Frustration

In this essay Richard Dawkins proposes the following:

In fact, natural selection is the very opposite of a chance process, and it is the only ultimate explanation we know for complex, improbable things… We need a better explanation [than design by space aliens], such as evolution by natural selection or an equally workable account of the painstaking R&D that must underlie complex, statistically improbable things.

An equally workable account? An “ultimate explanation”? R&D? R&D is research and development. R&D is design. The logic and terminology of design is inescapable, even by those who deny that design exists.

Richard Dawkins is certainly not a stupid person, but I find it amazing that he cannot see the obvious problem here. Natural selection is not random, but it does not create anything; it only throws stuff out.

The F-35 fighter aircraft (for which our company is designing a new pilot ejection parachute), did not come about by throwing out the Wright Flyer biplane, and then throwing out the Piper Cub, and then throwing out the F-16. The impotence of natural selection as a creative force is transparently and logically evident.

Dawkins:

[It is proposed that] He [God] was always there and always complex. But if you are going to resort to that facile cop-out, you might as well say flagellar motors were always there.

But flagellar motors were demonstrably not always there.

The who-designed-the-designer question is nonsensical when one considers that time had a beginning. This is also transparently and logically evident.

The physical universe (matter, energy, space, and time) came into existence at a finite point. Dawkins must surely admit this. Whatever caused the physical universe did/does not exist in time, because time did not exist. Therefore, the cause of the universe has no past and has no history. That which does not have a past or a history has no point of origin, and thus no designer or originator. The infinite regress logically stops at the origin of the time domain, which is well established by modern science.

None of what I’ve written here is hard to understand. In fact, it’s logically trivial. How the “Professor of the Public Understanding of Science” could have missed these obvious points astounds me.

Is my frustration understandable?

See here for an insightful rebuttal of Dawkins’ logic and argumentation.

  • Delicious
  • Facebook
  • Reddit
  • StumbleUpon
  • Twitter
  • RSS Feed

38 Responses to Frustration

  1. I think another point worth considering is, if natural selection really were capable of ‘creating’ so much of what we see in the natural world, it too would just be a collection of processes and systems within the natural world. In other words, one more tool far more plausibly developed by the mind of a creator than one which just happened to come into being by (there’s that word again) chance.

    What allows Dawkins to feel he’s an “intellectually fulfilled atheist”, if it can accomplish all he hopes and claims, would more neatly fit into a design-centered view of the world than a non-design view.

  2. I think, the actual “creative” force is mutation (including gene duplication). It provides the raw material on which natural selection can act.

  3. Disregarding that the universe had a beginning, we still might reasonably ask why it exhibits the properties that it has, i.e. those conducive to the support of life. I understand the odds seem to be very much in favor of it not exhibiting such properties. The improbability of the specified conditions make it look very much like design.

    Furthermore, I see no good reason at all why we can’t propose an undesigned and eternal creator even if he is complex. This argument of Dawkins’ has never looked very good to me.

  4. G’day crandaddy,

    It’s interesting to see Dawkin’s write of “He [God] was … always complex” The God of the Bible, as I have come to see, is not one of specified complex information. God is not complex but simple. This may seem opposite to many of the bloggers’ thoughts, and this in no way takes away from the complexity that has been and will be wrought in our universe, but as to the Designer’s specified complexity I don’t think God fills that bill … and I hope He doesn’t mind me saying that [thunderclouds start milling overhead ...]

    Dawkins has been here before and philospher Alvin Plantinga has written on this point:

    ” ‘Not much. First, is God complex? According to much classical theology (Thomas Aquinas, for example) God is simple, and simple in a very strong sense, so that in him there is no distinction of thing and property, actuality and potentiality, essence and existence, and the like. Some of the discussions of divine simplicity get pretty complicated, not to say arcane. (It isn’t only Catholic theology that declares God simple; according to the Belgic Confession, a splendid expression of Reformed Christianity, God is ‘a single and simple spiritual being.’) So first, according to classical theology, God is simple, not complex.

    ‘More remarkable, perhaps, is that according to Dawkins’ own definition of complexity, God is not complex. According to his definition (set out in The Blind Watchmaker), something is complex if it has parts that are ‘arranged in a way that is unlikely to have arisen by chance alone.’ But of course God is a spirit, not a material object at all, and hence has no parts. A fortiori (as philosophers like to say) God doesn’t have parts arranged in ways unlikely to have arisen by chance. Therefore, given the definition of complexity Dawkins himself proposes, God is not complex.
    ‘So first, it is far from obvious that God is complex.

    But second, suppose we concede, at least for purposes of argument, that God is complex. Perhaps we think the more a being knows, the more complex it is; God, being omniscient, would then be highly complex. Perhaps so; still, why does Dawkins think it follows that God would be improbable? Given materialism and the idea that the ultimate objects in our universe are the elementary particles of physics, perhaps a being that knew a great deal would be improbable—how could those particles get arranged in such a way as to constitute a being with all that knowledge? Of course we aren’t given materialism. Dawkins is arguing that theism is improbable; it would be dialectically deficient in excelsis to argue this by appealing to materialism as a premise. Of course it is unlikely that there is such a person as God if materialism is true; in fact materialism logically entails that there is no such person as God; but it would be obviously question-begging to argue that theism is improbable because materialism is true.”

    Sorry for the long post, but I think Plantinga has some salient points. Since this is a blog about complexity I envision some complex replies!!!

  5. I would say it can be argued that Natural Selection is evidence for intelligent design, for it takes intelligence to compare the various genes then “select” for the best choice given the current environment.

    This selection was programmed into the organism to allow its descendants to adapt to changing conditions.

    This is purposeful, intelligent design.

  6. biliiad, no beancan5000 is probably not a troll. He’s just confused. I’ve seen this mistake many times.

    Beancan5000, the whole point of natural selection from the perspective of many (probably most) scientists is that it is NOT an intelligent process. Remember, for any event there can three causes 1. law, 2. chance or 3. agency.

    Natural selection is “law.” It works on random mutations (“chance”) to produce new species. At least that’s the claim. By definition, a product of pure chance and law cannot be the product of “purposeful, intelligent design.” Indeed, the one excludes the other.

