Fake Banner
    Twelve Misunderstandings Of Evolution
    By Steve Davis | June 10th 2009 03:31 AM | 43 comments | Print | E-mail | Track Comments
    About Steve

    gadfly: noun (1) fly that stings horses and cattle. (2) (derog) annoying person, esp one that provokes others into action by criticism, etc....

    View Steve's Profile


    No. 1 “Evolution is the external and visible manifestation of the differential survival of alternative replicators.”

    This is my all-time favourite, the Dawkins Fallacy, the definition of evolution Richard Dawkins gave in The Extended Phenotype p.82. The fallacy it contains is so obvious I’m amazed that his colleagues have not drawn his attention to it. The survival of replicators is a result of evolution, an outcome, and therefore cannot be the definition of evolution. If we said that fire is the visible manifestation of the production of ash, the statement would be true but meaningless. It would not be a definition of fire and it would tell us nothing about fire, just as Dawkins’ definition tells us nothing about evolution.

    No. 2 Cooperation is a "problem" in evolution theory.

    The origin of life itself required cooperation between complex molecules as they began performing the functions that we now associate with life, i.e., metabolism, homeostasis and reproduction. Cooperation therefore preceded evolution.  It follows therefore that if there is a problem it is with the theory, not with cooperation.

    No. 3 Selfishness is the principal factor in evolution.

    Selfishness is an insignificant factor in evolution because it is an exaggerated and therefore less common form of natural self-interest. Self-interest can be explained as the assertiveness necessary for survival. The procuring of food for example is a manifestation of this assertiveness. Selfishness on the other hand is a learned behaviour found among organisms that are nurtured by parents. It is discarded as the organism matures. An eighteen-year-old first-time mother knows more about selfishness and cooperation than any selfish gene theorist. She knows that when her darling is introduced to other children it will have to give up any notions of preferential treatment. If sharing and cooperating do not come naturally it will be taught to share and to cooperate so it can smooth out the lumps and bumps of life’s journey. She also knows that during this process she will have to protect the child’s self-esteem if it is to have the assertiveness necessary for a full life. In short, she knows that a successful life (biological fitness) requires a balance between individual and social needs. If only the gene-centrics were as perceptive.

    No. 4 Group selection is only a theoretical possibility.

    All selection is group selection. The first life forms that survived while others failed were groups of complex molecules. The eukaryotic cell that became the basis of all modern life forms was/is a grouping of bacteria. An organism is a group. When an organism is favoured by selection, what we see is group selection.

    No. 5 Hamilton’s rule shows how altruism is spread genetically.

    Hamilton’s rule is nonsense. Proportional altruism based on degrees of relatedness is nonsense. The altruistic act is merely a refined or exaggerated form of cooperation; the focus by Hamilton and his followers on altruism was a deliberate distraction to prevent considerations of cooperation. (See Misunderstanding No.11) Their redefined version of altruism in which the altruist suffers a loss of fitness rarely occurs in nature and so cannot be a significant factor in evolution. Cooperation is the principle underlying survival, and therefore underlying evolution. Cooperation can be seen at all levels of life, between molecules, within cells, between cells, between organs, between organisms within groups, and between groups. And because cooperation within groups has assisted individual organisms in survival, reproduction, and just plain old comfort, it has brought about the further development of altruism. Altruistic acts occur when individuals see themselves as belonging to a greater entity. It is that reality that the gene-centrics cannot handle.

    No. 6 The study of animal behaviours gives a good picture of evolutionary theory.

    The study of animal behaviours gives a limited picture of evolution because most evolution took place at the level of micro-organisms. The animals that we are most fond of studying, that is, those closest to us in evolutionary terms, are late arrivals on the scene. This view also ignores the evolution of plants. For more on this see the work of Lynn Margulis on endosymbiosis.

    No. 7Alarm calls are motivated by selfishness.

    This is a typical misunderstanding spread by armchair experts, an example of the mistakes that flow when the original premise (No.1 above) is incorrect. The fallacy lies in the assumption that an animal giving an alarm call is drawing attention to itself, and is therefore increasing its own risk, but increasing the survival chances of its genes that reside in close kin, who will presumably escape. This clever little fabrication disintegrates when we move out of the air conditioning and into the real world. An animal in imminent danger of attack does not waste time or energy on alarm calls; escape is its only concern, so alarm calls are given by those that consider themselves to be relatively safe. This is therefore behaviour for the good of the group, not for the good of an individual’s genes. This misunderstanding came about as selfish gene theorists scurried to explain unselfish behaviours, and made contortionists of themselves in the process.

    No. 8 Members of the same species are competing for the same resources.

    This statement is true to such a limited extent that it promotes misunderstanding. In general terms, members of the same species share resources. When a resource becomes scarce they fight for it when all other options have been exhausted. In many cases they migrate, from butterflies to humpback whales to humans. And that migration is not a case of individuals moving off to satisfy selfish needs; it is a group activity featuring mutual aid and cooperation. For territorial species, once boundaries are established competition ceases, and the various groups share within the groups. Yes, there is competition within groups, but that competition is insignificant in comparison to the time and effort that individuals within the group devote to cooperation. Competition within species is an occasional behaviour; it is not a rule of nature.

    No. 9 Gene selection is a valid concept.

    Gene selection as a concept is almost irrelevant to evolution. Evolution is tied to natural selection. Natural selection targets organisms, not genes. An organism is a product of the genotype, (not individual genes) and the environment in which the genotype develops. This environment includes cultural influences and the effect of memory, that is, past experiences. And it’s these infinitely variable qualities, rolled into a package in which genes are invisible, that becomes exposed to natural selection. Gene-centrics are fond of talking, for example, of a gene for speed in a predator being favoured by natural selection. This alleged gene for speed might result in the greatest variation ever, but if it ends up in a predator that is noisy, or smelly, or poor-sighted, or stupid, or cannot mate, it will not be passed on. It is the total package that succeeds or fails. Richard Dawkins tried to get around this by claiming that when a single variation is being selected, as in the light or dark wings of the peppered moth, then a gene is being selected. Not so. A trait is being selected, not a gene. That trait could be the product of several genes or even several groups of genes. So Dawkins tried to get around these group influences by claiming that it’s common practice for biologists to talk of “a gene for X”, even when it’s known or understood that more that one gene can be involved. Not good enough. Careless language such as this leads to fallacies such as gene selection in evolution. But of course Dawkins cannot talk of groups of genes being selected because that would be, well, group selection, and that, he tells us, is a heresy.

    The case of speed in a predator is an interesting one. It is dependent on so many factors- physical structures and their relationships, metabolism for fast access to energy or the storage of it, emotional factors in the will to push to the limits, and the developmental factors behind all these, that speed cannot be regarded as a genetic trait, it has moved beyond that, through genotypic, to become a phenotypic trait for which mention of genes is useful only as background information. It follows from this that in discussions of natural selection and evolution, which also deal with phenotypes, mention of genes is again only useful as background information.

    No.10 Genes build organisms for the preservation of genes.

    Genes cannot plan or think; they come together randomly. If the random group has features favoured by natural selection it survives and an organism may be formed. If its features adversely affect its fitness it fails. Natural selection is the driving force; it directs development by eliminating the unworkable.

    No. 11 Selfish gene theory is a legitimate way of looking at evolution.

    Selfish gene theory is based on the definition of evolution given by Richard Dawkins that has been shown to be false. A theory cannot be built on a false definition. The theory as a package is baseless. The book that brought selfish gene theory into public awareness, The Selfish Gene, did not present a scientific view of evolution; it presented a world-view, a fact noted by John Maynard Smith in the London Review of Books. “The Selfish Gene reports no new facts. Nor does it contain any new mathematical models…What it does offer is a new world view.” Actually, what it offered was a recycling of an old world view. The Selfish Gene was a contribution to individualism, an idea that had captured the imagination of a section of the British intelligentsia since the Normanisation of England. The concept was first given written form by Thomas Hobbes in the 17th Century, and a remarkable parallel exists between Hobbes’ warped view of society and Dawkins’ warped view of evolution.

    Hobbes: Individuals are concerned only with their own self-interest.
    Dawkins: The primary function of genes is their own survival,

    Hobbes: Individuals compete with other individuals within society.
    Dawkins: Genes compete with alleles (that he calls rivals!) for a place in an organism.

    Hobbes: Society is an artificial construct built by individuals for their protection.
    Dawkins: An organism is a structure built by genes to facilitate their survival in a natural world fraught with danger.

    Hobbes: Society is secondary to the individual.
    Dawkins: The organism is secondary to the gene.

    If both these views were correct then we could reasonably argue that the similarity between the two is purely coincidental, with no connection at all. But the fact that both are wrong, yet similar almost to the point of being identical, suggests that a link exists. Hobbes was an advocate for individualism, and the fact that Dawkins regards group considerations as heresy, the fact also that he attempted without justification to present selfishness in a positive light, tells us that the link is indeed individualism. Still not convinced? Consider this. Selfish gene theory was a fringe theory without credibility until Hamilton’s equation for the genetic spread of altruism was dragged out of the archives to give an alleged mathematical basis for selfishness in nature. But Hamilton’s preference for modelling rather than evidence did not end there. He was so taken with mathematics that he wrote Geometry for the Selfish Herd, a completely fictional explanation for the herding instinct that was also seized with glee by the evolution-as-selfishness crowd. But Hamilton was not the first to attempt geometrical explanations for natural phenomena. Lo and behold, Samuel Stumpf tells us in Philosophy: History and Problems, that Hobbes “…saw in geometry the key to the study of nature. With a razor-sharp intellect and a fervour that caused him to exaggerate the possibilities of this method, (sound familiar?) Hobbes undertook to recast the whole gamut of knowledge in accordance with this single approach…He therefore set out an ambitious project, which was to recast the study of physical nature, the nature of man, and the nature of human society, using the same method throughout.”

    So there we have it. A world-view based on mathematics and selfishness. All roads lead to Hobbes. Of course, Hamilton was not alone in the modern era in seeing mathematics as the key to understanding nature. Haldane, Fisher and Maynard Smith were similarly afflicted. (With no math background, Dawkins had to be content with mathematics-envy.) But with what were they afflicted? They were afflicted by a meme. Ain’t that delicious? (For another discussion of this subject see Truth and Reconciliation for Group Selection VI, a blog article by D.S. Wilson.)

    No. 12 “Selfishness beats altruism within groups, altruistic groups beat selfish groups.”

    This was the conclusion reached by E.O. Wilson and D.S. Wilson in their joint paper Rethinking the Theoretical Foundations of Sociobiology. Much as it pains me to include a misunderstanding disseminated by two of the best, I’ve included this because it teaches us (and them I hope) an important lesson; never accept anything advocated by a selfish gene theorist. The assertion that altruism is locally disadvantageous was first made by Darwin. (Possibly his greatest mistake, but what a mistake! Which brings up another lesson; there’s no such thing as an authority; once a person is beyond criticism, you’ve created a deity.) It was gratefully picked up by the gene-centrics and accepted by the Wilsons, to then play a starring role in the statement above. The second of their conclusions is correct, the first is incorrect. It is incorrect for several reasons. There is no disadvantage associated with altruism because altruists give that which they have in excess. Their fitness is therefore not affected. (Acts affecting fitness are the acts of heroes, but I doubt that anyone would postulate heroism as an evolutionary imperative. It may well have played a role in human evolution by contributing to group solidarity, but not in evolution generally.) But just as importantly, when the two conclusions are paired they are seen to be contradictory. If selfishness beats altruism within groups then it’s most unlikely that an altruistic group would ever develop. Certainly not often enough, or effectively enough, to influence evolution. Furthermore, the first conclusion overlooks the fact that groups are simply cooperating individuals. Cooperation defines the group, so selfishness cannot play a dominant role in within-group motivations. It can exist in groups, but an individual cannot be exclusively selfish as this would result in rejection/ejection by the group. Selfish behaviours therefore have to be occasional only; the selfish individual engages in cooperation to a far greater extent than in selfishness. Cooperation wins.

    The conclusion the Wilsons should have reached is; Cooperation beats selfishness within groups and between groups. Now that is a solid foundation.

    Comments

    No. 9 "This environment includes cultural influences and the effect of memory, that is, past experiences." This point is a little far-reaching, because the definition of 'cultural influences' and the mechanisms of memory are still deeply disputed. Arguing that memories have an effect on evolution is (to me) less reasonable than declaring that some sort of recall will be integral to evolution. This recall can take the form of relatively less complex behavioural conditioning, where creatures have an unconscious reaction or reflex to an external stimulus, or our very complex forms of past recollections combined with abstractions and the ability to cross apply memories to relevantly similar situations. Culture and memory are relatively new, or at least fiendishly difficult to prove their age, and so I don't think they can really apply to classical evolution. Although, the rest of your points effectively lay Gene Selection to rest.

    No 11. Being a former Hobbesian I took some issues with your characterizations of how his principles stand up. There is a critical difference between selfish gene theory (Dawkins) and the selfish human theory (Hobbes). I hold to neither, but the major difference is that the Hobbesian theory includes awareness and decisions. Hobbes includes a weighing (later carried to its extreme by Bentham's Hedonistic Calculus). Because each of the creatures have awareness, they are able to make decisions about the paths they are to take. The genes are 'programmed' to react a certain way, and carry no such introspective determination of how such a behaviour would affect themselves. Basically, it is theoretically possible that the Hobbesian theory could be accurate while the Selfish Gene theory would remain preposterous.

    No 12. “Selfishness beats altruism within groups, altruistic groups beat selfish groups.” This statement is true, but it needs to be SEVERELY clarified. Selfishness and dissemblance beat naive altruism within community oriented groups. This requires that the selfish party not be known to be selfish. I recall an example in nature of bats sharing food with the thinnest amongst themselves, this meant that the bats which looked thinnest could get a lot of food for themselves, and then share it with their own offspring and no one else. This required however, that the selfish creature give the appearance of unselfishness. It was also predicated on the fact that the group was a community oriented group that encouraged sharing in the first place. If there was no sharing, then there could be no 'taking advantage of sharing'. But the concept that selfish creatures can excel in altruistic groups has merit, in fact we see it all the time around us. Next time you are caught in traffic, watch the drivers of the cars in the nearly empty merging lane.

    Group selection is considered a heresy at gene level? Really? I didn't think it went that far, although I'm not a biologist and could easily have missed it. Notably, it's only considered a heresy because there's a mathematical proof that genes cannot be stably selected for or against due to organism interactions. Subject to the assumptions of the model, it's undoubtedly true.

    No. 5: Could there not be genes that predispose an organism to see themselves as parts of certain kinds of entities? Would it not be the case, then, that these predispositions could be selected for?

    No. 7: (Typo: you missed a space.) Do you know if this is also true in humans? It seems that we emit, at least, single alarmed calls whenever we're startled, though this could easily be a special case.

    No 8: They may mean 'competing' in the economic sense, wherein anyone accessing the same resource is defined as 'competing,' no matter what the actual mechanism for dividing the resource. Again, the language is pretty fuzzy, as you're aptly pointing out.

    No. 9: I was going to say 'environments cannot be selected for' but it's not entirely true; you explicitly bring up migration.
    And so, as in any nature/nurture debate, the answer is both, and that moreover there's continual feedback between the two.
    Given a constant environment, you can tell which genes lead to which phenotypes, and so figure out how selection for phenotype leads to selections of gene groups. But, to use your phrase, outside the air conditioning, how many constant environments are there? (Six?) It might be fine over the lifespan of a bacteria, but for any mammal you have to start with what season they're born in and then start feedback calculations and then...

    I think (and I am not a scientist in this sector, it really is just my oppinion) that it is not correct, that cooperative behaviour wins every time against competetive behaviour. It depends on the surrounding situation. Sometimes you just give more as you ever could get back in reward, if you behave cooperative, even if your life does not depend on the outcome.
    In my eyes evolution does not aim at getting perfection, but as everything else in nature it just aims at getting the balance. In the moment that some lifeform is in disadvantage it will change to the situation and in an advantagous situation it will be confronted with some new problems (like overpopulation). So there is no advantage in beeing cooperative or competetive expect your surrounding situation favours one over the other right now.

    Gerhard Adam
    So there is no advantage in beeing cooperative or competetive expect your surrounding situation favours one over the other right now
    I think you're interpreting cooperation and competition as decisions an animal makes, instead of considering that there is a cooperative dynamic that precludes other behaviors.

    There is no such thing as a human that survives outside of society.  Therefore humans are absolutely cooperative and dependent on a social group for survival.  Within that group they may exhibit various behaviors amongst themselves or in specific situations, but they cannot avoid being cooperative (in the sense of being part of their society).

    The problem with the competition argument is that there is a vast difference in competing for resources necessary for survival versus competing for position within a social group.
    Mundus vult decipi
    Concerning your first 'misunderstanding': firstly, a single definition in a popular-scientific book does not constitute a general 'misunderstanding of evolution'. The title is misleading. Secondly, it's quite normal for definitions of processes to state the outcomes of the process, so your criticism fails. You may find it incomplete, but it is not wrong and certainly not a 'fallacy'. Thirdly, when criticizing a definition, provide a more accurate one, so we can compare whether there actually is a difference.

    I won't bother with the other 11 points, as they all fail on similar grounds. Suffice it to say that supposing that the likes of Haldane, Maynard-Smith and the Wilson's were all blind to their own prejudices, motivations and moral convictions constitutes an enormous underestimation of their intelligence and insight.

    logicman
    Another very stimulating article, Steve.

    Here's an analogy for a cell.

    Components are passed along 'conveyor belts' and placed on a 'template', where they are fastened together.  It is entirely illogical to view this as a 'black box' within which the template performs any active or purposive function whatsoever.

    The cell is a factory within which the genes function as both blueprint and template to serve the needs of the cell.  If DNA was ejected by a cell and grew its own new cell, then its function would be primary.  In practice, the cell division process creates two copies of the DNA which end up in the two new  cells.  The organelles in the cells are already duplicated.  The factory is ready to roll.  DNA is just one factor in the factory, not the boss.

    A query:  what is the role of selfish genes in aneucleate cells such as red blood cells?  It is my understanding that the nuclei are dissolved.  DNA suicide?  Altruistic genes?  Red blood cells keep us alive by delivering oxygen, directly by carrying it, indirectly by chemical dilation of blood vessels, and they help to protect us as part of the immune system.  Perhaps we need a theory of the altruistic anucleate cell - not!
    Gerhard Adam
    That's the flaw isn't it Patrick?  After all, the gene had the upper hand with bacterial conjugation which allows the transfer of genetic material that is disproportionate between cells.  On the other hand, sexual reproduction restricted the genetic exchange to half.  Therefore, half the genes must risk being excluded.

    As you know, my contention is that sexual reproduction was the one of the first fully cooperative evolutionary changes which invariably leads to numerous other cooperative adaptations.

    Doesn't sound like a premise for "selfishness" to me.
    Mundus vult decipi
    "a single definition in a popular-scientific book does not constitute a general 'misunderstanding of evolution'"

    If Dawkins didn't believe that, he shouldn't have written it down. Ergo, he's either deeply flawed writer, or else he's wrong.

    Take your pick.

    I also note that you don't dispute that the definition is meaningless.

    jtwitten
    Hell, I'll dispute it.  I'll also dispute that it is even a definition.  It is a description.  They are different things. 

    Do you only write things down with which you think constitutes a "general" understanding?  How does one persons opinion constitute a general misunderstanding?
    If Dawkins didn't believe that, he shouldn't have written it down. Ergo, he's either deeply flawed writer, or else he's wrong.
    For the record, this sentence is the most nonsensical thing I have read today.  And I've read this entire article. 

    . . .which brings back around to things that shouldn't have been written down.
    Steve Davis
    Hell, I'll join this dispute Josh. A description as precise, as concise, as this one, especially one that begins with "Evolution is..." would be regarded by most people as a definition. If Dawkins had said "Evolution can in certain circumstances be regarded as..." then that would be an acceptable desciption depending on the conclusions drawn from it. But he was firm. He was clear. He left no room for movement or for other considerations. It was a definition. And it was flawed. 
    jtwitten
    First, that statement is in now way precise.  It does not define a phenomenon in any precisely recognizable way.

    Second, you are the first to call Dawkins on the rhetorical flourishes that he so loves.  What is more likely?  That Dawkins is unaware of the technically correct definition of evolution or that he was being dramatic?

    Third, one can easily construct similar sentences that are grammatically correct using the word "is" that are not definitions.  For example, "The rose is red."  This is a description not a definition.   

    Fourth, the simple fact that we disagree means that his statement was not that firm, clear, or lacking in room for movement.
    Steve Davis
    Ahhh yes. Wordplay. I had almost forgotten how much fun it can be. But is it useful?
    jtwitten
    You seem to think so. 

    Much of your criticism of the research into evolutionary theory and altruism is based on transmuting the definition of altruism/cooperation used by the research community (requires the altruistic act to have a cost, by definition) into your own (based on the assumption that resources are rarely limited) and holding those researchers accountable to your personal definition, not the framework within which they worked.

    Since your critique involves no data, no math, no solid theory, and no logic to support your conjectures, it seems that all we are left with is wordplay.
    Gerhard Adam
    ...requires the altruistic act to have a cost, by definition...
    Which is where I have the problem, since it sets a requirement to be a fortune-teller.  Why should rushing into a building to save a stranger be defined as an altruistic act, solely based on whether I survive or not?  If I survive, then technically it has cost me little or nothing, yet if I die it has cost me the ultimate.  So why should such a simplistic criteria define what is considered altruistic? 

    (This is precisely why my criticism suggests that it should reflect the "potential" cost)
    Mundus vult decipi
    jtwitten
    This is precisely why my criticism suggests that it should reflect the "potential" cost
    It does.  The models deal with what happens to populations, not specific individuals.  The cost of running into burning buildings is the probability that you will die.  Ironically, use of population level probabilities is actually how "fortune-telling" (i.e., cold reading) is done.  Whether you perceive the cost is irrelevant.
    Gerhard Adam
    Well, if it does, then why should altruism be considered an "enigma"?  After all, according to Axelrod's "Evolution of Cooperation", the entire premise of the "Tit for Tat" strategy in the Iterated Prisoner's Dilemma specifically requires that the initial action be altruistic.  If both players follow up, then the result is cooperative, but the first action must always be altruistic.

    This clearly indicates that altruism isn't an enigma at all, but rather it is a pre-requisite condition that results in cooperation (obviously somebody has to take the first "chance").

    I'm also confused as to why every discription of altruism defines it as having a "cost" when, in reality, there may be no cost at all.
    Mundus vult decipi
    jtwitten
    Tautologically, altruism is defined as such, because that is the technical definition, whether or not it confuses the lay public or does not agree with popular usage.  It generally also refers to active behavior, which according to the second law of thermodynamics has at least some energetic cost.  The existence of behavior that gives benefit without cost would not be an enigma.  The existence of costly behavior that confers benefit to another is.

    Cooperation is slightly different than straight altruism.  In cooperation, the benefit gained by cooperating is greater than the potential benefit from acting individually.

    You are confusing work looking at how altruism might evolve (Hamilton and kin selection) and work looking at whether altruism is an evolutionary stable strategy (game theory: can a group of altruists be infiltrated and taken over by non-altruists/cheaters/defectors). 

    This whole line of appears to be based on the unsupported assumption that resources are generally unlimited.  Even if this were temporarily the case, do you reject the concepts of exponential growth and carrying capacity?
    Gerhard Adam
    I'm going to respond at the bottom, since I tend to be a bit wordy, I didn't want this crammed into a single column.
    Mundus vult decipi
    Steve Davis

    Dear oh dear Josh, there’s so many angles here I hardly know where to begin, so I’ll stick to one.
    You have set yourself up on this site as the resident champion of the scientific method, a move we can only applaud. But here you are now, all reactionary and prickly, defending a thought system developed by those who treat the scientific method with contempt.
    Let’s take a look at what you’re defending, and what you’re failing to defend.
    In Bill Hamilton’s paper Geometry for the Selfish Herd we find two blatant abuses of the scientific method. The first was Hamilton’s rejection of field data in preference for mathematical modelling that supported a pre-conceived notion. The second was his criticism of Galton for restricting his conclusions about animal behaviour to the animals Galton actually studied, and for not extending those conclusions to human behaviour. I criticised this in an article on this site, and this was the perfect opportunity for you to support your pet theme. But what response did I get? Silence. You have obviously followed my articles as you've provided us with a summary of my position, but when I stuck my neck out and took a well-deserved swipe at an icon, a time when I could have done with some support, you were nowhere to be seen.
    But that silence was informative.
    I've made the point elsewhere that one of the tragedies to flow from selfish gene theory (yes, there’s been more than one!) is that a whole generation of biologists has grown up thinking that the theory is serious science. They’ve been compelled to internalise an ideology.
    I think we are seeing the evidence of that here.

    jtwitten
    Steve, I am always prickly. 

