Mathematics Shows How to Ensure Evolution
quantamagazine.org
quantamagazine.org
I'm not sure how I can relate "self-loops" in my example.
Another way to understand this is to consider that you cannot distinguish "Darwinian" vs. "Lamarckian" in the evolution of single-celled microbes. Microbes don't grow old and die. Each Amoeba is billions of years old. (HeLa cells are called immortal.[1]) A change in the cytoplasm of an ancestor cell would be carried into both daughter cells.
In "Selfish Gene" Dawkins works out how death and birth have adaptive value for genes in multicellular species. The meta-evolutionary "choice" to reproduce by sex and children (rather than, say, budding, or metamorphosis, or some kind of mass mitosis) is "made" by the normal process of evolution, not some additional meta-process.
In short, there is no a priori reason to believe that the life of Earth has not evolved intelligence. (After all, we are intelligent.) What I mean is, the machinery of evolution itself will have evolved to be able to direct its own evolution, IF intelligently self-directed evolution is adaptively valuable.
If you want to prove that Nature is not intelligent you have to show that there is adaptive value to remaining "dumb". Given what we now know about adaptive systems, that's the only way Nature would have evolved without ambient intelligence: if it pays to act dumb.
[1] https://en.wikipedia.org/wiki/HeLa
- - - - -
edit to add: Trawling through the new links I see: https://news.ycombinator.com/item?id=17405096
> "Cells Talk and Help One Another via Tiny Tube Networks
Long-overlooked “tunneling nanotubes” and other bridges between cells act as conduits for sharing RNA, proteins or even whole organelles "
That's teh sort of thing I'm talking about: Life is one thing.
All unicellular life survives by growing (accumulating enough of the right chemical components) to replicate its genetic material (RNA or DNA), and then divide. Errors happen during duplication that provide mutations that natural selection can work on. In the general act of existing all these organism accrue broken proteins and waste that they cannot get rid of, these accumulate in one half of a dividing cell, which eventually results in one half of the separated cell failing.
In addition to replication errors environmental stresses can also just directly damage the genetic material which again provides mutations than selection can be applied to.
So yes we can distinguish between Darwinian selection and Lamarckism, as Lamarckism is not actually a thing - the evolution of single felled organisms is purely Darwinian, and can (and has) been extensively studied.
I suspect the misunderstanding here is that meiosis (or is it meitosis? I can never remember which is which) is only notionally a clone. Once you acknowledge that one of the two cells after a split is not necessarily identical to the original you can see a lineage of genetic variation that is the hallmark of evolution.
This is ignoring lateral gene transfer, but that’s essentially an effect of gene survival and statistics.
I've never heard that before. It seems strange to me because if the cell can accumulate its trash why not just throw it out? Why spend the energy to replicate the nucleus for a doomed cell? Can you provide a reference?
In any event, I'm not saying they don't die, I'm saying that every amoeba is a chemical process that has been unbroken for billions of years. The bag gets bigger and pinches in twain, over and over again, with nothing like the discontinuity between mother and child that we have as multicellular lifeforms.
Every amoeba alive today has been alive since... well the first cell, I guess.
> So yes we can distinguish between Darwinian selection and Lamarckism, as Lamarckism is not actually a thing
I'm talking specifically about the level of singled-celled organisms. We can distinguish between traits that have obvious correlations with DNA sequences, and traits that do not, but we cannot say with certainty that no cell has ever acquired a trait that it passed to its descendants sans DNA. In fact, there's a whole field of "epigenetics" opening up right now:
> Epigenetics is the study of heritable changes in gene function that do not involve changes in the DNA sequence.
~ https://en.wikipedia.org/wiki/Epigenetics
> the evolution of single felled organisms is purely Darwinian, and can (and has) been extensively studied.
Sure, "ignoring lateral gene transfer"... I wasn't going to mention that but it obviously complicates the traditional Darwinian dogma, doesn't it? You've got bacteria "inheriting" traits from each other!
But anyway, my point is that "the evolution of single-celled organisms [seems] purely Darwinian" because it evolved that way, not because of some intrinsic impossibility of Lamarck-ish "transmission of acquired traits".
Skipped LGT because it required multiple sentences being written on my phone. It is a misconception that bacteria /decide/ to take up or release plasmids into the environment. They don’t, it is a simple statistical process. F a bacteria is in an environment where there is a selective pressure that favors expressing a given gene (or group)then the bacteria that have those genes increases its proportional presence in the population pool. Now those bacteria have become dominant because of a particular set of genes, which means those genes are being expressed at a significant rate. Expressing the genes requires duplicating that section of the bacteria’s genome, and moving it of to the protein build molecules. So now we have an huge (proportional) amount of floating snippets of genetic material for the useful genes. That structure of these molecules tends to them wanting to connect their dangling ends, so many of them do just that (all expressed genes do can end up doing this, it is simply the selective pressure that increases the rate of expression).
Now as time goes by some of the bacteria will just die, and when their cell wall collapses any remaining genetic material just ends up floating in the environment. The most stable are those without dangling ends - the plasmids. The plasmids can also just escape through the wall generally - there is some ability for a bacteria to get rid of waste : ion pumps do the work, but there’s a limit to the size, which is why bacteria cannot simply get rid of all waste - the net result is that an environment in which a given set of genes is useful starts to accumulate plasmids that contain those genes.
For absorption of the plasmids, that is simply a product of how bacteria absorb all materials, so now the plasmid is inside another bacteria. In the event of any error by the when it cuts the plasmid it can end up including the plasmid material in its own genome
1. Meta-evolution happens as part of evolution, there's a logical "strange loop" there.
2. We see the kinds of macroscopic evolution among multicellular species that we do (such as the "choice" of having separate species, or the "choice" of having death-and-birth) not because it's somehow impossible to be otherwise, but because these modes of life have meta-evolutionary adaptive advantage.
When I contrast "Darwinian" vs. "Lamarckian" evolution I mean that the former does not have "inheritance of acquired traits" while the latter does. I might be using the term "Darwinian" in a non-standard way and I apologize if so.
When I'm talking about not being able to differentiate the two modes or kinds of evolution in the case of microorganisms, all I'm saying is that there's one process that has aspects of both modes.
Isn't what you just described about LGT obviously "Lamarckian" in the sense that a bacteria acquires a trait from "eating" a plasmid, incorporating it into its own genome, then dividing and passing the trait on to each offspring? I guess what I'm asking is, what do you mean by Darwinian?
Heads and that cell is healthy and the healthy cells win. Tails and that cell has cancer and is stillborn and cancer loses.
It's awesome that you're thinking about this stuff though. I think you'd get a lot out of this: https://www.khanacademy.org/science/high-school-biology/hs-r...
The mechanics of it, is like I said, the population bottleneck. Cancer cells are parasites and the bottleneck leaves them without a host.
A fascinating cancer is "Devil facial tumour disease". It has learned to spread from animal to animal, thus bypassing this purifying mechanism. But as it no longer consists of the animals' own cells it may be argued that it has become its own species and should no longer be considered a cancer. Supporting the speciation argument is the fact that the cancer cells even have a different amount of chromosomes from the animals.
This isn't a determinist "cancer will form or it won't" situation-- there's a lot of variance and luck. Furthermore many cancers form later in life, after people have already reproduced, so natural selection doesn't have that strong of a hold on it. If cancer worked the way you think it did, don't you think that cancer would have been naturally eradicated during our evolution?