What is life? (and why the answer doesn’t matter)

by Nathaniel Virgo

What is life?  Some people will say it’s obvious: life is reproduction.  But I may never choose to reproduce, and a worker ant couldn’t if it wanted to – does that make us dead?

Others will say life is evolution.  But on closer inspection, that doesn’t really stand up either.  Evolution is easy enough to implement on a computer.  You just store a bunch of random bit strings in memory, evaluate them according to some “fitness function”, and then “mutate” and “recombine” the best ones to produce a new generation.  By iterating this process you get what’s called a “genetic algorithm”, and this can be used to design robot controllers and all sorts of other things.  These things evolve, but are they alive?  Some might say yes, but anyone with any experience in genetic algorithms will say no.

At school I was taught that life was a combination of seven properties, all of which had to be present in order for it to be life. (The creators of our syllabus loved the number seven.  There were seven different types of energy as well.)  I can’t remember all seven signs of life, but they were things like reproduction, growth, movement and respiration.  But this isn’t a very enlightening definition, and one can find exceptions to at least some of those things.

People with a more thermodynamic point of view might say that life has the property of resisting the law of increasing entropy: it stays in an organised state by “feeding on negative entropy” in its environment, expelling all of the positive entropy that gets generated inside its body by exhaling and defecating.  This is a wonderful picture that deserves to be understood – but again, if we look, we can find other things besides life that have the same property.  Hurricanes are my favorite example, feeding on warm (low entropy) air from the sea surface and expelling that heat as low temperature (high entropy) thermal radiation into space.

Still others will point to the cell membrane as the defining feature, along with the self-maintaining network of chemical reactions we call metabolism.  But if we create such membrane-bounded self-maintaining structures in the laboratory (as seems likely soon enough), will we automatically call them alive?  Some will say, no, they will only be alive if they have the ability to adapt to their environment as well as having these properties.  But again, it’s not so hard to find inanimate structures that adapt to their environment in some sense or other, and the specific combination of membrane-bound chemistry and adaptiveness doesn’t seem any more fundamental than the seven signs of life I was taught at school.

And all of this ignores many other things that all known organisms on Earth have in common.  Every single living cell has nucleic acids (in the form of RNA or DNA), they all use ATP to carry energy, they’re all made of proteins, made from the same set of amino acids.  They all (with minor variations) use the same ATP-powered molecular machinery to read those nucleic acids, according to the same genetic code.  And of course, they’re all (more or less) membrane-bound, they all adapt to their environment, they all feed on negative entropy, they’re all the product of evolution and some individuals of every species are able to reproduce.  All of this makes life very easy to differentiate from non-life.  All of these things are 100% correlated with each other in living things, so you just have to throw in two or three of these shared properties and you’ve already drawn a sharp distinction between all living things and all non-living things on Earth.  The problem with defining life is that it’s too easy.

But why is life so different from non-life in so many specific ways?  The answer, at least to the question of specific molecules, is probably quite simple – it’s because all life on the present-day Earth shares a common ancestor.  Life in its present form long ago out-competed whatever alternatives existed, and we’re left with only the winning version.  The big question, if one cares about the definition of life, is which of these shared properties are fundamental to the very concept of life itself, and which are merely due to all life on Earth sharing the same common ancestor.  Of course we have strong intuitions about some of them, but we can never answer them all for sure until we’ve studied many separate examples of trees of life – life that evolved on other worlds and didn’t share a common ancestor with life on Earth.  With only one example of life it’s very difficult to say whether (for instance) proteins are the only way it can be achieved, or just the way it happens to have been done on Earth.

Of course, we might well find phenomena on other worlds (or even on our own) that are like life but not like life, in such a way that we’re not sure whether to call them alive or not.  I expect this to eventually happen.  If we discovered such a thing today it would force us to confront the blurred boundary between the living and the non-living, which only appears sharp to us because of the particulars of the world we inhabit.

Does any of this matter?  One reason it might matter to you is if you study the origins of life.  In order to know how life began, you might feel you need to know what life is.  If you decide that life is evolution you’ll spend your time wondering how evolution could have occurred in a lifeless world; if you decide that life is nucleic acids you’ll wonder what chemistry could have produced them; if you decide life is membrane-bound cells you’ll look for them, and so on.

But life on the present-day Earth isn’t any of those things, it’s all of them, and all of them must have arisen at some point.  Perhaps they all arose together, as the first fully formed organism suddenly popped out of an inert substrate (this is extremely unlikely). Perhaps nucleic acids arose first, with proteins, ATP and cell membranes appearing later; perhaps membrane-bound cells came first and nucleic acids came later; perhaps there were proteins and ATP and nucleic acids before there were cells – or perhaps early life was made from some entirely different set of chemicals altogether, which only later evolved into the ones we know today.

