|
Author
|
Topic: Principle of Causal Adequacy and Specified Complexity
|
Jules
Member
Member # 181
|
posted 26. March 2002 08:41
Jesse,
I think Dembski would be willing to waive the issue of complexity if you granted that specified complexity ceases to be improbable when there are intelligent agents around. Was it here or somewhere else where he explained that from a Christian perspective it would make sense that randomness really doesn't exist.
IP: Logged
|
|
Micah Sparacio
Member
Member # 6
|
posted 26. March 2002 09:40
Jesse:
I see two problems with your argument. First, a naturalistic explanation does not presuppose a chance explanation. Just ask James Barham, a participator on this board.
Second, even if we grant that the brain developed through chance processes, it would still be (and certainly is necessary for our every day survival) instructive for us to distinguish this unique natural cause (intelligence) from all other natural causes. In doing so, one would simply consider all other chance hypotheses besides intelligence when determining the reference class of chance hypotheses.
Finally, we really do not know enough about how the brain developed to say with certainty that there is a naturalistic explanation. Until we have that explanation (we certainly have candidates) I think this problem proves to be resolvable from a position of utility. Human intelligence is distinct from every day natural causes. If someone questions this principle, then I and she may be forever doomed to speaking past each other.
IP: Logged
|
|
Nelson Alonso
Member
Member # 52
|
posted 26. March 2002 11:31
Jesse:
Well, if we had a naturalistic explanation in hand for the behavior of intelligent beings, then presumably the "chance hypothesis" induced by this theory would yield a reasonably high probability that such beings would produce specified objects like art, novels, buildings, technology, and so forth. Since the amount of "complex specified information" is defined by Dembski as the probability of getting a specified event under the appropriate probability distribution (the chance hypothesis), this would mean that such things don't actually exhibit specified complexity. It is still possible there could be some unique thumbprint of intelligent action, but it wouldn't be specified complexity.
Nelson:
Actually this doesn't follow. A human intelligence builds something through information processing and through creative acts. Whether intelligence itself is due to a naturalistic origin is quite irrelevant. The distinction to make is whether one comes upon a machine and concludes design or a natural process (such as wind or natural selection).
Sandhoppers, which orient by the sun, for example, to head for the beach, was found to be partly learned, partly innate. They are born knowing the direction of the ancestral shore, but as they mature, they lose the instinct if it is not excercised, and they are easily trained to take a different heading. Arctosa spiders, are born with a direction sense but keep it only if itis practiced and can modifty it if necessary (see Robbert Wesson). Whether intelligence arose naturally or not, there is still a distinction to be made between intelligent causality and naturalistic causality. SETI would no doubt believe that the aliens communicating with them evolved naturally, and yet they conclude, from the specified complexity of the message they received, that an intelligence is speaking to them.
SETI, archeology, some forensic sciences, psychology and sociology use intelligent agency and various internal states of agents as part of their description of the causal entities, processes, events or actions cited as explanations for certain phenomena.
[ 26 March 2002, 11:35: Message edited by: Nelson Alonso ]
IP: Logged
|
|
Mister Pamboli
Member
Member # 175
|
posted 26. March 2002 12:52
quote:
A human intelligence builds something through information processing and through creative acts. Whether intelligence itself is due to a naturalistic origin is quite irrelevant. The distinction to make is whether one comes upon a machine and concludes design or a natural process (such as wind or natural selection).
But how do you distinguish between artifacts of intelligence and artifacts of complex natural processes? At what point does intelligence become intelligence? At what point do we have a creative act rather than a natural process? quote:
Whether intelligence arose naturally or not, there is still a distinction to be made between intelligent causality and naturalistic causality.
And how do we make this distinction? I suspect Jesse's point can be restated as a problem with the explanatory filter - that it cannot distinguish hard cases of natural processes in animal, human or alien minds exhibiting a near-intelligent causal relationship with an object or process under consideration.
If the answer is simply that when we see specified complexity we infer intelligent design we are caught in a circular argument as we get close to my speculation: is the ability to produce specified complexity a diagnostic of intelligence?
IP: Logged
|
|
Nelson Alonso
Member
Member # 52
|
posted 26. March 2002 12:58
Pamboli:
But how do you distinguish between artifacts of intelligence and artifacts of complex natural processes? At what point does intelligence become intelligence? At what point do we have a creative act rather than a natural process?
