Well… sometimes they can. But they requires some artificial life support. Or a new host. Cancer can do some weird stuff.
Evolution would imply the duplication of information in a way that changes some of the information. So the minimal requirement would be a molecule that polymerizes in a way that encodes information (a gene), and furthermore that an imperfect duplication could duplicate in spite of the imperfections.
RNA is the one that’s most often theorized to have actually happened, and seems most likely. It doesn’t need to be RNA, any polymer with the above qualities should work.
Nothing informs us this is not true. It’s not something we could do right now, and there may never be any practical reason for it. However, everything we know says it could be done. We use technology to make spaceships now. These certainly are more advanced spaceships than we can make now and there may never be any practical reason to do so, but there’s nothing impossible about it.
I’m sure there could be complex soupy chemical arrangements that would do all sorts of things equivalent to metabolism and photosynthesis and whatnot, whilst remaining an amorphous soupy puddle of chemicals, but one of the things that a cell membrane does, that isn’t especially chemistry-y in nature is that it defines the boundary of an organism - it organises those chemical reactions, at the most basic level, by constraining them in a collection that remains together.
In order for something to replicate, I think there has to be some replicable unit - cells seem like an obvious solution to that, but maybe that’s just because everything on Earth that is alive, is based on cells. Maybe there’s some other solution to that whole thing that we can’t even imagine because we are so cell-biased, but maybe not :-
The solution: ‘keep your stuff safe and distinct from other stuff by putting it in a bag’ has evolved over and over again on Earth, in different contexts:
- The nucleus and other organelles within cells have membranes enclosing them
- The cell itself has a membrane
- Multicellular organisms typically have internal organs that have membranes or other protective or boundary layers
- Multicellular organisms typically have some sort of skin differentiating their inside from their outside
- Some collections of organs or structures are enclosed in containers of skin and/or other tissues (scrotum, cranium, mammaries)
- Eggs/wombs/pouches (and I suppose host organisms, if you’re an endoparasite) are containers in which offspring develop
- Some organisms like caddis flies and hermit crabs construct or co-opt pieces of their environment as protective containers
- Humans keep money in purses and carry their shopping in bags, and wear clothes (which are just bags, really)
Keeping your stuff safe and together seems such a fundamental requirement for living things, I think even if life started off with some other solution (like, say, a really big, complex, tangly single molecule), it would stumble on the idea of membranes pretty quickly, and use them.
[remember to insert ‘ugly bags of mostly water’ joke here]
I disagree. We don’t even know how feasible it is to get a spaceship from one star system to another in reasonable time. The interstellar medium may be full of enough stuff that a trip to another star almost guarantees running into something. This may limit how fast we can travel. Or there may be other risks and limits we have yet to discover.
Consider a situation where a true self-replicating probe is only successful at finding a suitable planet one time in a hundred. The rest of the time, the probe winds up in a system that does not have the resources needed, or is lost on the journey. Or maybe the only planets have a gravity well too strong to get out of. That means each successful probe would have to make hundreds of copies of itself to keep the expansion going. It might take the resources of a planet to do that, making the whole purpose of expansion moot.
If all the risks and complexities mean that probes cannot find new star systems and replicate at higher than the loss rate, you won’t get geometric expansion, but a fizzle out.
Think of an analogy to a nuclear detonation. Unless you get to a critical density, the reaction always fizzles out. Maybe getting to a critical density of self-replicating robots that can sustain an exponential expansion is impossible for energy conservation reasons or something. We’re babes in the woods when it comes to this stuff.
How big would a self-replicating probe have to be? It has to carry fuel for an interstellar trip, along with fuel to land somewhere, and all the equipment needed to start high tech manufacturing on an unknown world.
It takes about 126,000 MWh equivalent to accelerate 1 kg to .1C. I assume a probe would be millions of kg including fuel.
