Recently, Hank Green posted a video discussing the definition of life.  He offered two definitions; that life acts in a manner to achieve a goal, and that life continuously decreases internal entropy.

There are problems with these definitions.  The first definition includes every machine that has a function.  The second one includes, for example, a machine that constantly reshapes parts of its body into a paperclip.

Other definitions of life are equally confusing, doing things like excluding viruses because they use other cells to reproduce, despite meeting the intuitive meaning we have for life.

So, dissolve "life."  Why do we care if something is alive?  To decide if its life has value.  Hank dances around the issue, showing that life has no inherent value by using mouthwash to kill billions of bacteria in his mouth.  But, he doesn't take this to its conclusion.  It doesn't matter if something is alive or not.  We won't suddenly care about the well being of viruses if a new definition of life comes along tomorrow pronouncing viruses to be living.  What we value is sentience.

Bugs are extremely low on the sentience scale, so we feel free to kill them.  Animals that are higher on the scale, such as cats, have laws preventing any sort of mistreatment.

tl;dr: Life is an ambiguous term, use sentience to describe a being's value.

New Comment
30 comments, sorted by Click to highlight new comments since:
[-]asr100

So, dissolve "life." Why do we care if something is alive? To decide if its life has value.

I don't think that's the only reason we care. There's a lot of mostly-true generalizations that we can make about living things -- they require energy to keep living, they have some means of internal self-repair, etc. If I ask whether a vaccine has live or dead bacteria, it isn't because I value the live bacteria more, it's because it lets me assess risks more accurately.

People routinely value non-living things over living things. For instance, many people wear fur. (Or for that matter, cotton.)

What we value is sentience.

This isn't really true either. We have lots of different values. For instance, many people are willing to kill mice to feed their pet snakes. I don't think they'd be dissuaded if it turns out that mice are "more sentient" than snakes by some metric -- they like their [distinctive] pet more than they like the mice, which from their point of view are[interchangeable and replaceable.

We also value rarity -- consider a rare lizard versus a common bird-of-prey.

And we value several kinds of instrumental utility -- foxes might be smarter than chickens, but just ask a farmer which they value more.

tl;dr: Neither life or sentience are strongly correlated with how much most people value something.

tl;dr: Life is an ambiguous term, use sentience to describe a being's value.

Do you suggest that sentience is a less ambiguous term?

[-][anonymous]50

I think it's not that hard to make a definition of life that includes everything most people consider alive and excludes everything else. You could just call life anything that stores instructions for making more copies of itself on nucleic acids. Maybe I'm being too glib, I dunno. But the better definitions of life don't really seem to be trying to determine whether edge cases should count as alive or not, which as you note is not really a pressing question.

It's more a matter of saying, "This thing we call life is interesting. What are the characteristics that make it so interesting?" and coming up with things like self-replication, homeostasis, metabolism, that stuff.

[-]asr10

I agree that it isn't that hard. But your definition doesn't quite do it. As written, your definition includes things that have died and things that aren't yet alive.

Dead plants and animals still have usable DNA for a while after death. If the tissues are preserved, sometimes it's a long while. And mostly, we don't think of spores or viruses as living things, but they certainly have DNA.

I would supplement your definition by saying that we refer to an organism as alive when its pieces are functioning in a coherent and mutually-dependent way to keep the organism as a whole alive. (Some fuzziness creeps in when you try to distinguish a composite organism from a collection of separate components...but I don't think that causes problems in practice.)

[-][anonymous]20

Oh yeah, good call, dead things. You know, the more I think about it, the more I think the concept of "dead" might be even more fraught with vitalism than "alive" is.

You could just call life anything that stores instructions for making more copies of itself on nucleic acids.

We could, for all the good it would do if we encounter organisms on other planets that store reproductive information on something else.

[-][anonymous]10

The point was to make a clearly after-the-fact definition that works, but doesn't say anything interesting. There may not be any exocritters that use some other macromolecule for their genetic material. If not, it would be a non-issue, right?

If you try to make a definition that says something interesting, I think it's got to be a whole checklist of interesting characteristics of life in order to exclude things most people wouldn't consider alive. Then you get people looking at objects that exhibit some, but not all of the characteristics and wringing their hands over whether or not it's really alive, as if that's a thing. Which I still don't think it is. I think what we're looking at here is lingering vitalism.

Life is just matter with interesting characteristics. We can talk about why it's interesting, but we can't explain why it's totally different from all the other matter, because it isn't.

