Problems with Richard Dawkins' arguments against God
Posted: Sat Nov 25, 2006 4:46 pm
I crossposted this on Dawkins' forum and thought that perhaps you might be interested in seeing what some non-theological problems with "The God Delusion" are.
-------
I apologize right away for the length of this post but this argument has to be critically analyzed.
Here I discuss Dawkins' "Ultimate 747" argument against God's existence (Ch. 4). While his earlier discussion of the anthropic principle is rather persuasive, there are some severe logical and conceptual problems with the “Ultimate 747” argument. These objections are on purely epistemological and scientific grounds, not on religious grounds. I have discussed my concerns with two other fellow astrophysicists, one at Harvard and one at CalTech, both of whom are not religious and both of whom come to the same general conclusions I do. I am also a non-religious, astrophysicist (at the Harvard-Smithsonian Center for Astrophysics). I have also passed this along to an atheist philosophy professor I respect, and the main charges I raise appear to pass his test as well. To be perfectly blunt (and I mean no disrespect), this is one of the shoddiest arguments concerning God's existence/non existence that I've ever heard.
I. Dawkins' argument (roughly) goes like this:
Theists (e.g. John Polkinghorne) invoke God the Fine Tuner to explain why the 6 fundamental constants (in particle physics/gravitation, etc.) are the way they are since the probability of them being different - or rather, the probability of having a universe much like that we know today with them different - is so vanishingly small that it is highly unlikely it could have happened by chance. However, multiverse theories or 'cosmological Darwinism may be able to explain the supposed 'fine tuning' of these fundamental constants. These theories, as I'm well aware of, are serious ones that are not only falsifiable in principle but (in many cases) falsifiable in practice since they some predictions for the state of the universe prior to recombination (e.g. when the universe was still optically thick to radiation).
Even if no cosmological model allowed a means by which we could explain some of the fundamental parameters of the universe, (direct quotes here) ... "the theist's answer is deeply unsatisfying because it leaves the existence of God unexplained. A God capable of calculating the Goldilocks values (the just right ones) for the six numbers would have to be at least as improbable as the finely tuned combination of numbers itself, ad that's very improbable indeed..."
This argument, though, not just undermines the fine-tuning argument for God's existence but it stands on its own as an argument against God's existence. This is because “ …God is so complex and statistically improbable (and because we have alternate explanations??), it is highly unlikely that God exists.
II. Refutation
There are three separate problems with this argument, any one of which completely undermines the argument. To understand these, it would be useful to have some scientific background and some background in epistemology (especially Kantian epistemology).
A. Applying Conditions to Unconditioned Realities
Dawkins says (no less than about a dozen times or so) that the God hypothesis is a scientific hypothesis. What does it appear that Dawkins thinks that God is a scientific hypothesis? If you take him literally, the reason appears to be that “God's existence is a scientific fact about the universe, discoverable in principle …” He continues: “f he existed and chose to reveal it, God himself could clinch the argument … in his favor”.
Scientific hypotheses, though, deal with statements about the regularities of objects, regularities that can be tested, and experiments concerning these regularities are repeatable. Moreover, all objects of scientific concern are objects of perception. To be an object of perception is to necessarily be, as Kant observed, phenomena whose existence is conditioned by space and time. Indeed, our knowledge about scientific objects is conditioned by the means by which we can acquire knowledge: we observe them, necessarily, in time and space (space being the condition of external experience and time being the condition of all experience, internal or external), and it is problematic, to say the least, to talk about 'objects' apart from their spatio-temporal characteristics. Any causal relationships, from which talk of 'probabilities of existing' have any meaning, conferred on them are not *in* the object but result from how we perceive them: an insight Kant gleaned from studying Hume's attacks on the cosmological argument. All objects of science are empirical objects, and as such, we know of them only by sensory experience. All sensory *information* is manufactured, by created by that process in the brain upon receiving it. However, we can never cut out the middle man in this process and just experience what yields this information *as it is*, that is to say as noumena.
God, though, is not like all the examples of phenomena, whereby we can make statements about the probability of this or that existing or this or that being a likely causal explanation. God is being outside the universe, is not a conditioned object like a proton and is not perceived in the way that protons are perceived: that is to say, not perceived at all. God is not a 'scientific fact about the universe' since 'scientific facts about the universe' are restricted to phenomena, objects of perception, which are conditioned realities.
