Michael Martin's disproof of God
Summary: Atheologians who are keen to disprove the idea of God often formulate disproofs of God based on analysing some of God's supposed attributes and demonstrating that they are inconsistent with each other or with themselves. In this essay I analyse one attempted disproof along these lines by philosopher Michael Martin. The problem with such disproofs, I argue, is that they set up definitions of God and then use those definitions to argue that God doesn't exist. But God is only the greatest possible being, and that is certainly coherent. So philosophers may be making up definitions that the greatest possible being cannot have, and thus essentially making up impossible characteristics that they use to disprove a being that could never have them, while ignoring the coherency of the greatest possible being, and what attributes THAT being has. That said, in the second section I put forward a hypothesis on why Michael Martin's definition of omniscience fails in disproving the greatest possible being. Finally I end on a more general note on all such arguments from an email discussion I had with someone where I mentioned this subject.
A philosopher called Michael Martin wrote an interesting attempted disproof of God's existence a while ago, available here, which I'd like to take a look at. Martin's argument, to rephrase it in my own words, is that the idea of omniscience (the way God can know everything) necessarily must include all kinds of knowledge. So, obviously, it would include propositional knowledge, like 'America is a country' but he argues also it must also include all other kinds of knowledge, such as experiential knowledge like how to swim. This wouldn't just be knowing how to swim as an intellectual idea, that you move your arms and legs in a certain way in the water, but to actually have knowledgeable experience of swimming. In the same way Michael Martin uses the example of experiencing, envy, lust, and other emotions that a God of complete love or goodness couldn't experience. If experiencing envy gives you a kind of knowledge of envy, and if God can't experience envy, then God can't really know everything and hence the idea of omniscience doesn't make sense. (Martin puts the argument a little differently, applying it only to the 'common person's' ideal of God, but since I think Martin's argument only has any force as long as it's applied to an orthodox Christian ideal of God, I am stating it against that. I am not sure of how that distinction is meant to work.)
I want to concentrate on this argument not just as a specific objection to God's existence, but as illustrative of a general type of argument I haven't yet addressed on this website. This will be done utilising two arguments that work against Michael Martin, and which also support each other. First of all, omniscience need not include all experiential knowledge to be omniscience, e.g. of swimming or lust, and secondly in some circumstances we intuitively reject the idea that omniscience does need to include all such kinds of experiential knowledge.
Martin's argument, I think, is ultimately based on the idea of contradictions existing in the attributes we give to God. God's non-physicality, his existence as a spirit, contradicts the idea that he can experience swimming, since a spirit cannot swim. In the same way, God's perfection contradicts with the idea God can experience envy, since envy is a morally bad emotion. As with lust, because God can't have sex, he cannot, it seems, experience lust. So Martin is showing that no matter what, God cannot really experience things that we can experience, through such contradictions. And through these contradictions we can see that God can't experience things humans can, he can't have, as Michael says all kinds of 'procedural knowledge'.
To make my first argument, before getting on Michael Martin's objection to God specifically, (I will get on to that in a second) I would like to point out how dangerous it is to argue that because two attributes of God are contradictory, then God doesn't exist. This point applies to all God-contradiction arguments generally. If someone points out a contradiction in two supposed attributes of God, then we should defend the idea of God as much as possible, but if two supposed attributes really are contradictory then it does not mean that God doesn't exist, but that we misidentified and mis-analysed the attributes he does have. The idea of 'the greatest possible being', which is a philosophical definition of God, certainly makes sense. So it may be that God can't really be omnipotent and omniscient in the way some people understand it, but that just means he has the next best thing, whatever that may be. And if that means God can't make two and two equal five, then he has the next best thing, which is to do anything logically possible (so, everything apart from making two and two equal five.) So, the greatest possible being can't make two and two equal five, but he can do everything else. And if tomorrow we discovered that God can't for example, make more than a certain number of planets in a second, then that will mean that the greatest possible being can't make two and two equal five, and can't make more than a certain number of planets. So we would define omnipotence differently, to cope with our new understanding, just as we would define omnipotence differently if we understood something new about omniscience that contradicted our understanding of omnipotence. But that doesn't mean the greatest possible being can't exist, only that he actually has fewer abilities that we originally thought. Philosophers need to be careful lest they fall into definitional confusions, and mistake those definitional confusions for real discoveries about God, or the non-existence of God.