    Not to bruise egos, but I do not follow your column as closely as you might hope.  In this case, I was only drawn into the debate by the comment by Alrenous, who took the lack of dispute about the so-called definition as evidence of lack of disagreement.  Ever helpful, I thought it might be useful to disabuse him/her/it of such a silly notion.  As we all know, "Absence of evidence is not evidence of absence."  The same goes for my silence on any other articles.  It cannot be construed as either support, opposition, or even indifference to a particular point.  I am a busy man.  

    There is plenty of debate in the field of evolutionary theory about the relative importance of genic selection, organismal selection, and group selection.  Dawkins advocates particularly strongly for one of these areas.  He is not, however, a leading light in the scientific field.  He is not as important a figure in the field as you fear.  In order to foster a true scientific debate on these issues, I have focused my criticisms on a few areas:

    1. The unstated major assumption of your own theories, that resources are usually unlimited.
    2. The use of personal definitions for terms that have specific technical definitions.  It is not appropriate to criticize the work of researchers using different definitions than those they used to frame their hypotheses.
    3. Over-concern with the value of specific words.  Altruism, cooperation, and selfishness are loaded words in normal usage.  This has no bearing on the scientific theories.  We could just as easily talk about behaviors A, B,&C.
    4. The defense of claims using rhetoric instead of solid logic, theoretical modelling, or data.

    If I feel compelled to do a complete refutation of your Twelve Misunderstandings, I will be sure to publish my own article on which you shall be free to comment and criticize.
    Steve Davis
    "I am a busy man."
    As we all are Josh, as we all are.
    "Dawkins ... is not, however, a leading light in the scientific field.  He is not as important a figure in the field as you fear."
    I fear you are wrong. With selfish gene theory now given credibility in text-books, encyclopaedias, even dictionaries, his influence is overwhelming and will continue to dominate unless biologists get active on the matter. Why should someone of no importance occupy such a dominant position?   
    jtwitten
    Steve, I'm actually involved in this field.  That statement is ridiculous.

    Find yourself a copy of Douglas Futuyma's Evolutionary Biology, 3rd Ed. (a standard text book in the field).  The 763 page text only contains 2 references to Dawkins.  See page 353-354 for a reasoned discussion of this debate.
    Steve Davis
    References to Dawkins are irrelevant Josh.
    Selfish gene theory dominates the field. 
    jtwitten
    Good, then it seems like we can move beyond issues with Dawkins quotes as a basis for criticism of evolutionary theory.   
    Selfish gene theory dominates the field.
    Instead of disputing this point, I will point out that relativity and quantum mechanics dominate the field of physics.  Why?  Because they are better models than other theories.  Similarly, displacing selfish gene theory's "dominance" requires a demonstration that it is wrong with a level of rigor equal to that demonstrating that it is a predictive theory.
    Steve Davis
    it seems like we can move beyond issues with Dawkins quotes as a basis for criticism of evolutionary theory.  
    But I already have Josh. I've gone straight to the source. It was Hamilton who provided the intellectual stimulus for all that followed, and I've shown that his work was flawed. That says plenty about the level of rigour you seem so impressed with.
    Hank
    Just as a semantic point - relativity and selfish genes are not really apples to apples.   Dark matter and selfish genes are probably better analogies.   Relativity and DNA are probably better comparisons to each other because we know they exist; there are just exceptional instances where neither of those can explain everything today, like at the quantum level in physics or personalities in people.
    Steve Davis
    Hank, that's a good perspective with which to conclude a discussion that was getting a little jaded. 
    Gerhard Adam
    Good article Steve.  As you know, much of the controversies are due to the abuse of words (which I'm sure Patrick has a good take on).  Selfishness is used, when self-interest is what's being described.  Altruism is viewed as some sort of aberrant behavior.

    Other concepts that muddy the water, are ideas like "individualism" and "collectivism", which both represent philosophical views that cannot exist as they are described.  Yet, the arguments are advanced from these perspectives, just like the "selfish" gene as if they had a basis in reality.

    It's almost as if there is an attempt to reduce biological processes to an engineering description, instead of recognizing that there are larger principles that will govern the group with variable behaviors among individuals within that group.

    In addition, another point that I don't think is emphasized enough, is that genes are not hard-coded single value traits.  They typically present as a range of possible outcomes, which will be influenced by the environment and/or circumstances.  It is also possible that variations in gene expression could occur that are completely unnoticed for generations until some external circumstance suddenly favors their existence. 

    In any case, I always like your articles because it tends to stir up one of my favorite areas.
    Mundus vult decipi
    adaptivecomplexity
    I have to disagree with a lot of what's said here. I'm not knocking group selection or any particular theory about the evolution of cooperativity, but there are some issues:

    1. Evolution is about outcome and mechanism. Evolution as outcome was accepted before process: Darwin convinced people that evolution happened long before people were convinced about Darwin's proposed process, natural selection.

    2. I'm not sure what the 'misunderstanding' is here - what evolutionary biologists are generally interested in understanding is how cooperation within or among species can evolve. Whether the first instance of cooperation ever happened before the first organism evolved is completely irrelevant to the question of how meerkat or cleaner fish cooperation evolved. So I don't agree with this:

    Cooperation is so intimately tied to life itself, in all its stages, in all its forms, that it’s almost synonymous with life, so a genetic connection to explain this particular behaviour is just not necessary.


    And this simply doesn't jive with the evidence:

    It’s unlikely that cooperation in any form has a genetic basis,


    There are many, many examples of scientists teasing apart the genetic underpinnings of cooperative behavior, mostly in microbes.

    The reason tigers hunt alone and lions hunt in packs is genetic. Anyone who doubts that should propose what kind of environmental change would make tigers, as they are now genetically, hunt in packs.

    3.
    Selfishness is an insignificant factor in evolution because it is an exaggerated and therefore less common form of natural self-interest.


    I'm not even sure what this means, but selfishness is widespread, from segregation distorter genes to the widespread practice of infanticide among mammals. Many species are social, but many more are not, and I'm not sure what the insights of a first-time mother have to do with selfishness among the millions of non-social species.

    4.
    An organism is a group. When an organism is favoured by selection, what we see is group selection.
    This is a little misleading here, because arguments about group selection are arguments about a population of organisms, all with their own independent genomes. An organism itself is not a population in the genetic sense - it's a population of cells harboring essentially one genome.

