Log in

The Synergetic Perspective on the Problem of the Level(s) of Natural Selection

My good friend Steve, over at ZARBI (it's linked on the right) has posted about the question of levels of selection, advancing a phenotype-centric viewpoint.

This is one of the most discussed and debated questions in the philosophy of biology – and many highly competent people have struggled with it over time. Based on my relevant previous study of philosophical questions, I have now studied the major peer-reviewed publications on these subjects. In addition, Steve (over the net) and my girlfriend have taught me so much about biology that it is hard for me to find words to express my gratitude.

From this position, I have begun to formulate my own takes on these questions. They are detailed in the comments to steve's blog-entry, but here is a summary of it, with some additional points. Have at it:

To put it in a nutshell – the problem with genotype- and phenotype-centric views as well as group-selection approaches to the question of the 'level of selection' is that they rely on a conception of the environment that is only something in and against which the success of the individual is measured, not as something the salient properties of which are at least partially determined by the actions of the organism itself, the group it is part of, their ancestors and other lifeforms past and present, because those have shaped and continue to shape the environment in which the organism and its group/relatives find themselves in.
In turn, these antecedent causal factors in the determination of the differential success of the individual organism, its traits and/or coding nucleotide-sequences are determined by factors ranging from purely environmental variables to the idiosyncratic contributions of phenotypes, specific phenotypic traits and also including specific coding nucleotide sequences. Each of these factors in turn has all these factors as its determining conditions.

This means that the causal conditions for differential success (a far better term than 'natural selection' because of its conceptual implications, as I argue in my comments to Steve's blog-entry) of a given identified explanans cannot be pinpointed to any one level of organization – neither the molecular gene-level, nor the levels of phenotypic traits, nor of idiosyncratic causal contributions by the behaviour of organisms or groups.

One approach that successfully eliminated some of the conceptual inadequacies at least in response to certain problems, was and is that of niche construction. It should be evident that it laid bare the crippling inadequacies of pure adaptationism and its inappropriate concept that all the environmental factors (including the contributions of other agents) that together with the genes determine the life of the individual and thereby its differential success are somehow a given for the individual, and not causally affected by it.

But in the long run, I think the most appropriate way to approach the issues is the following:

We need to effectively apply the realization that a hugely complex network of interactions between very different factors are causally responsible for differential success.1 This, and the omnipresence of problem of reductionism vs holism and levels of description vs levels of ontology – leads me to conclude that the only approach with a chance of be successful in finding detailed and reliable answers to any of these questions must take the form of describing through detailed mathematical-mechanistic models the patterns of organization on different levels of organization (atom - molecule - cell - organ - organism - groups - ecosystem) and how the organization on each level relates to the organization at the next higher and lower level. This is a huge, trans-disciplinary research-program, and it is one that would upon final success result in the complete unification of science. As such this final goal is unrealistic. But, seeing as we continue to find ever more pieces of this specific puzzle, it appears that this is not only an ideal to which we should aspire by means unknown – it as possible to get underway and improve our picture of the world along these lines.

