Life - an Unequivicol Definition

It's been awhile, but it's good to see you again!

One of the lines I like to use in my talks and discussions with colleagues is that modern ecologists are pretty well-trained to avoid giving clear answers to anything. This is mainly because the systems we try to study are inherently more complex than we can replicate with experiments or models, and there's a history of extended, high-profile disagreements lasting a decade or more, only to end with the realization that both sides are right under certain circumstances.

The thing that we're really starting to appreciate in modern ecology is that we don't have to develop rigid, unequivocal definitions and stances in order to make meaningful discoveries. In fact, focusing on rigid definitions more often than not distracts us from discovering the most important information.

It's a valid enough definition, I suppose. I feel like it's needlessly specific, though. For example, it's entirely conceivable that an organism could function just as well using GTP (guanosine triphosphate) instead of ATP, and it isn't outside the realm of possibility that such an organism could be discovered on Earth. It would be better to avoid committing ourselves to a specific definition that turns out to be based on rather arbitrary decisions like this.

Also, I'm not sure I understand what we would do with this definition. Is it for the purposes of delineating academic "territories" and earmarking grant funds? Or, is it for the purposes of making a baraminological argument about lines that evolution can't cross?

Neither one of these sounds like a particularly compelling reason to me.

I know you know what a tautology is.

I think you might be wrong about this. If we were to poll biologists with the question --- "What is the definition of 'life'?" --- I suspect that an "any of the above" or "it depends" option would be a very popular one.

Edited by Blue Jay, : Rewording

I was talking about ecologists specifically, not about all biologists. And it isn't really indoctrination: it's experience with a frustratingly complex and chaotic subject matter. The rest of biology doesn't have the same problem: genomics, for example, has a lot of highly quantifiable mechanisms that can be readily isolated and characterized in the lab.

As difficult and complex as climate is, it's still much more tractable and predictable than ecology. Consider that climate is actually one of the explanatory variables in ecology, so that an accurate prediction of future ecological dynamics would necessarily incorporate a climate model as one component.

Yes, needlessly specific. You're trying to define life in terms of a specific chemistry, when we don't really know that life couldn't exist on alternative chemistries. I suggested a hypothetical life-form that uses GTP instead of ATP. The two molecules are chemically very similar, and store the same amount of energy --- there's no reason why a metabolism couldn't exist that uses GTP.

If such an organism were discovered, would you entertain the notion that it wasn't alive, just because it doesn't fit your definition?

You said "this definition encompasses all known life." Since "life" is what the definition defines, this statement is true regardless of what the definition actually is.

Perhaps this is a nitpick, but the simplicity of your definition is an illusion. All you've really done is obscure the complexity of it behind a footnoted definition for "evolution." In all fairness, you should include the footnote as part of your definition, because without it, languages and stars,, which are also said to "evolve," can also be considered "life."

If we do that, here's what the definition looks like:

Anything capable of changes in the composition of hereditary traits, and changes to the frequency of their distributions within breeding populations from generation to generation, in response to ecological challenges and opportunities for growth, development, survival and reproductive success in changing or different habitats.

When phrased like this, it's pretty clear that "life," as you define it, has at least three fundamental characteristics: (1) hereditary traits, (2) a population, and (3) the ability to experience changes in the distribution of hereditary traits within the population in response to... etc.

If we look a little closer, we can see that "hereditary" presupposes reproduction, and "traits" presupposes organization. "Reproduction" and "organization" are two of the classical characteristics of life. So, "hereditary traits" is sort of just a glossed-over paraphrasing of a more substantive definition.

In fact, we can further evaluate your entire definition and conclude that it's basically just an obscurantist repackaging of the "classical" description of life:

Life is organized; it responds to stimuli, metabolizes energy, grows, reproduces, and adapts.

Simplicity is nice, but clarity is better.

I still think you're wrong about this. Ask any biologist for a definition of life, and they will probably either say, "it's complicated," or will give a definition that either excludes viruses or mules or some such, then will backpedal when said virus or mule us brought up in response.

I do not have a problem with AOkid's claim that biologists are equivocal on the definition if life, and I think it's kind of silly of you to make this claim when I'm pretty sure biology, as a field has already come to terms with this.