It is very easy to state a set of necessary conditions for life (self-organizing, capable of reproducing via transmission of structured information, consumes energy and resources) but very difficult if not impossible to create a set of sufficient and exclusionary criteria that cannot be satisfied by some kind of artificial system that does not display evolutionary development or be broadly recognized as ‘alive’ in the same sense that life-as-we-know-it is. It is clear that a certainly level of complexity is required to self-organize and replicate, but these are also (presumably) emergent phenomena arising from a highly complex system that does not follow simple rules even if it is the result of a fundamentally a quasi-deterministic physical system.
The o.p. may find this talk by University College London biochemist Nick Lane informative as this is the area that Dr. Lane’s research lab is working in. At around the 35 minute mark it details research into spontaneously self-organizing chemical systems. If you want to skip to the summary you can go to 52:45 but I would encourage you to watch the entire talk because while Lane goes into pretty deep detail on the biochemistry of life he does so in an accessible way even if you don’t know anything about organic chemistry or molecular chemistry. His work, along with others working on the topic of the biochemical fundamentals of living systems, informs the viewpoint that biochemical life is likely to be quite common in the universe in any environment in which there are mildly polar liquid substrates and some form of a thermal or electrical energy gradient just because there are so many opportunities for spontaneous organization to occur.
Whether complex life, much less one with some form of volition or cognitive abilities, may develop is still anyones’ guess, but regardless of what arbitrary factor you apply to that probability it is still almost certain that it has not just occurred once in the history of the universe, or even within our own galaxy. And of course, there is the possibility of life that does not have a biochemical origin, or even (hypothetically) is not made from chemical elements, although the circumstances which would give rise to that are obviously speculative and would likely result in something we might not recognize as a living system even if we could observe it.
Stranger
I think the question can be stated as “what is the simplest form of life imaginable”, so let me introduce LUCA:
Whatever the simplest form of life imaginable might be, it was simpler than LUCA. But it was still pretty complex. Very complex, in fact. That is the reason why so many people still believe it cannot have arisen without some divine intervention (those people are wrong, of course, IMHO).
I have no idea where the article was, but years ago I read of sulfur bacteria at ocean vents. They live on a sulfur chemistry and need no light. There was some speculation that they could do this chemistry in some habitat in tiny cracks of “rocks” or whatever the inorganic solid material out there was. Hydrothermal vent microbial communities - Wikipedia
Possibly off-topic as not biological, but there are self-sustaining chemical systems. This is the most famous:
I don’t think it is off-topic for not being biological, as the title of the thread does not mention biology either, but just Darwinian evolution. But it just seems that Darwinian evolution implies some kind of biology, or at the very least something we could call pre-life or quasi-biotic if you agree, and that is the part the Belousov-Zhabotinsky reaction you linked to does not show. Though it is fascinating in itself, it does not change with time to adapt to the environment.
Not necessarily. “Darwinian evolution” is generally interpreted to mean genetic adaptation at a species level by selective pressures, a.k.a. “natural selection”, and you can observe such selection in free markets (which are absent of undue influences) or in games with emergent success strategies, and in fact “natural selection” is fundamentally a game theory mechanism which accounts for evolution in terms of competing pressures and influences which can be produced in any stable system with negative feedbacks. Whether “natural selection” is a comprehensive explanation for evolution is subject to debate although that is largely framed in terms of how expansive you allow the term to cover and whether cooperative co-evolution and the “extended phenotype” qualify, and also the role of epigenetic mechanisms upon gene expression; however, essentially all biologists conversant with evolution agree that there is no evidence of teleological or ‘directed’ plans in evolution, and in fact it is trivial to demonstrate how many blind alleys and sub-optimized paths are evident in evolutionary development.
The Belousov-Zhabotinsky reaction and other chemical oscillators are interesting in the sense of demonstrating that non-equilibrium thermodynamic systems can be stable without any kind of supernatural élan vital of uniquely special living tissues, but while they can create stable patterns they do not, as you note, transfer that information to any kind of offspring or reproduce those structures, nor are they subject to evolutionary adaptation. Such mechanisms might be a path to the spontaneous development of truly adaptive systems although that is far from clear, and there are many hypotheses for abiogenesis of life on Earth for which we will probably never have a definitive answer.