As I said, we need to describe and relate mechanisms and patterns of organization in and between levels of organization. Here we have to realize that for any real system whose organizational features we seek to explain, we can expand the initial spatiotemporal boundary that our concept of the system implies (in a detailed or fuzzy way) – including every more of what was initially considered the environment of the system with respect to the features or set of features of which we seek to explain.2 While doing that, more and more of the causal determinants of specific phenomena, traits, features of the system (those we seek to explain) get included. At some point of extension, of inclusion of factors into our definition of the “system”, that which seek to explain can meaningfully be described as self-organizations, with the tools of synergetics, including control theory, thermodynamics, dynamical systems science in general.
Thankfully, this is already being done to great success in the higher-level sciences – from biology to computational neuroscience and psychology. By adopting this stance, we can describe organization in ever more precise (and ever more 'mathematical' terms) the lower we descend among the levels to the physical.
Some of the ever more abundant instances of successful application of system-theoretic, synergetic thinking I'm thinking of is the kind of organizational studies done by applying mathematics to describe the organization of the Belouzov-Zhabotinsky reactions, the cellular-automaton descriptions of population-dynamics, cell-growth, the deterministic-finite-automaton-model of ant-trail formation and following and even such things as the computational neuroscience models of features of biological systems.
The book “Self organization in biological systems” by Scott Camazine et al describes and explains some such successful applications in biology. The work of Teuvo Kohonen on self-organizing neural networks (self-organizing feature maps or SOMs) and the field of inquiry it has spawned (including the study of growing SOMs, or GSOMs) is where you can see the successful application of this perspective to problems in the fields of neuroscience, cogntive sciences and philosophy of mind. And finally, the authoritative work on the mathematical models, derivation of formulae for the description of complex situations by combining and re-forming the mathematical formulae for its constituents as well as the foundational work of systems-theory and synergetics in general is “Synergetics” by Hermann Haken. It's expensive, but if you're math-savvy enough, you'll be able to re-trace a considerable amount of important insights that are not commonly known as such, but are of essential importance.
From the system-theoretic, complexity-theoretic and (thus, effectively) synergetics-perspective, we can relate formal models - and once we manage to relate the models among and between the various levels of organization, we will be able to get to the answers we're looking for.

The relevant literature to judge the relative merit of the biological dimensions of my position can be found in the various University-Press 'Guide's, 'Companion's, 'Introduction's and 'Reader's on the Philosophy of biology. We have to see that the questions of levels of selection, adaptionism vs niche-construction etc, reducibility of the 'gene' concept and so forth – these are all philosophical questions. Biology, like so many other sciences that arose in that time or earlier, and especially evolutionary biology arose from natural philosophy – and many of the philosophical questions are actively researched by people in both fields. It's inter-faculty work, but it's not inter-disciplinary, because the actual research on these questions is made by people with degrees in biology and philosophy. Scholars of both disciplines are published by the same peer-reviewed journals on these subjects and are highly regarded by the same people. Here, as in some fields of the cogntive science, there is no operative distinction between philosophy and the special science. Or think about the evolution and evolvability of culture, language, and mentality in general as well as specific aspects of it – there is no operative distinction in these questions, either.
That's why I feel both that my congenial discussion with Steve on this subject can be very educational for me, and also that my own study enables me to conduct fruitful original research that is appropriate with respect to the intricate complexities of the these issues, as it underlies my comments on Steve's article.

1 I think the most appropriate conception of a causally responsible factor for something is that advanced by late, eminent J. L. Mackie: A causal determinant (in case of a real situation, a 'causal condition' in general) is an INUS-condition, that is an Insufficient but Necessary part of a conjointly Unnecessary but Sufficient set of conditions for a given particular effect. For more on this, see the wikipedia-entry on INUS-conditions and the relevant articles on the stanford online encyclopedia of philosophy.
2 Of course the expansion of the spatiotemporal boundaries are bound by the speed of light as it defining the section of spacetime that can causally affect a given event/system at a certain state


I think that although all kinds of aspects of an organism can determine differential survival, there is quite a simple criterion for determining at any given time which is the most important factor, and that is which factor can change the least before there is a noticeable difference in survival. That is the phenotype.

I get a feeling that what you are doing here is mixing up how a phenotype arises (in that case, things like self-organisation are certainly important), and what is actually selected: what makes for a difference in survival.
I get your general point, but my concerns are not dispelled.

for determining at any given time which is the most important factor, and that is which factor can change the least before there is a noticeable difference in survival. That is the phenotype.

I think I will have to respond in a two-part answer:

* I'm not so sure if this is true.
As I mentioned, the questions of natural selection can only be meaningfully construed as the question for the causes (INUS-conditions) of differential success of some property or other at some level in the domain of living things.

We can ask what the causally responsible factors are for the overall abundance of a certain nucleotide sequence, or what the causes are for the overall abundance of a phenotypic trait (again, frequency times population size, or each of these things considered separately).