Nelson:
This is Dembski's entire point. The difference is between specified complexity and unspecified complexity or specified simplicity. Dembski is saying that you can distinguish an intelligent cause from a naturalistic cause when you discover specified complexity.
Nelson:
Whether intelligence arose naturally or not, there is still a distinction to be made between intelligent causality and naturalistic causality.
Pamboli:
And how do we make this distinction?
Nelson:
See above.
Pamboli:
I suspect Jesse's point can be restated as a problem with the explanatory filter - that it cannot distinguish hard cases of natural processes in animal, human or alien minds exhibiting a near-intelligent causal relationship with an object or process under consideration.
If the answer is simply that when we see specified complexity we infer intelligent design we are caught in a circular argument as we get close to my speculation: is the ability to produce specified complexity a diagnostic of intelligence?
Nelson:
This is in no way a circular argument. How can it be? This is my point above. We observe intelligence producing specified complexity regularly, we do not observe natural processes producing specified complexity. We observe that it actually has a hard time with it. That is why we infer things like prime numbers comming from space as a sign of intelligence. This is why we infer messages in the sky inviting us to beach parties, we actually expect to go to beach parties.
IP: Logged
|
|
William A. Dembski
Member
Member # 7
|
posted 26. March 2002 16:44
Thanks for your many insightful comments on this thread. I had meant to jump in sooner, but now will have to content myself with addressing but a few of the thoughts taken up.
James Barham writes: "I certainly agree that present-day science does not have the conceptual resources to explain the teleological organization of life. However, the question is, What does this imply for the future progress of science? Can we confidently predict, on the grounds of a priori reasoning, that any particular phenomenon will forever remain outside the scope of future science? What is to prevent us from saying that biology today is in the same position as physics at the end of the nineteeth century? It is beginning to dawn on many people (in no small part thanks to your efforts) that there is a gaping hole in our knowledge of how life works, just as it dawned on Planck, Bohr, and the others that classical mechanics did not have the conceptual resources with which to explain the stability of the atom. I suppose someone could have argued then (maybe someone did?) that this means that a supernatural force must be supporting the atom externally. But in the event, a way forward was found, and a whole new formalism was discovered that was more adequate to the phenomena. What is there in principle to make us think that a similar way forward may not be found eventually in biology?"
I think there is good reason to be suspicious of the analogy with quantum mechanics. With quantum mechanics, scientists found a whole new range of phenomena with no known analogue to classical phenomena. By contrast, when biochemists find molecular machines, they are finding machines in every sense of the word. James sometimes raises the problem of information realism -- that information needs properly to be objective and independent of human inquirers to be a legitimate entity for scientific study, and that the information that design theorists claim to find in biological systems constitutes a subjective imposition. But with something like the bacterial flagellum, you've got an outboard motor-driven propeller. That's what it is. Humans invented such machines well before the 20th century. But it was only in the early 20th century that scientists figured out what the flagellum does. That sort of convergence indicates that the biological specification qua information of such systems is objectively given. That they are machines in the true sense of the word, and whatever account we give of life must account for the origin of such machines. I agree with James that life is more than machines but it at least includes machines. To look for an internal teleological principle that does the design work of producing biochemical machines is understandable if one has good reasons for rejecting the external teleology of intelligent design (James and I agree that some sort of teleology is necessary). But I have yet to see any good reasons for rejecting external teleology, save that it doesn't fit with our current conception of scientific rationality (and where did we get that?). But to come back to the Principle of Causal Adequacy, we do have evidence of external teleology producing the specified complexity of biochemical machines and we are still waiting for an account of how internal teleology can do it. If ID theorists have their work cut out for them (and we do), then it seems to me that teleological self-organizationalists like James have it even more cut out for them.
Pamboli addresses the need to balance the Principle of Causal Adequacy with Occam's razor. He is of course correct. As I've stressed throughout my writings, intelligent agency can mimic natural causes (all good murderers try to do that). The whole point of arguing for specified complexity as a mark of intelligence is to identify some feature of objects which intelligent agency adequately accounts for (thus invoking the Principle of Causal Adequacy to justify intelligence) but for which natural causes cannot adequately account for them (thus invoking Occam's razor to justify intelligence).