  7. Lee Spetner Addresses a theory of environmentally driven mutations, in his book Not By Chance,

    http://www.amazon.com/Not-Chan.....1880582244

    I find his theory has much support in empirics.

  8. The Origin of Theoretical Population Genetics (University of Chicago Press, 1971), reissued in 2001 by William Provine:

    Natural selection does not act on anything, nor does it select (for or against), force, maximize, create, modify, shape, operate, drive, favor, maintain, push, or adjust. Natural selection does nothing….Having natural selection select is nifty because it excuses the necessity of talking about the actual causation of natural selection. Such talk was excusable for Charles Darwin, but inexcusable for evolutionists now. Creationists have discovered our empty “natural selection” language, and the “actions” of natural selection make huge, vulnerable targets. (pp. 199-200)

    (HT to Paul Nelson)

  9. “1. law, 2. chance or 3. agency”

    I find it difficult to believe that “chance” has anything to do with causation whatsoever. Chance ( as I last recall in R.C. Sproul’s interview with Ben Stein ), cannot be used in any causal relationship since it has no ontology — it is simply a statistical relationship representing the odds of probability. But it has no power whatsoever since it has no being. Chance possesses no ability to cause anything.

  10. toc,

    I agree with you in principle. But I don’t think most people mean by chance that something happens for no reason at all. There is a reason, but we don’t know what it is. I think the point being made is that what happens by “chance” is something that did not need to happen. It could have been otherwise. Law, however, always produces the same results.

  11. BarryA,

    Why do you say natural selection is a law? I can see how it is not totally random, but I can’t see how it is totally fixed like a law is.

  12. 12
    JunkyardTornado

    BarryA: “Remember, for any event there can three causes 1. law, 2. chance or 3. agency. Natural selection is ‘law.’ It works on random mutations (‘chance’) to produce new species. At least that’s the claim.”

    Joseph: “[William Provine said:] Natural selection does not act on anything, nor does it select (for or against), force, maximize, create, modify, shape, operate, drive, favor, maintain, push, or adjust. Natural selection does nothing….Having natural selection select is nifty because it excuses the necessity of talking about the actual causation of natural selection”

    If natural selection is law then natural selection doesn’t do nothing. A computer program is law, and programs design things all the time. When we tell a program, “I have all this complex criteria and requirements here, and have to find a solution that adheres to it. I don’t have time, so you do it. Here’s your raw material to work with.” When the computer comes back and says, “Here’s your solution”, it was designing, and yet it operates purely according to law. There’s a question of what natural law can accomplish if its below a certain threshold of complexity, (in which case you’re just shifting the burden to the mutations) but certainly it accomplished something. Natural law would say, “of these n random variants that have produced, I’m rejected x of them, because I deem them unviable, the rest of them I’m keeping.”

    BarryA: “By definition, a product of pure chance and law cannot be the product of “purposeful, intelligent design”

    But that’s just a matter of definition and assertion and world view, and not anything that has been demonstrated in an way.

    Certainly its no more than a world view to say that what humans do is not mechanism. The designs of humans can quite rationally be viewed purely in terms of a physical process. We can talk about the sensory discrimination of humans. Their sensory organs work solely via chemical and physical processes. We can talk about the human brain and its size and the level of granularity at which it can store a model of the external world. We can talk about the fundamental drives of humans – food, shelter, reproduction, pain avoidance, danger avoidance, etc. and how human form societies and eventually are able to accumulate food surpluses enabling some of them to devote their attentions elsewhere. We can talk about the gradual accumulation of knowledge over time, as models of the world stored in the brain are transferred to different physical mediums such as paper and ink, and thus serve as leverage and are exploited by future generations.

    So the point is, if what humans are and do can be explicated purely in terms of a physical mechanism, then why can’t whatever created us have been a physical mechanism. Surely if you trace this process back to something you can’t find a cause for in this universe, then you’re stuck. You can just label whatever the ultimate cause was “agency”, but that doesn’t mean what humans do is “agency”, i.e. some mysterious process not of the physical world. To say what humans do is not purely a mechanism, is a religious position, and not even one that I think the Bible espouses.

    If humans design via some non-mechanistic (i.e. supernatural) “agency” capability (whatever that is) then it might be reasonable to assume that this supernatural agency existed elsewhere previously and accounted for our own agency capability. But there is no proof that we are such agency, nor is it common sense to assume that we are. Nor is it common sense to assume there wasn’t some set of physical conditions and natural laws at work that preceded mankind and resulted in mankind’s existence. Even so, whatever physical conditions (i.e. laws and input to those laws) there were that preceded us, if their operation resulted in our existence then they would equate to us, just as if f(x) = y then f(x) equates to y, so you’re just pushing back what needs to be explained.

    Not meaning that humans or brains were physically present, but if a bunch of stuff is floating around invisibly out there, and some laws as well, capable of translating all that stuff into a human being, then its no different than having to explain a human being. Einstein didn’t become a genius the instant he finished the last sentence on the “The Theory of Relativity”. He was already a genius when he was sitting in the Patent office.

  13. 13
    JunkyardTornado

    toc: “Chance ( as I last recall in R.C. Sproul’s interview with Ben Stein ), cannot be used in any causal relationship since it has no ontology — it is simply a statistical relationship representing the odds of probability. But it has no power whatsoever since it has no being. Chance possesses no ability to cause anything.”

    Seems to be an insightful statement.

  14. 14
    JunkyardTornado

    Except that, the whole point of sexual reproduction is to produce huge numbers of random variants. Why would that be so important. Just think about that – randomness is integral to the whole setup.

  15. RD: “… natural selection is the very opposite of a chance process …” hold on a minute! In classic neodarwinian faith chance (“pure dumb luck”) is where all the variation comes from—without variation there could be no selection.

    In Einstein’s time the folks were moving away from chance toward absolute determinism, with the Darwinists seemly oblivious as they defified chance. The irony of it all—physics worshipping determinism and biology genuflecting to chance—and neither much noticing what the other was doing. Then along came quantum theory which put chance back into physics, with Einstein—the consumate determinist—asserting that God does not play dice with the Universe.

    Maybe Einstein was wrong and randomness, which is imaginable, exists along side determinism (2 + 2 = 4 in all possible worlds) at the most fundamental level.