Alternatively, we might ask for the causes of differential succes of certain individual traits in populations, or in the environment (again the overall replicative success in the biosphere), or we might ask for the causes of changes in frequency of a certain trait in a population, or a nucleotide sequence. Or we might ask why one organism produces more surviving offspring than other organisms of the same species living at the same time in roughly the same situation. These are all different questions - and the last one (and some of those before) again seems to me to be only meaningful upon similar analysis. The causal conditions for an organism in a group in an environment to produce more offspring than rivals is not a meaningless one - but again I don't see how the exact causal factors responsible for this differential reproductive success of an individual can be limited only to phenotype-properties. Even if that were so, any property at the phenotype level is again caused by complex interactions in and between different systems. The same is true for all similar questions.

In every case where there is a phenotypic traits that facilitates survival because a change in conditions occurs that makes this trait favourable, the cause of the differential success is not localizable solely in the trait and its causal antecedents. The other INUS-condition, the other cause for the differential success is the change in environment that makes having the trait more beneficial than not having it. Without this, there would have been no differential success.
Thus it is a major causal condition not localizable on the phenotype level.

So, given that the question of differential (reproduction-)success is only meaningful when we can specify it to questions of the above kinds - I don't see how the answer to any of these questions can be said to work only on the level of the phenotype without oversimplification.

Peter Godfrey Smith explains some of the things I'm trying to say better than I can - for example in the essay "The Replicator in Retrospect" on his page at Harvard: http://www.people.fas.harvard.edu/~pgs/Philosophy%20of%20Biology.html

(the second one)

My position is a little different from his, especially in my insistance on formulating the question as a specifc question of the form "what are the causal conditions for differential success of A", where we need to specify what we mean by success (frequency of X in population P, or frequency times population size etc) and what A is - a 'thing' like a nucleotide-sequence or a complex macroscropic trait of individuals, or a certain change in frequency or abundance etc). Except where the consquences of this clash with his view, I think I can agree with Peter Godfrey-Smith.
Damn, while writing, the 'two part answer' became a multi-part answer and I forgot to change the sentence below the quote... Sorry.

I think I ought to mention that when I say I find the centric points of view inadequate, I mean (in virtue of what I have explained above about the specific forms the questions have to take to be meaningful) that localizing something on some level might be appropriate, but only to very specific questions, and not the general one of 'what are the causes of differential success'.

The 'levels-of selection question' becomes inadequate. It asks 'what are the kinds of 'things' that nature selects for, and on which level of organization are they found'.

This is a meaningless question, because although relatively inconspicuous, the problems arise becaue it impless that there is something that selects among one kind of 'things', even if biologist will vehemently deny that they affirm this when asked (and for good reason).

In this world, there are changes in certain factors that concern the biosphere. These factors may be very high-level, complex regularites that undergo changes, or rather simple ones, or anything in between. Some traits are more abundant than others, some nucleotide sequences are, some properties of groups are more successful than others.

For every such high-level property, we can ask meaningfully how it came to be established - and we can ask meaningfully what the causes of any change in any of these factors are.

In light of this, the insight of evolution by 'natural selection' comes down to what it was initially understood by Darwin to mean - traits are at least partly regulated through mechanisms of inheritance and themselves contribute to the passing on the herediatary information in the organism having any trait. The result is diversification of life, and the hypothesis of common ancestry.

But we can ask meaningfully about the causes for changes in frequency and abundance of any regularity at any level of biological organization, and for every one such question there will be a different answer.
"The 'levels-of selection question' becomes inadequate. It asks 'what are the kinds of 'things' that nature selects for, and on which level of organization are they found'."

I think it is a perfectly reasonable question, and the answers in many cases are very clear. We see such answers in convergent evolution, such as the close shapes of sharks and dolphins. We know what is selected, and we know the organisational level at which it is selected.
I still think you are drifting away a bit from the key idea of selection (or differential survival).

I think you are making a subtle mistake in bringing in changes in the environment in the way you do. What we need to consider in this matter is what can and can't change in an organism in each generation to enhance survival.

"The causal conditions for an organism in a group in an environment to produce more offspring than rivals is not a meaningless one - but again I don't see how the exact causal factors responsible for this differential reproductive success of an individual can be limited only to phenotype-properties."

