Friday, January 25, 2008

Marconomic Theory of Regressive Evolution

I will start with a generalised genetic disproof of inheritance of acquired characteristics from Wikipedia.Genetic disproof
There are many formulations of the genetic disproof, but all have roughly the same structure as the following:

Acquired traits do not affect an organism's genome.
Only the genome is passed to the offspring.
Therefore, acquired traits cannot be passed to the offspring.
While this proof may be logically valid, it suffers from the material fallacy of begging the question, since no one who believes in inheritance of acquired characters would believe both assumptions.


My theory throws these assumptions away and replaces them with this observation. The inheritance of acquired characteristics is not evident because the genome has embedded within it thousands of generations of experience as to what genetic traits are worthwile and those that aren't. In *general*, genetic mutations mimicking acquired traits do *NOT* make the organism any *fitter* to survive in nature.

Instead, the following rules of regressive evolution apply:

-Genetic mutations in offspring are a result of stresses on the parent(s).
-Mutations have a directed element that depends on the exact stress involved.
-The directed element is programmed into the genetic code via association between the stress which is acting as a trigger, and the mutation which has been "proven" to be appropriate to survive better under new conditions indicated by the stress.
-The strength of association between the trigger and the mutation is entirely dependent on the reliability of the trigger being an indication that the mutation is appropriate to the new "environment".
-Many of the mutations concerned are simply switching on or off complex functions that are permanently programmed in the genetic code whether they are active or not.
-Certain combination of stresses may create combinations of mutations without precedent in itself, but each mutation would have been "field tested" individually in distant or more recent history.

Examples that should be quite clear-cut:

Darkness stress: Species adapted to cave conditions or permanently underground (moles etc.) are uncannily blind whereas their closest relatives have full sight. The stress of being in darkness your whole life would cause a considerable proportion of your progeny to have genetic vision defects - all other things being equal.

Wetness/dryness stress: Kamerrers experiments on the midwife toad (Chapter 7 Panda's thumb) had some merit. Clearly, one must separate the selection pressure from the stress trigger to prove the point that mutations to switch back on genes that helped in water are not evident without the wetness stress. Experiments need to be formulated with this in mind. River-courses changing their path makes sudden changes in local environments.

Radiation stress: In Earth's history every single large scale event would have radiation associated with it. Volcanoes erupting, asteroid impacts, solar activity, and possible magnetic field fluctuations would all send rare radioactive elements in the atmosphere or increased radiation. Is it any wonder then that the most obvious, largest scale mutations are caused by radiation? It is a clear signal that radical mutations in any direction might cause some to survive better.

"Regressive evolution" as a term has been almost exclusively used in describing reduction of features, such as in the loss of eyesight and pigmentation in cave dwelling fish. In "Marconomic" terms, reduction in features is also the clearest example of it in nature. Switching off something that was there appears a backwards or downwards move, while the re-switching on of a previously inactive feature appears to be either a leap or not a truly new feature, depending on whether it had been observed before in related species. Philosophically speaking, backwards, downwards and leaps are artificial concepts based on thinking of evolution (falsely) as a ladder rather than the reality of it being a "bush". Switching on or off of features that are then passed on is adaptive either way. Thus I group losses and gains of such features in the same "regressive" boat.

I call it "regressive" evolution because the process almost always borrows tricks that have developed in the past with selection pressures optimising them. Switching them off saves resources when conditions dictate they are unnecessary. However, they are there for when they are required, and perhaps almost all of the evident "faster" evolutionary processes have borrowed perfectly working genetic tricks and combined them in new ways.

At first glance regressive evolution is not really evolution because it never introduces truly *new* changes. However there is such a large number of possible combinations of trigger-responsive genes that the number of species in the Earth is dwarfed by it. If there is a million genes that can be switched on or off depending on triggers, that makes two to the power of a million possible combinations. Virtually all of them will be truly new changes. Also triggers will almost always produce a spectrum of mutations - This is because the future cannot be predicted perfectly. Chance, probability and statistics is a big part of the process. Truly random mutations, however are almost certainly an evolutionary dead-end.