The definition of what a virus even is (which is touched on in your linked article) is subject to debate, but if you accept that the fundamental basis for life is that it has to be capable of self-reproduction (or in the case of sexual and parasexual species, reproduction by exchange of genetic material) then virions (viral particles) are not living organisms; they are at most ‘spores’ carrying genetic material and a mechanism to be uptaken by its targeted host cell. Some virologists define the ‘virus’ to be the infected host cell whose behavior (and in the cases of retroviruses, its actual genome) are modified to reproduce virions, but I think that is stretching to try to find some way to define a virus as living. However, viruses certainly demonstrate that it is possible to transmit genetic information between different organisms and reproduce that code (often with modification as viruses are subject to high rates of mutation and genetic drift) even by things that do not otherwise satisfy the broadly accepted criteria of ‘alive’ and therefore show that there is no special sauce that distinguishes one bit of sufficiently complex chemistry from another in terms of being subject to evolutionary pressure.
There are also prions (“proteinaceous infections particles”) that are most notably responsible for transmissible spongiform encephalopathies such as Creutzfeldt-Jakob disease, chronic wasting disease, and scrapie, which can be transmitted between organisms (typically by ingestion of contaminated tissue) and cause the progressive malformation of specific neural tissues by catalyzing conformal protein structure even without modification to the genome or epigenome. Prions aren’t alive by any definition since they are just long chains of peptides, and it is arguable whether reshaping conformal proteins is in any way transmitting information, but they are an example of a complex biomolecule reproducing its pattern within a living matrix even though it is not alive.
Stranger
FWIW, the original poster is most interested in non-biological or non-cellular examples self-sustaining chemical systems capable of Darwinian evolution. So very much on-topic.
Stranger: very interesting video. I was particularly struck by the claim that the basic biologic processes were conserved across all organisms (presumably including all Archaea). While this wasn’t exactly new to me, it’s striking that we haven’t observed something akin to convergent evolution, where various physical processes evolve in parallel. (I understand that similar chirality implies common ancestors, which is also intriguing).
That may be, or it might be the case that there is some as-yet-not-understood property of left-handed amino acids which make them preferred for living systems. and would be used even in the case of independently developed living systems and their descendants. We can be pretty confident that all extant life that we have so far found has a common origin at some point just because there are so many basic proteins that are shared, and are (presumed to be) too complex to have evolved repeatedly by coincidence, although it is also entirely possible that biochemically independent single-cell biotic or pre-biotic systems evolved in distinct spontaneous emergence and then coevolved together until they were essentially a single evolutionary branch of life in the same way that streams come together to form a river.
Stranger
Yes, but are you not introducing life through the back door? All those abiotic systems you mention are “played” by living beings (basically: us humans). Can you come up with an example were the entropy is “defeated” by other external, non-living actors?
The same objection can be made with viruses and prions: they may not be alive, but the mutations and the genetic drift happen in living cells, where the viruses multiply. Without cells to infect viruses do not reproduce or mutate, as far as I know.
Simulations use cellular autonata create and reproduce identifiable structure from basic rules and random perturbation without any input from a living user. Conway’s Game of Life is the most well-known example of this, which I think anyone would acknowledge is not actually alive. And this isn’t just a computer simulation phenomena; essentially all persistent turbulent systems such as tropical storms, flowfields around bluff bodies, et cetera, while evolve under non-equilibrium conditions to identifiable patterns even though turbulence is a fundamentally chaotic behavior that is not predictable at the individual particle level.
When you say “…entropy is ‘defeated’”, it isn’t clear what you mean. Living systems can no more ‘defeat’ entropy than any other non-living mechanism. Living systems will mediate and organize energy flows to do ‘useful’ work for self-organization but ultimately the entropy with any closed boundary per the Clausius inequality regardless of whether it involves a living system or not, and in fact living systems often increase entropy by dint of optimizing the use of available energy gradients.
Stranger
I see the input introduced in advance by way of the program, which is made by living humans. We put it there to start with. We are the deux ex machina. You can argue that the idea expressed in the program is platonic, eternal and independent from living beings, but I would not be convinced. Conways Game of Life would never work without living beings to build the computers to run it, program those computers, and put the energy in place to run them, even if the idea behind CGoL is independent of the idea being put into practice. Life through the back door again.