In case that argument went over your head, consider that treating God as a conditioned object also runs counter to the very definition of God that Dawkins claims he is employing in the first place: God is a “superhuman, supernatural intelligence …who created the universe and everything in it”. Putting aside the rather impoverished concept of 'created' implied here, if a being is somehow 'supernatural' how could its existence be conditioned by nature (that is to say be an object of scientific inquiry)? Furthermore, how could the existence of God possibly be a 'scientific fact about the universe' if God's existence transcends the universe? But even if Dawkins' definition was completely without problems, the problem of misidentifying God as a conditioned, perceived object for which ideas of 'probability' have any meaning whatsoever still stands. Is there way in which we could possibly have knowledge of God? Perhaps, especially through negative theology, but there's no clear answer in my opinion. We can have knowledge of 'some' things without recourse to synthetic analytical propositions (e.g. definition of a triangle) or empirical propositions (about objects of perception), both of which would be about conditioned objects. Rules of logic, such as commutative and identity properties are candidates. But even if we couldn't come up with examples it only demonstrates that we are limited to conditioned realities in our knowledge pool. Whether or not there's something more beyond this is not a question we can answer.
Since God is something which is unconditioned, that is to say not an object of perception, God's existence cannot possibility be a scientific hypothesis, and the whole argument about probability comes crashing down. Amusingly, Dawkins makes the same mistake as Aquinas (whom he ridicules) in that he applies conditions to unconditioned realities.
B. Smuggling the 747 Argument Outside of the Context in which it has Meaning
This problem is logical as opposed to epistemological.
The entire argument Dawkins poses is predicated on the assumption that we have to take the Fine-tuning argument seriously, and then the argument is turned on its head. I, of course, think the 'Fine Tuning' argument is silly. The anthropic principle is perfectly sufficient. The universe is the way it is because if it weren't the way it is then we wouldn't be here to argue about it. I find the argument answered analogously to going to the parking lot, finding the license # of the first car and not concluding that because the probability of that combination is so low then the car with that # must have been put in front of me on purpose. So the need to demonstrate God's existence by the universe's supposed fine tuning is a desire dead to me.
However, Dawkins' entitles the chapter "Why There Almost Certainly is No God" and proceeds on with the next section as if he's demonstrated that God almost certainly doesn't exist. The title is not “I've almost certainly shown why the Fine Tuning argument doesn't work.” As it appears in the text, the entire argument about God's probability of existing was invoked in the first place to counter the theist argument, not as a standalone argument. Imagine what the chapter would look like if the argument for God's existence is never discussed and he suddenly foists the probability of the 6 constants as being the probability of God's existence. You would think it would be completely irrelevant.
What if I reject the reason for needing the counterargument in the first place? One would think that any assignment of God's probability of existing from the section on Fine Tuning, however well/poorly motivated, would be a moot point if the supposed basis for discussion it in the first place is never brought up/irrelevant. To put it another way, if the entire point about Fine Tuning one of the constants is irrelevant (since there's a naturalistic explanation or the entire question is misguided), why still tether the irrelevant fine tuning probability to God as being the probability of God's existence?
And by the way, which 'unlikely probability' should it be? The combination of 6 constants is one choice. Another possible choice comes from the 'Flatness Problem' in cosmology. In this problem, if the universe had an energy density, Omega_o, that departed by more than 1 in 10^59 from being =1 at Planck time then the universe either would have collapsed into a Big Crunch in a nanosecond or expanded too fast to allow large-scale structure to form (e.g. no galaxies). The chances of the universe having Omega_o within the desired range are ridiculously slim. The naturalistic answer, of course, is inflationary big bang cosmology. But why isn't Dawkins using this probability? It seems to be just as legitimate and fundamental to the universe to me. For many natural phenomena (e.g. the existence of the Earth, of the eye, of DNA, etc.) explained by naturalistic processes you can conceivably assign a probability for its appearance being 'chance'. So which is it: the probability of DNA being the way it is? the 6 constants?
