Reductionism Redux

As I’ve mentioned, I’m a great fan of Denis Noble, and recommend his book, the Music of Life, but you can get the content pretty well in total in this video of a lecture:

Principle of Systems Biology illustrated using the Virtual Heart

and there’s other material on his site.

So I was interested to see Ann Gauger making a very similar set of points in this piece: Life, Purpose, Mind: Where the Machine Metaphor Fails.

She writes:

In a recent essay Steve Talbott highlights the inadequacies of our current way of thinking and speaking about biology. He points out that organisms are more than the sum of their mechanisms. In fact, he rejects the machine metaphor as completely inadequate to describe living things. Living beings are adaptable and responsive to their environments, changing their behavior based on external cues and their own requirements. They are transformative, existing as entities that are much more than the molecules that compose them. They are not what they eat — they make what they eat into themselves. Living beings are integrated wholes that come from other living things. And they are more than their DNA. DNA requires a functional cellular environment to be properly read and interpreted, just as a cell requires DNA to be able to sustain itself. In order to understand the whole picture you have to look at the cell from many points of view, not just a gene-centric one.

Yes indeed.  After all, it is at the phenotypic level (extended, or otherwise – Dawkins might be famous for his Selfish Gene concept, but he also deserves credit for the Extended Phenotype concept) that “natural selection” takes place, in other words that when, as I tend to prefer to put it “self-replicators replicate with heritable variance in reproductive success”, it is the successful phenotypic variants that become more prevalent, carrying with them their load of genetic and other heritable stuff.

This is really important, in practice, when we find that although many human disorders are highly heritable (often well over 50%), it is rare for any one allele to have more than a tiny effect – the odds ratios risk SNPs found in GWAS studies tend to be extremely small – the vast majority of people carrying the “risk” SNP can are perfectly healthy.

But far from this making Darwinian evolution less likely, as Gauger thinks it might:

Maybe biological systems do reflect intelligent agency, because intelligent agents are the only known source capable of designing, assembling, and then coordinating so many interrelated sub-systems into a functional whole. And maybe, by acknowledging this, we can come to understand biology better.

it seems to me to make it far more intuitive.  After all, as Petrushka keeps pointing out, human intelligent agents are not very good at “designing, assembling, and then coordinating so many interrelated sub-systems into a functional whole” – what we tend to do is to go through a series of prototypes and upgrades, where we add things here and take off things there, until we end up with a Prius instead of a Horseless Carriage – something much more akin to evolutionary processes, where we discard the dud designs, and retain what works – breed from the winners, in effect.

And it also lets us view a genotype as a distribution of heritable chunks, where the phenotype reflects the summed effects, and where variation is pretty smooth.

 

38 thoughts on “Reductionism Redux

  1. Well thanks for the nod. My point is that humans don’t handle high-dimensional problems very well.

    Look at out buildings. We tend to build rectilinear boxes having known components with known properties. Construction gets very expensive when you try to build the Sydney Opera House (from the original plans), or the Gaudi Cathedral.

    When we design complex shapes we need to test them in wind tunnels.

    And the stuff we design is incredibly simple compared to something like a living cell. (Does Barry Arrington even read his own stuff?) One week he posts videos marvelling at the complexity of cells; the next week he is asking rhetorically if the space station is simple.

    Here’s a clue, Barry. the answer is yes. Space stations are pretty simple compared to biology. The analogy from human design is a non-starter.

    When we want to solve high-dimensional problems we turn to evolutionary algorithms. Just this morning I ran across Ant Colony Optimization algorithms.

    Biological evolution is pretty much a continuous process of optimization.

  2. This is really important, in practice, when we find that although many human disorders are highly heritable (often well over 50%), it is rare for any one allele to have more than a tiny effect – the odds ratios risk SNPs found in GWAS studies tend to be extremely small – the vast majority of people carrying the “risk” SNP can are perfectly healthy.

    It would not be unreasonable for the risk alleles still around in the population to be the ones that only have a tiny effect!

