Doesn't Natural Selection lead to Specified Complexity?

Perhaps I now understand where my misunderstanding lies -- thankyou for that explanation.I was focussing on adaptation but thinking of it as gene distribution within a population -- I know understand them to be different things.I would still expect there to be some non-random element to drift though ... surely a purely random set of mutations would have no trend.Oh wait though ... if population size was very small and then expanded I suppose that would have an effect ... but then only if the individuals with a specific non-adaptive mutation were the most prevalent breeders -- and there should be some reason for that.By ignore drift I meant 'in the context of natural selection's capability for designing apparently specified entities' not ignore it as though it didn't exist.I think the focuss on drift has led to a drift away from specified complexity NOT being a marker for intelligent design at all.

That depends what you mean by trend.Consider the simplest case where you have two alleles a₁ and a₂ of some gene, both equally fit, with respective frequencies f₁ and f₂ in the population, such that f₁ + f₂ = 1.Will these frequencies stay the same over time? They will not. Instead they will undergo what mathematicians call a "random walk", so that the graph of f₁ against time will fluctuate up and down randomly.Or rather, it will do so until either f₁ = 0 or f₁ = 1, when it will stick in that position, the allele a₁ having been either eliminated from or fixed in the population. And it is certain (i.e. having probability 1) that it will end up doing one or the other. The only question is which. In fact it is the case that at any given time in the process the probability that a₁ will go on to achieve fixation is just equal to the value of f₁ at that time.So there is no underlying trend for the frequency to drift one way rather than the other, but because it is subject to random fluctuation it will drift one way or the other, and end up at one of the two extremes.

quote:

---

I think you are confusing natural selection with evolution as a whole.

---

I see no reason for you to think that.

quote:

---

Since I am interested in the specifying capability of natural selection, and since drift can be viewed as 'chance' then for the purposes of THIS discussion we can neglect drift.

---

You went beyond that with your claim that natural selection was deterministic, which required you to mix drift in with natural selection. That made it relevant.

quote:

---

Check back -- an answer you gave to one of my questions means that you acknowledge the determinism of natural selection

---

More correctly the determinism of natural selection plus drift....

quote:

---

Your third option for drift being 'purely by chance' is insufficient for explanation.

​

If a particular mutation becomes prevalent then there WILL be a reason for that.

---

Essentially the reason is the statistical inevitability that SOME neutral mutations will eventually achieve fixation, while most vanish. Simply calling it "insufficient" is like arguing that the lottery is rigged because nobody could win by "chance".

It is the determinism of NS that makes the difference in evolution.Given the SAME set of mutations, but a different set of selective pressures the result will be different -- regardless of drift.So it is the selection that specifies the outcome.I don't remember mixing drift in with NS -- I added it as a kind of offset.Why is it a statistical inevitability that some non-selectable mutations will become fixed?If a mutation is purely random and not subject to selective pressure, why would it appear more than others?

I think I see what I have been missing -- thankyou.I was missing the fairly obvious point that the variants of a particular allele are mutually exclusive within a single individual.

quote:

---

It is the determinism of NS that makes the difference in evolution.

---

Again, NS is better conceived as a statistical bias than inevitable determinism, for the reasons I have already explained,

quote:

---

Given the SAME set of mutations, but a different set of selective pressures the result will be different -- regardless of drift.

---

Not necessarily. It is very unlikely, but drift could in principle produce the same outcome. How unlikely depends on the relationship between the mutations and the selective pressures.

quote:

---

I don't remember mixing drift in with NS -- I added it as a kind of offset.

---

You do it implicitly every time you say that NS is deterministic - and when you defined the environment to include ALL events, even those which have no selective effect.

quote:

---

Why is it a statistical inevitability that some non-selectable mutations will become fixed?

---

Dr. Adequate has already explained it. See Message 62 above.

Saw and replied to Dr. Addequate's post -- get that bit now.If drift could cause the same outcome, despite selective pressure then natural selection is a non-starter as an explanation for diversification in biological systems.I cannot implicitly include drift in natural selection while I am explicitly excluding it.I have also said that I segregate genetic mutation from all other events -- thus they are not included in my concept of environment -- they are included in my concept of genetic mutation.

quote:

---

If drift could cause the same outcome, despite selective pressure then natural selection is a non-starter as an explanation for diversification in biological systems.