It’s fair to say that none of these questions have been answered.  People have opinions, but the debates are not over.  How each of these things happened, and in what order, is an important question for science, and by focusing on some narrow definition of life we risk ignoring some important parts of the story.  Instead we should embrace the strange diversity of specific features shared by life as we know it, and admit to ourselves that in the universe at large the boundary between living and inanimate is probably not as sharp as we might like to think.

About these ads

9 Responses to “What is life? (and why the answer doesn’t matter)”

  1. I remember the seven signs of life thing too. I also remember only realising at quite a late stage just how meaningless it was, and wondering why they bother teaching it…

    Apparently some modern syllabuses have 8 signs of life! http://answers.yahoo.com/question/index?qid=20080918115850AACeH2t

  2. A definition of life allows one to talk about its purpose.

    People like things to have a purpose.

    If you think the 7 signs of life are arbitrary, try this quiz:
    I won first time!!!

    • That’s pretty hilarious. I got them all wrong twice in a row – but then, they’re patently a *different* seven signs of life from the ones I was taught in school.

      (I just noticed that at the top of this page my post has been juxtaposed with a link to James’ previous post, “just because a thing has a name, doesn’t mean it’s a thing.” Which I suppose is really my point too: perhaps in the wider universe, life’s not a thing.)

  3. I assumed “right” was the same as “in the order of the mnemonic I was taught at school”.

    Anyway, I prefer the three Fs – I think I could guess most peoples ordering of those with two attempts aswell ;)

    Full of life.
    Is there life on Mars.
    Take a life.
    Life is special.
    Three lives remain. (like ‘three sleeps to christmas’, alive, asleep)
    Cats have 9 lives.

    Alive and kicking.
    ITS…. ALIVE (muhahaha)

    It’s definitely used as a noun (some/a life) as well as a property (to be alive), but to argue that it’s been reified I think one needs argue that it 1) is a definitely an abstract concept and 2) not an actual thing, i.e. 3) not just bollocks. I’m no authority on the first two or what constitutes them, so I don’t feel I can say if ‘life’ has unduly become thing-like, but knowing that there are two different words, I expect reification, if it did happen, happened a long time ago.

    Actually, this recent discussion of reification has really confused me, the examples on wikipedia make sense to me – as in the sense of saying how I think something works is _actually_ how it works, but I don’t see how this applies directly to Spearman’s g or life any _more_ than it applies to the coffee I’m drinking, the chair I’m sitting on or any other philosophical cliché of a thing one can name.

    I guess in science, reification happens, people take things as read, theorize, take as read, theorize etc to the point where they loose touch with the reality of the system they are studying and end up missing (or even outright denying) important facts. I think it probably happens pretty quickly too, can’t say I haven’t caught myself doing it. So it is theories around ‘g’ and around life or around whatever that become reified, not them themselves. In the case of ‘g’ it is its connection to biometric measurements was reified.

    Lets not confuse the thing/property distinction with reality/model distinction.

    On this note, I think life is more like a thing than, say, the number 42. I started a post on that actually, then I realized that I was going slowly mad and stopped. In one gnarly sentence: Ill-definedness seems to be a property of things we think are real (dare I say analytic… seems I might, even though I know nothing about Kant) and well-definedness seems to be a property of things we don’t (synthetic things… perhaps). So, the fallacy of reification seems to me to be equivalent the assumption that the world, or some portion of it, works in some kind of idealized, well-defined way.

    • Hmm, well, yes, fair enough I suppose. If I understand correctly, you’re saying that when I give a name to, or even try to talk about *anything* in the physical world I’m indulging in this kind of reification process, because things in the world are really just [collections of atoms | correlations in sense data | whichever other route you want to take to attack the notion of thing-ness]. I can’t disagree with that.

      However, I guess the point is that doing this is sometimes useful and sometimes not. I wouldn’t get very far as a bricklayer if I didn’t reify the notion of a brick. In the case of life, it’s useful for biologists to talk about life as a thing, because on the present-day Earth there *is* a very sharp distinction between life and non-life. My point is just that this distinction/reification/whatever is probably not as useful to those who study the origins of life (or astrobiology for that matter), because then you’re likely to be dealing with cases where that nice neat distinction breaks down. Any sharp distinction you do draw will be more or less arbitrary, and thus not particularly helpful.

      This is a simple point but a difficult one to get across easily, and for that reason I collect metaphors. My latest one concerns supercritical fluids. It’s (more or less) sensible to teach children that water comes in three “states of matter”: liquid, solid and gas. But if you compress steam to far greater pressures than children ever experience, then cool it down, then release the pressure again, you’ll have gone from steam to water without ever passing a phase change. Liquid water and steam are clearly distinct if you’re concerned with their behaviour at atmospheric pressure, but in the universe as a whole they’re just two different points in the same region of a phase diagram.

    • Also: yes, a long time ago. Centuries if not millennia – but there’s nothing wrong with challenging a few long-held preconceptions here and there.

    • P.S. I believe you’re using the terms synthetic and analytic correctly, though I’m no Kant scholar either.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 483 other followers

%d bloggers like this: