Dr. Joshua Swamidass, a theistic evolutionist, joined us recently at TSZ. I think the following comment of his will lead to some interesting and contentious discussion and is worthy of its own thread:
Third, if we drop “Darwinian” to just refer to the current modern synthesis of evolutionary theory, you are right that the scientific account does not find any evidence of direction or planning. I agree with you here and do not dispute this.
So the question becomes, really, is it possible that God could have created a process (like evolution) with a purposeful intent that science could not detect? I think the answer here is obvious. Of course He could. In fact, I would say, unless He wanted us to discern His purpose, we could not.
In my view, then, evolution has a purpose in creating us. Science itself cannot uncover its purpose. I find that out by other means.
If you see something good (and not too hairy), would you mind providing a link? Thanks.
I reject determinism as an accurate model of reality. One reason is that unstable nuclei decay stochastically which to me is synonymous with unpredictably. I’m not arguing for or against “libertarian free will” As to free will, I’ve never understood the “libertarian free will” argument that seems to occupy those with an inclination for theological arguments. It seems about as relevant to reality as the colour of unicorns. (I don’t have a dog in that fight either.)
Thus I have decided to pragmatically accept reality as I (with the added input of shared experience) perceive it until something comes along that requires me to re-evaluate reality.
Nice to hear from you Walto. 🙂
Brief summary of Dembski’s paper from a practical standpoint, and why I was enthusiastic about the paper.
For many practical applications from computer security, to modeling of physical and chemical systems, to analyzing casino games, to analyzing the financial markets, to analyzing biological systems, etc…Computer scientists and mathematicians are interested in building deterministic algorithms that look random.
That is to say, if you actually look at the computer program that implements a random number generator, you will see that it generates a predictable set of numbers, but to the outside world it has to look random. The way this is often achieved is that the program is “seeded” with some number (like the time of day), and then presumably with each run, the deterministic predictable nature of the program is concealed. The major uncertainty is rooted in what “seed” is used as to prime to the deterministic recipe.
So usually in using a computer program as random number generator, I invoke it with a statement like:
rand(1)
or
or something like
or whatever where what is in the parentheses is some “seed” that I choose. Hence, strictly speaking the random numbers aren’t really random, they are deterministic.
When I’ve run various financial and casino simulations, I have random number generating algorithms where I’m able to control the “seed”, thus the “random” number generator provides the same identical sequence of “random” numbers each run if I choose. This is so I can debug the software.
We have examples of TSZ using random number generators for Keiths Drifting Weasel for example. It is desirable sometimes to get the identical “random” number used by a variety of versions of software to see if there are bugs in the software. After we are assured the software works, we can then choose a variety of seeds at random (pun intended) like the time of day to run the software.
But if random number generators for computers are ultimately deterministic with respect to the “seed”, it raises the question whether reality is ultimately deterministic and we are just not privy to the details of the algorithm creating reality. Lewontin framed the question nicely, is randomness owing to our uncertainty about reality or is reality really capricious!
As far as scientific practice, the question doesn’t immediately have to be answered, but it is of interest for computer scientists to make deterministic systems look as if they are not designed, but look capricious. But this raises the philosophical question as to why something looks capricious versus designed, and how something can be deliberately designed to look capricious.
Right. But then there is also evolution to man-so my side gets better odds than yours.
walto,
I wonder which one of those two I am supposed to be.
Must be someone else you mean.
Wow!
Not sure how that works. I don’t have a side; I’m just me. Who’s on your side? If they think humans evolved then we’re in agreement.
I see it’s only been cited 12 times, a couple by Dembski himself. There’s a mountain of rebuttals of Dembski’s papers: Elsberry, Shallitt, Perakh, Matzke and so on.
Not forgetting Tom English!!!
You believe in libertarian free will don’t you?
Thanks. FWIW, I think your summary goes a bit farther than Dembski’s paper–although he may have had something like those conclusions in mind for his down-the-line research. As I read it, his conclusions are more modest than that, simply divorcing randomness from chance/probability. There may be profound inferences to be drawn from that….but they aren’t obvious (to me, at least).
Although I have been mildly critical of some of Bill Dembski’s work, it is understandable why on his journey to writing the essay on Randomness, he became highly interested in intelligent design.
Setting aside the question of whether ID is real, the computer scientists at the conference he referenced, as well as mathematicians and computer scientists in general, raised interesting questions.
To create a good random number generator that is at the heart actually non-random, one must have some mathematical notion of what “looks random” vs. what “looks non-random”.
So there are mathematical properties that one can build a spectrum where on one end we have “looks random” vs. “looks non-random”. This is question in terms of appearances I suppose can be codified mathematically without resolving any ultimate philosophical questions.
Finding 500 fair coins 100% heads on a table “looks non-random”. It is a mathematical property of the system.
Bill tried to extend these formalisms to biology to argue biology “looks non-random”. In the minds of his detractors, what probably was unforgivable was to argue, “it looks non-random, therefore it is intelligently designed.”
In evolutionary theory, marsupial mammals and placental mammals evolved independently — meaning, important similarities were not the result of immediate/proximal common descent as a matter of principle.
Like the casino cheating scandal the FBI uncovered, the patterns of similarity between marsupial and placental mammals looks non-random. This fact does not go unnoticed by evolutionary biologists, they have words for this non-random phenomenon: “convergent evolution” or “homoplasy via convergent evolution”. It is an empirically detectable, somewhat mathematically describable property of a system of objects, namely, they “look non-random.”
Hence, it is understandable what motivated Bill’s interest in ID as he pondered these things.
My mild criticism of the rest of his work has been the formalisms he uses aren’t very practical and are so overly complicated as to be un-usuable. One does not need such Dembskian formalisms to recognized the non-randomness in the relationships between placental and marsupial mammals.
I think we are all agreed that the resemblance between various pairs of placentals and marsupials is not random. The argument is over potential reasons. You think it’s evidence for ID. Why? Others (like me) think it’s due to similar selection regimes acting on similar populations. A mammalian pursuit carnivore ought to have certain characteristics if it’s going to be successful at its job. And thus the resemblance between Tasmanian wolf and real wolf. Why does it need to be more than that?
I think it might be more accurate to say that “on his way to argue in favor of intelligent design, he became highly interested in randomness”.
Why? I don’t understand what you are getting at here.
The unregenerate want to sin because they like like it. There is no compulsion here at all.
They like sin because they are selfish and think that their wishes are more desirable than God’s.
God is not a slave-master their own nature is.
peace
not every sort of predetermination only those where we do what we want because we we to.
peace
My comments on Dembski’s paper “Randomness by Design”
I cannot follow all the math, but he does make some interesting observations about how we look at randomness.
He makes several statements that could easily be misconstrued or quotemined by those of the ID mindset:
“Suffice it to say, randomness, to be randomness, must be designed.”
“…generating randomness involves forethought and design.”
“Randomness is fundamentally a question of design.”
But if you read further, he makes it fairly clear that what he means is that randomness has multiple definitions, and if you pick one, it takes some work to program a computer to make it generate output that matches that particular definition. This sounds reasonable to me.
I think anyone reading the paper should take care not to project “randomness is designed” onto nature and conclude that all chance occurrences in nature, or any phenomenon appearing to be random, is actually designed. Dembski takes care to distinguish between randomness and chance.
To his credit he acknowledges the importance of the Law of Large Numbers – something that evolution depends heavily on. I think everyone should read “The Improbability Principle” by David Hand, to get a good grounding in this.
Note Dembski’s concluding statement:
“Similarly, the random objects I advocate reflect a changed point of view. In
times past random objects were random because they mimicked chance. Forgeries they were. As long as the counterfeit looked specious, one could pretend it was the product of chance. But the technology for uncovering these forgeries was always improving. The latest statistical test was ever threatening to expose the “well-established” random object. However, within the new framework, the “conditions for the possibility” of such objects, to use a Kantian phrase, henceforth rests with the patterns that render these objects random, and not with the objects themselves. Patterns become strictly prior to random objects. Without patterns, objects are just objects, not random objects.”
This is similar to my earlier statement that intent and design are not inherent properties of objects, and that intent and design can only be inferred by comparison with a known pattern.
I can paraphrase his last sentence to apply to design as well: “Without
patterns to compare them against, objects are just objects, not designed objects or natural objects.”
I guess he felt this somehow did not apply to living organisms when he began his work on Complex Specified Information.
dazz:
Fair Witness:
walto:
Fair Witness:
LoL!
Fair Witness? Really?
Mung,
Well, to be fair to fair witness, s/he only criticized Sal’s quotation from the paper, which s/he no doubt HAD read. And I do think that Sal’s quotation gives a somewhat misleading picture of the paper’s focus.
Joe F. may be right that Dembski had ID in the back of his mind when writing it, but if one knew Dembski’s work only from this paper, or if it were all he’d ever written, I’m not sure you could suss religion out of it. (Though, again, I’m hardly an expert.)
All I was saying is that doing what one wants is insufficient for free will. One must also perform the actions BECAUSE one wants to. That’s actually consistent with what you say in your subsequent post.
[I’m still waiting for someone to come and save my “half-compatibalism” from the abyss! Why is it that keiths is only around when I DON’T want him around?? I don’t want to have to *gulp* REVISE MY VIEWS!!!!]
