Print Story I've gotten my books
Diary
By ucblockhead (Sat Oct 04, 2008 at 08:38:59 PM EST) (all tags)
Also, Natural Selection


The Tao is infinite, eternal.
Why is it eternal?
It was never born;
thus it can never die.
Why is it infinite?
It has no desires for itself;
thus it is present for all beings.

The Master stays behind;
that is why she is ahead.
She is detached from all things;
that is why she is one with them.
Because she has let go of herself,
she is perfectly fulfilled.


I got my books for the Books4Obama  thing.  The results were one copy of Summerland, which I have not read, and nine books by people I have never heard of.  I'll read that, at least, though the "young adult novel" thing is always a turn off.  Of the other ten, one is a kids book that the FoML pronounced good, and there are two others that look potentially interesting.  So not worth the money, though that was hardly the point, obviously.

There's some that we'll likely give to a library or something.  They are mostly thrillers, which neither my wife or I have any interest in.


The recent diary by TheosophileEscargot has gotten me again thinking about a subject that always annoys me, that is, the misunderstandings and fallacies people have about Darwin's Theory of Natural Selection.  These are not particularly related to anything anyone posted in that diary.  It just got me thinking.  These are the errors I often see people make, even (and especially) people who are not "creationists":

Evolution is not directed

Evolution is not an example of progress from simpler forms to more complex forms.  This can happen, sure, but we can also see more simpler forms replace more complex ones. 

You very often see notions of some ultimate goal to evolution in SF, for example, notions of evolving into some higher form of being.  (Bad Star Trek novels and even good SF like Childhoods End.)  This is entirely inconsistent with the actual theory of evolution by natural selection.

Darwin did not discover "Evolution"

The idea that animals evolved from earlier forms was not knew to Darwin, but rather, was nearly the conventional wisdom among biologists before he came along, based on the very obvious similarities between the physical structures of different species.  Darwin didn't discover it.  What he did was produce a coherent theory that explained how it worked.  His contribution is not the "Evolution" part but the "Natural Selection" part.

Natural Selection does not produce creatures perfectly adapted to their environment

One of the very real problems you deal with when you write genetic algorithms on a computer is "local minima".  That is, the algorithm slowly makes improvements until it cannot make minor improvements that help matters.  If there is a large gap between an "ok" result and a "great" result, the algorithm can easily find itself stuck in the "ok" result because all the minor improvements are worse.

The situation in actual biology is the same.  If it would take a long chain of mutations to reach a better result, and the links in that chain are all worse results, then the species will not evolve in that direction.

As an example: there are many, many bird and reptile species with green skin.  Clearly it is advantageous to have green skin in certain environments.  No mammals have green skin?  Why is this?  It is because mammalian pigments are all based on melanin and there are no simple mutations that can be made to that that will produce green.  So just because it is advantageous does not mean that it will inevitably appear.

Not every feature of an organism must have a reason

There is a tendency to say "well, the creature has X, so because of natural selection, X must confer some advantage.  This is an error.  It is perfectly possible for a feature to have no benefit at all.  As long as it confers no disadvantage, natural selection will not remove it.

For instance: why do men have nipples?  The most likely answer is that because of pure random chance the mutation that led to them was not sex linked and since they confer no disadvantage, there is no evolutionary pressure to remove them.

(This is what Stephen Jay Gould called a "spandral".  This is a feature that has no benefit to the organism but exists merely as an offshoot of another, selected for feature.  These can form the basis for future traits that may have a benefit, but do not themselves have any benefit.)

Evolution is slow

To an evolutionary biologist, a significant trait appearing in a thousand generations is monstrously fast.  (Fast enough to be considered controversial in theories like "Punctuated evolution".)  Humans, for instance, are almost certainly only partially adapted to the rigors of agricultural civilization.  The idea that we could observe evolutionary changes in humans over the last century, or even the last millennium, is ludicrous.

Environments change constantly

There is a tendency to look at a creature in its environment and say "well, natural selection says it must be well adapted to the environment we find it in".  But environments are constantly changing, sometimes dramatically nearly overnight.  Some species we find may be poorly adapted and on the way out.  (For instance, there's some evidence that the cheetah would be headed for extinction regardless of human presence.)

Evolution is not a story of superior forms taking over from inferior ones

By most objective standards insects are more successful than mammals.  Our own family, the great apes, was barely holding on before people figured out how to bash things with rocks.  There is strong evidence that the human race itself almost went extinct, having been reduced to only a thousand or so members.


That isn't very well organized, but I'm done.

< Public Servants and the Free Market | October 9 >
I've gotten my books | 20 comments (20 topical, 0 hidden) | Trackback
Link? by ammoniacal (2.00 / 0) #1 Sun Oct 05, 2008 at 09:22:33 AM EST
to discussion of this cheetah-like, human near-extinction event.