So I have said that in *general*, genetic mutations mimicking acquired traits do *NOT* make the organism any *fitter* to survive in nature. However, this begs the question: Is there any *specific* genetic mutations that do? (Mimick acquired traits that triggered the mutation)?

The trait/mutation would have to fit the following conditions:
A) The organism must be in a state of life or death stress or reproduction-challenging stress(1)
B) The "situation" must be very analogous to situations repeated in the genealogical history (2).
C) The mutation must clearly help progeny in analogous stressful situations more than just being shown/trained/helped by the parent(s) etc. (3)
D) The disadvantages of the mutation must be clearly outweighed by the advantages given assumed environmental constraints. (Needs a clear cost/benefit advantage)

(1) It is not enough, for instance, that a giraffe stretches their neck reaching for higher food. For a trigger situation, the giraffe must see other giraffes dying of starvation, be short in food itself, and endure the frustration of seeing lots of leaves just out of reach. (Using a well known Lamarckian example) If it isn't a life or death situation, the mutation will be both unecessary and resource-depleting.

(2)The "knowledge" associated with the trigger must be on firm statistical ground. The trigger will never be something only associated with proximal individual adaptation. For instance one 1930's experiment devised to investigate inheritance would chop of an insect's antenna that generally grows back. Any number of generations "experience" would not offset the natural experience of the insect, which is - that a more ideal regrowth of an antenna is unnecessary for survivability and reproduction in any conceivable natural environment. Even if it was - chopping off an antenna would not be an indicative trigger.

(3) It is tempting with certain features, to assume that if strength in one arm is important to survivability, and a person strengthens that arm for a greater part of their life to suit their job, and that their children are likely to have the same constraints - that a genetic mutation that would strengthen that arm without as much training would have evolutionary merit. However, if the feature is being passed on to the child by training from the parents, and very few are dying due the the arm being too weak, a mutation may be less reliable than just letting the parents train their children specifically.

A Rundown of evidence that supports my theories:

1) Experiments show that general mutation rates increase as a result of stress.
2) Humans and bacteria have some parts of their genome that are extremely similar. Even though these same parts are quite resistant to random mutations (ie. random mutations do no noticeable change to the organism). As with most parts of the genome, the specific functions of these segments are unknown.
3) Natural selection needs mutations to select from. Stressful situations lend themselves to particular mutations. This article lays out a clear example of adaptive mutation. Because almost no research is looking for similar adaptive mutations, it is unclear whether this is the exception or the rule.
4) My theories do NOT rule out random aspects to the process. Genes that *Are* under constant selective pressure will have (seemingly random) mutations and will use natural selection as a process of refinement, quality control and mutations that don't affect the purpose of the gene will build up in that gene over time -*at well researched rates*. Selective pressure is the error-correcting mechanism in this case and certain critical sequences can stay the same for Billions of years by this mechanism. This fits in perfectly well with my theories.
5) The apparent extension of this randomness into virtually all genes fits the data quite nicely. However, this should not rule out that the apparent randomness overall, is due to the randomness of the factors that cause mutation triggers. The orthodoxy is that the apparent randomness rules out environmental feedback causality of most of the overall apparent mutations. As pointed out in (3) there is some clear counter-examples to the orthodoxy.

23 comments:

Dr. Clam said...

Both parts of the syllogism can be pedantically quibbled with:

'Acquired traits do not affect an organisms genome'?

Consider the phage. It toils not, neither does it spin, but it can be acquired and can transfer blocks of code to the genome.

'Only the genome is passed to the offspring'?

Consider maternal effect genes. You aren't just given your software: your Mum installs it, decides which bits you should turn on and off (no annoying animated paper clip) and watches over your shoulder to see what you are doing until you are ready to go off on your own.

Marco said...

Sorry - This is a work in progress. I will keep adding to it as I go. I seem to have struck a "Eureka" point (in my mind at least) and I'm going to go with it. I have a few examples of triggers and likely mutation responses.

Dr. Clam said...