    And what ID now adds is that agency is also fundamental and not emergent.

    It is in the heart of the reductionist—which is not an entirely bad thing—to reduce everything to the fewest and simplest principles. Everyone must begin with something, unless, as they say, we think it’s turtles all the way down.

    So what ID is bringing to the table is that agency doesn’t reduce—it’s elemental. For what machine can you imagine (put this wire here, that circuit there, this pulley here …) that outputs consciousness and free will and thus agency?

    Chance, necessity, agency—all three are imaginable—and the evidence suggests that all three exist on a fundamental level.

  16. 16

    BarryA: true, I am not a troll and true, I was confused.

    I was under the impression that Natural Selection was a process where “the force of Natural Selection” examined the environment, compared genes for the best fit, then selected the best genes to fit the environment. This would be the act of a purposeful intelligence.

    After some study I found I was mistaken regarding how Natural Selection works.

    In a population all the various genes are expressed randomly throughout a population, such as (using Darwin’s finches) birds with long beaks and short beaks. The genes for short and long beaks are in the gene pool but depending on the environment the trait that increases the odds for survival will appear with greater frequency due to more members surviving.

    So, should the environment become hostile to short beaks then only the long beaks will manage to acquire food and survive thus being “selected” for survival, the short beaks being “deselected” because they won’t find food to survive.

    Perhaps “Natural Selection” was a poor choice of words for Darwin to use, for to select something implies the ability to choose which requires an intelligence to compare and analyze the different choices available.

    Maybe “favorable survivability trait vs. lethal trait” would have been better.

    An example of a lethal trait: no fur in an arctic zone.

    However, I still think this ability for an organism to survive changes in the environment via its “toolshed of genes” is evidence that an intelligence knew the environment could change thus added the necessary genes to allow survivability.

    Most species would be extinct now if these genes to adapt were not available.

  17. I find it odd that evolutionists claim natural selection is not random when the “selecting” is based upon randomly changing environmental conditions (meaning not solely in the climate sense).

  18. The Darwinists keep treating natural selection as if it were some sort of “probability multiplier”, which through a feedback process somehow makes the on-the-face-of-it improbable more probable. But how can this be? Wouldn’t it be easier for chance to stumble across a certain design if there weren’t natural selection? Imagine an idealized world in which all “viable” organisms reproduced. By “viable” I mean having the mere capability to reproduce regardless of how competitively “fit” they are compared to their peers. Imagine that none of these organisms faces competition from any other organisms. Also imagine no limitations on resources, and that mutation is still operative. Wouldn’t such a world more quickly produce any particular design than one in which natural selection is operating? It seems to me that NS should actually make the arrival of designs less likely than pure chance. NS seems tantamount to occasionally tearing the paper out of a monkey’s typewriter and throwing the sheet away no matter how close the monkey was getting to outputting Hamlet. It seems that NS can only hinder the process of design discovery, not help it. So why is it regarded as a crucial and creative force? Why is it regarded as something that “as the very antithesis of chance” increases probability rather than decreasing it?

  19. “Natural selection is therefore a result of three processes, as first described by Darwin:

    Variation

    Inheritance

    Fecundity

    which together result in non-random, unequal survival and reproduction of individuals, which results in changes in the phenotypes present in populations of organisms over time.”- Allen McNeil

    “Natural selection is the simple result of variation, differential reproduction, and heredity—it is mindless and mechanistic.” UBerkley

    “Natural selection is the result of differences in survival and reproduction among individuals of a population that vary in one or more heritable traits.” Page 11 “Biology: Concepts and Applications” Starr fifth edition

    Results are usually dependent on the inputs. Variation is allegedly random. Inheritance of any mutation is not a sure thing- so how is NS non-random?

  20. “Richard Dawkins is certainly not a stupid person”

    Depends how you define intelligence. How does educated fool sound?

    He certainly is becoming an intellectual dinosaur. His ‘Delusion’ is prime evidence of that.

    If intelligence is measured by the ability to reason lucidly and succinctly then he’s lost it big time.

    “Bragging of being wise they became fools” (the Greek uses – mo?raino? or morons), comes to mind.
    The Greek moros from which the word derives = dull stupid or insipid

    So the good news is that the bible calls people like Dawkins morons which of course justifies us calling them the same! ;-)

  21. “Richard Dawkins is certainly not a stupid person”

    Depends how you define intelligence. How does educated fool sound?

    He certainly is becoming an intellectual dinosaur. His ‘Delusion’ is prime evidence of that.

    If intelligence is measured by the ability to reason lucidly and succinctly then he’s lost it big time.

    “Bragging of being wise they became fools” (the Greek uses moraino = morons), comes to mind.
    The Greek moros, from which the word derives, means dull, stupid or insipid

    So the good news is that the bible calls people like Dawkins morons which of course justifies us doing the same! ;-)

  22. oups, got posted twice! Hmmm… maybe it’s a sign from God. :-)

  23. Joseph 19-
    Natural selection is not random in the sense that the genes that are selected to survive are those that are best suited for the environment. As long as the environment stays the same, the changes will continue in the direction of becoming more fit for the environment.

  24. Matteo:
    It seems that NS can only hinder the process of design discovery, not help it.

    Correct. Natural selection is likely to throw out a bunch of worthwhile stuff along with the bad. At best, NS is information-preserving; at worst it is information-destroying — the antithesis of a creative process. This is also not very difficult to figure out, but Darwinists seem to be mysteriously immune to, and perplexed by, simple logic.

  25. In this video we see Dawkins, Miller, and Myers, in all their comedic glory, with reference to His Holiness Pope Darwin.

    Nobody expects the Spanish Inquistion!

  26. Gil in your blog you wrote:

    The who-designed-the-designer question is nonsensical when one considers that time had a beginning. This is also transparently and logically evident.

    The physical universe (matter, energy, space, and time) came into existence at a finite point. Dawkins must surely admit this. Whatever caused the physical universe did/does not exist in time, because time did not exist. Therefore, the cause of the universe has no past and has no history. That which does not have a past or a history has no point of origin, and thus no designer or originator. The infinite regress logically stops at the origin of the time domain, which is well established by modern science.