    6.
    The animals that we are most fond of studying, that is, those closest to us in evolutionary terms..


    I disagree here too. Fish and insects are common subjects of evolutionary study. When it comes to vertebrates, fish are the most distant from us phylogenetically, and insects are very distant. The fact that we find common evolutionary principles in insects and fish suggests that there is a lot to learn about evolution by studying animals.

    7.
    When a resource becomes scarce they fight for it when all other options have been exhausted.
    That's not even true of humans, and that's not true of territorial species - some Galapagos finch species are great examples. That's not true of mates either, in species where there are harems - one male gets most of the matings.

    9.
    It is the total package that succeeds or fails.


    This is true, but it's hard to study all of the genes at once. Geneticists have well-developed methods for looking at the effects of single genes on a trait. Sometimes that includes looking at the effects of a gene in an inbred lab strain - a homogeneous genetic background, but we can also look at the effects of a gene in natural populations, either by averaging out the background (what essentially happens in genome-wide association studies), or through more classical quantitative genetics techniques.

    Genes work in packages, but individual genes (or alleles) come down to you with their own unique evolutionary history - they've been passed through different genetic backgrounds to get into your genome.
    Mike
    Steve Davis

    Hi Mike, here's my thoughts;

    1. Evolution is about outcome and mechanism. Evolution as outcome was accepted before process: Darwin convinced people that evolution happened long before people were convinced about Darwin's proposed process, natural selection.
    I don’t quite know what your point is there Mike. Natural selection is about the survival and reproduction of phenotypes. Replicators should only be introduced as background information. And was it Darwin who convinced people? I thought that it was already accepted, it was the process that was unknown.

    2. I'm not sure what the 'misunderstanding' is here - what evolutionary biologists are generally interested in understanding is how cooperation within or among species can evolve. Whether the first instance of cooperation ever happened before the first organism evolved is completely irrelevant to the question of how meerkat or cleaner fish cooperation evolved. So I don't agree with this:
    Cooperation is so intimately tied to life itself, in all its stages, in all its forms, that it’s almost synonymous with life, so a genetic connection to explain this particular behaviour is just not necessary.

    And this simply doesn't jive with the evidence:
    It’s unlikely that cooperation in any form has a genetic basis,

    There are many, many examples of scientists teasing apart the genetic underpinnings of cooperative behavior, mostly in microbes.

    The reason tigers hunt alone and lions hunt in packs is genetic. Anyone who doubts that should propose what kind of environmental change would make tigers, as they are now genetically, hunt in packs.
    I did not rule out a genetic connection, I probably overstated some points a little, and understated the misunderstanding, which is that there is a widespread view that in evolutionary terms cooperation is a problem that needs to be explained in genetic terms. That view is incorrect for the reasons outlined.

    3.
    Selfishness is an insignificant factor in evolution because it is an exaggerated and therefore less common form of natural self-interest.

    I'm not even sure what this means, but selfishness is widespread, from segregation distorter genes to the widespread practice of infanticide among mammals. Many species are social, but many more are not, and I'm not sure what the insights of a first-time mother have to do with selfishness among the millions of non-social species.
    If you are not sure what the statement means then you have not been reading Gerhard’s series on biology. I did not deny the existence of selfishness, my point is that the influence of selfishness is overstated. Its easy enough to grasp.  And did I link a first time mother to “millions of non-social species”? I linked her to selfish gene theorists who are blinded by ideology, a reasonable point.

    4.
    An organism is a group. When an organism is favoured by selection, what we see is group selection.

    This is a little misleading here, because arguments about group selection are arguments about a population of organisms, all with their own independent genomes. An organism itself is not a population in the genetic sense - it's a population of cells harboring essentially one genome.
    This was thrown in to provoke thought , but I’m not alone in this view, the Wilsons have made comments on similar lines.

    6.
    The animals that we are most fond of studying, that is, those closest to us in evolutionary terms..

    I disagree here too. Fish and insects are common subjects of evolutionary study. When it comes to vertebrates, fish are the most distant from us phylogenetically, and insects are very distant. The fact that we find common evolutionary principles in insects and fish suggests that there is a lot to learn about evolution by studying animals.
    There is indeed Mike, but there is a tendency to draw overstated general conclusions about evolution from the study of animal behaviour. Nothing wrong with that in principle, as long as conclusions are reasonable, but some biologists have demonstrated that they are incapable of sticking to principles.  

    7.
    When a resource becomes scarce they fight for it when all other options have been exhausted.

    That's not even true of humans, and that's not true of territorial species - some Galapagos finch species are great examples. That's not true of mates either, in species where there are harems - one male gets most of the matings.
    You can always pick out exceptions Mike, but that is not good science. The dominant influence in fitness is cooperation, not aggression. Check out Gerhard’s explanation of harems in his Biology series. And it is true of humans. Yes, there are confrontations going on at any time somewhere in the world, but the percentage of the total population engaged in competition for resources is almost negligible. Most of us are quietly going about our business sharing resources through trade.

    9.
    It is the total package that succeeds or fails.

    This is true, but it's hard to study all of the genes at once. Geneticists have well-developed methods for looking at the effects of single genes on a trait. Sometimes that includes looking at the effects of a gene in an inbred lab strain - a homogeneous genetic background, but we can also look at the effects of a gene in natural populations, either by averaging out the background (what essentially happens in genome-wide association studies), or through more classical quantitative genetics techniques.

    Genes work in packages, but individual genes (or alleles) come down to you with their own unique evolutionary history - they've been passed through different genetic backgrounds to get into your genome.
    I don’t have a problem with that

    Steve Davis
    Come on now Mike. You can't get all excited about the sideshow then tip-toe around the main event. The substantive issue to come out of the article was that cooperation wins over selfishness as an influence in evolution. You were keen to comment on the lead-up points, what are your thoughts on the main issue?.
    adaptivecomplexity
    what are your thoughts on the main issue?.
    Sorry, I've been traveling, so I'm replying late.
    But my response is brief: nowhere here do I see any empirical evidence for the claim that 'cooperation wins over selfishness'. Other than citing cooperation among the parts of cells and multicellular organisms, and endosymbiosis, I see no discussion of any field studies, genetic experiments, etc. 