I've re-read Chapter 7 of the Panda's Thumb, and note that SJG is with me in denying the syllogism- with reference to the same phenomenon of horizontal gene transfer. However, he points out the the key feature of Lamarckism is that adaptations are directed: new changes are preferentially made in useful ways. Your rules of regressive evolution are directed, but they are not evolution, in that they can never introduce *new* changes: they amount to shuffling around a pre-determined volume of phenotype space in response to environmental cues. (See this post). Am I not right?

Marco said...

Yes, you are partly right in that it isn't evolution as it is understood by biological "experts". However, I would wager that the rapid speciation of the last few million years is at least 99% just exactly this type of shuffling.

Dr. Clam said...

I thought you might be hanging out for a more extensive critique of your ideas... if you were me, you certainly would be. And as, if Klaus Rohde has critiqued them, he's done it on his own blog which I haven't visited lately, I thought I would have a go.

You say that the ‘inheritance of acquired characteristics is not evident’, but according to this model no inheritance of acquired characteristics is involved at all. If a genome has within it thousands of generations of experience as to which traits are worthwhile and which aren’t, that experience must be genetically encoded in the organism. Under environmental stress, these traits are not acquired, but expressed. This is what you say with: ‘Many of the mutations concerned are simply switching on or off complex functions that are permanently programmed in the genetic code whether they are active or not.’ This would imply ‘Some of the mutations concerned are not simply switching on or off complex functions that are permanently programmed in the genetic code whether they are active or not.’, but later on you say: ‘regressive evolution is not really evolution because it never introduces truly *new* changes’.

Thus the point should read: ‘The mutations concerned are simply switching on or off complex functions that are permanently programmed in the genetic code whether they are active or not.’

This is plausible. There may be cases of apparent ‘selection’ or ‘speciation’ which are really switching to a latent phenotype.

However, if a gene is not being expressed, than mutations in that gene do not impair the fitness of the organism carrying that gene. The longer it is unexpressed, the more these mutations will build up, and the more the gene will be degraded until every individual carries and unusable copy of the gene. Hence, the places to look for these latent phenotypes are under conditions where environmental stresses occur relatively frequently, with not too many generations between events. They are unlikely to be present in the large numbers of species that live in relatively static environments. Hence also, it is very unlikely that there would ever be a ‘million genes’ that can be switched on or off independently. We would expect no more than a few phenotypes consisting of large clusters of genes switched on or off by a single environmental cue. As the clusters were degraded by genetic drift, the consequences of the environmental trigger would become more and more unpredictable, and introduce new combinations, but this is a consequence of random mutation, not of the regressive evolution model. ‘Large numbers of possible combinations’ is not implicit in any embodiments of regressive evolution that are likely in the real world.

I will now take a stick to your examples:

(1) Dark/Light. There is a perfectly general and non-uncanny model for the blindness of dark dwelling animals. The maxim of entropy in evolution is ‘use it or lose it’ and any characteristic not actively selected for will tend to disappear in time, because it imposes a metabolic cost and no benefit for the organism. There is no selection pressure against defective eyes in the dark, so animals that would normally be eaten or fall over precipices survive to reproduce and mutations that impair eye function can build up in the population. To the extent that having functional eyes imposes any stress on an animal that finds itself permanently in the dark- which I doubt, after the first generation- then these populations will tend to be selected against and disappear.
If I understand you model correctly, it implies that a population raised in the dark may in toto, after a few generations, lose its sight due to the switching on of a ‘blind’ phenotype. Because of genetic drift, as described above, I think this is unlikely. But it should be amenable to experiment.

(2) Wet/Dry. I do not really need to take a stick to this example, only give it a gentle tap on the bum, like I was Harbhajan Singh and it was Brett Lee. The cycling of wet/dry on a multigenerational but not too multigenerational scale is something that is very likely in nature and the ideal environment to look for regressive evolution.

(3) Radiation. I will make up for my gentleness toward the last example by whacking this one really hard and then jumping up and down on it. There are very clear, very well attested chemical mechanisms for how radiation damages DNA and leads to mutations. The sensitivity of different base pairs to radiation damage is not the same, but in general the selectivity you could get this way does not amount to a hill of beans. The fact that mutations are caused by radiation is not a clear signal of anything but the stringency of the conditions required for these chemical reactions which we call ‘life’ to tick along happily. You could argue, ‘Ah, but the genome has evolved in order to be sensitive to radiation in this way’, but remember the principle of ‘use it or lose it’. Unless a chemistry of life evolves under conditions where it was under strong selection pressure to resist radiation, it will end up being sensitive to radiation, because radiation is by definition high energy and reactive.

Dr. Clam said...

I haven't made any comments on your last five paragraphs because I don't understand them. What are you trying to say? Please explain! ;)

Marco said...

However, if a gene is not being expressed, than mutations in that gene do not impair the fitness of the organism carrying that gene.

This is exactly the point where our views diverge. It is critical for an organism's long term future fitness that these unexpressed genes are passed on extremely faithfully. My point is that all mutations are *unrandom* and are genuine attempts at trying something new that could demonstrate possible improvements. Thus, any genes that are completely switched off are the most important to copy identically - because any new mutations on the unexpressed gene cannot be "field-tested" in that generation. I envisage that there is sophisticated error-correcting algorithms on un-expressed genes!

In the last five paragraphs, I am contrasting the naive Lamarckian view of some well known examples, with the extensive preconditions for it to actually happen given my "Regressive evolution" model.

Marco said...

(1) Dark/Light.

My model is clear on this one. It predicts a much higher percentage of general genetic eye defects due to "darkness stress". Selection pressures will then kill off many more of those without the "defects". I am not talking about a single switch that switches off the eyes. I am talking about a multitude of genes that are necessary for clear eyesight. The stress triggers mutations that switch off some of the ones with the highest metabolic cost *first*. In theory the process is reversible. In practice, the cave fish would have to "fall" into a cave with lots of light, but crucially *with no competing light adapted fish species*, for at least some sight to mysteriously return.

To the extent that having functional eyes imposes any stress on an animal that finds itself permanently in the dark- which I doubt, after the first generation-

Try spending one day in a cave and not get stressed! How is the next generation *not* going to be stressed. Most or all of the instinctive actions involve eyesight!

Marco said...

(3) Radiation.

I don't think I am particularly naive in this, but seeing mutations due to radiation as "errors", just means that organisms in a high radiation environment just need more sophisticated error-correction algorithms. More important genetic functions need it too. The infrastructure that detects these radiations both uses these error rates as an automatic system of mutations and needs exceptional long term error correction in itself for these very rare radiation events to generate mutants with half a chance of surviving its mutations.

Dr. Clam said...

This is exactly the point where our views diverge. It is critical for an organism's long term future fitness that these unexpressed genes are passed on extremely faithfully.

Hmm, hrm. Rather, this is exactly the point where you diverge from the experimental evidence.

Point on darkness stress unto the seventh generation taken. But, but, if your model is not talking about switching to a new phenotype but an increased rate of mutations due to darkness stress:
(1) If these mutations are random, then the bad-eye frogs (f'rinstance) will still have the good-eye frog instincts and will continue to suffer 'darkness stress'.
(2) If these mutations are non-random- which is the same as re-inserting the 'many ' in that sentence I took it out of, and recusing yourself from the 'never' in that other sentence- then you are on thin ice. In order for your theory to be a scientific theory, and not a crackpot theory, you now need to suggest a *mechanism* consistent with our knowledge of molecular biology that would explain preferential mutation in the required direction.

Marco said...

Hmm, hrm. Rather, this is exactly the point where you diverge from the experimental evidence.

No. My theory is entirely consistent with the experimental evidence you sighted. It is entirely inconsistent with the inherent assumptions of that experiment. The experiment assumes that the mutations they ae studying are random. In my model, the irreversibility of certain mutations has more to do with the context. A particular genetic sequence in a Hyena in Africa might no longer be appropriate under any circumstances to a wolf in Canada. Technical reversibility of the mutation in question is moot because of the virtual impossibility of the sequence of reversing triggers required. A million years is also an exceptionally long time, even in evolutionary terms.