    I still think Noble misses the mark aiming at Dawkins. Dawkins’s genes lurk deep within the germline, mostly silent. The copies of those genes that contribute to complex multigene phenotypes by actually being expressed are not acting as ‘reduced’ evolutionary entities, but in their phenotypic role: switches, proteins, cofactors etc as part of the broader whole, entirely linked and with common cause, for this life at least. The fact that the interplay in a life is complicated and co-ordinated does not mean that a gene can no longer have a selective advantage all of its own, by virtue of its overall net effect from participation in a series of individuals, and profit or lose as an allele accordingly.

    Dawkins never says (to my knowledge) that individuals can be understood from the bottom up; it’s a strawman. But they are built from the bottom up … aren’t they? Well, Noble reads Dawkins’s statement “they build us, body and mind”, then immediately says “of course they don’t …” ! Yet epigenetic markups and wipes are generated ultimately from DNA; switches are DNA-encoded and bound by protein or RNA made from DNA, all of which are themselves ‘genes’ in Dawkins’s terms – stretches of DNA with independent evolutionary persistence. This is what organisms are made from: DNA is both source of product and source of control – including the products and control that enable its own replicative synthesis.

    To borrow Noble’s own metaphor, no one note gives a chord all of its richness. It is the interplay between them that counts. A new note may add discord, or a pleasing piquancy, which the selective milieu – the record-buying public, say – may favour or reject. But to argue that chords are not therefore built from notes would be perverse. Any gene has an environment – every other gene in the organism, and its own place in the wider environment. That environment is too complex to compute, but it is not too complex for the individual contribution of a stretch of DNA to have the result of advancing or retarding its representation at its locus, as a direct consequence of its own actions in that complex milieu. It either affects net offspring production or it doesn’t.

  3. Complex organisms are built from the bottom up – certainly. However, underlying structures serve only as part of a template, not a program. There is little that one can do to predict emergent behavior from bottom-up construction.

    Emerging properties occur so rapidly and in such profusion that self-interactions and interactions with the environment in which these structures are immersed start influencing subsequent development very early on.

    I go back to a point that is so familiar to those of us in condensed matter research; it is a routine part of our experimental procedures. Temperature dependence is an important clue to what is going on inside complex systems. That is why we do studies on condensed matter by changing temperature.

    The sinus rhythm that Noble mentions is temperature dependent. So is all neural activity. We are dealing with soft matter systems in which kinetic energies are comparable to binding energies. Kinetic energies are determined by the background temperature in which the system is immersed.

    How a system behaves with temperature is not only highly dependent on temperature; it is also very sensitively dependent of small changes in the concentrations of various atoms, ions, or molecules.

    I have never liked the “programming” metaphor. It doesn’t even work with much simpler systems in which we have been able to make connections between levels of complexity.

    Template is a better word than program; but even then, we cannot predict from bottom up what will happen as new phenomena emerge. Everything is interacting; and as “everything” gets more complex, so do emergent phenomena. New templates are emerging at every level.

    Place the same system in a different environment, change its temperature, change the concentrations of certain atoms or ions, and you get something that behaves entirely differently.

  4. His annoyance at Dawkins comes through – he seems at times to give vent to ‘exceptionalist’ bristling at Dawkins’s (deliberately?) provocative prose – the famed ‘lumbering robot’ bit. But rather ironically, his attempt to turn Dawkins’s prose on its head is simply to return to the biological view Dawkins was trying to shake – to stop thinking of the ‘good of the organism’ all the time. Important insights can be gained on evolutionary matters by looking at things from the gene’s point of view. Which is true, not ‘nonsense’ as he dismissively states.

    The fact that no-one could devise an empirical test to see if DNA or protein is more important in the modern cell also has the whiff of ID argumentation about it – Biped would find common ground! The evidence points clearly to the fact that nucleic acids predate the modern protein translation system. DNA is no prisoner; it is the only thing that escapes, even if in a cell provisioned with (DNA-derived) protein.

