Life is short

Continuing with rather philosophical musings about life, Ed Trifonov has recently suggested a new approach to defining life:  let’s just vote on the definition.
So how does that work? And why should it work in the first place?
Note that I am diving straight into the subject, and not prefacing this post with a review of the various definitions of life. I assume that this blog’s readers have been exposed to some aspects of the debate on how to define life. Wikipedia and the references therein are a decent starting point, in case you want to refresh your memory. But just so we have something, here is the definition from the American Heritage dictionary:
LIFE:  the condition that distinguishes organisms from inorganic objects and dead organisms, being manifested by growth through metabolism, reproduction, and the power of adaptation to environment through changes originating internally.
Trifonov’s rationale for doing what he does is as follows:
The definitions [of life, IF] are more than often in conflict with one another. Undeniably, however, most of them do have a point, one or another or several, and common sense suggests that, probably, one could arrive to a consensus, if only the authors, some two centuries apart from one another, could be brought together. One thing, however, can be done – sort of voting in absentia – asking which terms in the definitions are the most frequent and, thus, perhaps, reflecting the most important points shared by many. Such analysis is offered below, revealing those most frequent terms that may be used for tentative formulation of the consensus.
Where to start?  Trifonov decided to take two book chapters which together list 123 non-redundant definitions of life. He then counted the words in those chapters, omitting connecting words and grouped them by meaning , then ordered them by the definientia (the words serving to define another word or expression) frequency (click to enlarge):

By word count, seems like life has mainly to do with living. Well, no surprises there, but somewhat tautological and less-than-informative.  However, rejoice O system biologists, for SYSTEM is the second most frequent keyword grouping. Then we have organic stuff, CHEMICAL, COMPLEXITY, REPRODUCTION with ENERGY and ABILITY trailing. Trifonov continues:

Thus, the consensus of the life definition patched from these nine definientia would be: Life is [System, Matter, Chemical (Metabolism), Complexity (Information), (Self-)Reproduction, Evolution (Change), Environment, Energy, Ability,…] where the square brackets correspond to some compact expression containing the words listed within. For example, one possibility is:

Life is metabolizing material informational system with
ability of self-reproduction with changes (evolution),
which requires energy and suitable environment.    

(I added the underlines.) Hm.  Actually, not bad for a definition culled from a simple exercise in word counting. Of course, to put these definientia together one would need some knowledge of life, this the exercise is not completely automatic and unbiased, nor does it profess to be so.

But Trifonov wants to condense this definition even more. To quote Hemingway: “boil it down; know what to leave out; tell a story in six words”. Is there still some redundancy in the definientia themselves that would let us boil it down and tell the story in only six words? Trifonov argues that metabolism implies the existence of energy and materials.  Whereas the existence of materials already implies a suitable environment. But self-reproduction subsumes all the above, as it requires metabolism, energy, materials and environment. However, variations and self-reproduction  are actually mutually exclusive. Both must be noted. The boiled down, Hemingwayan definition would therefore be:

 Life is self-reproduction with variations.

And, to top it off, six words! Hemingway achievement unlocked.

Of course, this succinct definitions renders all sorts of problems. trifonov admits to that:

One unforeseen property of the minimalistic definition is its generality. It can be considered as applicable not just to “earthly” life but to any forms of life imagination may offer, like extraterrestrial life, alternative chemistry forms, computer models, and abstract forms. It suggests a unique common basis for the variety of lives: all is life that copies itself and changes.

Here is where I think things go a bit too far: is self-reproducing (and mutating) software alive? Is the Weasel program alive? Are viruses alive? All of those examples fulfill, at least technically, the above definition. In a previous post, I talked about going from life to non-life on a scale roughly correlating to size and thus the amount of information and sustaining materials life can carry with it. The difference between life and non-life seems to be not only in self-reproduction with variations, but the ability to do so at some level of autonomy. When adding the caveat of autonomy, viruses are not alive, since they require the transcriptional and translational machinery of their host cells. Neither are organelles such as mitochondria, since most of their proteins are encoded by the nucleus. But requiring autonomy raises another problem, which I find hard to solve: how far does this requirement of autonomy go? After all, all heteretrophs are, to some degree non-autonomous, as they require basic materials produced only by autotrophs. So the definition becomes fuzzy again: humans are alive, although they cannot self-sustain without plants. Plants are alive, but they cannot fix nitrogen and require bacteria to do so. So are we to say that  autotrophic nitrogen-fixing bacteria the only living species on earth?  So maybe the autonomy criterion be limited to self-reproduction rather than metabolism? But those are hard to separate: without sugar, there is no DNA. Without essential amino acids, which most heretrophs acquire by consuming other organisms, there are no proteins to effect reproduction.

despite the difficulties,  my definition would be (seven words, unfortunately):

 Life is autonomous self-reproduction with variations.