In order for your theory to be a scientific theory, and not a crackpot theory, you now need to suggest a *mechanism* consistent with our knowledge of molecular biology that would explain preferential mutation in the required direction.

It astounds me that the assumption of mutations being random is so heavily entrenched in evolutionary science that research, for instance (1988), sighted by this Wikipedia entry on adaptive mutation is seen as a possible exception to the rule that mutations are always random rather than an indicator of a mechanism that would work just as well for the spotted hyena as for E. coli that has a frameshift mutation. Almost every other research goes in with the blind assumption that in nature, mutations are always random. I ask why? Has the randomness of mutations been proven? Has the randomness of weather been proven?

Dr. Clam said...

We are talking about two different things when we use the word 'mutation' and that is causing problems.

When I talk about a mutation I mean any event that causes any particular base pair to be replaced by another base pair in a sequence of DNA. This happens. It is a consequence of entropy, the great watchspring of the universe. We can compare the sequences of the gene corresponding to the same protein in different species, and find that they are different in places that make no difference to their function, and that the more closely related the species are, the fewer the differences. These changes are *random mutations*. They accumulate over time as genes are passed down. There is very good experimental evidence for this. This the basis for the 'molecular clock' with which the 'tree of life' has been estimated.
If a gene is not expressed at all, there will be no selective pressure to remove those of these *random mutations* which will destroy the ability of the expressed protein to do what it is meant to do. This is my point. It is not credible to have big collections of unexpressed genes passed down inerrantly from generation to generation.
And a million years is not a long time.
Species can remain static for much longer times.

Secondly, we can use mutation to refer to any heritable change in the organism.

My challenge for you was to explain how mutations (1)- these events on the molecular level- could possibly happen so as to direct mutations (2)- observable changes in the organism- preferentially in one direction rather than the other. If you can do this, you will be a scientist, my son. If you cannot, you will be a pseudoscientist.

"It's all in Dawkins, all in Dawkins: bless me, what do they teach them at these schools?"

Marco said...

I'm a believer in entropy just like the next crackpot scientist :). The law of entropy says that over time my room will get messy. I concur that it does. This does not mean, however, that I place things randomly. Every item that ends up on the floor has a proximal cause that made sense at the time. Any particular base pair changing can be a random event, but it need not be and still appear to follow laws of entropy like my room. DNA, like other sources of knowledge and information, does not necessarily degrade over time. Make enough copies of the information (redundancy) and the initial information tends to persist indefinitely (somewhere at any rate) There is a lot we do not understand about the function of RNA. Recent research has hinted at RNA going back and altering DNA depending, among other things, the environment and stress factors. I believe my theories to be consistent with all biological evidence.

Dr. Clam said...

DNA, like other sources of knowledge and information, *necessarily* degrades over time. This is the #1 empirically verifiable fact about the universe. If you reject it you place yourself firmly in crackpot territory and we have no common language to talk about these things.

Things happen to damage DNA. This is proven.
DNA-repair mechanisms are not 100% efficient. This is proven.
If there are no selection pressures to remove errors from the population, they will accumulate.
Therefore, random mutations will build up and make unusable any unexpressed genes over time.

My new challenge- since you are wriggling about rejecting the premises of my old challenge- is to re-write your theory in a new post such that it is clear what you are trying to say.

Marco said...

A Rundown of evidence that supports my theories:

1) Experiments show that general mutation rates increase as a result of stress.
2) Humans and bacteria have some parts of their genome that are extremely similar. Even though these same parts are quite resistant to random mutations (ie. random mutations do no noticeable change to the organism). As with most parts of the genome, the specific functions of these segments are unknown.
3) Natural selection needs mutations to select from. Stressful situations lend themselves to particular mutations. This article lays out a clear example of adaptive mutation. Because almost no research is looking for similar adaptive mutations, it is unclear whether this is the exception or the rule.
4) My theories do NOT rule out random aspects to the process. Genes that *Are* under constant selective pressure will have (seemingly random) mutations and will use natural selection as a process of refinement, quality control and mutations that don't affect the purpose of the gene will build up in that gene over time -*at well researched rates*. Selective pressure is the error-correcting mechanism in this case and certain critical sequences can stay the same for Billions of years by this mechanism. This fits in perfectly well with my theories.
5) The apparent extension of this randomness into virtually all genes fits the data quite nicely. However, this should not rule out that the apparent randomness overall, is due to the randomness of the factors that cause mutation triggers. The orthodoxy is that the apparent randomness rules out environmental feedback causality of most of the overall apparent mutations. As pointed out in (3) there is some clear counter-examples to the orthodoxy.

Dr. Clam said...

The reference (3) results can all be explained by across-the-board increases in mutation rates due to stress, as in (1) (which point I have no problems with whatsoever.) The article makes this pretty clear.

As for your point (4), selective pressure *cannot* work on an unexpressed gene. You cannot say: 'Selective pressure is the error-correcting mechanism in this case and certain critical sequences can stay the same for Billions of years' for the unexpressed genes that need to be shuffled for your theory to work.

But, I have almost completely lost sight of what your theory actually is trying to say. I have a feeling we are talking at cross purposes. You seem to be responding to my mild critique of the plausible reading of your text ‘The mutations concerned are simply switching on or off complex functions that are permanently programmed in the genetic code whether they are active or not' with a very strong and to my mind incomprehensible assertion of adaptive mutation at the molecular scale which contradicts your assertion ‘regressive evolution is not really evolution because it never introduces truly *new* changes’.

Please confirm that my understanding of your theory as outlined in my second (27/1) and third (2/2) comments is correct. (Or not!) Your initial post is apparently self-contradictory and becomes more so when read in conjunction with your comments- so please consider having another stab at outlining your theory from the beginning ;)

Marco said...

DNA, like other sources of knowledge and information, *necessarily* degrades over time. This is the #1 empirically verifiable fact about the universe. If you reject it you place yourself firmly in crackpot territory and we have no common language to talk about these things.

My hard-drive degrades over time. I back it up. The degradation of the actual information depends on lack of effort or systems to copy and error correct it. Natural selection is a well proven error-correcting mechanism. My theories postulate the need for other, as yet unproven error-correcting mechanisms (virtual selection pressures?). The marginal metabolic cost of these mechanisms is bound to be many orders of magnitude less than the metabolic cost of having the gene be expressed. There is NO need to rewrite the laws of the universe.

Marco said...

Please confirm that my understanding of your theory as outlined in my second (27/1) and third (2/2) comments is correct. (Or not!)

Your understanding is perfectly correct. I think in some respects you are (unusually) following the orthodoxy with regards to DNA degradation. There is an implicit assumption that the bits of DNA that are stable over billions of years are only stable because of constant selective pressure.

Your initial post is apparently self-contradictory and becomes more so when read in conjunction with your comments- so please consider having another stab at outlining your theory from the beginning ;)
Not likely - I keep re-reading it and thinking it says exactly what I mean - Adaptive mutations are the rule rather than the exception.

Dr. Clam said...

All right. I give up. If you're not going to make the effort to redraft your theory in a form more comprehensible to us unwashed masses, there is no point. I am not going to wrestle with Proteus. I think the current form fatally conflates microscopic and macroscopic mutations, muddies the water dreadfully by saying "Many of the mutations concerned..." instead of "The mutations concerned...", and I still have no idea what the last five paragraphs are on about. (Now penultimate five).

The experimental evidence on the development of unexpressed stretches of 'junk' DNA suggests strongly that there are *no* "other, as yet unproven error-correcting mechanisms" to keep it from changes. Your theory requires the existence of such mechanisms. Therefore, experimental evidence suggests strongly that your theory is wrong. Furthermore, information theory- obscure to me, but pellucid to folks such as Androoo- has demonstrated that the *only way* to ensure that code is bug-free is to run it. In biological terms, to express the gene. Therefore, information theory also suggests strongly that your theory is wrong.