  5. In his book, he makes it clear (I thought he did in the talk as well, but maybe I was reading into it from the book) that he regards both the Dawkins version and his own inversion of it as polemic – neither one truer than the other.

    He is selling the value of the higher-level view, not arguing that it is more correct.

  6. I’d agree about the programming metaphor being misleading (they all are if taken too far!).

    But one of the things that mark biological systems out is that – broadly – their emergent behaviour tends to be the same. That is partly because they go to some lengths to generate a homeostatic internal environment, and also because external environmental fluctuation tends to be much slower than generation time. And, the genes that survive are more likely to be those that generate a consistent effect. Natural Selection can only happen if an allele finds itself consistently favoured, to give at least a directionality to fitness, if not an absolute value. If a gene tends to give phenotypes that are all over the map, that could be its undoing.

  7. Perhaps – it’s the sniffy dismissal as ‘nonsense’ that might lead the casual viewer to a different conclusion!

    But, as I say, I think he is mistaking the evolutionary viewpoint – which is definitely ‘more correct’ with DNA at the centre 🙂 – with the developmental/physiological one, where a whole menagerie of different interacting physical molecules is involved.

  8. Allan Miller:
    Perhaps – it’s the sniffy dismissal as ‘nonsense’ that might lead the casual viewer to a different conclusion!

    But, as I say, I think he is mistaking the evolutionary viewpoint – which is definitely ‘more correct’ with DNA at the centre :)– with the developmental/physiological one, where a whole menagerie of different interacting physical molecules is involved.

    I don’t think it is “more correct” with DNA “at the centre”. DNA is important, but it’s not even the only vector for inheritance. And natural selection – the core of Darwin’s idea – acts at the level of the phenotype, not the genotype.

    We downgrade the phenotype at our peril 🙂

  9. Lizzie,

    I think Dawkins regarded the gene as the unit of evolution. Variation takes place primarily in the genome. Most developmental variation caused by the environment does not get inherited and therefore cannot be regarded as important to evolution.

    I don’t think Dawkins meant much more than this.

    Edit to add:

    I don’t know enough to know if that sounds stupid, but I do think that the standard dogma is and always has been that evolution can only operate on heritable traits, and aside from the non-nuclear parts of the egg, the genome is what varies and gets inherited. Cellular mutations are much more likely to be fatal.

  10. petrushka:
    Lizzie,

    I think Dawkins regarded the gene as the unit of evolution. Variation takes place primarily in the genome. Most developmental variation caused by the environment does not get inherited and therefore cannot be regarded as important to evolution.

    I don’t think Dawkins meant much more than this.

    Edit to add:

    I don’t know enough to know if that sounds stupid, but I do think that the standard dogma is and always has been that evolution can only operate on heritable traits, and aside from the non-nuclear parts of the egg, the genome is what varies and gets inherited. Cellular mutations are much more likely to be fatal.

    Well, I’m thinking of things like epigenetics, symbiosis, and even culture. Language, for instance.

  11. Ok, I know this will sound like a really dumb question, but what I can say? I’m stupid.

    The “design inference” is supposed to take “it’s complex!” as a premise and yield “therefore, probably designed!” as a conclusion. (Never mind how the “inference” is supposed to work.)

    But what Petrushka said above caught me up short, because it seemed to me as if he/she were saying that the things we know are designed tend to be relatively simple, compared to, say, organisms. So this looks like a potential defeater for the design inference, that can take “it’s simple!” as a premise and yield, “therefore, probably designed!” as a conclusion.

    The Achilles’ heel of all this design stuff — from the Stoics all the way through Leibniz down to “ID theory” — has always been, “organisms and artifacts are exactly the same, except for all the ways in which they’re different”. On Petrushka’s view, now, it seems that “we designers” build complex things by emulating evolutionary processes, and that complexity is no guide to design at all.

    Is this what you folks have been trying to say all along, and I’m just kind of dense in only getting it now?