---

You're wrong again. Just because drift COULD explain it, it doesn't mean that drift is a good explanation. Drift is not a good explanation for adaptions (but it is a warning against assuming that a feature is an adaption). We need selection to adequately explain what we observe, because drift is just too damned unlikely to come up with anything useful.

quote:

---

I cannot implicitly include drift in natural selection while I am explicitly excluding it.

---

Indeed, you should not. That is why I object to you doing so.

quote:

---

I have also said that I segregate genetic mutation from all other events -- thus they are not included in my concept of environment -- they are included in my concept of genetic mutation.

---

I strongly suggest that you drop the idea of mutations directly causing drift to any significant extent - it's been explained to you why that is wrong often enough, by myself and Dr. Adequate.

My understanding of neutral theory and genetic drift is that they are driven by neutral and nearly neutral mutations, so I don't see how there couldn't be a direct relationship between mutations and drift. What am I missing?--Percy

Percy, the appearance of mutations is required for drift. The spread of those mutations, however is controlled more by reproductive success rather than the same mutation independently occurring again and again.

Right, but it was suggested that Peter drop the idea that mutations directly cause drift, and even taking into account the italicized word it's hard to see why you feel this way. I thought maybe you were trying to emphasize that drift is the way that existing neutral or nearly neutral alleles change frequency in a population, as opposed to the appearance of new alleles through mutation, but I wasn't sure and I'm still not sure.Also, while it's true that the spread of mutations is controlled by reproductive success, by definition the mutations responsible for drift have little to no impact on that. Neutral and nearly neutral alleles can be linked to other alleles that are not neutral (I was just over at Wikipedia, turns out there's a term for this, genetic draft), but drift alleles do not exercise any significant influence over reproductive success themselves. And once any alleles find themselves in circumstances where they do impact reproductive success then by definition they're no longer drift alleles.I'm not yet familiar with Peter's level of understanding, but if the distinction you're trying to draw has merit it may be at too detailed a level to be worthwhile in this discussion.--Percy

I think if you look back over the posts it should be very clear that I'm talking about the spread (or decline) of mutations after they appear. Just as with selection, mutations are the driver, but not the immediate cause of the actual frequency change (which is differential reproductive success).

I don't really see this. Genes that have a selective benefit or detriment can still be subject to genetic drift, otherwise all beneficial mutations would be maintained in populations once they arose, which would make the effects of selection for a given allele in a given environment truly deterministic. The stochastic forces that produce genetic drift can affect any allele.Drift is an effect of essentially random sampling with regard to allelic fitness. The tack you are taking seems to take the contextual nature of mutations to a ludicrous extreme such that the fitness of any allele would be determined by the specific situation of the individual organism carrying it. So any allele of an organism that died before reproducing would have a neutral fitness. But these are population genetics concepts and they only really work in terms of populations.There are no 'mutations responsible for drift', drift is not a mutational phenomenon it is a phenomenon of random sampling. Mutation simply creates the alleles, whatever their fitness, upon which drift and selection operate.TTFN,WK

Hi WK,I'm sorry I can't say things the way you'd like, but I can't help believing you do actually understand what I really mean most of the time. I'm not going to respond, it would be a distraction from the topic anyway.My point was that it is misleading at best to just state that the idea that mutations directly cause drift is wrong. The obvious interpretation is that this is a claim that mutations should be disregarded as a factor in drift. If there's more context elsewhere so that Peter knows and understands what was actually meant then that's great. For example, if Peter saw what PaulK wrote and understood it as what you say here then that's great: But in many cases people don't distinguish between alleles and mutations, it often being the case that the term allele isn't familiar and so the term mutation stands in for allele, and while I don't know if that is the case here, what I fear is that Peter is being subjected to this same sort of nit-picky treatment where something stated insufficiently precisely from your point of view is criticized as if it were wrong.There's a guy in my group at work who is brilliant, true genius level, but he demands an impossible level of precision in all technical communication and is widely despised.--Percy

There aren't any alleles responsible for drift either. My whole point was that the fitness of any given allele is incidental to drift, drift occurs regardless of the fitness of alleles. Drift and selection are two separate factors that can work independently, in accord or in competition to alter the frequency of alleles in a population.If you are trying to say that genetic variation, i.e. more than one allele, is required for drift or selection to operate then of course I agree. In a perfectly homogeneous population there will be no effect from drift or selection and a de novo mutation will be required before they can operate. Is that your point? I'm not that smart but I'll try my best to compete, 'never knowingly under nit-picked' that's my motto.TTFN,WK