You should know that I seek always and everywhere to follow the orthodoxy, in the correct etymological sense of the term: from Gk. orthodoxos "having the right opinion," from orthos "right, true, straight" + doxa "opinion, praise," from dokein "to seem,"

Marco said...

All right. I give up. If you're not going to make the effort to redraft your theory in a form more comprehensible to us unwashed masses, there is no point. I am not going to wrestle with Proteus.
Thank you for your candour. Redrafting is still on my backburner, and you have given me lots of scientific context that has refined my views.
I think the current form fatally conflates microscopic and macroscopic mutations, muddies the water dreadfully by saying "Many of the mutations concerned..." instead of "The mutations concerned...", and I still have no idea what the last five paragraphs are on about. (Now penultimate five).


(re macro mutations)I had originally envisaged that the scope of my theory would be all-encompassing. However, even within the bounds of my theory, unexpressed genes would plausibly lose context in most cases within a million year time-frame. The multitude of triggers I had in mind would present themselves one way or another within that time-frame.

Furthermore, information theory- obscure to me, but pellucid to folks such as Androoo- has demonstrated that the *only way* to ensure that code is bug-free is to run it.

Not quite true: NASA used to have a saying - If at first you don't succeed, you obviously haven't done enough simulations. I don't know how a simulation engine would work in DNA, but I suspect the more likely scenario would be a multi-generational cycling in of unexpressed (but potentially important in the future)genes for testing purposes. To the orthodoxy, this might seem like random macro mutations (or random micro mutations of the switches that turn on the macro genes to *expressed*)

Dr. Clam said...

Okay, then I don't give up, if you have one simmering away on the backburner! :)

I'm very happy with the limited time horizon you are now considering- even if you reject my (orthodox) argument for why that is the sort of time horizon we should expect.

As for the testing mechanism, I think the only practicable way to do this is still to express the genes, but this could be done in 'laboratory' cells localised in a small organ concerned with passing on genetic material. Apropos of which, you would think there would be strong evolutionary pressure to *not* carry around your genetic material in a place where it could conveniently be bitten off by shepherds. What teleological reasons have been postulated for the stupid way male mammals are constructed? I would google this myself, but I should do some real work...

Marco said...

I'm very happy with the limited time horizon you are now considering- even if you reject my (orthodox) argument for why that is the sort of time horizon we should expect.

Try as I might I can't think of any recurrent trigger that would only recur in longer time-frames. Stresses like darkness and wet/dry cycles would conceivably be well within it. Mineral balance of food, or gas balance of the air, or astronomical cycles would usually be within it. Really big asteroid strikes might be rarer, but every strike will favour different mutations, so a generalised increase in mutations is enough without specific directionality.

I like your "Laboratory cell" model, and if I recall correctly, the RNA to DNA effects recently discovered were noticed in sex cells.

Marco said...

The (alleged) disproof of "Lamarckism" (inheritance of acquired traits) goes as follows:

Genetic disproof
There are many formulations of the genetic disproof, but all have roughly the same structure as the following:

Acquired traits do not affect an organism's genome.
Only the genome is passed to the offspring.
Therefore, acquired traits cannot be passed to the offspring.
While this proof may be logically valid, it suffers from the material fallacy of begging the question, since no one who believes in inheritance of acquired characters would believe both assumptions.


from among other places:
http://www.nationmaster.com/encyclopedia/Inheritance-of-acquired-traits

This proof is scientifically sloppy because the two "facts" they are based on have "exceptions". ie. There are adaptive mutations see:

http://en.wikipedia.org/wiki/Adaptive_mutation

that counteract the first "fact".

and further evidence that mechanisms exist that the environment can possibly affect DNA via RNA:

http://cueldee.blogspot.com/2007/06/rna-really-new-advances.html

and things like maternal effect genes are not part of the genome but are passed on.

Thus biologists have convinced themselves that Lamarckism is the wrong way to think without disproving it, but just based on the discredit it had received and by ridiculing the wrong way it was approached in the first place (such approaches as training oneself will mean your offspring won't need to as much)