The point was to make a clearly after-the-fact definition that works, but doesn't say anything interesting. There may not be any exocritters that use some other macromolecule for their genetic material. If not, it would be a non-issue, right?

It would still be an issue if we started making synthetic life. A good definition ought to cause as little inconvenience as possible. I agree that life isn't fundamentally different from all other matter, but the reason we have the word at all is because it's handy to be able to encapsulate it as a reference class. Individuals with a solid grounding in natural sciences may at least be at a state of "I know what I mean when I talk about it," but if we are going to define it at all rather than simply taking the Justice Stewart approach, the definition should encapsulate what we actually mean.

[-][anonymous]20

I agree that life isn't fundamentally different from all other matter, but the reason we have the word at all is because it's handy to be able to encapsulate it as a reference class.

I think the reason we have the word is more to do with historical vitalism than with carving reality at its joints. These days it's most often, or at least most rigorously, used to indicate "that thing biologists study, with all the membranes and nucleotides and amino acids." If you want to be more abstract than that, I don't think trying to define "alive" is really a good approach at all, because it means too many different things. But you can certainly look at all the interesting things it can mean and figure out which one you're currently talking about. You might want to give it a new name, though.

There exists a particular cluster in thingspace which we call "living things" and we have invented a magical quality called "life" to apply to members of this cluster.

Virii are peripheral members of the cluster, just like penguins are atypical birds, so there's confusion. There's also general confusion about which dimensions should be considered important, i.e. complexity, intelligence, etc.

Discussions of the meaning of life are confused for the same reason as discussions about the morality of coffee tables.

You could say the same thing about any concept so this is hardly a dissolution IMO.

He offered two definitions; that life acts in a manner to achieve a goal, and that life continuously decreases internal entropy.

Every control system does these things. It acts in a manner to achieve a goal (keeping a perception close to a reference value), and if successful, that implies that the entropy of the thing it is controlling decreases.

That does not make all control systems living things.

So, dissolve "life."

Done long ago. Vitalism is dead.

It's not obvious to me why "every machine that has a function" shouldn't count as alive for these purposes. If someone suggested those definitions in the first place then presumably he's looking for somehting broader conceptually than the carbon based replication we usually mean. (didn't actually watch the video)

Tangential, but the term "sentience" should really be deprecated. It seems to serve no purpose save for enabling conflation between consciousness and intelligence.

I'm not sure everyone spends much time thinking about sentience. But people seem prone to empathy and the more an animal seems like a person, the more empathy we have.

An animal could be more "personlike" on different dimensions including sentience, but also including facial expressions.

Biologists have looked into this. E.g.: J. Maynard Smith and Eors Szathmary - in "The Origins of Life", p.3:

What is life? [...]

An alternative is to define as living any population of entities posessing those properties that are needed if the population is to evolve by natural selection.

This might just be nitpicking, but that would exclude artificially created life that doesn't have mutations.

It would also exclude humans if we enhanced ourselves beyond mutating. This seems to me a much stronger counter example.

Evolution requires variation. In the real world, there's no such thing as enhancing a living system so it doesn't vary. A living system that doesn't vary doesn't stay alive for very long.

Clarification: Do you think it would be impossible to bring humans to the point that we no longer have mutations, or that it would lead to our extinction, or neither?

That it isn't going to happen. The future surely contains massive variation, as adaptive strategies are explored on ever-larger scales. Endless stasis just isn't how evolution operates. Attempting to defend completely against mutations is pointless and futile.

I'm not sure if you're right, but in any case I expect variation in the distant future to come from design rather than random mutation.

Randomness was never one of the "properties that are needed if the population is to evolve by natural selection" in the first place. I never mentioned it, and nor did Maynard Smith and Eors Szathmary.

FWIW, my preferred version is that life is that which persists via copying.

[-][anonymous]10

Like chain letters?

That's correct. Chain letters - and the rest of human culture - is literally alive.

Here is Dawkins (1976) on the topic:

As my colleague N.K. Humphrey neatly summed up an earlier draft of this chapter: "memes should be regarded as living structures, not just metaphorically but technically".

[-][anonymous]20

Seems a bit perverse to use a definition that says a chain letter is alive, but a mule isn't. Why not just make up a new term?

Mules are alive. Mule cells reproduce and persist via a copying process. Being sterile does not mean you are not alive, just that you are near the end of the line. Chain letters are alive too - it is unfortunate that few people realise that.

IMO, a new word is not needed: we don't need multiple terms for practically the same basic thing.