Entropy is “defeated” is not my best analogy, sorry. I mean that the information to transmit from one generation to the next, which is necessary to improve the system, so that it shows Darwinian evolution, which requires some kind of memory (RNA and DNA in living beings, computer hard disks for simulations) is something only living beings are capable of. Living beings do this by taking energy from outside the system (the sun, mostly, or chemical energy in hydrothermal vents) to reduce entropy locally, while still increasing it on the bigger system. When life “learns” to harvest external energy sources it can become wasteful and still thrive: plants, for instance, are much worse than PV modules when converting sun radiation into energy, but it is enough to sustain life on Earth. Can any non-living system harvest energy to improve with time apart from simulations or games that require living beings to set them up? I don’t know of any.
The evidence that all life on Earth is related isn’t just that we all use the same chirality, or the same proteins. We all use the same genetic code. In every known living thing, the codon CGU in RNA always codes for the amino acid arginine, and UGG is always tryptophan, and so on. There are a combinatorically-large number of possible genetic codes, and yet everything uses the same one.
That is, of course, physically impossible, which is why it’s not what the paper you linked suggests.
While it is true that computers, the operating system, and applications that run upon them are all artificial constructs, the rules in Conway’s Game of Life are objectively simple rules that in no way define a teleological path of development or enforce organization, and in fact every game with a unique set of inputs will come up with different systems of organization over sufficient time. Of course, these will not every develop sufficient complexity to become self-aware, or even escape the system they exist in, and the game is really just a very simple simulation of cellular automata, but systems that are identifiable as cellular automata are seen at all scales all over the universe, and may well be an organizing principle behind the abiogenesis of life in any medium that can develop adequate complexity for the emergence of energy-moderating systems. How likely this is to occur and whether it is really the correct explanation for the abiogenesis of life on Earth is unknown, and will likely remain unfalsifiable barring access to a time machine, but it is a credible hypothesis and Conway’s shows that systems with even extremely simple interactions can produce unexpectedly complicated patterns of behavior.
Really?
Now, a trio of U.S. researchers proposes a novel explanation. Today in Nature, they report that by monitoring the formation rates of amino acid pairs, called dipeptides, they’ve found multiple mechanisms that ultimately promote dipeptides whose two members share the same handedness.
“It’s quite convincing,” says Gerald Joyce, a pioneering origin of life chemist and president of the Salk Institute for Biological Studies who was not involved with the study. Researchers next hope to learn whether the same mechanisms skew larger peptides and proteins toward left-handedness—and whether it can explain the opposite bias in RNA and DNA, whose bases have sugars that are inevitably right-handed. If so, the new mechanisms could explain how life itself took on one mirror-image form and not the other.
…
One clue comes from recent work by Matthew Powner, an origin of life chemist at University College London, and his colleagues. Over the past 5 years, Powner’s group has discovered a set of sulfur-based molecules that likely would have been present on early Earth and shown how they readily link individual amino acids to amino acid precursors, called aminonitriles, forming dipeptides. Because these reactions take place in water and work with all the amino acids found in living organisms, they offer a plausible route to how the first proteins may have formed.Powner’s team didn’t check whether its sulfur-based catalysts had a chiral bias. That’s where Donna Blackmond, an origin of life chemist at Scripps Research, and her colleagues Min Deng and Jinhan Yu grabbed the baton. They tested two of Powner’s sulfur compounds to see whether the catalysts were sensitive to chirality as they formed dipeptides. They were, but not in the way Blackmond had expected. The catalysts created about four times as many “heterochiral” dipeptides—those pairing a left-handed amino acid (L) with a right-handed (D) one—as fully chiral products. “We thought it was bad news,” Blackmond says, because it suggested that even if amino acids on early Earth started with a bias, it would have been scrambled as proteins formed.
Stranger
Yes, really. That paper is proposing a mechanism which would cause all amino acids to be the same chirality. It’s not proposing a method that would favor one chirality over the other. Rather, a random fluctuation causes one to be slightly more common than the other, and then whichever one won that coinflip would take the lead and become dominant.