Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.
I agree that random is a slippery term that's best avoided. However, it seems to me that as far as processes are concerned, there is no logical room left after we include deterministic and random factors. That makes random a synonym for nondeterministic. Am I missing something?
Yes - the way the term "random" is used, it is not synonymous with "non-deterministic". "Random" often implies "flat distribution", "directionless", or "equiprobable". As one example, consider the lottery where the numbers are drawn from bouncing balls. Is that process "random"? Certainly most people would agree that it generates random numbers, and yet the process itself might be deterministic.
Another example is suggested by your dictionary definition. Consider a measurement in a physics experiment, say of the length of some object. The length is not precisely defined due to quantum fluctuations, and therefore the result of the measurement is, strictly speaking, non-deterministic. And of course in addition there are the usual types of measurement errors, which may or may not be deterministic.
And yet, no one in their right mind would say that the results of measuring something with a tape measure are "random". They might say the errors are random (even though they may be deterministic), but they would never say the result is. Why not? Because the result has a non-zero mean and small fluctuations around that mean. We don't usually characterize distributions like that, distributions that are fairly sharply peaked around some non-zero value, as "random".
I presume we agree that a stochastic process is one that has some random generators.
"Stochastic" is close to synonymous with non-deterministic, but carries a slightly different connotation. In physics it's often used for systems in which the non-deterministic element is some kind of noise, thermal or otherwise. More generally it usually implies the presence of some kind of direction, force, or tendency, one that is perturbed but not destroyed by the noise. I'd say it's the best term to describe evolution.
Why don't you actually respond to evidence that I provided that "stochastic" and "random" are used as synonyms by the very people whom you insist do not use them as such?
The only true intellectual dishonesty here is your own continued willful ignorance of the fact that "stochastic" is used by scientist to mean what I have said it means.
Depends on how you define "algorithm". If you define it meaningfully let's say as "an effective method for solving a problem using a finite sequence of instructions." then no. It isn't.
Therefore, it is the exact opposite of completely random (as in dice-rolling) chance and "happy accidents".
Depends on how you define "algorithm". If you define it meaningfully let's say as "an effective method for solving a problem using a finite sequence of instructions." then no. It isn't.
Well, it is a farily finite sequence of steps, though in this case, it happens to occur naturally.
It would not be "solving a problem", except (perhaps) in how reproductive success could be abstracted as a "problem".
Snowflakes also build according to a naturally occuring algorithm, that is a finite series of steps, though it is more difficult to say that they are "solving a problem".
Yes, but the steps would not be random. That is what I mean.
The predictive power of evloution rests on the idea that the steps of natural selection are not random, even if some of the elements in it, such as the units it selects from, are.
Why don't you actually respond to evidence that I provided that "stochastic" and "random" are used as synonyms by the very people whom you insist do not use them as such?
The only true intellectual dishonesty here is your own continued willful ignorance of the fact that "stochastic" is used by scientist to mean what I have said it means.
1) If "random" means ""[o]f or relating to a type of circumstance or event that is described by a probability distribution", then all processes in the physical world are random, and the statement "Darwinian evolution is random" is a tautology.
2) The dictionary definitions you gave include another definition - that outcomes are equiprobable - which does not apply to evolution. Therefore if you insist on using it, you must qualify the term "random" to make it clear which definition you mean. If the options are 1) or 2), the statement "evolution is random" is either tautological or false.
3) Technical books may define "random variable" or "random process", but they rarely if ever define "random".
4) Every time a substantive issue arises in these threads, such as your false assertion that unfalsifiable statements cannot be confirmed, you lose the argument and resort to the same tactic: you retreat back to semantics. It's extremely boring, and I've had enough of it.
The issue here is that, even if an allele has a selective advantage, the odds against its eventually becoming fixed in a population are enormous. For instance, if an allele allows individuals possessing it to produce on average 1.000001 offspring for every offspring produced by individuals not possessing the allele, the probability that the allele* will become fixed is ~.000002, or 1 in 500000, hardly an efficient process.
For the distributions of selective advantages in natural populations see Nielsen (2003).
*despite being more likely to become fixed than the disadvantageous allele that allele allows individuals possessing it to produce on average .999999 offspring for every offspring produced by individuals not possessing the allele
For instance, if an allele allows individuals possessing it to produce on average 1.000001 offspring for every offspring produced by individuals not possessing the allele, the probability that the allele* will become fixed is ~.000002, or 1 in 500000, hardly an efficient process.
If the selective advantage of a mutation is tiny, obviously the odds any given instance of it will end up fixed are small, at least in a large population. But in a fixed population model taking into account mutation, the allele is certain to be fixed eventually (unless something alters the genome in such a way that it ceases to exist, or so that its rate of mutation goes to zero or its advantage vanishes).