In biology, there’s a long history of supposedly “unequivocal definitions” turning out to be oversimplified approximations with limited utility for real-world biology.

For example, the word "carnivore" is a problematic term. It's debatable whether any animal can be truly regarded as a “carnivore”: it seems that nearly every putative “carnivore” we examine closely turns out to be more appropriately classified as an “omnivore.”

So, does the term “carnivore” still hold enough meaning to justify its continued use? If so, when should we use it? For example, is it okay to categorize wolves as “carnivores,” even though they obtain a small percentage of their calories from eating berries? At what percentage of plant-derived calories should we start using the term "omnivore" instead, and why should we use that percentage?

Also, we have to ask whether trying to impose a rigid definition on things is helping our science progress. It is possible that we're actually hindering our progress by masking important information behind an arbitrary categorization schema?

It really isn't the simple matter you make it out to be.

I see plenty of reasons to be less specific. Here are a few:

1. Your definition draws the line between "life" and "non-life" in pretty much the same place as several other definitions that are less specific (e.g., "life has cells and metabolizes"). So, what exactly does the added specificity bring to the table in terms of classifying entities?

2. Would you seriously consider excluding an entity from the definition of “life” if it used GTP, instead of ATP, for its metabolism? If not, what’s the point of proposing that the identity of the metabolic molecule be used as a defining characteristic of life?

3. Your definition still has the “virus problem”: there are ways in which viruses are more like “life” than like other types of “non-life.” In this way, your definition doesn’t perform any better than any other definition when it comes to grouping things by their phenomenology.

This opens a rather large can of worms, and I'm not sure where to begin with it.

As AOkid and NoNukes have pointed out, evolution is a population phenomenon: individuals are all incapable of evolution. So, maybe you could get away with this if you said, "anything whose population is capable of evolution."

Of course, we then have to ask what the mule's population is (does it include horses and donkeys? Or are mules a separate population?); and what happens if that entire population is incapable of reproduction. Surely an entire population that is incapable of reproduction cannot evolve, since there can be no differential reproductive success (change in fitness) within that population.

Also, I'm having a little trouble justifying this in light of what you wrote in Message 56:

This muddies the waters. A mule's inability to reproduce counts as a "process of evolution," thus letting it meet the requirements for "life," but a crystal's ability to reproduce without change disqualifies it?

I think you need to provide some better justification for the dividing line you've chosen.

Intuitively, any valid definition of life should probably include mules and exclude crystals, but you'll need to elaborate on your definition a bit before this is clear.

...which brings us back to the the point I made a long time ago that we have to consider the reason for coining a term.

Why would we want to use terms is such a way that they don't reflect reality?

The life sciences have been trying, for some time now, to shift away from thinking in term of dichotomies toward thinking in terms of continua, because it better represents reality. We don't need or really even want an unequivocal definition of "life" because it facilitates and justifies all kinds of false thinking. For example, read what you just wrote:

"We don't have a continuum"? Yet, in the very next paragraph you say, "That does not destroy the continuum from chemicals to life in any manner".

You can't have it both ways: if there is a continuum, there is no clear dividing line. You blame us for it, but the very first responses you got in this thread were people expressing skepticism toward the idea of "defining" life unequivocally. How on Earth are you blaming us for the dichotomy that we've said, from the beginning, we don't like?

Yes, I see the difference in the words, but all you're arguing is words. Reality doesn't have to conform to linguistic conventions. 'Life' is a vague, imprecise term with no clear, unequivocal definition, so the dividing line between 'life' and 'non-life' is fuzzy.

Some things could be placed in either 'bin,' so despite the way it molests the rules of linguistics, 'life' and 'non-life' are not necessarily mutually exclusive categories.

-----

Anyway, have a good weekend, and Merry Christmas!

Edited by Blue Jay, : fixed codes

Instead of yammering on about the shortcomings of your definition, how about I make a couple constructive suggestions to improve on it?

Based on your proposed definition, you seem to have focused in 3 major characteristics of life, which I'll call "containment," "metabolism" and "information."

You have explicitly constrained "metabolism" and "information" in terms of specific types of molecules, but have left "containment" more vague. Perhaps if you also constrained "containment" in a similar fashion, your definition would be more appealingly self-consistent.