Dawkins strategy begs a further question. The structure of all these probability arguments contains naturalistic alternatives to the God as an explanation. The structure is 'complicated outcome A is supposedly explained by God but is really explained naturalistically, therefore God's existence is unlikely'. Dawkins puts up an artificial barrier between the 6 constants and all other 'unlikely' outcomes/conditions like Omega_o, the appearance and structure of DNA, etc. This is done with good reason because breaking down the barrier between this 'wild' physics and physics more typical of everyday experience reveals some rather absurd conclusions one is forced into. The best example is the following. I take my cup of coffee and release it. Knowing nothing about physics there seems to be an equal probability of it going up, down, left, right, not moving, etc. The possible vectors for motion are enormous in number, yet the cup always falls to the floor. It seems highly unlikely that this situation, given all logically possible outcomes, this could happen by chance. Furthermore, I discover that the motion of my cup of coffee is consistent with a r^-2 law. The (ignorant) theist concludes that this is 'designed' or 'willed' by some Intelligence that 'wants' this to be true while the physicist appeals to the theory of gravity. The physicist then argues that because a being able to so exquisitely fine tune the cup's motion and the behavior of the cup to fall exactly the same way each time would have to be even more improbable, so such a being (God) is unlikely to exist. This is essentially what Dawkins is doing. The fact that he appeals to the 6 constants as opposed to gravity is completely irrelevant. The r^-2 relationship for the force of gravity is just as fundamental as those 6 constants. Both arguments have good naturalistic causal explanations, and a minor change in both would result in a 'universe different than what we have'. Taken to the absolute extreme, Dawkins' argument morphs into “there are naturalistic explanations in the universe, therefore God is unlikely to exist”. This is a very strange, almost 'Anselmian' position to be in and is completely unconvincing to me. But even if you don't buy the argument presented in this paragraph, Dawkins still smuggles his argument out of the context of which it has meaning and decides that it stands alone as an argument against God's existence. This is not acceptable.
C. Conflating Different Definitions of Probability and Application of Probability where it is Meaningless
This is argument exploits a scientific problem in Dawkins' argument.
We have a means by which to assign probability to the universe being the way it is if we tinker with the fundamental constants or some other cosmological parameter. This probability, though, is a mathematical one, with well-described initial conditions and some sort of implicit distribution function. The clearest example I can think of is the Saha equation in solar physics which describes the fraction of free particles vs. bound particles (e.g. the 'ionization fraction') that depends on initial conditions such as the transition energy and the fluid temperature. Note that these are probable outcomes, not prior probabilities.
The error in Dawkins' analogy is that, granting that there could be a need for a 'Designer', I have no idea how you 'assign' a probability for God, specifically how you could assign the prior probability for something a probability (that of the 6 constants) that is not a prior probability but a probability of outcomes. The only sense in which God's existence could be a probability is a Bayesian probability since that deals with prior probabilities. But this is absolutely not the probability talked about with the 6 constants. The probability for the 6 constants is not a probability of them being the way they are (that probability is =1!) but a probability of the universe being 'the way it is' if the constants were different. It would be one thing if by "God" we meant the possible outcome among many outcomes at some initial condition ( initiated by a "meta-God/universe??"), and that we had a range of initial values (or some sort of distribution function: possible alternative outcomes would be “Schmod” and the “FSM”) for which "God" would be realized then at least in principle one could establish the probability of God's existence in the way Dawkins wants to. But this is not the notion that theists have of God for the reasons presented in A. and B. and also according to Dawkins' own definition. The probability of God's existence is not the same kind of probability describing outcomes of the universe if we tweak constants. Citing the probability of the universe being the way it is and, in the same sentence, pinning this probability on a non-empirical, non-outcome, non-object (God) is an abuse of the word 'probability'.
-----
I regret to have to make such negative comments. Richard Dawkins is a very gifted scientist, and I have no doubt that he is brilliant person on the whole. This “Ultimate 747 argument”, however, betrays his brilliance. As to why the theologians at the Templeton-sponsored conference couldn't get the supposed brilliance of his argument, there is one thing to say. It is highly unlikely that they were trained scientists, let alone astrophysicists, and are also unlikely, unless they follow Rahner, to argue from Kantian epistemological standpoints, which are necessary to expose some of the argument's problems. Besides, if theologians as a lot are really as unsophisticated in their thinking as Dawkins tends to imply (Keith Ward and Bonhoeffer are apparently exceptions) then should he really have felt confident that by surviving their critiques his argument has passed some serious hurdle? Maybe he should have talked to my colleagues.