If the selective advantage of a mutation is tiny, obviously the odds any given instance of it will end up fixed are small, at least in a large population. But in a fixed population model taking into account mutation, the allele is certain to be fixed eventually (unless something alters the genome in such a way that it ceases to exist, or so that its rate of mutation goes to zero or its advantage vanishes).
The results of that paper (which I'm familiar with) are fully consistent with what I said, and in fact with a small (and obvious) extension contradict what you asserted.
I take issue with several points in the following post.
Firstly, the bit where I think we have common ground:
I would say that over "moderate" timescales, with "sufficient" populations, in environments with constant section pressures, then the effect of any randomness would be minimised, or could be ignored.
This is the situation that is easiest to comprehend, and is used most often in discussing how evolution works. Unfortunately it is not appropriate when discussing what has happened in the history of life on Earth, which is of significant interest in the context of evolution. This is because over long enough timescales, catastrophic events completely alter the environment. Even if you ignored these, individual organisms are co-evolving, and constantly altering the environment for other organisms in the ecosystem
Again, we don't notice a lot of this because it is microscopic, but the "battle" between pathogens and immune systems is certainly *not* predictable, except in the pretty trivial observation that one or another will get the upper hand for a while (and this might sometimes wipe out one set of organisms).
I see your level of comprehension of this discussion hasn't changed at all.
"Evolution is predictable in the long run" - what does that mean to you? To me, it means that certain things can be reliably predicted in the long run (weird that it would mean that, huh?). That paper doesn't affect that conclusion.
For example: bacteria in petri dishes with some citrate. At the start of the experiment, none of the bacteria can metabolize citrate. The ability to metabolize citrate would be a strong advantage given the environment. So, let multiple colonies in multiple petri dishes live and reproduce for a long time.
Again: the claim at issue is that the theory of evolution by natural selection is predictive in the long term; that is, that it can predict some results of some experiments. So let's apply it to this. According to the theory, there is some non-zero probability per generation that a bacterium will mutate in such a way that it can metabolize citrate. Moreover, the theory tells us that that bacterium and its descendants will have an advantage in the sense that they will reproduce more rapidly than bacteria without that ability. More specifically, the theory predicts that eventually, all bacteria in that petri dish - and therefore eventually, in all petri dishes - should have the ability to metabolize citrate (modulo a few details, like that said ability doesn't come at a cost so significant it cancels out the benefit).
In principle the theory will also tell us how long we need to wait on average before that happens in a given petri dish, but to extract that prediction requires knowledge of the probability I mentioned, and that's hard to estimate (although one can certainly try). Regardless, the above is a definite, solid prediction for the long term, and we can test it.
So someone tested it. Guess what the result was (after a decade or so)?
"Evolution is predictable in the long run" - what does that mean to you? To me, it means that certain things can be reliably predicted in the long run (weird that it would mean that, huh?). That paper doesn't affect that conclusion.
But that is not what the paper is saying, the authors are saying that this experiment is support for the idea that if one could "rerun the tape of evolution" one would get (significantly) different outcomes. How is this "predictable"? The experiment also supports my point that evolution is less predictable over long timescales than short timescales, as there is more chance that an important random event will alter the outcomes. In the case of citrate+ metabolism, nothing interesting happened for 15,000 generations, and nothing *observable* happened for the first 30,000 generations.
More generally, we suggest that historical contingency is especially important when it facilitates the evolution of key innovations that are not easily evolved by gradual, cumulative selection.
At its core, evolution involves a profound tension between random and deterministic processes. Natural selection works systematically to adapt populations to their prevailing environments. However, selection requires heritable variation generated by random mutation, and even beneficial mutations may be lost by random drift. Moreover, random and deterministic
processes become intertwined over time such that future alternatives may be contingent on the prior history of an evolving population.
These accidents of history may even determine the survival or extinction of entire lineages, given the capricious and sudden nature of some environmental changes.
Stephen Jay Gould maintained that these historical contingencies make evolution largely unpredictable. Although each change on an evolutionary path has some causal relation to the circumstances in which it arose, outcomes must eventually depend on the details of long chains of antecedent states, small changes in which may have enormous long-term repercussions.
I read the bit below as being the position that you are taking.