There are nothing but phenotype-properties that are exposed to the environment. That is my point! The environment does not react to DNA sequences, because it can't read them! It can only react to consequences of those sequences - the phenotypes(*). Or, rather, changes to them.

(*) Actually, there are some cases where sequences directly have an influence, such as in the physical stability of DNA, but they are rare.

I'll respond a bit more after I have read that paper. I like what I have read so far. I agree that a failing of Dawkins is that he allows too much use of language that implies agency.

Again I have to apologize for presenting my thoughts in such a disjoint manner, but I just remebered that I forgot to add an important point to my answer to your passage "for determining at any given time which is the most important factor, and that is which factor can change the least before there is a noticeable difference in survival. That is the phenotype."... namely what I initially intended to be the second part of the response to that:

Sometimes, major causal factors for differential success (as differential reproductive success of the individual or some other specificiation) cannot be traced to phenotypic, genetically inherited traits, but to learned cultural behaviour (i.e. socially inherited).

Perhaps you remember the recent story about a dolphin population where females had learned simple tool-use in hunting and passed this on to their offspring. The result is, naturally, that the set of individuals who have learned this can exploit the food-resources more effectively. The population size of this group, and the group whose survival is affected by their success (usually but not necessarily family-members) and the dynamics therein will benefit from this developed and passed-on learned behaviour, as will the frequency of the trait. Both, of course, relative to any similar-sized population in an environment with similar food-resources and otherwise relatively stable environemental conditions.

Because once we think in terms of specified differential success, there is a specific application of the term 'fitness' to each specific target of the question. For everything under considerations, factors can be specified that contribute to an increase in frequency and/or overall abundance of that specified target 'thing'(/property) at the designated level of organization.

So there is no general 'fitness', only conditions relevant to increase of frequency and/or overall abunance in some specified target - differential success in becoming replicated. This, again, applies no matter what target we chose - nucleotide sequence, macroscopic trait of individuals or properties distributed over groups and populations.

We can see, most prominently in ourselves, but also in the observed developments of proto-cultural behavioural strategies of various complexity in other species that the evolutionary rise to success of centralized nervous-systems, especially the evolution of the neocortex has payed off. Among other things it has lead to learned and socially passed-on adaptive behavioural strategies that affect the dynamics of population-size positively - and this effect is relatively stable over generations.

As such, in the above specified way, we can say it increases the 'fitness' both of the individuals having the ability and the whole set of individuals that benefit from their improved success in dealing with certain circumstances.

Of course, all of this has consequences for all the debates where the (as I hope to have explained, only partially meaningful) 'levels of selection'-question has been thought to be relevant, like the question of altruism. Hmm... I think I'm gonna deal with the special cases later.

I justed wanted to make the general point that non-phenotypic, but learned and socially passed-on traits can play a major causal role in increasing the differential success of specified targets from individuals' chances of survival to positive changes of traits and population-size.
Allow me to address your points from the various comments in one comment of mine:

"We know what is selected, and we know the organisational level at which it is selected."

Yes, and in virtue of that, these are specific cases. But depending on the case in question, the organizational level will be different, and the contributing causal factors will be different, too. That means that there is no one level of selection. And that's why the general question doesn't make sense. I agree with what you say - but what you say shows that the question can be answered and the organizational level can be identified.

The question "at which level of organization can we most readily locate a minimal set of changes that form the causal 'nexus' of factors for phenomenon X" is specified and makes sense because of that. So the facts you so rightly mention confirm what I tried to say.

I think you are making a subtle mistake in bringing in changes in the environment in the way you do. What we need to consider in this matter is what can and can't change in an organism in each generation to enhance survival.

That specific question is valid, but there is no mistake in recognizing that is one (kind of) question of a large group of relevant questions - and I see no reason to only consider survival of individuals - we need to consider various things, like frequency of traits of individuals, abundance thereof, changes in population size relative to external factors etc.