Jesse makes an interesting point that has been discussed on this thread. Let me add my two cents. Jesse writes: "Most cognitive scientists and neuroscientists suspect that intelligence can, at least in principle, be explained naturalistically, which would imply that things like art, technology, architecture, discussion board posts, etc., do not really exhibit specified complexity in the first place. Of course our understanding of the brain/mind relation is still pretty limited and this issue is up for grabs, but it is at least true that we should not assume that intelligent agents can produce specified complexity."
In NFL I distinguish between producing specified complexity and generating it. When an embossed sign falls over in a snow storm and inscribes its text onto the snow, specified complexity was produced in the snow. But there was no generation of novel specified complexity. It seems that when we are dealing with information processing systems (computers and humans both qualify), we can ask whether the system just in virtue of its structure exhibits specified complexity. We can also ask whether the system in its behavior outputs or produces specified complexity. Now, a standard digital computer operates deterministically. In its construction, it exhibits specified complexity. But what about its behavior? Does it produce specified complexity. It produces it in the same way that a sign falling over in a snow storm does. That is, it takes existing specified complexity and reexpresses it. It does not generate it de novo. What about humans? Do humans generate specified complexity? Well, if we are just computers (wetware) or naturalistic mechanisms, then we don't. What's more, if we are computers or naturalistic mechanisms, then the computers we manufacture do not exhibit specified complexity that was generated de novo.
Even so, they exhibit specified complexity which they got from us which we got from … where? The Darwinist wants to say from natural selection. But the beauty of specified complexity is that backtracking it under the operation of natural causes only makes the specified complexity problem worse. Humans have more specified complexity than the computers they manufacture. If we try to explain the specified complexity in humans naturalistically, we end up at some being that exhibits even more specified complexity than humans. And if the causal chain ends somewhere (rather than going back in an infinite regress), then we end up at some primal intelligence (there's an information-theoretic proof of God here which someone may want to write up). My point, then, is that specified complexity cannot be limited exclusively to derived intelligences but at some point requires primal intelligences that cannot be reduced to natural causes. This is not to assume that humans are intelligent agents in the primal sense (that needs to be argued), but at least to allow the possibility. I therefore don't think that Jesse's worry that design theorists are slipping in some untoward assumption is warranted. It's also worth pointing out that cognitive scientists are nowhere near establishing that human intelligence reduces to known mechanisms operating in known ways. How brain causes mind, if it does so in a mechanistic sense, is an unsolved problem. To claim that it does is for now sheer metaphysics, not science.
IP: Logged
|
|
Nelson Alonso
Member
Member # 52
|
posted 26. March 2002 17:13
Dembski wrote:
quote:
If we try to explain the specified complexity in humans naturalistically, we end up at some being that exhibits even more specified complexity than humans. And if the causal chain ends somewhere (rather than going back in an infinite regress), then we end up at some primal intelligence (there's an information-theoretic proof of God here which someone may want to write up).
Are you saying that if something contains specified complexity, one can obtain the minimum specified complexity of the designer, so the first designer of it all had the most specified complexity of all designed systems in the universe combined?
IP: Logged
|
|
Jesse
Member
Member # 112
|
posted 26. March 2002 17:25
William A. Dembski:
Even so, they exhibit specified complexity which they got from us which we got from … where? The Darwinist wants to say from natural selection. But the beauty of specified complexity is that backtracking it under the operation of natural causes only makes the specified complexity problem worse. Humans have more specified complexity than the computers they manufacture. If we try to explain the specified complexity in humans naturalistically, we end up at some being that exhibits even more specified complexity than humans. And if the causal chain ends somewhere (rather than going back in an infinite regress), then we end up at some primal intelligence (there's an information-theoretic proof of God here which someone may want to write up). My point, then, is that specified complexity cannot be limited exclusively to derived intelligences but at some point requires primal intelligences that cannot be reduced to natural causes. This is not to assume that humans are intelligent agents in the primal sense (that needs to be argued), but at least to allow the possibility. I therefore don't think that Jesse's worry that design theorists are slipping in some untoward assumption is warranted.