  12. Kantian Naturalist:
    On Petrushka’s view, now, it seems that “we designers” build complex things by emulating evolutionary processes, and that complexity is no guide to design at all.

    Is this what you folks have been trying to say all along, and I’m just kind of dense in only getting it now?

    No, you aren’t stupid, stupid – I’m really enjoying your posts at UD 🙂

    Petrushka has been saying for ages that the problem for ID is that the only way we know of for designing things anything like as complicated as an organism is to use evolution-like processes, including trial and error.

    I’ve been saying something parallel – that the way we design things has a lot in common with evolutionary processes – that brains themselves operate by a kind of “neural darwinism” by which networks that excite positive feedback loops “breed” – in both the longterm and short term, while those that excite negative feedback loops weaken. ( And like you, I don’t think brains “think” – organisms do, by means of their constant feedback interaction with their environment – by taking actions that bring new inputs into the system).

    My hunch is that the thing that scares – or incenses – ID proponents, is the word “random”. Which can mean so many things, from “unintended” to “flat distribution” (a meaning that seems to be current on UD right now) to “unmodelled” to “stochastic”, and probably a lot more besides. To me the big difference between design processes and evolutionary processes is that designers do things on purpose – in other words they are able to forward-model the potential outcomes of their design decisions, and feed back the virtual consequences as input. In other words they can simulate first, and act later. Evolutionary processes can’t simulate.

    And, I’d say, it shows.

  13. Yes, that all makes sense to me now.

    And for my part, allow me to also say that I agree substantially with a point you’d made — I believe you made, at any rate — about the whole “law of non-contradiction” fiasco. Which was this: the LNC is something one is committed to if one is talking about objects. If is not talking about objects, then one isn’t committed to the LNC. That’s why I kept on bringing up Buddhism: Buddhist logic isn’t committed to the LNC, and you know what? Buddhist metaphysics isn’t committed to the notion of an object, either!

  14. Kantian Naturalist:
    Yes, that all makes sense to me now.

    And for my part, allow me to also say that I agree substantially with a point you’d made — I believe you made, at any rate — about the whole “law of non-contradiction” fiasco.Which was this: the LNC is something one is committed to if one is talking about objects.If is not talking about objects, then one isn’t committed to the LNC.That’s why I kept on bringing up Buddhism: Buddhist logic isn’t committed to the LNC, and you know what?Buddhist metaphysics isn’t committed to the notion of an object, either!

    Well, I’m not sure I said it that clearly, and in any case, I was banned first and indicted later 🙂 I hadn’t even weighed on on the LNC shibboleth at the time I was banned (in fact I think the whole thing post-dated my banning). But I said later, here, I think, that, essentially, it depends what the meaning of is, is.

    So that provided the executioner with his post hoc charge 🙂

    I like what little I know of Buddhism. I still have a referent for the G word, it’s just not a super/non-natural creator.

  15. Actual intelligence seems to involve imagination. As you say, the ability to simulate action and imagine consequenses. Mammals in particular seem to learn through play. This may carry into dreams. Apes, including naked ones, seem capable of learning through daydreaming.

  16. In the bible it says Gods spirit gave life to the universe. So biology has a greater equation of glue then mere physics.
    Its not just parts working with well done connecting parts.

    I liked the idea that medicine etc needs to find new facxts and not just use old facts to make more intimate equations.
    I think creationism could better lead medicine in some presumptions.

  17. Robert Byers: I think creationism could better lead medicine in some presumptions.

    Given that you’ve had 2000+ years will that be happening any time soon do you think?

  18. Well, lots of creationists appear to think that.

    But none of them will get on and actually DO anything to prove their point, will they?

    And that’s because they know they can’t. Creationism has never added any knowledge to the heap – in fact, it is dedicated to the suppression of knowledge and learning.

    Which is, in turn, why it’s slowly withering away. Too slowly (but you’re helping, Mr Byers, every time you put finger to keyboard, so do carry on) , but definitely withering.

  19. Robert Byers:
    In the bible it says Gods spirit gave life to the universe. So biology has a greater equation of glue then mere physics.
    Its not just parts working with well done connecting parts.