    I think the above is debatable. We cannot prove that time and space came into existence at some point. Nor can we prove that matter/energy came into existence at some point. It is perfectly reasonable to theorize that all of them have always existed. The universe in it’s present shape and form (elements, planets, life, stars, galaxies, etc) can be said with certainty to have had a beginning because it is apparent they are the product of intelligent design.

    One thing we can be sure of is that something has always existed. This is because something cannot come from nothing. Another thing we can be sure of is that there is an intelligent person who can control what we see of the universe. This is because we see can see apparent design from the micro to the macro within our universe. The probability of that apparent design being caused by something other then intelligence is so miniscule as to be impossible.

    We cannot say with certainty whether or not matter/energy in it’s most fundamental form has always existed or not because there is no way to verify that information. For all we know matter/energy in it’s most fundamental state may be part and parcel of God. We do not know what God is comprised of. We can say with certainty that God is spirit or consciousness, with a mind, because intelligence only exists as a property of a mind of conscious being. But that may just be one aspect of God.

    God may be far more complex then we can possibly imagine. But that doesn’t give credence to Dawkin’s claim that there is a philosophical infinite regress problem with the concept of God. Just because something is very complex doesn’t mean that it couldn’t come into existence without intelligent causation. Dawkins thinks he is being clever with his “who designed the designer” routine, but really it just betrays the lack of intellectual rigor on his part when he decided to use such sophomoric rhetoric. He doesn’t see the irony in his belief that incredibly complex humans, with intelligence who can create all sorts of amazing things, including life through procreation, have come to exist without being designed; yet God could not exist without being designed? Heads I win, tails you lose.

    Dawkins is like a politician who turns complex scientific and philosophical ideas into simple mindless sound bites for use in his propaganda war against belief in God. Nothing he has to say should be taken seriously due to his take-no-prisoners agenda. God can be extremely complex and not the product of intelligent design. That doesn’t automatically support the claim that complexity in nature can also be without intelligent causation, which is what Dawkins is trying to prove by his feeble argument. We can be assured that intelligent causation is needed to explain what we see in nature because we can use the scientific method to show that what we see in nature is not enough to explain how nature got to be the way it is. If I throw some paint buckets in a room and a paint brush and a blank canvas, no matter how much time passes we know that without intelligent design we won’t find the Mona Lisa gradually come into existence. We know that the laws of nature preclude that from occuring. So we know that a designer is necessary for our world to exist in the way that it does. The laws of nature preclude matter/energy from self organizing into what we see all around us.

    So just because God is complex and couldn’t be the product of intelligent design, that doesn’t change the fact that our world has to be the product of intelligent design.

    About space and time. Space and time also may have always existed as inherent properties or aspects of God. What is space? God exists, therefore God exists somewhere. That area where God exists takes up or defines the concept of space. Space exists because God exists as it’s ground of being. To say that time had to begin and that time exists only in relation to matter/energy, cannot be said with certainty. What is time? Time is a continuum of reality. We can measure points in time in relation to each other. If God always existed then time also always existed as a continuum of God’s existence.

  27. 27
    JunkyardTornado

    Rude: “And what ID now adds is that agency is also fundamental and not emergent…

    So what ID is bringing to the table is that agency doesn’t reduce—it’s elemental. For what machine can you imagine (put this wire here, that circuit there, this pulley here …) that outputs consciousness and free will and thus agency?

    Chance, necessity, agency—all three are imaginable—and the evidence suggests that all three exist on a fundamental level.

    To the best of my knowledge the existence of ‘agency’ as defined by ID is not something that has been demonstrated or proven in anyway in ID, via logic, physical evidence, or anything else. Rather, the existence of agency was considered to be self-evident in ID from its inception. Dr. Dembski says in the foundational papers words to the effect, “When we look at things in the natural world, we generally ascribe one of three possible causes, “Chance necessity or agency.” It was merely considered self-evident that this third category called agency existed, which was definitely not mechanism, but in many way seemed indistinguishable from chance, in that you couldn’t predict what it would do.

    As far as the explanatory filter, my understanding of it is as follows. (And here I am asking for clarification if my understanding is incorrect. I haven’t actually discussed this before.): There is a means by which you can rule out something happening merely by chance in ID, which from what I know of it, I actually accept. It involves an item of a certain complexity which also has a detachable pattern, plus you take into consideration the number of possible particle interactions in the universe, and conclude this thing could not have happened by chance. Fair enough. This is not even a problem, because science never uses chance as an explanation for anything. Chance is where you are going in, without knowledge, and the objective of science is always to identify mechanisms that would account for phenomena.

    But back to the EF – the rest of the argument goes, if you’ve formally ruled out chance, then if there’s also no known mechanism that could have caused this thing, then the only remaining alternative is this thing we merely assume exists called “agency”. Here’s the problem(s) (as I see it): Whereas a formal method was given for ruling out chance, no formal method is given at all in ID for ruling out mechanism. We can never say a mechanism did not cause something. If something is specifiable, then there is always a mechanism theoretically that could output it. And It seems it would take omniscience to prove a mechanism did not in fact cause it. ID gives no method for establishing even that it is more likely that agency caused something rather than a mechanism. It just says if we formally rule out chance and there’s no mechanism either then the only thing left is agency. But I say you can never rule out mechanism, and agency doesn’t exist. If we trace back to some mechanism that wasn’t caused by anything in the physical universe, then I think we can just stop there and say that mechanism has always existed and is in fact part of God. There’s no reason to invoke the magic undefinable property “agency”, either as an explanation for the universe, our own behavior, the behavior of animals, or anything else.

    I know I’ve monopolized band-width a little bit the past few days (or longer) so I’ll try to lay off for a while. Thanks for your patience.

  28. JunkyardTornado:

    You touch two very important, and related, points, about which I would like to comment:

    1) Agency. Agency can easily be defined empirically, and in that sense it requires no special “demonstration”. Agency is the principle which originates actions, and actions are easily defined as the specific output of conscious beings. In other worlds, conscious beings, particularly humans, interact with outer reality in a special way: inputs from reality which are represented in consciousness (perceptions); and outputs to reality from conscious representations (actions).In both cases, consciousness is in the middle. I am not discussing here what consciousness is, just accepting it as an empirical fact: consciousness “is” a fact, because it is constantly observed (as personal consciousness). Actions are therefore observable outputs of an empirical reality (consciousness), and agency is the word to describe the principle or power which originates actions.

    2) Because actions are easily observed, they can easily be analyzed. One important characteristic of a specific subset of human actions is that they are able to generate CSI (let’s call them “intelligent” actions). That’s an important point. Using Dembski’s definition of CSI, we have a very stunning empirical abservation:

    a) Humans, with their actions, generate CSI very easily and constantly (any post in this thread, for example, is a good example of CSI).

    b) We know of no other example of CSI originating in any other way.

    In other words, there are only two kinds of CSI in the world, as far as we know: human artifacts, and biological information. As we don’t know directly how biological information came into being, we can safely affirm that the only examples of CSI whose origin we know for certain are the products of agency.

    3) Let’s go to the problem of how the generation of CSI can be explained. You seem to accept Dembski’s point that it cannot come into existence by purely random causation. That’s good, because so I need not deal with that.
    Your point, instead, is that CSI, both the human and the biological, could be explained by what you call “mechanisms”.

    You say:

    “Whereas a formal method was given for ruling out chance, no formal method is given at all in ID for ruling out mechanism. We can never say a mechanism did not cause something. If something is specifiable, then there is always a mechanism theoretically that could output it. And It seems it would take omniscience to prove a mechanism did not in fact cause it.”

    With the word “mechanism” you seem to refer to necessity, in other words to specific deterministic causation. Unfortunately, your reasoning is not valid, neither for human CSI, nor for biological CSI.

    There are many reasons why necessity cannot account for CSI, and I will discuss here the two main aspects:

    a) The first reason is empirical: no deterministic mechanism is known which has that power. That, of itself, should be enough. You may say that, in theory, there could be some unknown deterministic explanation of CSI which could be discovered in the future. Even if that were true (and it is not, see next point), it would anyway not be relevant from a scientific, empirical point of view, because science is only about the best available explanation, and the vague possibility of some unknown explanation is no reason to reject existing explanations. So, empirically, you are left with only two choices: either admit that the origin of biological information can be explained in terms of design, and that that is the best (indeed the only) available theory, or just affirm that it is a complete mystery (but in that case, you have to show why the design hypothesis is wrong).

    b) But, luckily, there is another fundamental reason why CSI cannot be explained by necessity, now or ever. It’s a logico-mathemathical reason, not an empirical one.

    The reason is that necessity, for its own nature, cannot be the vehicle of complex information, because it has not the “flexibility” and power to express it. You can find that point very well argued in Abel and Trevors’important paper “Three subsets of sequence complexity and their relevance to
    biopolymeric information”, easily available online.

    The concept is very intuitive, in its essence: if you need to write a complex work with meaning, let’s say Shakespeare’s Hamlet or any equivalent work, you cannot derive it from a mathemathical rule (necessity), because no rule in the world can describe the exact succession of letters which make up the drama. In other words, from all points of view Shakespeare’s drama appears as a mainly random, and not deterministic, series of characters; but, in some way, those characters happen to have a special meaning in the english language and for the human mind. That’s exactly what CSI is.

    In the same way, you will never find any deterministic rule which can account for the nucleotide order in the gene for myoglobin. It’s simply impossible, because that order has a functional meaning (it will translate into a functional protein, given a definite context and symbolic code), but there is no mathematical rule in the whole universe to generate it from law.

    Specification is one aspect of meaning. Meaning can be defined only in relation to consciousness.

    And consciousness? You will say that consciousness itself must be a byproduct of mechanism. But that is true only if you accept a purely materialistic and deterministic view of all reality, including consciousness. But that’s only a theory, and, in my opinion, a very bad theory indeed.

    Consciousness, instead, is a fact (it is observable). As such, it exists, either you can explain it or not. And the same is true for its product, actions; and for those specific actions which exhibit CSI.

    So, unless and until you can really explain out consciousness in terms of mechanisms (and I really believe that you can’t!), ID has all the rights in the worlds to use consciousness and agency (empirical realities) in a theory to explain other aspects of reality (biological CSI).

    It’s really as simple as that.

  29. GP:

    Well said.

    I think JT needs to observe that we refer to mechanical necessity to explain natural regularities — e.g. that heavy objects tend to fall. Such are of low contingency, and are explained per EF as “law.”

    When we see high contingency, e.g. the falling (and tumbling object) is a die, and its uppermost face on settling can come from the set {1,2,3,4,5,6} we see that chance or intelligence may control the value to one extent or another — fair vs loaded dice for instance.

    But when we see functionally specified complex information, we see very high contingency associated with function, where the config space is so large that random chance or the like [cf Dembski-Marks on active information improving over random search] will easily exhaust available probabilistic resources. So, intelligence — givne our longstanding and direct experience of it — becomes the best explanaiton.

    Also, we know THAT intelligence, form our own direct knowledge of our activities,. This fact of intelligent consciousness is fact no 1 and if certain theories in science and associated worldviews struggle with it and what it routinely does, so much the worse for them!

    For, they stand exposed as empirically inadequate and sometimes even self-referentially incoherent. [Cf my appendix 6 here after quite an exchange with Frosty, who is therefore credited as effective co-author.]

    Thanks again GP for keeping up the good work.

    Cheerio

    GEM of TKI

  30. Richard Dawkins is certainly not a stupid person, but I find it amazing that he cannot see the obvious problem here. Natural selection is not random, but it does not create anything; it only throws stuff out.

    I don’t know if he said it in the text you omitted, but I know he’s said it elsewhere: it is natural selection acting on mutations that leads to complexity.

    But flagellar motors were demonstrably not always there.

    Interesting — how exactly do we know this? I mean, it certainly sounds reasonable, but how is it demonstrated?

    (Sorry if I repeat something already said above; I don’t have time to read through it all.)

  31. JT(#13): “So the point is, if what humans are and do can be explicated purely in terms of a physical mechanism, then why can’t whatever created us have been a physical mechanism.

    Come on. For this big “if” to be considered plausible the truth of reductionist materialism and mind=physical brain neuroscience theories must be blandly assumed, thus it is a classic example of assuming an important factor of what is actually in debate.

    “To say what humans do is not purely a mechanism, is a religious position, and not even one that I think the Bible espouses.”