This coomenat has be n soidnsord by hurricanbe ice malt liqur


Toba by yicky yacky (4.00 / 1) #3 Sun Oct 05, 2008 at 11:19:27 AM EST

is one of the ones that's frequently mentioned, but the jury is still well and truly out on whether it was a bottleneck event or not. Much of the evidence for human population bottlenecks comes from mitochondrial DNA and the Y chromosome but, as these represent a relatively tiny part of the entire genome, it's debatable as to how valid extrapolations from this data are.


----
Done.
[ Parent ]

Evolution by ucblockhead (4.00 / 1) #9 Sun Oct 05, 2008 at 01:14:14 PM EST
The idea is compelling, because given that the human species was marginal before the event and massively successful after, it is easy to assume that the severe selective pressure of this event caused some very fundamental changes in the human species.
----
ウセーバラケダ
[ Parent ]

Evolution is slow by wumpus (4.00 / 1) #2 Sun Oct 05, 2008 at 10:04:41 AM EST
Depends on the traits selected for.  I am willing to bet there was a huge jump in smallpox survival genes in Indians just during the 16th century (and to a lesser extent, black plague resistance in Europeans in the 14th).  Sexual selection can go pretty fast as well.  Note that all of these are "graded" pass/fail while "adapting to agriculture" is not.

Wumpus





More than one time-scale by Alan Crowe (2.00 / 0) #4 Sun Oct 05, 2008 at 11:42:18 AM EST
You can select fruit flies for more bristles on their wings and it goes well at first. 10, 12, 15,... more and more, if I remember correctly you can eventually get flies with over a hundred bristles. Then you find that they are a pure breed. You get degenerates with fewer bristles, but you never get heritable variation towards more. You have used up the stock of variation already in place and must sit twiddling your thumbs, generation after generation, waiting for mutations.

I think this is the point at which researchers irraditate their fruit flies to try and get more mutations and speed things up. Irradiation used to be a popular idea in crop breeding, but I think that genetic engineering has supplanted it.

So I think there is both a long-march time scale with major changes taking millions of years and a sprint time scale with heavy selective pressure able to bring about rapid changes in some features but not others.

This could matter a lot. Think about women liking sex and liking children. Nature hasn't had to bother about the distinction between liking sex and liking children. If you liked sex, the children came along soon enough. Sheltered from selective pressure, one anticipates wide variation in how much women like children.

Along comes contraception and the rules of the game change suddenly. Women have more choice. Some remain childless. Some stop at one. Some long for number five and six. Women who like sex but don't like children will drop out of the gene pool. I think it is extremely hard to say whether this will happen quickly, in thousands of years, or slowly, in hundreds of thousands of years.

My prediction is that children in the future will have much happier childhoods because they will be more definitely wanted, but I don't know whether to date this to 5000 A.D. or 500000 A.D.

[ Parent ]

Time scales by ucblockhead (2.00 / 0) #7 Sun Oct 05, 2008 at 12:46:06 PM EST
The time scale in question isn't "years" but "generations".  That is, evolution is "slow" for both bacteria and humans, but for bacteria, a single human can observe a thousand generations.  The point I am trying to make is that the "sprint" timescale sees major changes over a thousand generations, not two.  This might be a year for bacteria and 25,000 for humans.

My prediction for the future is that technology and societal change will completely change things to the extent that we can't really say what humans will be like.
----
ウセーバラケダ
[ Parent ]

nipples by LilFlightTest (2.00 / 0) #5 Sun Oct 05, 2008 at 12:28:06 PM EST
from a fetal development standpoint, the structures in the human body are analogous between males and females. There's a "default" that's used up til a certain point, then sex determination kicks in. as you said, males have nipples because they're included in the default, which tends to swing more toward female. then non-sex-specific things become sex specific. the tissues are the same, just their degree (and direction) of development varies with sex.

---------
if de-virgination results in me being able to birth hammerhead sharks, SIGN ME UP!!! --misslake


Yep. by iGrrrl (4.00 / 1) #16 Mon Oct 06, 2008 at 02:46:30 PM EST
 Basically, it's "cheaper" to do it that way. For example, one there's a mechanism for creating a vascular system. There aren't specific genetic programs for putting the arteries and veins in an arm vs. a leg. Tumors highjack this system and get a nice blood supply...

"I don't have time for martial law, I have to get to the gym!" zarathus
[ Parent ]

i love biology by LilFlightTest (2.00 / 0) #17 Mon Oct 06, 2008 at 04:05:31 PM EST
i still love it, even though i no longer do it for a living.