    It's an empirical question, and to answer it, you probably have to define it in a more limited way - maybe starting with cooperation vs. competition among single-species populations. 
    Mike
    Steve Davis

    Mike, your response was just what we would expect from a professional in the field. Careful, considered, tending towards sceptical, in short, a legitimate response. Can I safely assume therefore, that you directed the same level of professional care and scepticism at The Selfish Gene on your first reading of it or hearing of it, and that you are yet to be convinced of its basic premise, that "A predominant quality to be expected in a successful gene is ruthless selfishness. This gene selfishness will usually give rise to selfishness in individual behaviours."

    adaptivecomplexity
    Can I safely assume therefore, that you directed the same level of professional care and scepticism at The Selfish Gene on your first reading of it or hearing of it,
    Actually, what Dawkins writes in popular books plays very little role in my thinking about evolution - which is true of just about all active researchers. I carefully read the technical literature, which is where the real scientific debates are taking place.  Even well-written popular books avoid the technical details, and in this field that includes mathematical details, as well as field observations, that are really necessary to evaluate the merits of these arguments about selfishness and cooperation.
    I would suggest starting with one of these technical papers - this one by Axelrod and Hamilton (you may have to do a web search to find a free copy). Read that and some of the subsequent papers that cite this one, and write a post about your take on it.
    Mike
    Steve Davis
    Mike, once again you've provided a response crafted with professional care and consideration, but this time instead of asking a question, I'll venture some comments.
    Let's be frank Mike, the question I asked was not a difficult one to answer, and as you are a professional in this field I'm sure that readers would be interested in your answer. You were keen to come into the discussion with a lengthy comment, so I can't understand the coyness now.
    You say that "what Dawkins writes in popular books plays very little role in my thinking about evolution, which is true of just about all active researchers." This begs the question as to why all these active researchers have allowed Dawkins to hijack the public discussion of these issues, and why they have allowed his version to become text-book dogma.
    You've suggested that I write an article on Axelrod and Hamilton's handling of cooperation. There's no need. Why hack at the branches of evil when you can strike at the source? (Sorry Hank, that's from Thoreau. I know you think some of his stuff to be a bit quaint, but he also came out with some rippers.)
    All of the work done on cooperation in the last few decades has has been heavily influenced by Hamilton. Even DS Wilson, who has no trouble seeing through the Dawkins facade, still seems to be strangely in awe of Hamilton if his latest blog is anything to go by. I've presented the argument in other articles that Hamilton's work was flawed. That's about as much as I can do.
    Steve Davis
    Mike, thanks for your thoughts on segment 2, I've made the correction.
    "No. 1 "Evolution is the external and visible manifestation of the differential survival of alternative replicators." which is supposedly wrong because "The survival of replicators is a result of evolution, an outcome, and therefore cannot be the definition of evolution. "

    The criticism hinges on the idea that he supposedly doesn't independently describe what a replicator is, as "proven" by quoting a single sentence.

    The thing is, he does describe what a replicator is, and then he uses that term in sentences like the quoted one.

    "No. 2 Cooperation came about through evolution." "Cooperation has evolved, but did not originate as an outcome of evolution. The origin of life itself required cooperation between complex molecules as they began performing the functions that we now associate with life, i.e., metabolism, homeostasis and reproduction. Cooperation therefore preceded evolution."

    When people talk about cooperation having come about through evolution, they're clearly talking about the cooperation within and between living entities, and they clearly don't mean to say that 'every possible thing that could be construed as a type of coordination must have come from evolution' which is what this criticism is taking them to mean. I.e. the criticism is ridiculous nitpicking based on the meaning of words, dressed up as highlighting some major mistake.

    "It’s unlikely that cooperation in any form has a genetic basis, although such a connection is possible."

    That's just ignorant.

    I stopped reading after these two.

    "No. 3 Selfishness is the principal factor in evolution."

    This shows an alarming ignorance of what the selfish-gene view is.

    The selfish-gene view is a metaphor saying that evolution builds things that benefit the genes not necessarily the organism. It doesn't say that the genes are somehow selfish (as if they could have intentional attitudes). Nor does it say that evolution builds systems that are selfish.

    Even though the criticism is meant to be of this view (and it _is_ trying to criticise the selfish gene view specifically "An eighteen-year-old first-time mother knows more about selfishness and cooperation than any selfish gene theorist.") the criticism says nothing about that actual view and instead focuses entirely on criticizing the idea that a selfish _attitude in living creatures_.

    "No. 4 Group selection is only a theoretical possibility.

    All selection is group selection. The first life forms that survived while others failed were groups of complex molecules. The eukaryotic cell that became the basis of all modern life forms was/is a grouping of bacteria. An organism is a group. When an organism is favoured by selection, what we see is group selection."

    As with No 2, all this is doing is redefining a term (here it's "group") to mean something totally different to what people actually mean when they talk of the phenomenon (when they talk of group-selection), in order to make out that there is some mistake when in fact there's no such thing.

    Gerhard Adam

    The existence of behavior that gives benefit without cost would not be an enigma.  The existence of costly behavior that confers benefit to another is.

    While there may be a cost, this is much more probabilistic than generally discussed.  In other words, if an altruistic act fails, then the costs may be quite high (including death) however, if successful the payoff may be substantially higher than the investment in the act itself.  Consider the case of my running into a burning building to save my child.  There are four possible outcomes:  (1)  The child could die (same result as no action),  (2) I could die (which is my potential cost),  (3) We could both die (highest cost), or (4) we could both survive. 

    Case #1: Clearly option one represents no decision, since it is the product of no action. 

    Case #2: In the second case, it may be a trade-off that I consider viable (especially if I'm older and the child has good prospects for survival, if not, then my benefit may not be achievable anyway - although I'm no worse off than option #1). 

    Case #3: In the third case, the ultimate cost is paid and we simply lose all benefit. 

    Case #4: In the fourth instance not only do I have the benefit of surviving (to engage in future reproduction), but my offspring also survives (which may also convey additional benefits within the group - i.e. status, etc.). 

    Therefore while there is a cost to the act, the benefit may far exceed my costs, so altruism is essentially a gamble with large potential payoffs.  Refinements in assessing the costs would depend on the age of the child, viability of survival into the future, the number of other children, etc., but you get the idea.

    In addition, there may be other benefits that are less tangible when non-related individuals are involved and they will depend on how the group recognizes or rewards such behavior.  Altruism cannot be assessed from the presumption of failure, but rather the perception that an individual, in all likelihood, does not really assess the cost against the perceived benefit.  In those cases, the risks are considered acceptable with failure marking the cost of altruism.   