Also, I don't think it's a good idea to try to define "life" and "a living organism" with the same definition. "Organism" is a term about how you define separate "units" of life, and I don't think it's necessary to open that can of worms in the basic definition.

So, try this modification:

quote:

---

Life is a self-contained entity comprised of chemical reactions including a metabolism (based on ATP synthesized by enzymes) and a genetic process (requiring the transfer of information from DNA to RNA), and a set of enclosing membranes consisting of lipid bilayers.

---

That is a perfectly valid and reasonable definition of "life." As I've said before, I think it's too restrictive on a number of fronts (viruses, hypothetical aliens, etc.). But, there is one implication of your definition that I find interesting: it tells us exactly what we have to do in order to make artificial 'life' in the laboratory.

I'm not a big fan of this approach. I mean, I agree that it's a valid and meaningful perspective, and that it's by no means irrational of you to pursue that angle; but I also think it opens some new baggage, such as the stuff that RAZD has been going on about.

Here's how I see it. I think of the definition of 'life' as delineating the minimum limits of 'life,' and the definition of 'living organism' as delineating a sort of maximum limit (i.e., the point at which you distinguish between 1 unit of 'life' and 2). If you want to define 'living organism' without separately defining 'life,' you'll have to account for both 'limits' in one definition, and I'm worried that it will (and, in this thread, has) become a bigger and more convoluted mess.

I figured this was the reason. But, in leaving it vague, you sort of made it into a vulnerability. Earlier in the thread, I attacked your definition for the highly restrictive specificity of the other two criteria; and here, RAZD attacked this criterion for its highly ambiguous lack of specificity. I think consistency among the criteria will go a long way toward reducing its vulnerabilities.

Consider: is there anything that has a shell, fur or chitin, but doesn't have lipid bilayer membranes? If not, why try to account for them all when a smaller, more specific 'container' is available and more universal?

Off-Topic Side Note: One if my hobbies is speculative biology, in which total nerds like to try to invent plausible life-forms that inhabit other worlds. In fictional situations like these, where you may have many different biospheres, it becomes possible to adequately test hypotheses about what criteria are required for 'life' and 'life'-like phenomena, and you can justify being specific. However, in real life, where we don't have such a broad perspective, I don't think the specificity is justifiable.

Remember that we didn't set the criteria for 'life,' then go out looking for things that fit the criteria: that wouldn't be very scientific of us. Rather, we let our observations dictate to use what criteria we should use. So, the criteria are living hypotheses, and we're still testing them to decide whether they are all appropriate.

Go back to RAZD's table, where he attached a percentage value to each entity based on the number if 'yes' answers. Rocks, and sulphuric acid and dust clouds don't meet any of the criteria (except, maybe 'organization'). But, viruses meet 3 or 4 of the 7 traditional criteria for 'life'*.

Viruses stand out a little bit from the crowd of 'non-living' things. They certainly seem to behave a lot like 'life' in at least some ways, even though they don't exactly fit our expectations for what 'life' is. They possibly even 'evolved' from things that originally met all 7 criteria. This causes us to question whether all of our criteria/hypotheses are valid or necessary.

*I disagree with your assessment that they do not respond to stimuli: I believe the chemical cascades initiated during the processes of penetrating a host-cell membrane and inserting the viral genome into the host genome should be considered a response to a stimulus, but it's admittedly marginal.

That's fair enough, I suppose. But, remember that the criteria are still untested hypotheses, so the idea that 7 criteria = life is a tentative conclusion. There are cases, such as mules, where it seems that 6 (or 6.5) criteria is enough, so there is some uncertainty about where 'white' actually is.

And, viruses may have marginal claims to meet some of other criteria. For example, the process of building a capsid coating could be considered a rudimentary form of 'growth.' So, that might justify calling it 4.5 criteria met.

But, all that is immaterial, really. I don't know where the phrase 'edge of life' came from, but I think you're getting too hung up on the semantics. But, 'almost alive,' 'edge of life,' 'life-like,' 'quasi-alive,' 'sort of alive'... they're just ways to express that viruses are like 'life,' but don't meet all the traditional criteria.