Lenski's paper said:
Simon Conway Morris countered that natural selection constrains organisms to a relatively few highly adaptive options, so that ‘‘the evolutionary routes are many, but the destinations are limited’’ (16). He and others point to numerous examples of convergent evolution as evidence that selection finds
the same adaptations despite the vagaries of history. Evolution may thus be broadly repeatable, and multiple replays would reveal striking similarities in important features, with contingency mostly confined to minor details
Again: the claim at issue is that the theory of evolution by natural selection is predictive in the long term; that is, that it can predict some results of some experiments. So let's apply it to this. According to the theory, there is some non-zero probability per generation that a bacterium will mutate in such a way that it can metabolize citrate. Moreover, the theory tells us that that bacterium and its descendants will have an advantage in the sense that they will reproduce more rapidly than bacteria without that ability. More specifically, the theory predicts that eventually, all bacteria in that petri dish - and therefore eventually, in all petri dishes - should have the ability to metabolize citrate (modulo a few
details, like that said ability doesn't come at a cost so significant it cancels
out the benefit).
Citrate+ metabolism was (virtually) unknown in E.Coli, so much so that the inability to metabolise citrate+ was considered to be an identifying feature of E.Coli. This was a very rare event, in fact the paper calculates the upper bound for this mutation rate:
Lenski's paper said:
With no more than three mutations among the 8.4 x 1012 cells tested here and in the third replay experiment, the upper bound on the ancestral mutation rate to Cit+ is 3.6 x 10-13 per cell per generation (Fig. 4). To the best of our knowledge, this value is the lowest upper bound ever reported for a mutation rate that has been experimentally measured. It is also probably far too high because no mutations were actually observed for the ancestor, nor were any found among another 9.0 x 1012 cells of 60 clones sampled through 15,000 generations; and because some cell turnover and other DNA activity probably occurred during the many days that plates were incubated.
You are missing out the fact that *eventually* this trait would evolve, if the population hadn't been wiped out in the meantime, or the environment hadn't changed, or if another incompatible trait hadn't evolved first. This is also a very simple ecosystem. With more different types of interactions, and with longer time, there would be more scope for more random events to alter the course of evolution.
As soon as you are talking about real organisms outside laboratory environments, you can make predictions, but over sufficiently long timescales apart from "the organsims will be adapted to their environment" the predictions won't be valid. Not because of any inaccuracy in your initial data, but because subsequent random events will have affected the course of evolution, and the whole ecosystem.
If the selective advantage of a mutation is tiny, obviously the odds any given instance of it will end up fixed are small, at least in a large population. But in a fixed population model taking into account mutation, the allele is certain to be fixed eventually (unless something alters the genome in such a way that it ceases to exist, or so that its rate of mutation goes to zero or its advantage vanishes).
But these are nontrivial qualifying statements. Only a tiny minority of any organisms successfully reproduce. Most organisms that have lived have no living descendants. Most species that have existed have no living descendants.
The issue here is that, even if an allele has a selective advantage, the odds against its eventually becoming fixed in a population are enormous. For instance, if an allele allows individuals possessing it to produce on average 1.000001 offspring for every offspring produced by individuals not possessing the allele, the probability that the allele* will become fixed is ~.000002, or 1 in 500000, hardly an efficient process.
You are not taking several factors into consideration. Your probabilities do not work in isolation. They work against the backdrop of other evolving entities in the fitness landscape. Your sources seem to forget that. (And, so do most Creationists who cite these things.)
Don't forget: Members of species compete against each other for resources. A small advantage, for example, in picking berries, grants that member access to a lot more berries than anyone else. The entity's chances of finding a good mate increase accordingly.
This is more dramatic if you consider even the slightest advantage in avoiding stealthy predators. Any tiny improvement, there, could yeild a tremendous chance of becoming fixed in the population.
Granted, the probability might not be exactly 1, as Sol claimed. But, it does turn out to be fairly high.
The ULTIMATE, broader causes and effects of evolution would be the same, if you reran the the tape of evolution. In a broad sense, you would likely get organisms that fill relatively similar niches in relatively similar manners.
There are just sooo many different ways one can fill a niche, and physics restricts the sorts of niches we would find in any given ecosystem.
However, this implies that the PROXIMATE, more specific causes and effects will likely be very different. (Though, we would also expect a certain amount of "coincidental convergence" of some superficial features, too.)
If we looked at the life forms of an alternative "run" of evolution, we would find that they would look very different from what we are used to, but will probably function and behave, in a broad sense, in recognizable ways.
For example, host/parasite relationships would probably work the same. Though, the existence of a parasite might not be recognized until it is spotted moving to a new host. (As is usual for novel parasites, even on Earth.)
Given enough time, multicellular life would likely develop, since its occurrence follows from host/parasite and other related models. Though, the exact manner in which it happens could differ, as we see in the variety of ways that sponges work.
I would disagree with you on the "general function" argument.
I would say that the evolution of grasses was one of the most significant events of the last 100-million years. There were whole niches that could not exist before then. There could be no niche for "herds of wildebeest" in the Permian Period.
Without Savannah or Tundra, whole ecosystems would be grossly different.