I think you may be right. I probably misinterpreted Bill’s statement. Apologies to the readers:
Selection can’t select for non-existent traits. Furthermore, selection is a barrier not a facilitator of evolving features requiring simultaneous existence of many parts before they become beneficial. As Gould said, “what good is half a wing”.
That selection maintains an existing trait is not necessarily evidence selection evolved it especially if the intermediate steps would be selected against.
Not to mention, depending on the number of changes involved, if it requires a large amount of changes in the genome and glycome, selection may not have the population resources in the first place to achieve a transformation from an ancestral form.
Ergo, I don’t view natural selection as unequivocally an explanation for convergence.
But I’m glad you agree the features placental and marsupial mammals are non-random. Progress!
None of this seems in any way a response to the argument. What nonexistent traits are you thinking of? What simultaneous existence of many parts? What half-wing equivalent? Let’s remember the context: convergent placentals and marsupials. I don’t think there are any new parts involved at all.
Speculation based on nothing, but I do appreciate the gratuitous introduction of another -ome word.
Let’s face it: you desperately want to rule it out, and I have explained that to destroy your argument it only has to be plausible. “Unequivocally” isn’t necessary.
But I’m glad you agree the features placental and marsupial mammals are non-random.Progress!
It isn’t progress if everyone had already agreed before you said anything. Progress would require a change.
swamidass:
Well, then your priorities are hopelessly out of whack. Screw the papers and the grants; TSZ is more important. 🙂
Right, and it at least seems logically possible, as far as I can see. The problem is the next hurdle: it needs to be more likely than the alternative explanations.
Which is more likely?
a) evolution looks undirected and contingent because it’s undirected and contingent; or
b) evolution looks undirected, but actually it isn’t, and in fact it was fully intended to produce us, a particular primate species, even though the scientific evidence points in the opposite direction. God’s “best possible world” just happens to be one in which evolution gives every appearance of being undirected, with no purpose.
Option (a) comports with the scientific evidence; option (b) runs counter to it. You’d need some very strong reasons to prefer (b) to (a). What could justify such a choice?
How? I actually think it undermines Plantinga’s “free will defense”, and I’ll be doing an OP on that topic shortly.
It solves the problem of libertarian free will vs. predestination, but libertarian free will is incoherent for other reasons, so that isn’t much of a victory.
Compatibilist free will was already reconcilable with predestination, so it doesn’t help there either.
Yes, it can be purposeful, though I’d scratch the words “and predetermined”, because 1) “pre” implies “before”, and that’s emphatically not what we’re talking about here, since God is timeless in this scenario, and 2) “determined” implies determinism, and that’s also not what we’re talking about.
Agreed, though it does raise the question of why God would choose to work so indirectly in this case but not in others.
I think it actually allows the opposite — a completely chaotic and random world, lacking natural laws, that nevertheless achieves God’s purposes purely randomly and is selected by him, out of all possible worlds, for that reason.
Thanks. I’m actually an engineer by profession, but with a fascination for philosophical issues.
Indeed, Sal. You’ll struggle to find anyone here (or any evolutionary biologist – or anyone seriously interested in evolutionary biology) who has suggested the evolutionary processes that produced wolves and thylacines were random. Selection is not random. Remember the niche, Sal. Remember the niche!
Let it go and wave it good-bye. We don’t live in a deterministic universe!
keiths:
dazz:
How so? What part of that cannot be accomplished timelessly?
Which is okay. The events are still random — non-deterministic — even if the world in which they occur was chosen based on their outcome.
He’s guaranteeing the desired outcome, but he’s not doing it via determinism.
Maybe it will help if I specify exactly what I mean by “random” and “deterministic” here. I see them as opposites — there is no randomness in a fully deterministic system, and no determinism in a fully random system. The next state of a fully deterministic system is a function of its current state and nothing else. If the next state can’t be predicted (at least in principle) from the current state, then at least one random event is occurring.
I hope you’ll agree that a created world can contain random events by that definition. If so, then how does it make a difference that the choice to create a particular world was based on the timeless knowledge of the outcomes of random events within it? They’re still random, not deterministic.
Indeed. Allow the possibility of one random event and determinism disappears: compatibilism disappears.
dazz,
The guarantee is not internal to A and C. If A has libertarian free will, then there is literally nothing about the state of A and C at time t0 that guarantees that A will choose X at time t1. The choice is not deterministic.
We’re talking about cases in which the choice is not fully determined by the initial conditions. God timelessly knows the outcome of the choice — even if it is random, not deterministic — and uses that knowledge to decide whether to create the world in question.
If you find it difficult to swallow the idea that God can timelessly know the result of a libertarian choice or the outcome of a random event, consider the alternative. If God doesn’t know these things, then he doesn’t know them period. He’s timeless, and so his lack of knowledge can never change to knowledge. He’s stuck in ignorance.