    It is not a religious position, it is empirical based on observation and experience. To say what humans do is purely a mechanism is a huge assumption, ignoring among other things the mystery of human creativity and the phenomena of parapsychology. Of course to the materialist it’s just a matter of time for an explanation (promissory materialism), or the evidence really doesn’t exist.

    “If humans design via some non-mechanistic (i.e. supernatural) “agency” capability (whatever that is) then it might be reasonable to assume that this supernatural agency existed elsewhere previously and accounted for our own agency capability. But there is no proof that we are such agency, nor is it common sense to assume that we are.”

    Here we go again. This assumes it is reasonable to simply assume that the human mind must be ultimately material mechanism, ignoring all the evidence. The example previously offered was of a computer design program coming up with a solution, based purely(?) on mechanism. But you can’t get around the unfortunate fact of the need for the creativity and insight of the programmer.

  32. 32
    JunkyardTornado

    Gpuccio, in that paper you provided (Abel and Trevors’“Three subsets of sequence complexity and their relevance to biopolymeric information”) the author discusses all the incredibly complex things a cell does and emphasizes that its only the appearance of choice, and actually how its behavior determined by a complex program. He talks about FST being the output of algorithms. He says the genes are literally a program. So obviously programs do very complex things. He himself characterizes the algorithms in biology as “necessity”. The author himself is somewhat confused because he will intermittantly shift directions entirely and say the program itself could only be generated by something making “uncoerced” choice, even though he’s shown repeatedly how programs (i.e. “necessity”) can accomplish very complex things. So all in all, its muddled thinking, obfuscated by a torrent of arcane verbiage intended to bolster an implicit argument from incredulity.

    But anyway, you have a complex program producing a human being (via epigenesis and so forth). Why couldn’t human behavior itself be detgermined a program. And if it not, what is it then? Well, its magic according to ID, or to use their synonym, “agency”. But to go in the opposite direction, since a complex physical mechanism (genes, epigenesis, so forth) creates a human being, why couldn’t there be a physical program implicit in the universe itself that originated biological life? It would be in a different form than DNA, and what its form would be is hard to determine at this point, but to look out at the complex gargantuan universe of energy out there and say that it cannot possibly be a program for us defies credulity (to use my own argument of incredulity for the moment.) That is, if something as comparitively simple as a human brain drives human behavior (and it does.)

    Of course if you keep tracing stuff backwards, you’ll hit an incredibly unlikely set of initial conditions in the universe that is responsible for our existence. If you want to label the cause of that as supernatural “agency”, go right ahead. But its outside the realm of science, and it doesn’t mean that humans operate on a day to basis via supernatural magic (which is what “agency” is). Just take humans out of the picture, since you think they operate via magic. If all you had was a world of insects, could you see such a world as product of physical mechanisms in the unverse (or maybe you think insects have non-physical free will too.)

    I did read the entire article, and the other posts in this thread, and had intended to right a more thorough response here, with a lot of quotes from the paper you mentioned, etc. However, the fact is I’m facing some difficulties that are preoccupying me somewhat, so I’m conflicted about how much time I shoud be spending here. In fact, I would definitely ask for prayer from anyone who would care to.

    Regards.

    I will add the following – if you read God’s own accounts of creation in Job 38, he characterizes his direction over elemental physical forces in a lot of metaphoric detail. However, There is not any detail at all or any specific description of how animal and biological life was created. Furthermore, inanimate objects are personifed repeatedly in scripture, and you can dismiss this as pointless metaphors if you want, or see some deeper truth conveyed here – that the entire universe and its elemental physical forces are “alive” and by that I don’t mean they have immaterial souls anymore than an amoeba does (which BTW is also alive).

    (Psa 98:7-8) Let the sea roar and all it contains, The world and those who dwell in it. Let the rivers clap their hands, Let the mountains sing together for joy

    (Job 38:7) …When the morning stars sang together And all the sons of God shouted for joy…

    (Isa 55:12) …The mountains and the hills will break forth into shouts of joy before you, And all the trees of the field will clap hands.

    (Luke 19:39-40) Some of the Pharisees in the crowd said to Him, “Teacher, rebuke Your disciples.” But Jesus answered, “I tell you, if these become silent, the stones will cry out!”

    (Gen 1:24) Then God said, “Let the earth bring forth living creatures after their kind: cattle and creeping things and beasts of the earth after their kind”; and it was so

    Pax.

  33. 33
    JunkyardTornado

    Some hasty adddendums are necessary here. I had said the author himself characterized the algorithms of biology as “necessity”. I went back to look for that quote and couldn’t find it. The following may have been what I had in mind:

    Algorithms are processes or procedures that produce a needed result, whether it is computation or the end-products of biochemical pathways. Such strings of decision-node selections are anything but random. And they are not “self-ordered” by redundant cause-and-effect necessity.

    So first he says algorithms produce a “needed” result, and I guess I was thinking of ‘needed’ as a virtual synonym for “necessary”. Its all the more confusing because the word “necessity” is in the same sentence but its to assert that algorithms are not “cause and effect necessity.” What an algortithm is other than complex causes producing a necessary effect (or a “needed” effect to uses the author’s words), I don’t know. But its emblemattic of the sort of ambiguity thats rampant in this paper.

    Here’s some of his quotes:

    “Other complex structures are the product of digital recipe (e.g., antibodies, signal recognition particles, transport proteins, hormones). Recipe specifies algorithmic function. Recipes are like programming instructions. They are strings of prescribed decision-node configurable switch-settings. If executed properly, they become like bug-free computer programs running in quality operating systems on fully operational hardware. The cell appears to be making its own choices. Ultimately, everything the cell does is programmed by its hardware, operating system, and software. Its responses to environmental stimuli seem free. But they are merely pre-programmed degrees of operational freedom.”

    “The digital world has heightened our realization that virtually all information, including descriptions of four-dimensional reality, can be reduced to a linear digital sequence. Most attempts to understand intuitive information center around description and knowledge [41,63-67]. Human epistemology and agency invariably get incorporated into any model of semantics…

    “FSC is a succession of algorithmic selections leading to function. Selection, specification, or signification of certain “choices” in FSC sequences results only from nonrandom selection. These selections at successive decision nodes cannot be forced by deterministic cause-and-effect necessity. If they were, nearly all decision-node selections would be the same” [The last part is erroneous. Cause and effect is contingent on environmental external conditions as well and these external conditions causes differing results.]