    I liked the idea that medicine etc needs to find new facxts and not just use old facts to make more intimate equations.
    I think creationism could better lead medicine in some presumptions.

    How do you think creationism might do this, Robert?

  20. Why would an omniscient designer ever “imagine”, “model”, or “design” anything? Why would an omniscient creator ever be in doubt about the outcome of some system? How could an omniscient engineer ever make anything that was not optimal?

    These are questions that ID does not wish to face in public.

  21. Lizzie

    I don’t think it is “more correct” with DNA “at the centre”. DNA is important, but it’s not even the only vector for inheritance.

    The instances where it isn’t are rather trivial, IMO. Membrane, for example – there is an ‘immortal membrane’ scaffold into which, when membrane growth is required, subunits are placed. Some experiments have shown that perturbations can be carried over for a few generations. But eventually this fizzles out. The subunits are made by proteins made by DNA. And Noble’s pointing to the fact that the egg is provisioned initially by maternal proteins is, again, trivial. These proteins perform the first bits of expression of the new haploid pair, until the results of that expression take over. Yes, bits of the mother extend into the child – see also: milk, the placenta. If you mean cultural transmission, that’s a whole different ball-game, which Dawkins addresses via the concept of memes.

    And natural selection – the core of Darwin’s idea – acts at the level of the phenotype, not the genotype.:

    Yes, of course – Darwin was a bit hazy on the old genetics thing 🙂 But Dawkins knows all about phenotype. Selection looks at genes through a glass, darkly. But still, what ultimately survives or not as a result of this are genetic sequences. And they either affect offspring production or they don’t, as segments in a series of lives, however complex the interplay. If they elevate it, they are being selected (for). If they depress it, they are being selected (against).

    Take the music analogy – the sound of an orchestra playing a score is conditioned by the complex interplay, and each performance affected by the environment and the occasional unfortunate bum note. Audiences like it or don’t, and if the performance were the thing being mutated and passed from instance to instance, that would be different from a setup in which the score were being mutated and copied. And in this instance it is, clearly, the score. There are many cases where it makes no sense to look at things in this way – I am not advocating ‘reductionist physiology’, and nor is Dawkins. But I think there are strong mechanistic reasons to consider that nucleic acid, or something with similar properties, started the whole thing, and there is no point at which complex diploid phenotypes ‘took over’ to become the point of it all.

    Well, I’m thinking of things like epigenetics, symbiosis, and even culture. Language, for instance.

    However much Noble may put up slides showing various Weltanschauung redrawings of the interactions of genotype and phenotype, it remains my opinion that, partly thanks to the central dogma *** (which has a strict biochemical underpinning), interactions almost invariably reach outward from DNA. They may reflexively return to it – obviously, certain key functions such as DNA and RNA polymerisation have nucleic acid substrates, and the same goes for binding and initiation factors, or proteins that relax histone, or methylate and demethylate key sequences. But I’m not aware of any ‘epigenetic’ factor that does not have its root in a nucleic acid sequence. This notion of ‘downward causation’ is simply looking at the back half of a loop. It all traces back to a nucleic acid allele of some kind.

    *** the only half-exception I can think of is the telomerase sequence, though even here it is not an exception on Crick’s original formulation – it is not protein sequence that dictates the telomerase sequence.

  22. Looking at the musical analogy, there is the score and the performance.

    Performances have many factors: the skill of individual musicians, the conductor, the hall, the concertmaster, the quality of the instruments themselves, the audience, and so forth.

    Many of these factors can be inherited as fashions and so forth, but at some level, a piece is defined by the score and not by fashions in performance.

    Scores have been known to mutate. Sometimes scores speciate. There are multiple versions of Petrushka, for example.

  23. Something I found fascinating when I heard it was the idea that top piano players don’t actually know which finger it is that presses each note. Someone had to be filmed to work it out.