---------
if de-virgination results in me being able to birth hammerhead sharks, SIGN ME UP!!! --misslake
[ Parent ]

Me too by ucblockhead (2.00 / 0) #19 Thu Oct 09, 2008 at 11:24:59 AM EST
I sometimes wish I'd gone into that instead of rushing out into the computer industry.
----
ウセーバラケダ
[ Parent ]

funny that you say that by LilFlightTest (2.00 / 0) #20 Thu Oct 09, 2008 at 06:35:14 PM EST
i'm in IT because i couldn't find a bio job (at least not a bio job i wouldn't hate).

---------
if de-virgination results in me being able to birth hammerhead sharks, SIGN ME UP!!! --misslake
[ Parent ]

Hmmmm by yicky yacky (2.00 / 0) #6 Sun Oct 05, 2008 at 12:29:26 PM EST

Evolution is not directed

This is still a subject under a great deal of debate. You're right in that it's not the case that any given organism is being explicitly "aimed-at", or that evolution mandates a given and predictable result (although I guess we could devolve into philosophical quagmires on determinism at this point). On the other hand, it's not true in the sense that evolution isn't entirely directionless. There are only certain and finite options for the potential development of any given genome and these lines can't be broken-out-of easily; on an information-theory level, the set of future paths is bounded (and many of these are detrimental). This means that it is "directed" in some nebulous sense, but it's not the sense that most people mean when they use the term (an auteur's existence isn't implicit, for example).

On top of this, you have the current arguments over clade selection, the "evolution of evolvability" (some organisms' physiology allowing for a larger tree of future pathways) and the idea of complex systems providing a larger palette of hooks for future evolution (again, with many branches being detrimental). There's an interesting discussion between PZ Myers and Dawkins which touches on some of these.

Some theorists also take the view that evolution is directed in a more concrete sense very seriously (in the sense of a kind of anti-entropy, not that it's being 'conducted'). Robert Wright believes that, given enough time, resources and competition there is an innate bias towards complexity and even intelligence inherent to evolution. Dennett disagrees with him; interestingly not on the existence of the bias, but on the notion that there's any guarantee of "getting there"; one of the points being that the bias may exist for seemingly relatively stable ecosystems, but there's nothing inherently stable about evolution (the evolution of one microbe has the potential to destroy the most complex civilisations etc.). The interesting part is that, for the bias to exist, it seems as if there has to be a certain amount of instability in the system, but not too much. Too much instability means that the basis for complex systems never arrives as the system is too dynamic for a given strategy to gain sufficient advantage. Entirely stable systems (which are arguably theoretically impossible, although we have examples that are "fairly close") don't seem to facilitate the bias either, as there's little to motivate change. The "biasists" think that, over time, a system can't help but end up within these bounds. [Dennett vs. Wright]

There's also a time component to these views. Whether an advantage turns out to be beneficial is dependent on what the ecosystem around it is doing. Some studies have shown that "being first" is better than "being best" in "evolution of evolvability" terms. An adaptation with the potential for a bright and diverse future may well never get to see it as it's out-competed in the short term by an adaptation with less potential for future complexity but a better immediate-term advantage (this being higher-dimensionally analogous to the point about local maxima versus absolute maxima). The "biasists" don't dispute this, but assert that, over time, a sufficient number will get it right, given enough resources.

So the issue of whether evolution is directed or not depends very much on your definition for the term. I've noticed a reluctance among some to acknowledge the more subtle and systemic definitions for fear of giving certain factions the wrong impression. The truth is that evolution is feedbacky as all hell; and feedback's a bitch to quantify correctly.

Evolution is slow

This simply isn't true. Evolution can be fast or slow as circumstances and adaptations dictate, and depending on the scale at which an adaptation is being examined. The human eyeball didn't develop quickly, as it's a complex aggregation of equally complex subsystems, but when you look at smaller-scale evolution, it can happen very fast. Microbes and molds that can metabolise human-synthetic-materials have been discovered in the last ten years. This is significant as these materials never occurred within nature until we made them, and within 50 years nature has adapted to take advantage of them. That's pretty darned fast; scary, in some sense.

Evolution is not a story of superior forms taking over from inferior ones

This is almost semantically meaningless as it depends entirely on the definition of "superior" and "inferior" in question. If we're using the selfish gene analogy, in terms of raw numbers, insects are superior to humans; in terms of the probability of a given organism passing its genes on, we're superior to insects. It's this duality that's at the heart of the "directed / undirected", "superior / inferior" debate.


----
Done.


Stuff by ucblockhead (2.00 / 0) #8 Sun Oct 05, 2008 at 01:11:06 PM EST
As I said in another post, "slow" is in terms of generations, not years.  50 years for microbes is equivalent to 50,000 years for humans.