    “Genes are not analogous to messages; genes are messages. Genes are literal programs.”

    “However, DNA programming instructions may be stored in nature (e.g., in permafrost, bones, fossils, amber) for hundreds to millions of years and be recovered, amplified by the polymerase chain reaction and still act as functional code”

  34. JT:

    Re your:

    he [T & A] says algorithms produce a “needed” result, and I guess I was thinking of ‘needed’ as a virtual synonym for “necessary”. Its all the more confusing because the word “necessity” is in the same sentence but its to assert that algorithms are not “cause and effect necessity.” What an algortithm is other than complex causes producing a necessary effect (or a “needed” effect to uses the author’s words), I don’t know. But its emblemattic of the sort of ambiguity thats rampant in this paper.

    First, I have read the T & A FSC, RSC, OSC [peer-reviewed] paper with great interest and do not at all find it ambiguous or hard to understand. (It says things that cut straight across the ideas of those who would imagine that the functionally organised complexity of cell-based life could originate by chance + necessity acting on matter + energy, but that is a very different thing.)

    Here is the T & A abstract:

    Genetic algorithms instruct sophisticated biological organization. Three qualitative kinds of sequence complexity exist: random (RSC), ordered (OSC), and functional (FSC). FSC alone provides algorithmic instruction. Random and Ordered Sequence Complexities lie at opposite ends of the same bi-directional sequence complexity vector. Randomness in sequence space is defined by a lack of Kolmogorov algorithmic compressibility. A sequence is compressible because it contains redundant order and patterns. Law-like cause-and-effect determinism produces highly compressible order. Such forced ordering precludes both information retention and freedom of selection so critical to algorithmic programming and control. Functional Sequence Complexity requires this added programming dimension of uncoerced selection at successive decision nodes in the string. Shannon information theory measures the relative degrees of RSC and OSC. Shannon information theory cannot measure FSC. FSC is invariably associated with all forms of complex biofunction, including biochemical pathways, cycles, positive and negative feedback regulation, and homeostatic metabolism. The algorithmic programming of FSC, not merely its aperiodicity, accounts for biological organization. No empirical evidence exists of either RSC of OSC ever having produced a single instance of sophisticated biological organization. Organization invariably manifests FSC rather than successive random events (RSC) or low-informational self-ordering phenomena (OSC).

    On the strength of your remarks above though, I think some notes on points may be helpful:

    1 –> In our direct observation and experience, a program requires an underlying algorithm, codes and language, to set up a process that step by step executes a procedure to solve a problem or achieve a result.

    2 –> Such is therefore a matter of high contingency — not a natural regularity — and thus is either chance or intelligent action. [Just as that a heavy object falls is a natural regularity, that if it is a die it ends up with a particular face uppermost is chance and/or agency; dice -- notoriously -- can be loaded. Similarly the orderly pattern of ions in a crystal of Na Cl reflect the strong forces at work, but through the high ordering preclude information storage by coded patterns of the Na+ and Cl- ions.]

    3 –> In our further direct observation, programs are functionally specified and complex, being invariably the result of an intelligent actor; hence the “no empirical evidence” in the abstract. This is further supported by the scope of the configuration space that is required to implement any program long enough to do really interesting things, thence the isolation of islands of function in the space. [GA's predictably won't work if we expand the space to the equivalent of say 300 - 500,000 contatenated 4-state elements.]

    4 –> E.g. look under the hood of your friendly neighbourhood PC: programs indeed are not physical, mechanical necessity. Nor — pace many rumours on how Microsoft wrote Vista — are they a matter of a million monkeys banging away at keyboards over in Redmond. Programmers use intelligence, knowledge and skill to target and control particular steps to achieve results based on understanding of materials, forces and structures used [in this case the PC's design and the Intel/ AMD microprocessor, which in turn rest on the underlying solid state etc physics and are designed structures], not chance collocations of matter and energy acted on by blind physical forces. Sand is natural, microprocessors and microcontrollers are not.

    5 –> So, invariably, when one sees a cybernetic system of known origin, we see that highly skilled designers have made the components and have created the programs and structures that control how they act to achieve the goals. FSC is reliably the product of intelligence, per our empirical observation. So, per induction, FSC is a reliable sign of intelligence.

    6 –> In the cell, we have a functionally specified, information-rich complexly organised, algorithmic system, one far beyond the complexity of what we can do to date over in Redmond — on the testimony of one certain William Gates. We know that the search space of the observed cosmos is utterly insufficient to get to such colocations and coordinated function or even to the DNA codes [500 kbases and up ~ 9.91*10^301029 config space states] by chance + necessity per known laws of molecular behaviour. Laws that are very well tested and reliable thank you.

    7 –> The proposition that such a system could spontaneously organise itself by chance and/or mechanical necessity on the gamut of the universe is therefore of very little credibility, and this is the root of the utter breakdown of evolutionary materialist OOL research in our time. [Cf Sections A and B and Appendix 1 the always linked.]

    In short T & A are precisely correct to say that to get the needed functionality for life [i.e. relative to purpose] via algorithmic processes, we are not dealing with mechanical necessity or blind chance to produce the required functional sequence complexity, as opposed to random sequence complexity or orderly sequence complexity.

    Algorithms, as physically instantiated, and per observations, are complex actuating causes set up by intelligent agents, to achieve their ends. That is not at all ambiguous, it is just that you evidently are hoping that such can set themselves up by lucky noise.

    That is analogous to the example in Appendix 6 the always linked, where by chance stones falling down a hill form up in the configuration “WELCOME TO WALES.” That would be bad enough — it is probably next to impossible to happen by chance + necessity though it is strictly possible.

    Now, there is an additional twist: lo and behold, this happens right on the border of Wales, next to a railroad, too!

    By contrast, it is very reasonable that railroad workers will be able to arrange stones intelligently to send such a message. (Just like the arrangement of plants outside my neighbourhood announcing its name are the product of intelligence; not chance seeding.)

    GEM of TKI

  35. PS: When I have written machine code programs, the machine did not have a choice itself about how it acted — sometimes, in debugging, I wished it could and would do what I wanted not what I [inadvertently] told it to, which made it crash . . . — but I sure did.