I think you've slipped in an untoward assumption right here--namely, that specified complexity exists in the first place. You're saying that even if we had a natural explanation for how the brain works, then that would not dissolve "the" specified complexity of things like art and technology, but just push it back to the design of the brain itself. OK, but what if we have a natural explanation for how complicated brains could have a reasonably high probability of evolving, given starting conditions involving a planet containing nothing but single-celled life? Then perhaps you'd say that just pushes "the" specified complexity back to the design of these single-celled organisms. What if we have a naturalistic explanation for how a planet with the right initial conditions has a reasonable probability of undergoing abiogenesis? And we also have a naturalistic explanation for the formation of such planets and solar systems from nebulas, etc., all the way back to the initial uniform particle soup that physicists think was present right after the Big Bang?
In this scenario, it seems that relative to all the natural laws operating in the entire universe, we would have shown that nowhere in the causal chain was there any specified complexity at all--every step of the way, we had "reasonably" high-probability specified events (probability greater than 1 in 10^150) flowing from other reasonably high-probability specied events. You can only claim that "specified complexity" is being shuffled around if one of the links in the causal chain had a very low probability relative to the prior conditions + the natural laws at play. If no event in the sequence had such a low probability, there is no specified complexity here to begin with, and thus no "infinite regress". Of course we do not have complete naturalistic explanations for abiogenesis or the workings of brains etc., so it's certainly possible that there is such a "missing link" here. But you cannot simply assume that there is--if you do, you're just begging the question by taking for granted that somewhere in the causal chain leading from the initial conditions at the Big Bang to the current specified phenomena produced by intelligent agents, there has to be some event that was too improbable to have happened by natural law. If this were known to be true, ID theorists would have already won the battle, but of course no one who favors naturalistic explanations is going to accept this.
[ 26 March 2002, 17:35: Message edited by: Jesse ]
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 26. March 2002 18:53
Bill:
Certainly, organisms and machines are similar in that neither category can be explained by appealing to chance or to any laws of nature that we have so far identified.
But there is nevertheless an obvious and fundamental asymmetry: machines are built by organisms, and not the other way around. The reason for this asymmetry should be clear. Machines are made of parts that are inherently inert, having no intrinsic tendency to maintain themselves in the "functional" configuration, which is only functional with respect to the maker, not to the machine itself. Organisms construct themselves, strive to maintain themselves in existence, act intelligently on the world in order to achieve this goal (and innumerable sub-goals), repair themselves, etc. In short, organisms value their own existence. Pace Rodney Brooks, Marvin Minsky, and their disciples, no machine does these things in any but a Pickwickian sense.
So, to say that the bacterial flagellum is identical to an outboard motor seems to me unwarranted. Certainly, we human beings have stumbled upon the general form of many of the contrivances of the living world in our effort to extend our functional capabiliites via technology. But this is not too surprising, since we are organisms ourselves, and are striving to do similar things exosomatically that our bodies must do on their own. For example, we stumbled upon pumps already in Antiquity before we knew the heart was a "pump." But this does not mean that the living heart is not a very different sort of thing from a manmade pump. For one thing, disaggregated myocytes placed in a culture medium will seek each other out, re-assemble, and begin pulsing rhythmically all on their own. If you smash an artificial heart to bits with a hammer, the pieces of dacron and titanium will not do that! So I think we ought to be more circumspect with these machine analogies.
Now, I of course acknowledge that we self-organization theorists have our work cut out for us. I certainly have no idea of the details of how the bacterial flagellum came into being, although presumably it must have been by a sequence of transformations, just as human technological progress takes place piecemeal. This sounds like good old Darwinism, I suppose, except that I would insist that none of these transformations was by "accident," but all were guided by the functional striving of the organisms in questions. Not by "clairvoyance," of course, but by intelligent search via trial and error. (I have suggested the term "tatonnement" to describe this, to replace the Darwinian concept of "bricolage".)
But, yes, you are right, we definitely have our work cut out for us.