 Fuzzy? You betcha. That’s life.

PS: as you can see from the article’s Pubmed page, it generated a flurry of comments. Those make for a great read too. Enjoy.



Trifonov EN (2011). Vocabulary of definitions of life suggests a definition. Journal of biomolecular structure & dynamics, 29 (2), 259-66 PMID: 21875147

Share and Enjoy:
  • Fark
  • Digg
  • Technorati
  • StumbleUpon
  • Facebook
  • Reddit
  • Twitter
  • FriendFeed
  • PDF
  • email
  • Print
  • Google Bookmarks

69 Responses to “Life is short”

  1. Somebody says:

    And last one “conclusion” 😉
    – even chair have “his” “metabolism”. In “his” own life cycle, and yours too…

  2. TJ Radcliffe says:

    The six-word definition is incomplete: “Life is self-reproduction with imperfectly inheritable variations.”

  3. Mr. E. says:

    At any ‘frozen’ instance of time you cannot tell whether something is alive or not. It is only after observation over a period of time that you notice that certain forms persist in an environment and that you cannot explain their existence through some random and spontaneous aggregation of elements from that environment. These forms mutate over the generations.

    I think that a crucial element to life is the information which governs the process of consumption and the process of proliferation of this information through replication. The adaptability aspect is certainly key to sustaining these processes in a changing environment. I would phrase it as follows:

    “Self preservation and propagation of information through adaptable exploitation of the environment.”

  4. Sandy Stewart says:

    These definitions are good, but they assume that something is either “life” or “non-life.” In my view, this is enforcing a binary categorization on something that is inherently continuous. Perhaps the best thing to do would be to agree that there are things that are obviously living (bacteria, trees, cats) and things that are obviously dead (a piece of aluminum), and recognize that there is a continuum of states in between (prions, viruses, ?).

  5. Iddo says:

    Sandy: I agree. You can check out my Life on a Gradient post (two posts ago or so) for that.

  6. fastpathguru says:

    “Life is evolution’s self-sustaining byproducts.”

  7. Budd Riley says:

    Life is the oposite of death

  8. Abbey Normal says:

    I don’t think basing a definition of life on gravity makes much sense, given that at different scales different forces become much, much more important than gravity. At small scales, forces like surface tension have *far* more importance. Though I do see an underlying point – that of being able to act in opposition to outside forces.

    Personally, I like Francisco Varela’s idea of autopoiesis (self creating/producing) as central to what make living systems alive.

    With an autopoietic system you get reproduction, metabolism, action in opposition to outside forces, and adaptation to outside forces all as inherent parts of the kind of system it is. That’s a lot of ground you can cover in just 2 words. Perhaps some notion of the variability is needed – though much variablity fall out of not defining it as one static “thing” in the first place, but instead a process that leaves room for variation. So, to get the variablity across generations you can include a notion of history/lineage.

    So my definition (paraphrased from Varela and others, not my original ideas):

    Life is an historical autopoietic system.

  9. Abbey Normal says:

    Oh, and the definition I give above also doesn’t have to be either/or. As many have pointed out above, we may want to leave room for a continuum of life. I would argue that a system can be more or less autopoietic, so the above definition would permit degrees of “aliveness”.

  10. Dadster says:

    This is a most sensible approach to defining ” life ”
    I don’t think ” life ” is some vague  undefinable quality.

    Life is that “quality of cosmos that preserves information”.

    This takes care of the requirement of “continuity of the spectrum of differently living entities” to include  not only insects, viruses prions bacteria and animals , plants and rocks and streams , galaxies and black holes etc

    The various changes at various time- scales ( at femto scales and some like a galaxy systems , black holes etc that changes at slower rates such as,evolution) , self organization, free- will , reproduction , instinct of survival , space traveling viruses and bacteria are just multifarious processes of “information- preservation”.