-------
I apologize right away for the length of this post but this argument has to be critically analyzed.
Here I discuss Dawkins' "Ultimate 747" argument against God's existence (Ch. 4). While his earlier discussion of the anthropic principle is rather persuasive, there are some severe logical and conceptual problems with the “Ultimate 747” argument. These objections are on purely epistemological and scientific grounds, not on religious grounds. I have discussed my concerns with two other fellow astrophysicists, one at Harvard and one at CalTech, both of whom are not religious and both of whom come to the same general conclusions I do. I am also a non-religious, astrophysicist (at the Harvard-Smithsonian Center for Astrophysics). I have also passed this along to an atheist philosophy professor I respect, and the main charges I raise appear to pass his test as well. To be perfectly blunt (and I mean no disrespect), this is one of the shoddiest arguments concerning God's existence/non existence that I've ever heard.
I. Dawkins' argument (roughly) goes like this:
Theists (e.g. John Polkinghorne) invoke God the Fine Tuner to explain why the 6 fundamental constants (in particle physics/gravitation, etc.) are the way they are since the probability of them being different - or rather, the probability of having a universe much like that we know today with them different - is so vanishingly small that it is highly unlikely it could have happened by chance. However, multiverse theories or 'cosmological Darwinism may be able to explain the supposed 'fine tuning' of these fundamental constants. These theories, as I'm well aware of, are serious ones that are not only falsifiable in principle but (in many cases) falsifiable in practice since they some predictions for the state of the universe prior to recombination (e.g. when the universe was still optically thick to radiation).
Even if no cosmological model allowed a means by which we could explain some of the fundamental parameters of the universe, (direct quotes here) ... "the theist's answer is deeply unsatisfying because it leaves the existence of God unexplained. A God capable of calculating the Goldilocks values (the just right ones) for the six numbers would have to be at least as improbable as the finely tuned combination of numbers itself, ad that's very improbable indeed..."
This argument, though, not just undermines the fine-tuning argument for God's existence but it stands on its own as an argument against God's existence. This is because “ …God is so complex and statistically improbable (and because we have alternate explanations??), it is highly unlikely that God exists.
II. Refutation
There are three separate problems with this argument, any one of which completely undermines the argument. To understand these, it would be useful to have some scientific background and some background in epistemology (especially Kantian epistemology).
A. Applying Conditions to Unconditioned Realities
Dawkins says (no less than about a dozen times or so) that the God hypothesis is a scientific hypothesis. What does it appear that Dawkins thinks that God is a scientific hypothesis? If you take him literally, the reason appears to be that “God's existence is a scientific fact about the universe, discoverable in principle …” He continues: “f he existed and chose to reveal it, God himself could clinch the argument … in his favor”.
Scientific hypotheses, though, deal with statements about the regularities of objects, regularities that can be tested, and experiments concerning these regularities are repeatable. Moreover, all objects of scientific concern are objects of perception. To be an object of perception is to necessarily be, as Kant observed, phenomena whose existence is conditioned by space and time. Indeed, our knowledge about scientific objects is conditioned by the means by which we can acquire knowledge: we observe them, necessarily, in time and space (space being the condition of external experience and time being the condition of all experience, internal or external), and it is problematic, to say the least, to talk about 'objects' apart from their spatio-temporal characteristics. Any causal relationships, from which talk of 'probabilities of existing' have any meaning, conferred on them are not *in* the object but result from how we perceive them: an insight Kant gleaned from studying Hume's attacks on the cosmological argument. All objects of science are empirical objects, and as such, we know of them only by sensory experience. All sensory *information* is manufactured, by created by that process in the brain upon receiving it. However, we can never cut out the middle man in this process and just experience what yields this information *as it is*, that is to say as noumena.
God, though, is not like all the examples of phenomena, whereby we can make statements about the probability of this or that existing or this or that being a likely causal explanation. God is being outside the universe, is not a conditioned object like a proton and is not perceived in the way that protons are perceived: that is to say, not perceived at all. God is not a 'scientific fact about the universe' since 'scientific facts about the universe' are restricted to phenomena, objects of perception, which are conditioned realities.