  36. Gil

    We don’t know that the universe or time had a beginning. First of all we can only talk about the portion of the universe that we can observe. We don’t know what, if anything, lies outside that region. For the portion we can observe we think we can trace its history back to a point in time where it was much smaller, denser, and hotter than it is today but we can’t trace it back to a moment of creation. Our understanding of physics breaks down under those conditions. As well, there’s always the possibility that our physics, even the part we believe is correct, is substantially wrong. The big bang theory is not written in granite and in fact is still somewhat controversial.

  37. Dave — The big bang theory is not written in granite and in fact is still somewhat controversial.

    OTOH, it has succeeded in making predictions (microwave background radiation). Like the 4.5 billion year old earth, it is the best we got when it comes to objective science.

  38. JunkyardTornado:

    Thank you for your detailed answer. I think kairosfocus has already given some answers to your remarks, but I would like to add some personal comments:

    1) About the Abel and Trevors paper: the problem you perceive is not really a problem at all. You say:

    “He himself characterizes the algorithms in biology as “necessity”. The author himself is somewhat confused because he will intermittantly shift directions entirely and say the program itself could only be generated by something making “uncoerced” choice, even though he’s shown repeatedly how programs (i.e. “necessity”) can accomplish very complex things.”

    Where is the difficulty? Programs do complex things because they are complex. They work algorithmically, but they do not create new CSI, they just utilize the information they own in various, pre-determined ways. If you read the papers by Dembski and Marks about information (yes, the censored ones), you will see that the concept is that you can indeed improve the power of random search, but only if you incorporate specific useful information in the algorithm. In the same way, an algorithm, or a machine, can do complex things, because it already has that information in its structure.

    But the information in the algorithm or in the machine can be produced only by agency. Only agency can produce new CSI.

    You have certainly witnessed computer programs who “speak” to you, give you answers in english language, and so on. But, as we know well, those answers are already written somewhere in the program, and their output is purely mechanical. There is no consiousness, there is no perception of meaning. Therefore, there is no new CSI. If a computer tells me “there is an error in your input”, that phrase was already written. And who has conceived it and written it? An agent. Who has conceived and written the algorithm which decides when to output it? An agent.

    Computers are not agents. That’s not because they are too simple. There has been a huge increase in computer power and complexity since Turing conceived his proverbial test, and no real improvement has been made toward a machine who can really imitate human behaviour. It’s not a problem of complexity. Computers will never be conscious. They will never represent things. They will never perceive meanings. They will never make choices.

    Consciousness is the real difference. When the modern folly of believing that consiousness can arise from software, or from adding parts to our personal abacus, will be over, then we can really begin to study the properties of consciousness and its relatioship to matter, brain, and neural activities. Until then, foolish theories about how consciousness can arise from loops, neural networks or parallel computing will keep us engrossed in idle thinking.

    Just to be completely clear: if I (an agent) design the DNA sequence for a new protein (let’s say a new enzyme), on the basis of my understanding of how proteins work, of the genetic code, of the context of the cell, and of the laws of physics and biochemistry, and if I have the protein synthesized in a cell (by molecular engineering), then the protein will certainly be able to accomplish a complex task (its enzimatic activity). Is the protein creating CSI? No. “I” created CSI, when I designed its DNA sequence, which is therefore functionally specified (I wrote it so that the corresponding protein could be functional). Could a computer do that? No, unless I (an agent) had inputted in the program all the CSI which is necessary to algorithmically obtain that result.

    So, are we similar to computers? No. We are conscious, computers are not. We represent, computers don’t. We feel, computers don’t. We choose, computers don’t. That brings us to the next point.

    2) You say:

    “But anyway, you have a complex program producing a human being (via epigenesis and so forth). Why couldn’t human behavior itself be detgermined a program. And if it not, what is it then? Well, its magic according to ID, or to use their synonym, “agency”. ”

    No program produces a human being, neither genetically nor epigenetically. What we know is that human beings express themselves through human material bodies, which, in some way (largely not understood) derive from other human beings through complex stages of development, from the zygote to the adult body.

    I have to remark again that the concept of “human being” includes, in a very relevant role, the concept of human consciousness. Even if we admit (which I am not so ready to do) that the human “body”, including the brain and nervous system, develops in a purely algorithmic way (through “necessity”), just the same we have not explained what is consciousness, where does it come from, and how does it express itself through the human body. Therefore, we have not explained how consciousness can produce new CSI.

    Because, you see, consciousness does produce new CSI. In no way all the products of human thought can be traced to a reshuffling of existing information, as it happens in computers. In no way are the products of human consciousness purely algorithmic.

    Although the previous statement should be rather self-evident, I will cite here an extreme demonstration of it, due to one of the best mathematicians and physicians of our times, Roger Penrose. If you read his books, The Emperor’s New Mind and Shadows of the Mind, you will find his famous argument from Godel’s theorem, which shows how human cognitive processes are not purely algorithmic, not even in the strict field of mathematics. In other words, humans can do “cognitive” mathematical things which no computer will ever be able to do. And if that is true for mathematics, how much truer will it be for disciplines like phylosophy, art, psychology, religion, and so on?

    So, you see, again consciousness makes the difference. Human cognition is unique in the world we know because it expresses both consciousness and intelligence. That’s why we are the only known source of CSI. That’s why our intelligent agency is the best model for the origin of the only form of CSI (biological information) whose origin is not known. You can call consciousness, intelligence and agency “magic”, if you want. But it’s very empirical magic, magic which can be observed in action everyday, everywhere.

    So, to sum up:

    - A complex algorithm acts through necessity, but is never written through necessity. Only agents write complex algorithms.

    - Agents are always conscious entities.

    - No computer or algorithm will ever become conscious.

    - Consciousness cannot be explained algorithmically. Consciousness can do things which no algorithm can do. One of them is producing new CSI.

    - Algorithmic machines cannot produce new CSI, although they can certainly transform the information they have, so that they are apparently generating CSI. In reality, they are just algorithmically transforming the CSI which has already been inputted in them.

    Finally, I will not address your specifically religious arguments, because I never do that here, to respect the principle that ID is about science, not religion.

Leave a Reply