    If actually true, and not a dream I had, fascinating. Not totally sure why, or what it implies apart from (to me) the idea that consciousness is not a thing that I experience as a being e.g. “I am conscious of playing the piano” or “X+PIANO = XPIANO” but rather that you are a consciousness playing the piano and nothing else.

    “PIANO”.

    That iteration does not know of “keys” and “fingers” I suspect.

  24. An old fable:

    A spider was trying to catch a millipede, but the millipede outran him. Finally the spider called out to the millipede, “With so many legs, how is it that you are able to move so many of them all at once and run so fast?”

    The millipede began to think about it, and suddenly he couldn’t keep his legs moving in the right order; so there he sat.

    And the spider caught up with him and ate him.

  25. Something I found fascinating when I heard it was the idea that top piano players don’t actually know which finger it is that presses each note. Someone had to be filmed to work it out.

    Even crap ones like me don’t know! I never thought I’d get the hang of hand independence, but thought I’d try when my mum gave me the bin bags of music she’d amassed, due to arthritis. I’d sit there for hours trying (perhaps starting with a Mozart sonata was over-ambitious!), go away frustrated, but a few hours later something had happened in the brain, and my fingers seemed to find the notes in counterpoint almost by auto-pilot. Weird. I enjoy playing just for that sensation, of doing something puzzlingly complex without ‘knowing’ how.

    I saw a documentary about a guy whose long-term memory was shot, didn’t recognise his wife and swore blind he couldn’t play, but sit him down at the keyboard and away he’d go. Learned motor skills, I guess; different part of the brain.

  26. Something I found fascinating when I heard it was the idea that top piano players don’t actually know which finger it is that presses each note. Someone had to be filmed to work it out.

    There are many examples from music that demonstrate concepts like dynamic emergence.

    For example, Chopin’s “Fantasie-Impromptu Op. 66” sounds entirely different when played up to speed than it does when played slowly.

    There are similar “musical illusions” on the classical guitar. For example, the tremolo in Agustin Barrios Mangoré’s ”Un Sueño en la Floresta” is only three of the same notes in quick succession with the “ring”, middle, and index fingers of the right hand (a, m, i).

    Similar accompaniments to melodies can be achieved by underlying patterns that separate the melody from the accompaniment, such as in Stanley Meyer’s “Cavatina”.

    Or as in Fernando Sor’s “Etude Op. 6, No 11.” (Here a little slower than I think it should be played; but it’s ok).

    The melodies in these last two emerge from a rather consistent set of patterns of (i, m, a) or (p, i, m, a) or (a, p, i, m) with the right hand.

    I discovered after I learned these last two pieces that there were suddenly a lot of other things I could do that I didn’t know would follow from having learned these. It made learning other pieces much easier.

    As I am sure anyone who plays Chopin’s Fantasie-Impromptu (I cannot) will understand, when you get to a stage where such patterns flow effortlessly, the player can actually feel the separate parts and automatically dynamically adjust the pressure on a key or string in order to maintain balance between accompaniment and melody.

    If you think too carefully about it after getting to this stage, you loose the ability to balance the parts and bring out the expression and melody.

    You do the tough, analytical work during practice; and you focus on only particularly difficult parts. And it turns out that working out efficient fingering and transitions for both the left and right hands pays enormous dividends when the piece is finally played up to speed. It becomes so automatic you no longer know which finger is doing what while playing; you just feel the parts and can blend them at will.

  27. Allan Miller: Even crap ones like me don’t know! I never thought I’d get the hang of hand independence, but thought I’d try when my mum gave me the bin bags of music she’d amassed, due to arthritis. I’d sit there for hours trying (perhaps starting with a Mozart sonata was over-ambitious!), go away frustrated, but a few hours later something had happened in the brain, and my fingers seemed to find the notes in counterpoint almost by auto-pilot. Weird. I enjoy playing just for that sensation, of doing something puzzlingly complex without ‘knowing’ how.

    I saw a documentary about a guy whose long-term memory was shot, didn’t recognise his wife and swore blind he couldn’t play, but sit him down at the keyboard and away he’d go. Learned motor skills, I guess; different part of the brain.