As to multicellular life, well we only have one planet to check on but it seems that life has existed for about 3-3.6-billion years (depending on interpretation of fossils). Multicellular life has only existed for about half a billion years. Using dodgy reasoning (as used by xenobiologists), I would say that we might assume that on average life might take 3-billion years (3.6-0.5) for multicellular life to evolve on an Earthlike planet. If this looks anything a Poisson curve, then you would see a significant probability (say 20%) that multicellular life wouldn't evolve on an Earthlike planet within 6-billion years, which is pretty close to the length of time that such a planet is actually habitable.
You are not taking several factors into consideration. Your probabilities do not work in isolation. They work against the backdrop of other evolving entities in the fitness landscape. Your sources seem to forget that. (And, so do most Creationists who cite these things.)
Don't forget: Members of species compete against each other for resources. A small advantage, for example, in picking berries, grants that member access to a lot more berries than anyone else. The entity's chances of finding a good mate increase accordingly.
This is more dramatic if you consider even the slightest advantage in avoiding stealthy predators. Any tiny improvement, there, could yeild a tremendous chance of becoming fixed in the population.
The fixation probability you're referring to is the probability that if a mutation to an advantageous allele occurs once, it will fix. That is completely different from the probability that said allele will eventually fix, because mutations are guaranteed (with the caveats I already mentioned) to bring it about over and over again until that happens. As someone that has evidently "thought" about these issues for years, if you cannot understand that I very much doubt I can help you - it's completely obvious.
Incidentally this fact was essential to the exchange we just had about citrate etc., so you evidently didn't understand that discussion either. It's no wonder you can't see why it keeps being brought up.
But that is not what the paper is saying, the authors are saying that this experiment is support for the idea that if one could "rerun the tape of evolution" one would get (significantly) different outcomes. How is this "predictable"?
I explained precisely how, and gave an example of an experiment that verifies that prediction.
You are missing out the fact that *eventually* this trait would evolve, if the population hadn't been wiped out in the meantime, or the environment hadn't changed, or if another incompatible trait hadn't evolved first.
I didn't miss that - I already commented on it, in fact.
This is also a very simple ecosystem. With more different types of interactions, and with longer time, there would be more scope for more random events to alter the course of evolution.
So what? Every scientific theory in the history of human thought is subject to exactly the same criticism. They are all simplifications, models for reality that do not take every possible event into account. They tell us what will happen under certain limited and controlled conditions, not simply what will happen.
But these are nontrivial qualifying statements. Only a tiny minority of any organisms successfully reproduce. Most organisms that have lived have no living descendants. Most species that have existed have no living descendants.
The fixation probability you're referring to is the probability that if a mutation to an advantageous allele occurs once, it will fix. That is completely different from the probability that said allele will eventually fix, because mutations are guaranteed (with the caveats I already mentioned) to bring it about over and over again until that happens. As someone that has evidently "thought" about these issues for years, if you cannot understand that I very much doubt I can help you - it's completely obvious.
Incidentally this fact was essential to the exchange we just had about citrate etc., so you evidently didn't understand that discussion either. It's no wonder you can't see why it keeps being brought up.
Even if the a mutation is guaranteed to occur there is still a non-zero probability that the mutation will never fix in the population:
[LATEX]1-(1-u)^{N_\mu}[/LATEX]
where u is the ultimate fixation probability of a single mutation and Nμ is the number of time that the mutation occurr de novo.
By the way, you still have not explained how the observation of fixed deleterious mutations supports your hypothesis that eventually all advantageous mutations will go to fixation.
It's interesting that you're able to parrot these results while evidently completely failing to understand what they mean... that equation confirms what I just told you (and by the way, it follows trivially from elementary probability theory).
By the way, you still have not explained how the observation of fixed deleterious mutations supports your hypothesis that eventually all advantageous mutations will go to fixation.
It's interesting that you're able to parrot these results while evidently completely failing to understand what they mean... that equation confirms what I just told you (and by the way, it follows trivially from elementary probability theory).
You assertion is only true if the mutation an infinite number of times. If the mutation only occurs a finite number of times, there will still be a non-zero probability that the mutation will go extinct.
I provided you with a paper that documented the existence of fixed deleterious mutation in a population, a phenomenon that contradicts your assertion that beneficial always, eventually go to fixation.
You assertion is only true if the mutation an infinite number of times. If the mutation only occurs a finite number of times, there will still be a non-zero probability that the mutation will go extinct.
Which is exactly what I've been telling you for the last 15 posts. Good to see you've finally caught on.
I provided you with a paper that documented the existence of fixed deleterious mutation in a population, a phenomenon that contradicts your assertion that beneficial always, eventually go to fixation.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