In case that argument went over your head, consider that treating God as a conditioned object also runs counter to the very definition of God that Dawkins claims he is employing in the first place: God is a “superhuman, supernatural intelligence …who created the universe and everything in it”. Putting aside the rather impoverished concept of 'created' implied here, if a being is somehow 'supernatural' how could its existence be conditioned by nature (that is to say be an object of scientific inquiry)? Furthermore, how could the existence of God possibly be a 'scientific fact about the universe' if God's existence transcends the universe? But even if Dawkins' definition was completely without problems, the problem of misidentifying God as a conditioned, perceived object for which ideas of 'probability' have any meaning whatsoever still stands. Is there way in which we could possibly have knowledge of God? Perhaps, especially through negative theology, but there's no clear answer in my opinion. We can have knowledge of 'some' things without recourse to synthetic analytical propositions (e.g. definition of a triangle) or empirical propositions (about objects of perception), both of which would be about conditioned objects. Rules of logic, such as commutative and identity properties are candidates. But even if we couldn't come up with examples it only demonstrates that we are limited to conditioned realities in our knowledge pool. Whether or not there's something more beyond this is not a question we can answer.
Since God is something which is unconditioned, that is to say not an object of perception, God's existence cannot possibility be a scientific hypothesis, and the whole argument about probability comes crashing down. Amusingly, Dawkins makes the same mistake as Aquinas (whom he ridicules) in that he applies conditions to unconditioned realities.
B. Smuggling the 747 Argument Outside of the Context in which it has Meaning
This problem is logical as opposed to epistemological.
The entire argument Dawkins poses is predicated on the assumption that we have to take the Fine-tuning argument seriously, and then the argument is turned on its head. I, of course, think the 'Fine Tuning' argument is silly. The anthropic principle is perfectly sufficient. The universe is the way it is because if it weren't the way it is then we wouldn't be here to argue about it. I find the argument answered analogously to going to the parking lot, finding the license # of the first car and not concluding that because the probability of that combination is so low then the car with that # must have been put in front of me on purpose. So the need to demonstrate God's existence by the universe's supposed fine tuning is a desire dead to me.
However, Dawkins' entitles the chapter "Why There Almost Certainly is No God" and proceeds on with the next section as if he's demonstrated that God almost certainly doesn't exist. The title is not “I've almost certainly shown why the Fine Tuning argument doesn't work.” As it appears in the text, the entire argument about God's probability of existing was invoked in the first place to counter the theist argument, not as a standalone argument. Imagine what the chapter would look like if the argument for God's existence is never discussed and he suddenly foists the probability of the 6 constants as being the probability of God's existence. You would think it would be completely irrelevant.
What if I reject the reason for needing the counterargument in the first place? One would think that any assignment of God's probability of existing from the section on Fine Tuning, however well/poorly motivated, would be a moot point if the supposed basis for discussion it in the first place is never brought up/irrelevant. To put it another way, if the entire point about Fine Tuning one of the constants is irrelevant (since there's a naturalistic explanation or the entire question is misguided), why still tether the irrelevant fine tuning probability to God as being the probability of God's existence?
And by the way, which 'unlikely probability' should it be? The combination of 6 constants is one choice. Another possible choice comes from the 'Flatness Problem' in cosmology. In this problem, if the universe had an energy density, Omega_o, that departed by more than 1 in 10^59 from being =1 at Planck time then the universe either would have collapsed into a Big Crunch in a nanosecond or expanded too fast to allow large-scale structure to form (e.g. no galaxies). The chances of the universe having Omega_o within the desired range are ridiculously slim. The naturalistic answer, of course, is inflationary big bang cosmology. But why isn't Dawkins using this probability? It seems to be just as legitimate and fundamental to the universe to me. For many natural phenomena (e.g. the existence of the Earth, of the eye, of DNA, etc.) explained by naturalistic processes you can conceivably assign a probability for its appearance being 'chance'. So which is it: the probability of DNA being the way it is? the 6 constants?