    Clive Wearing. I knew him (slightly) before it happened – a great musician. More than just motor skills, I think – the music is all still there too.

  28. “what we tend to do is to go through a series of prototypes and upgrades, where we add things here and take off things there, until we end up with a Prius instead of a Horseless Carriage ”

    Brilliant, all the work of Nicolas Carnot was found over here and over there and put toghether to make a horseless Carriage with an engiene.
    It can´t fail no matter how smart a darwinist seems sonner or later will show his blindness.

  29. That was a description of how human designers work. It’s not what we observe in biology, is it?

    There are no “jumps” from ‘horses’ to ‘cars’ in the fossil record. Feel free to point one out, that presumably would be when your “intelligent designer” intervened?

  30. Blas:
    “what we tend to do is to go through a series of prototypes and upgrades, where we add things here and take off things there, until we end up with a Prius instead of a Horseless Carriage ”

    Brilliant, all the work of Nicolas Carnot was found over here and over there and put toghether to make a horseless Carriage with an engiene.
    It can´t fail no matter how smart a darwinist seems sonner or later will show his blindness.

    Don’t over interpret my comment! In fact, I do think there are key differences between the work of intentional designers and evolution, and one of them is that intentional designers can make intuitive leaps (from horse to Internal Combustion Engine, for instance) and also jump lineages (make an Internal Combustion Engine drive a generator, for instance, or a lawn mower, or a chain saw). But my point is that to get a really complex, optimised version, we go through lots of prototypes and iterations, discarding the least successful ideas and “breeding from” the better ones – in other words, our optimization processes bear a fair resemblance to evolutionary optimisation processes.

  31. , our optimization processes bear a fair resemblance to evolutionary optimisation processes.

    Also the more counter intuitive designs like the aerodinamics in F1 cars are made by planned trial and error in devices like wind tunnels and computerized simulators and that is far from the darwinist optimisation process.

  32. Blas: Also the more counter intuitive designs like the aerodinamics in F1 cars are made by planned trial and error in devices like wind tunnels and computerized simulators and that is far from the darwinist optimisation process.

    Well, trial-and-error is very like “the darwinist optimisation process” – in fact that’s exactly what it’s like, and leads to “counter-intuitive” designs!. The design process you mentioned sounds interesting – can you tell me more?

    Remember, I’m not trying to say that human design processes are like evolutionary processes – I think there are distinct differences that are apparent in products, for example, in the distribution of features over lineages. But not a million miles away from each other either.

  33. One of the techniques used is entirely Darwinian – optimisation of aerofoils can be performed by randomly varying from a poor start point to create a population of marginally modified variants, and selecting from the variants those that outperform the others according to some metric. Do this for a number of iterations and you have allowed “RM + NS” to do your designing for you. Wonder where they got the idea from?

  34. I’ve read Derek Noble’s document that Dr Liddle referenced in another thread on the “reductionist problem” but now I can’t find it. Please indulge me by posting the link again.

  35. timothya:
    I’ve read Derek Noble’s document that Dr Liddle referenced in another thread on the “reductionist problem” but now I can’t find it. Please indulge me by posting the link again.

    Here is the one I linked to, but there are more listed here.

  36. Lizzie: Well, trial-and-error is very like “the darwinist optimisation process” – in fact that’s exactly what it’s like, and leads to “counter-intuitive” designs!. The design process you mentioned sounds interesting – can you tell me more?

    Remember, I’m not trying to say that human design processes are like evolutionary processes – I think there are distinct differences that are apparent in products, for example, in the distribution of features over lineages. But not a million miles away from each other either.

    Well, they could equally well be made by UNPLANNED trial and error, constrained only by what variations are actually physically possible in the system; there would still be an optimised results, but there would have been a lot more “wasted” effort and time.

    Biological evolution is not generally constrained by time available – when it is so constrained, say by rapid environmental change, extinction may result.

Leave a Reply