The question from this in my mind is whether a clear, well-thought out, comprehensive analysis of the greatest possible being would show the greatest possible being can't have the attributes of the kind of God described in the Bible, because it turns out the Biblical God really can't have attributes we think he should have. So, the real question is, not whether two attributes of God are contradictory, since ultimately that can't disprove the greatest possible being, it is whether on closer philosophical analysis, the greatest possible being is not only limited in making two and two equal five, but in, for example, how much love he can give his creatures at any one time. If he was limited in such a way, that would undermine Christianity.
I think specifically in regards to Michael Martin, that he is putting the cart before the horse in this aspect. Martin is arguing that a personal God must necessarily have experiential knowledge, but that he can't have experiential knowledge of how to swim, for example, so he can't have as much knowledge as humans, and hence can't be all-knowing. But, remember, omniscience is only the best omniscience any being can have. Michael is putting the cart before the horse. Our idea of omniscience shouldn't define how how we think of the greatest possible being, rather the idea of the greatest possible being should define how we think of omniscience. It may be that the best kind of omniscience God can have, and that is actually possible, is one in which a supreme being doesn't have all kinds of experiential knowledge, just like the best kind of omnipotence doesn't include the ability to make two and two equal five. To be sure, I'll grant to Michael Martin that God doesn't know some things that we know, like lust or envy. But then again, you'd think that the problem would be that omniscience is nonsensical, NOT that God can't experience lust or envy. In other words, I see how it undermines Christianity that God can't be omniscience, I don't see how it damages Christianity to say that God can't experience lust. But I think Martin's argument only applies on the weaker level.
That said, I would like to make a second argument that Michael Martin's objection isn't that strong, supporting my one above - based on the idea that we can make analogies with Michael Martin's argument where it doesn't seem fair that he should have to know some things to be omniscient. And I would argue that this is because the concept of omniscience, properly understood, automatically excludes experiences God can't have by definition. And this can be shown in examples where we feel such knowledge can be excluded from God without undermining his omniscience.
In some sense, I can know propositionally how to eat, and I can know how to eat because I have experienced eating firsthand. In the same way I can know propositionally that a person called Will G exists (myself), and I can know propositionally that I am Will G. But the propositional knowledge that 'I', and only I am Will G, is a kind of propositional knowledge that no one else can have. God can't know 'I am Will G' as a statement about himself, nor can you, nor can anyone else who reads this. Only I can. But does this mean that God isn't omniscience in a propositional sense? Of course not. That would be very unfair on God; to say he can't know all facts just because He can't know He is Will G. Actually no atheist philosopher I know of has said such a thing, because it is so ridiculous to demand that of God. This is because, by definition, God can't know 'I am Will G', so it's unfair to suppose that he should know it in order to qualify for omniscience. But, as I will argue, experiential knowledge like how to swim can be contradictory in much the same way.
Is it possible to have experiential knowledge of how to be Will G? For example, I have swum before, so I know how to swim, as an experience that I've had. But could you argue that I also have experience of being Will G, as I am Will G? Is being yourself something you can have experience of? In other words, can I know 'how' to be Will G as a kind of procedural knowledge? I would say there is no good reason to think just like swimming, or playing tennis, that you can't experience and get experiential knowledge of being yourself. But if this is true, then Michael Martin's argument can be reduced to absurdity. Demanding that God must have all kinds of experiential knowledge, including the experiential knowledge of how to be Will G, is like saying that because God hasn't experienced being George W. Bush, he can't know everything. That sounds like some kind of joke on atheistic arguments. But it's really just the experiential version of the propositional argument that since God can't know propositionally 'I am Will G' or 'I am George Bush', then he doesn't have the right amount of knowledge to be qualified as omniscient.
The thing is, I'm not saying that experiential knowledge of how to swim is not knowledge, just as knowing how to lust is not knowledge. Just as I'm not saying that propositional knowledge that 'I am Will G' is not a kind of knowledge. But it doesn't matter to our idea of God or our definition of omniscience if God doesn't know some things. Just as we intuitively know that for some reason, it doesn't matter to God's omniscience that he can't have all propositional knowledge, like 'I am Will G', so it doesn't matter to God's omniscience that he can't have all experiential knowledge, like how to be Will G. There is something absurd about the example that since God can't know I am Will G, he can't know everything. And this absurdity that makes us reject that argument should make us reject Michael Martin's. The difference I think that makes it more absurd to say this than what Michael Martin was saying, is that many people can know how to swim, but not many people can know how to be Will G or George Bush, just one, and it is in fact logically impossible for all those other people to know.