    Cooperation is slightly different than straight altruism.  In cooperation, the benefit gained by cooperating is greater than the potential benefit from acting individually.

    Altruism must occur first, since there must be an initial interaction whereby an individual willingly takes the risk of being cheated.  This is why the first action in "Tit for Tat" is to cooperate.  It is always possible to be cheated on the first interaction.

    You are confusing work looking at how altruism might evolve (Hamilton and kin selection) and work looking at whether altruism is an evolutionary stable strategy (game theory: can a group of altruists be infiltrated and taken over by non-altruists/cheaters/defectors). 

    Actually altruism is a pre-requisite to cooperation since every initial act of cooperation is an altruistic one.  The Prisoner's Dilemma example indicates that the only rational decision is to defect.  In the iterated case, it is the possibility of future encounters that make the initial "altruistic" decision to cooperate viable since it would result in higher payoffs.  Therefore the initial decision in the Iterated Prisoner's Dilemma is irrational unless it is weighed against the long-term results (assuming the opponent cooperates).  If the opponent fails to cooperate, then the entire relationship degenerates into a minimal return of selfish responses (so there is some degree of "self-protection" built in for the altruist).  As an ESS, altruistic individuals are not obligated nor confined to interacting with selfish invaders, so the net result is that selfishness tends to become isolated.  Given the higher return of interacting with other altruistic individuals, selfishness tends to be self-limiting in those cases where individuals have choices (I don't remember whether this was demonstrated by Axelrod, or by John Maynard Smith). 

    There is little doubt that some selfish traits may have invaded groups in the past, and that selfishness exists, but in those cases they would represent animals that we consider asocial.  It makes no sense to consider selfishness within the context of a social group, because such invasion is destructive for ALL parties (including the selfish instigator).  In my view such a strategy is fundamentally unstable and can only exist in very limited amounts, such that social groups are not stressed by having to support some degree of cheating.  In other words, any "invaded" group would cease to exist as a group (for all we know Grizzly Bears may have been cooperative in the past, but we would never know it now).

    This whole line of appears to be based on the unsupported assumption that resources are generally unlimited.  Even if this were temporarily the case, do you reject the concepts of exponential growth and carrying capacity?

    Not at all.  While resources may be not be unlimited, they must be relatively unconstrained or else you're dealing with a population in crisis (in the long-term).  Limited resources have farther downstream ramifications of offspring competing with parents, etc.  There are certainly strategies that exist within many species to deal with these changes (especially if they're cyclic), but essentially there is a difference in behavior during varying levels of resource availability.  This is one possibility when it comes to considering behaviors like the cannabilistic spiders since the size of the male is proportionate (larger = less likely) to the probability of being eaten, it has been suggested that the relative size is also an approximate indicator of how resource rich the environment is.  The larger the male, the greater the amount of available resources, the greater the likelihood of the male surviving to mate with more females.  This is still largely supposition, but it can illustrate the kind of mechanism I'm referring to.

    Mundus vult decipi
    jtwitten
    While there may be a cost, this is much more probabilistic than generally discussed.
    General discussion is irrelevant to the theoretical underpinnings of evolutionary theory.  Professionals deal with probabilities as a rule.

    You fail to understand the point of the game theory exercise.  What you are referencing is a test of whether altruism is not an evolutionary stable strategy.  Essentially, one assumes a population of all altruists and then ask if a non-altruist can invade the population and spread.

    You are positing that populations will not expand to use all of the resources available to them.  Limited resources are a necessary conclusion of the interplay of the second law of thermodynamics and observed exponential growth rates.
    Gerhard Adam
    What you are referencing is a test of whether altruism is not an evolutionary stable strategy.  Essentially, one assumes a population of all altruists and then ask if a non-altruist can invade the population and spread.
    No, I'm actually referencing John Maynard Smith "Evolution and the Theory of Games", pg. 164, although I did neglect to qualify it as reciprocal altruism.


    Thus suppose an individual plays the Prisoner's Dilemma repeatedly against the same opponent.  A reciprocal altruist starts with the choice 'cooperate', and in subsequent games, continues to cooperate against opponents who have cooperated, and to defect against those who have defected.  If we assume a long sequence of games against each opponent, so that the payoff in the first game is a negligible part of the total payoff, we get the payoff matrix shown in Table 30.  Reciprocal altruist is the only ESS of this game."

    While "Tit for Tat" was shown to be an ESS, "Always Defect" was also, although it could be invaded by "Tit for Tat" while the reverse was not true.  There are variations where invasion can occur by defectors such as "Generous Tit for Tat" which is an unabashed cooperator.

    Since then, work by Martin Nowak and Karl Sigmund have shown that "Tit for Tat" wasn't quite as stable as originally thought, so it was replaced by a strategy called Pavlov ("Win-Stay, Lose-Shift") which was a bit more forgiving but responded better to errors.  It cannot invade an "Always Defect" strategy for which "Tit for Tat" is better suited, but it responds better overall to more "real-world" type of situations.

    As for positing that all populations will "expand to use all the resources available to them", the missing component is "if they could".  The fact that there are numerous other factors and species that are also competing for resources prevents the unconstrained growth that would lead to such shortages (precluding humans and domestic animals).  While exponential growth is the objective of many reproductive strategies, the reality is that most populations remain relatively stable.


    "Though these and all other species engage in massive overproduction (or superfecundity) and therefore could in principle expand exponentially, in practice they do not(1). The reason is simple: Most offspring that are produced do not survive to produce offspring of their own. In fact, most population sizes tend to remain relatively stable over the long term."

    "Understanding Natural Selection: Essential Concepts and Common Misconceptions", T Ryan Gregory April 2009

    (1) Humans are currently undergoing a rapid population expansion, but this is the exception rather than the rule. As Darwin (1859) noted, Although some species may now be increasing, more or less rapidly, in numbers, all cannot do so, for the world would not hold them.

    Mundus vult decipi
    Gerhard Adam
    Just FYI, I am aware that there are numerous other game theory strategies that can be involved, including the possiblity that individuals may not behave rationally, or that risks may not be equivalent between players.  In those cases, selfishness may provide a higher payoff if the individual can more readily absorb a risk that the other can't.  I've simply chosen the Iterated Prisoner's Dilemma as a simplying example.
    Mundus vult decipi