    Information cannot be destroyed nor created , but only preserved.
    Preservation is the defining quality of information itself.  
    Every question has it’s answer within itself and every answer has it’s question built into itself !
    Therefore I have a three -word definition for “life ”

    “Life is information”.

    Where information is absent there, life too does not exist.

    Phenomena and Noumena are nothing but ” information”.

  11. Peter Kinnon says:

    Although we can make reasonably comprehensive definitions of life in the biological sense it is interesting and, I believe, useful, to postulate an extended “life process” in terms of a continuum extending at least from the formation of the chemical elements in the stars right through to the current evolution of technology. And a presumed extension into the future.

    This is the main subject of my latest book “The Goldilocks Effect: What Has Serendipity Ever Done For Us?”. It is a free download in e-book formats from the “Unusual Perspectives” website

  12. Peter Kinnon says:

    Quite correct, Llama.

    Now, what drives negentropy?

    If you’re not sure check out chapter 8, section 3 of “Unusual Perspectives”.

    It is a free download.

  13. anti-entropy says:

    Life is anti-entropy.

    It has exactly the opposite effect as entropy.
    Life is a tendency towards order, where other physical things just tend to fall apart, i.e. increase entropy.

    It looks like the effect of a `strange attractor´, like you have in fractals, but for information. If there are multiple outcomes of a quantum experiment, with equal energy levels, the non-living thing will have random outcomes, but in a living system will (in the mean) act as if it makes choices.

    It is like Plato’s world of ideals is actually real, this would be the strange attractor.

  14. RobvS says:

    So if you do not get children, you are dead? Is the Pope dead?

  15. 3278 says:

    Dawkins’ definition from Blind Watchmaker – replication with heredity – has always been my preferred one. The addition of the word “autonomous” is a welcome one, though; many things replicate with heredity but do so only as directed. Certainly reductionist definitions like these tend to bring up uncomfortable moral questions, but I believe that’s one useful purpose of them: feature, not bug.

  16. Peter Kinnon says:

    Robus: I think you will find that the community of cells that happen to constitute the Pope have done a whole lot of replicating, and still are!
    Furthermore, I think you will find very few (if any) of that population were there at the time when that person was born.
    He is, too, a member of a replicative species.

    Anti-entropy: Negegentropy, dysentropy, syntropy are, of course, used interchageably with anti-entropy.
    Having used the concept to define life in my previous writings, I am now leaning away from the term because of ambiguities in our perceptions of order/disorder.
    So I have come to the conclusion that entropy is best left within the domains of thermodynamics and information theory where such ambiguities do not arise and for my next (more formal) book I am adopting different terminology which I believe is conceptually more precise.

  17. Gopal says:

    @Archimedix says: 12-February-2012 at 10:21 AM
    “Self-awareness as we commonly understand it is a pretty sophisticated phenomenon and should undoubtedly be a sufficient condition for life, as is “true” intelligence by itself already (intelligence is the capability to adapt by learning and to infer, conclude, act and react using learned knowledge). Intelligence is probably necessary for “true” self-awareness (I assume that by self-awareness you do not mean any type of self-perception but also the ability to obtain new knowledge from self-perception; any cybernetic system is able to receive input originating either directly from itself or from input caused by itself).”

    First, apologies for not reverting earlier.

    I think ‘intelligence’ presumes ‘self-awareness’ i.e., ‘self-awareness’ is necessary for ‘intelligence’ to emerge. Thus, a being that is ‘intelligent’ is necessarily ‘self-aware’, but the corollary is NOT necessarily true. For example, a virus may not be ‘intelligent’, but is certainly ‘self-aware’ at a rudimentary level – it may not ‘consciously’ enter a host, but becomes ‘aware’ (chemically/non-intelligently?) of its surroundings as conducive for reproduction and proliferates.

  18. Paul Agapow says:

    I dug into this subject in my thesis oh-so-many years ago. It’s an interesting (i.e. endless) subject, but you have to ask what purpose a definition serves. I’m not sure that it serves any.

    I’m also uneasy about the word “autonomous” after a more senior biologist schooled me on this point years ago. Certainly, you could argue that a virus is not autonomous as it requires a host genome. But there are a plethora of obligate parasites that also can’t survive without specialised environments. And I’d have trouble naming any organism that can “live” fully autonomous from it’s environment.

    Having said that, I do like the idea that life is that which does life-like things.