That leads to an absurdity: we humans can know the outcomes of libertarian choices and random events, simply by waiting until they happen and observing them. God can’t.
The absurdity is avoided if you accept that God can know these things timelessly.
dazz:
walto:
I’m with dazz. Compatibilism is simply the idea that determinism is compatible with free will. Whether God is involved is irrelevant.
walto:
dazz:
Right. Determinism is actually a bit of a red herring in the free will debates, because randomness is just as problematic for libertarian free will as determinism is.
I have no idea. Not familiar with the interface but nevermind.
If god causes A timelessly, and A causes B randomly, timeless causation rules apply to God -> A, but random rules must apply to A -> B. I really don’t want to get stuck in this loop where we both keep repeating the same arguments. I think “guaranteeing the desired outcome, via non-determinism” is still an oxymoron, but I’ll keep reading with interest in case someone else can clarify all that stuff
I have been persuaded by walto that knowledge doesn’t determine or cause things to happen, but a God that creates an undeterministic process knowing everything about it doesn’t mean he can create it with the purpose of creating the outcome: he doesn’t create the outcome, the random process does, he has no say on what the process will produce, he just happens to know it.
Well, that looks a lot like the point I’ve been trying to make all along. Timelessness is irrelevant, random is random
Agreed, in that case, middle knowledge breaks apart and becomes useless.
As I said above, that finally sunk a few days ago. I’m not disputing the idea that knowing something will happen (or knowing it timelessly) doesn’t cause or determine it.
My objection to middle knowledge is not about the knowledge, it’s about the causation chain and how it relies on determinism. Knowing the outcome of a deterministic process doesn’t hinder free will, it’s causing the initial conditions that in turn must cause the decision what does. I googled it an of course, that’s an obvious objection to MK made before
And what appears? Libertarian free will or no control or power over our actions at all?
Right. I think that’s what Alan is missing.
Re my half-caf compatibalism, If someone designed and programmed a robot to bark and reset its internal timer at precisely 2 PM EST yesterday and it did so, can we make sense of the claim that it did so freely?
It turns god into a timeless casino owner
Exactly, randomness destroys Libertarian free will. It does not facilitate it
It also destroys science
peace
Please, stop pontificating about science. It’s even more embarrassing than usual
I would like to see a detailing of your views and an explanation of why you find a Sovereign God to be so objectionable when a sovereign nature is not?
peace
Nature exists
If randomness is real prediction is impossible. It’s nothing to be embarrassed about it’s a fact.
The universe is a complex entity in which no part is disconnected from another. This has become even more apparent with our understanding of entanglement.
Our understanding of chaos theory makes it clear the even a tiny effect can influence the entire system
If one part of the universe is random the entire shebang is unstable and unpredictable.
peace
When you capitalize it it’s pantheism
peace
I’m not sure there’s a coherent view there to detail. Dazz said the distinction I was trying to make seemed ridiculous, keiths concurred, and they may very well be correct. But the intuition behind it is reflected in my question about the barking robot above. If MY motives, desires, etc are secondary to somebody else’s in the great chain of being it seems weird to me to say I’m acting ‘of my own accord.’
I get that’s not a great argument, but it may take me some time to adjust to this realization….
What’s wrong with that?
Or, you know, the start of a sentence.
You might want to consider the difference.
Glen Davidson
Casino owners seem to make a lot of money out of being able to predict in the presence of randomness.
Part of what Dembski is getting at in that paper are the different meanings of “random.”
When you consider all the various living creatures that exist, from the noble to the nasty, it would seem that if evolution had a purpose, it is not to create humans, but instead it is to simply explore all the possibilities.
The acceptance that evolution is real depends upon understanding that nature has ways of generating and selecting among those possibilities without the need for conscious intent. Some people’s desire to choose between consciousness and chaos is a false dichotomy, as most of nature lies in-between the two.
It really manages to avoid exploring all possibilities, instead exploring rather nearby possibilities for the most part.
Glen Davidson
dazz:
Right. I’m trying to figure out why you see that as a problem.
Note that what I actually said was:
The non-determinism doesn’t guarantee the outcome. What guarantees the outcome is God’s choice to instantiate a particular world — one in which the desired outcome happens non-deterministically.
That’s the point. By choosing to create a particular world, he gets what he wants — a desired state of affairs — without undermining the randomness of the events leading to that state of affairs. The events still meet the definition of randomness I gave above, as you can verify.
Neil Rickert,
Is randomness real, or is it just a concept in our minds?
keiths:
dazz:
I don’t see that at all. Could you explain why you think it “breaks apart”?
In the scenario we’ve been discussing, the initial conditions do not cause the decision. The decision is non-deterministic.