Dawkins strategy begs a further question. The structure of all these probability arguments contains naturalistic alternatives to the God as an explanation. The structure is 'complicated outcome A is supposedly explained by God but is really explained naturalistically, therefore God's existence is unlikely'. Dawkins puts up an artificial barrier between the 6 constants and all other 'unlikely' outcomes/conditions like Omega_o, the appearance and structure of DNA, etc. This is done with good reason because breaking down the barrier between this 'wild' physics and physics more typical of everyday experience reveals some rather absurd conclusions one is forced into. The best example is the following. I take my cup of coffee and release it. Knowing nothing about physics there seems to be an equal probability of it going up, down, left, right, not moving, etc. The possible vectors for motion are enormous in number, yet the cup always falls to the floor. It seems highly unlikely that this situation, given all logically possible outcomes, this could happen by chance. Furthermore, I discover that the motion of my cup of coffee is consistent with a r^-2 law. The (ignorant) theist concludes that this is 'designed' or 'willed' by some Intelligence that 'wants' this to be true while the physicist appeals to the theory of gravity. The physicist then argues that because a being able to so exquisitely fine tune the cup's motion and the behavior of the cup to fall exactly the same way each time would have to be even more improbable, so such a being (God) is unlikely to exist. This is essentially what Dawkins is doing. The fact that he appeals to the 6 constants as opposed to gravity is completely irrelevant. The r^-2 relationship for the force of gravity is just as fundamental as those 6 constants. Both arguments have good naturalistic causal explanations, and a minor change in both would result in a 'universe different than what we have'. Taken to the absolute extreme, Dawkins' argument morphs into “there are naturalistic explanations in the universe, therefore God is unlikely to exist”. This is a very strange, almost 'Anselmian' position to be in and is completely unconvincing to me. But even if you don't buy the argument presented in this paragraph, Dawkins still smuggles his argument out of the context of which it has meaning and decides that it stands alone as an argument against God's existence. This is not acceptable.
C. Conflating Different Definitions of Probability and Application of Probability where it is Meaningless
This is argument exploits a scientific problem in Dawkins' argument.
We have a means by which to assign probability to the universe being the way it is if we tinker with the fundamental constants or some other cosmological parameter. This probability, though, is a mathematical one, with well-described initial conditions and some sort of implicit distribution function. The clearest example I can think of is the Saha equation in solar physics which describes the fraction of free particles vs. bound particles (e.g. the 'ionization fraction') that depends on initial conditions such as the transition energy and the fluid temperature. Note that these are probable outcomes, not prior probabilities.
The error in Dawkins' analogy is that, granting that there could be a need for a 'Designer', I have no idea how you 'assign' a probability for God, specifically how you could assign the prior probability for something a probability (that of the 6 constants) that is not a prior probability but a probability of outcomes. The only sense in which God's existence could be a probability is a Bayesian probability since that deals with prior probabilities. But this is absolutely not the probability talked about with the 6 constants. The probability for the 6 constants is not a probability of them being the way they are (that probability is =1!) but a probability of the universe being 'the way it is' if the constants were different. It would be one thing if by "God" we meant the possible outcome among many outcomes at some initial condition ( initiated by a "meta-God/universe??"), and that we had a range of initial values (or some sort of distribution function: possible alternative outcomes would be “Schmod” and the “FSM”) for which "God" would be realized then at least in principle one could establish the probability of God's existence in the way Dawkins wants to. But this is not the notion that theists have of God for the reasons presented in A. and B. and also according to Dawkins' own definition. The probability of God's existence is not the same kind of probability describing outcomes of the universe if we tweak constants. Citing the probability of the universe being the way it is and, in the same sentence, pinning this probability on a non-empirical, non-outcome, non-object (God) is an abuse of the word 'probability'.
-----
I regret to have to make such negative comments. Richard Dawkins is a very gifted scientist, and I have no doubt that he is brilliant person on the whole. This “Ultimate 747 argument”, however, betrays his brilliance. As to why the theologians at the Templeton-sponsored conference couldn't get the supposed brilliance of his argument, there is one thing to say. It is highly unlikely that they were trained scientists, let alone astrophysicists, and are also unlikely, unless they follow Rahner, to argue from Kantian epistemological standpoints, which are necessary to expose some of the argument's problems. Besides, if theologians as a lot are really as unsophisticated in their thinking as Dawkins tends to imply (Keith Ward and Bonhoeffer are apparently exceptions) then should he really have felt confident that by surviving their critiques his argument has passed some serious hurdle? Maybe he should have talked to my colleagues.