You are completely correct, of course, about the semantic meaningless of "superior" and "inferior".  That goes to the heart of why it is a misconception.    The misconception being that there is a "ladder of life" with humans at the top and slime molds at the bottom.

Whether an advantage turns out to be beneficial is dependent on what the ecosystem around it is doing.

This gets back to why the notion of species adapting to a static environment is so completely wrong.  The bulk of the "environment" that a creature adapts to is made up of all the other creatures in the world that it competes with.  But since those creatures are also adapting, that environment is itself constantly changing.  You can get feedback loops, yes, but also cycles, and more complex patterns.

"Being first" and "being best" are, I think, both complete misnomers because they imply that it is a race that ends.  Instead, it is a never ending battle.  It is more accurate to say that the entire ecosystem goes through periods of chaos in which everything is changing rapidly and periods of stability, where things stay relatively static.

In terms of complexity, I think it is merely a matter of a high water mark.  If you've got a system in random flux, the high water mark will slowly increase.  We certainly see simpler forms holding their own in many, many niches.

"Intelligence" is a sticky issue.  In general, it doesn't seem to me that intelligence is that much of a benefit.  We certainly don't see "intelligent" creatures like the greater apes dominating environments.  I personally think that there is a threshold effect in which increased intelligence is only very marginal effect until you reach a certain threshold, and the creature gets tool using and language, which utterly changes the game.  But I think that's more of an accidental thing, and not a guaranteed result.  When I've run 5000 Earth simulations, I'll let you know for sure.

It is very true that certain directions are more possible than others...what I am arguing against is that there is a single direction.

But even so, I don't think that is a matter of "directed".  If you've got a top doing a random walk next to a cliff, we don't say it is "directed" to the cliff even if over the course of infinite time it will almost certainly go over the edge.

In that sense, I suppose I disagree with both Wright and Dennet.  I think there may well be a guarantee of getting there without there being any sort of bias.  It's just the drunkard's walk in action. 

Plus, it could be that species like us inevitably blow the crap out of everything, resetting the entire ecosystem to some simpler state.
----
ウセーバラケダ
[ Parent ]

Well by yicky yacky (2.00 / 0) #10 Sun Oct 05, 2008 at 01:40:08 PM EST

"Being first" and "being best" are, I think, both complete misnomers because they imply that it is a race that ends.

It can be a race that ends at the level of individual species, though never well.


----
Done.
[ Parent ]

there are no winners, by codemonkey uk (2.00 / 0) #14 Mon Oct 06, 2008 at 09:17:01 AM EST
only loosers

--- Thad ---
developer of ... ?
[ Parent ]

`survival of the fittest' is itself misleading by lm (2.00 / 0) #11 Sun Oct 05, 2008 at 09:27:31 PM EST
I prefer the term `survival of the adequate.'

There is no more degenerate kind of state than that in which the richest are supposed to be the best.
Cicero, The Republic


The phrase is bad no matter what by ucblockhead (2.00 / 0) #12 Sun Oct 05, 2008 at 09:34:52 PM EST
If phrased right, it is a tautology, because "fittest" is really just "able to survive".   Even more so because words like "fittest" or even "adequate" imply innate qualities when in truth it is all entirely situational.  Asking whether a polar bear is "fitter" than a crocodile is ridiculous on its face.

The essence of the theory isn't who survives, but who breeds, but I don't know of any pithy phrases that describe that.
----
ウセーバラケダ
[ Parent ]

"The getting laid of the fittest" n/t by gpig (4.00 / 1) #13 Mon Oct 06, 2008 at 06:26:59 AM EST

---
(,   ,') -- eep
"This option is deprecated, as it is conceptually flawed." -- man psql
[ Parent ]

fecundity of the freshest? by Merekat (4.00 / 1) #15 Mon Oct 06, 2008 at 09:27:40 AM EST
Hm, yes. I see what you mean.


[ Parent ]

Random stuff by duxup (2.00 / 0) #18 Tue Oct 07, 2008 at 05:52:39 PM EST
From my understanding of evolution it is that #### happens and if it is advantageous, awesome, if not too bad. 

For the most part evolutionary examples are all of stuff that worked out for the best as we study animals that are still alive or were successful.  I think with most of the examples we see providing positive examples people almost get the impression that evolution has some sort of positive slant, logic of its own, or something like that and in fact as far as I understand it, it is just random crap happening and is just as likely to #### up a member of a species with some sort of mutation as it is to bless it and even if it is blessed who knows if it will catch on. 

____


I've gotten my books | 20 comments (20 topical, 0 hidden) | Trackback