Each view - the gene-centric, kin-selectional and phenotype-centric view considers hugely different aspects, all of which valid, but - as they are conceived as answers to the 'levels-of-selection'-question, they cannot be upheld because they are too narrow, because different organizational levels are important for different legitimate questions.

There are nothing but phenotype-properties that are exposed to the environment.

If we understand a phenotype-property to be a property for which specific genetic information (and the mechanisms expressing it) is responsible, then I'm sorry, but I have to disagree vehemently - In light of my example demonstrating how learned and socially passed-on behaviour without being genetically inherited can strongly affect differential success - as chances for individual survival, as chances for long term success of the individuals who are taught the behavioural strategy and as chances for the stability or growth of popoulation size.

Aside from that, the statement is trivially false in that the individual organism interacts as a whole with the environment. All its properties play a part in the interaction with the environment and not all properties of the individual arise from gene-expression. There are no 'isolated' properties - because that would mean attributes without effects, which is a contradiction.
I would also disagree with me if I thought phenotypes were purely due to genetic information! But I take a FAR wider view. Genes can clearly influence the phenotype, but certainly don't determine it in any way fully - sometimes not even significantly.

Perhaps I have been using the wrong term all along, but I understood phenotype to mean "traits and physical characteristics of an organism or group".

What I was trying to distinguish between was the information in the genes, and what an organism ends up like, much of which can be nothing to do with genes! That is where things like self-organisation can come in.

It makes it even less sensible to talk about gene selection when much of which is selectable isn't produced by genes anyway.
Great papers on that site you mentioned.

I have been reading:


"Earlier I discussed Dawkins and Dennett as explanatory adaptationists. What might
their replies to this argument be?
As I understand both writers, they aim to take the high ground in this debate; they
hold that apparent design is, as a matter of objective fact, a special phenomenon whose
explanation is central to the task of biology. For both writers, this view has two components.
The first is an argument that apparent design is a real phenomenon, and the second is a
claim that this phenomenon presents special problems for scientific, secular world views."

That last sentence is precisely what I object to about Dawkins' "Mount Improbable", and his repeated claim that life and complexity is "improbable". Those are, I think, skewed ways of looking at things.


totally off topic

I need the opinion of a philosopher.

I argue that "integrative medicine" is not rational (http://tuftedtitmouse.blogspot.com/2009/03/warning-reeding-bad-fur-iq.html). In brief:

1. "Conventional medicine" = set of therapies proven safe and effective.
2. "CAM" or "complementary and alternative medicine" = therapies with insufficient evidence to be deemed safe and effective.
3. "Integrative medicine" = combo of conventional medicine + CAM therapies shown to be safe and effective.

The above seems obviously stupid and contradictory. But apparently not so obviously stupid to leading medical schools in the US over the past 10 years, which have formed departments of integrative medicine.

The only way I can make the above work is by weasling around with the word "proven." Maybe the CAM in #3 isn't the same as the CAM in #2. Maybe #3 is the cream of the CAM crop, yet still not backed by sufficient evidence to count as "conventional medicine."

Well, the cream of the crop includes homeopathy and Reiki. So I'd hate to see the evidential basis for the real dogs in that set.

If a smart philosopher guy like you were to confirm that "integrative medicine" is irrational, surely the world would listen. Eh?


Resume Writing Service

My friend suggested me to visit your blog. Very well explained. I would like to say that it is very interesting to read your blog.

Resume Writing Service (http://www.resume-service.biz/)


great essay about modern furniture: http://www.squidoo.com/modern-furniture-max

Hello Friends!

Found this great piece on contemporary furniture here : Enjoy


also a great youtube channel on the subject:


and some more articles





vjrjlsdc mvqy

masj mknou porno (http://www.miasporn.com) fdjabu g ze o hsi


Trial, righteous a check up on

Hello. And Bye.


Trial, righteous a evaluate

Hello. And Bye.
This article was vey helpful to me. It helped me to be informed and more aware. The details were such a blessing, thanks.

Beneficial info and excellent design you got here! I want to thank you for sharing your ideas and putting the time into the stuff you publish! Great work!