so the list of traits necessary for something to be deemed “living” continues to grow. A book published in 2004, Between Necessity and Probability: Searching for the Definition and Origin of Life, collected over 300 different definitions of life. Many more have been published since then, some of which completely eschew the distinction between the biological and technological, and instead define life purely in cybernetic terms. Others focus on obscure commonalities of living systems, such as the “handedness” of molecules found in biochemistry, a concept known as chirality. In my opinion, most of these attempts fall victim to what physicist Sean Carroll calls the yellow-taxi problem:

… in a city where all cars are blue except for taxis, which are yellow, it’s tempting to define “taxi” as “a yellow car.” But that doesn’t get anywhere near the essence of taxi-ness.

In 2011, a biologist named Edward Trifonov, decided to take an alternate approach. Instead of generating his own list of traits, he looked for commonalities in the lists of other researchers. After examining 123 definitions of “life” collected from a variety of sources spanning centuries, he concluded that the most common concepts that were cited across the majority of these definitions of life was simply:

Life is self-reproduction with variations

Huh. Self-reproduction with variations. A grand total of three words – so much for centuries of debate!

Trifonov’s simplistic conclusion understandably ruffled a lot of feathers. Some critics brought up the fact that variation is already implicit in self-reproduction, since the laws of thermodynamics guarantee that variation will occur in the form of mutation. In this case, a better definition would be Self-reproduction with error-correction. Others have concluded that a single, coherent definition of life is both impossible and pointless.

Carl Zimmer, an award-winning science writer for the New York Times and National Geographic, wrote a fantastic summary of this paper for, but took his review one step further. He turned to his social network, primarily inhabited by scientists, journalists, and sci/tech-junkies, and posed this question to them: Can you define life in three words?

Heaps of people chimed in, but one of them really struck me as a legitimately novel perspective to this age-old question. It is no surprise, that it came from one of my scientific heroes, MacArthur “Genius Grant” recipient and phylogeneticist extraordinaire, David Hillis.

By the time I met David Hillis at a conference in Lincoln, Nebraska, in early 2009, I was already a big fan of his work. He literally wrote the book on molecular phylogenetics, the topic of my master’s thesis, and pioneered many of the techniques that we use to infer the evolutionary relationships between species through examining changes in their DNA. He has be famous in some circles for producing this image of the great tree of life, which he constructed using the ribosomal RNA genes from over different 3,000 species.

This tree shows the evolutionary relationships between thousands of organisms representing every branch of life. Humans, as you’ll notice, occupy a single tip of a single branch of this tree. Seeing the entirety of the human species clustered among the other primates, which themselves are nestled beside the other mammals, which in turn are aggregated with the rest of the animals, which is finally just a branch among many of the great big tree of life, gives me an unexpectedly satisfying feeling of unity with the rest of the planet.

Building trees like this has given Hillis a uniquely evolutionary perspective on life. During a fascinating back-and-forth on Carl Zimmer’s facebook wall, he raised the compelling idea that attempting to generate a complete list of necessary and sufficient traits is fundamentally the wrong way to conceptualize life.

Let me give you an example. Some things are like bicycles. You can easily define a set of traits that determine if something is a bicycle or not (two wheels, pedals, handlebars, etc.). Anything with those traits is a bike, no matter where it comes from or who made it. This is an intensional definition.

Other types of things are like my family. It wouldn’t make sense to define my family by their unifying characteristics (for example, our strong jaw line and propensity for interjecting obscure movie lines into serious situations). This is because my family is a historical entity. Instead of characteristics, you point to examples of the thing and trace them back to their beginning. In this case, this would be my parents’ wedding on Boxing Day in 1971. You could do something similar for the nation of Australia, not by generating a list of traits that define the country itself, but by pointing to the Australian Federation and following all that stemmed from it. This is a type of ostensive definition, or a definition by example.

So is life like a bicycle, or like a Sprockett?

Many of the conceptual issues with defining life simply fall away when you realize that ostensive definitions are much better at capturing what most of us mean when we speak about “life.” Take a look back at that massive tree of life. If you point to any spot where two or more taxa meet, you can define all of the species extending from that point as a taxon. For example, the branching point that leads to all bats and humans, dogs and horses, represents the most recent common ancestor of all mammals. That point is defined as the origin of the “mammal” taxon.

Now, look at the root of the tree in the very middle of image, which represents the most recent common ancestor of everything and the beginning of all life on this planet. If you define that point as the taxon “Life,” then everything following from it is therefore “Living,” with no need to list additional restrictions in order to exclude bushfires, misfolded proteins, or computer viruses.

However, this has some interesting implications. When Craig Venter finally succeeds in bioengineering an evolving, self-replicating, synthetic cell as a I discussed last week, then no matter the result, it won’t be “living” because it doesn’t fit anywhere on that tree of life. No matter its characteristics, we would have to define it anew as Life2 or something similar, since it would have an origin that is entirely different than our origin. The same would apply to organisms found elsewhere in the universe, assuming that one life taxon didn’t seed the other via panspermia. If this were the case, then tree of life would have to be re-rooted to reflect our shared evolutionary origin. Hillis then clarified the real issue at hand:

The question people actually want to ask is “Are there entities in the universe that are similar to the Life we know about here on Earth?” The answer, of course, depends on what people mean by the arbitrary meaning of “similar”. One person might answer “I mean ‘self-replicating with variations’.” Then, the answer is yes: humans have created imperfectly self-replicating systems (“artificial life”) here on Earth. But then someone else says “But that is not what I meant by similar…I meant that they had to have metabolism and cellular structure and a nucelic-acid-based genetic system.” OK, then we have to keep looking to find something that similar. But then someone else says “But that’s pretty arbitrary…I’d still consider it alive if it didn’t have cellular structure.” Exactly…it is indeed arbitrary to argue over how similar something has to be to consider it “similar” to Life. [Emphasis added].

I think this ostensive definition of life removes one of the central problems with building a single, coherent definition of life: it relies on arbitrary distinctions that don’t reveal anything about the fundamental nature of life. This sentiment has been echoed by James Van Etten, a distinguished virologist and member of the U.S. National Academy of Sciences. When a member of the public asked him if viruses were really alive or not, he simply responded that viruses, like all biological entities, are what they are, and that reality doesn’t change based on what label we assign it. These distinctions are only useful in as much as they reveal something true about the nature of our universe. In the end, it is our language that must evolve to reflect biological realities, and not the other way around.



Daniel Sprockett is a researcher at the Case Western Reserve University School of Medicine in Cleveland, Ohio. He currently resides in Double Bay with his wife, Andrea, while she completes a Master’s of International Public Health at the University of Sydney. Dan will return to the United States in September, when he begins his PhD in Microbiology and Immunology at Stanford University.



Read more of Daniel’s articles here.


(Top image from Chase Clark)