IP: Logged
|
|
RalphW
Member
Member # 116
|
posted 26. March 2002 19:57
Jesse said:
quote:
Most cognitive scientists and neuroscientists suspect that intelligence can, at least in principle, be explained naturalistically
I presume that this can be taken to mean that the universe was loaded "from the Big Bang" with the capability to produce intelligence without any outside intervention. Given that OOL research to date hasn't demonstrated the ability of unassisted natural processes to produce organic compounds much more sophisticated than amino acids, it seems reasonable to doubt the causal sufficiency of these processes to produce intelligence. But even if we could demonstrate that natural processes could produce life and then subsequently drive an increase in biological complexity sufficient to support intelligence, it does not follow that they would be adequate to account for intelligence. We are committing a category error if we assume that because we understand the mechanics of the brain we can explain intelligence. Just as the chemistry of ink and paper does not account for the meaning of books, the functions of interlocking neurons and synapses do not account for intelligence, even though they are sufficient to convey that intelligence just as books convey meaning.
From another perspective, if we assumed that intelligence was ultimately the product of natural processes, then the question we are left with is why it is of any use. Dawkins' "memes" are based on this model of intelligence; our thought processes are based entirely on natural functions and can be explained by the same laws that explain nature. Thus ideas likewise are created and propagated by something akin to chance plus natural law, and the survival of an idea can be explained in terms of the survivability (or infectiousness) of its mental constructs. He goes on to argue that many of our ideas (e.g. religions) are false yet have survived because they are highly infectious. But if the brain is merely an organism for passing around infectious memes, then all we know about what we know is that it is infectious, not that it is true.
It could be that this is the crux of the entire ID debate. If natural processes are causually sufficient to account for intelligence, then the term "intelligent design" is meaningless. The use of terms relating to intelligence (e.g. purpose, meaning, design, etc.), including as it relates to our own behavior, is merely a shorthand for a particular class of natural events that are completely determined by the natural laws that fully account for all that happens. If, on the other hand, natural processes are not causually sufficient to account for intelligence, then they do not account for the universe as we know it because the universe contains intelligence, which must therefore be accounted for in a non-naturalistic fashion. And if there is something (or better Some One) who is outside of the universe and who accounts in part for our intelligence, then there is no necessary reason why that individual could not also account for other things in the universe, things that in fact are also intelligently designed.
IP: Logged
|
|
Jules
Member
Member # 181
|
posted 27. March 2002 07:28
Jesse,
Perhaps I'm missing your point. However, I think you would agree that we have a set of entities that require intelligent agency as a causal explanation -- buildings, airplanes, computers, etc. Even if we were able to explain in non-teleological terms how this intelligent agency originated, we would still require intelligent agency to explain that set. If I understand Dembski's argument, there exists another set of things -- biological entities-- that resemble human-made things to such an extent, that intelligent agency is required to explain their origin. He claims that what they both have in common is a degree of improbability that eliminates chance and law as direct causal agents. Even if we were able to explain -- in non-teleological terms -- the origin of the intelligent agency responsible for that set of biological entities, we would still need that intelligent agency as a causal explanation for it.
If such is the case, then your arguing that intelligence may be explicable in non-teleological terms is irrelevant.
[ 27 March 2002, 07:30: Message edited by: Jules ]
IP: Logged
|
|
Jesse
Member
Member # 112
|
posted 27. March 2002 15:12
Jules:
Even if we were able to explain in non-teleological terms how this intelligent agency originated, we would still require intelligent agency to explain that set. If I understand Dembski's argument, there exists another set of things -- biological entities-- that resemble human-made things to such an extent, that intelligent agency is required to explain their origin. He claims that what they both have in common is a degree of improbability that eliminates chance and law as direct causal agents.
My understanding of Dembski's argument is that there is no general distinction between "intelligent" and non-intelligent causes as you say--only a distinction between events that can be shown to have a high probability through some combination of chance + law (basically, anything that could be simulated on a sufficiently powerful computer with a random number generator built in) and those that have a low probability relative to all such "chance hypotheses." If the actions of an intelligent being can be explained through a combination of chance and law, and the origin of this intelligent being can likewise be explained, then the creations of this being will not posess specified complexity under the appropriate chance hypothesis. You may still wish to distingish between events of low specified complexity that require intelligence and events of low specified complexity that don't, but then you're moving beyond the scope of Dembski's argument, unless I've misunderstood something.
IP: Logged
|
|
|
Printer-friendly view of this topic