So, it seems that if it's unfair in a propositional sense to declare that if God can't know 'I am Will G', or else he gets kicked out of the omniscience club, then it must be unfair in the experiential sense to demand God experience lust, swimming, or building a house with his hands or he gets kicked out of the experiential knowledge club. God can't know 'I am Will G' by definition, he can't experience being Will G by definition, he can't experience lust or swimming by definition. And if we say God is omniscient even though he can't know he is Will G, because God can't know that by definition, then it also must be unfair to demand God experience lust, blind hatred, swimming, and building a house if he cannot by definition.
The key point is what you can't know by definition. And what you can't know by definition is automatically excluded from the concept of omniscience as it can apply to you.
So, if an atheist does not wish to make the point that, because God can't experience being Will G, therefore he can't experience everything and is hence not omniscient, he cannot make the same point with regard to other experiential knowledge, like how to build a house, how to feel lust or how to swim. So whatever we may have thought at first, God gets off the philosophical hook in some way. And so it really doesn't undermine His existence at all to say He can't experience lust, envy, or how to swim.
1/26/07
Post Note
This is part of an email discussion I had with someone where I summarised some relevant points relating to this issue.
XXZ wrote,
Hey!
Have you seen this argument I found on a website of a philosopher?
[Link]
I wrote in reply
I had a very quick look at the article, just skimming it. It's interesting, because I think in a way it's similar to Michael Martin's disproof of God that I wrote about. No one that I know of has been convinced by Michael Martin's disproof of God, I personally cannot imagine anyone being convinced by it, because to me at least, it has so little force. Yet for me, actually figuring out why I probably thought this took a long time.
Ultimately I think a lot of atheistic arguments that attempt to disprove God are based on definitional confusions. For example, God is conceived of as being omnipotent, which means he can do anything, and omniscient, which means he knows everything. But he can't know how to lust, so he isn't omnipotent or omniscient. But the problem here is that omnipotence has been defined wrongly, as has omniscience. The argument arises because someone has made a *plausible* definition, that means God should know how to lust, but this plausible definition is just that: plausible, it is not the truth.
So in relation to this philosopher's argument I think he is using plausible definitions of spatial unlocatedness, space, spatial points etc. that have served him and philosophy well. But I think he is making definitions that are just plausible and do not encompass the whole truth of the concepts he is using. And thus he is getting caught in a definitional confusion, because he's using those 'only plausible' definitions to argue God doesn't exist, just like those who make up attributes of God and argue that since those attributes are contradictory, God can't exist.
Exactly how are his definitions just plausible and not the truth, you might ask? Well, first of all, philosophers have a really bad track record at defining things properly (we still don't agree on what 'mind' is, what 'person' is, what 'good' and 'evil' are, etc). Secondly, I think that his argument is only as strong as the plausibility not only of the definitions he is using, but of the precise delineation of those definitions, which I find implausible for any philosophical definition, and not only philosophical but just about any definition of anything generally except something really simple like, 'what is a tent.' And thirdly I don't find his argument at all persuasive I have the strong suspicion that his definitions do not encompass the whole truth of the concepts he is using.
I think these points should make us very cautious about making strong conclusions regarding ANY arguments that primarily rely on definitions about stuff. Whether it's an atheological argument for the nonexistence of God, or whether it's some elaborate theological argument regarding God's nature or something like that.
A philosopher called Michael Martin wrote an interesting attempted disproof of God's existence a while ago, available here, which I'd like to take a look at. Martin's argument, to rephrase it in my own words, is that the idea of omniscience (the way God can know everything) necessarily must include all kinds of knowledge. So, obviously, it would include propositional knowledge, like 'America is a country' but he argues also it must also include all other kinds of knowledge, such as experiential knowledge like how to swim. This wouldn't just be knowing how to swim as an intellectual idea, that you move your arms and legs in a certain way in the water, but to actually have knowledgeable experience of swimming. In the same way Michael Martin uses the example of experiencing, envy, lust, and other emotions that a God of complete love or goodness couldn't experience. If experiencing envy gives you a kind of knowledge of envy, and if God can't experience envy, then God can't really know everything and hence the idea of omniscience doesn't make sense. (Martin puts the argument a little differently, applying it only to the 'common person's' ideal of God, but since I think Martin's argument only has any force as long as it's applied to an orthodox Christian ideal of God, I am stating it against that. I am not sure of how that distinction is meant to work.)
I want to concentrate on this argument not just as a specific objection to God's existence, but as illustrative of a general type of argument I haven't yet addressed on this website. This will be done utilising two arguments that work against Michael Martin, and which also support each other. First of all, omniscience need not include all experiential knowledge to be omniscience, e.g. of swimming or lust, and secondly in some circumstances we intuitively reject the idea that omniscience does need to include all such kinds of experiential knowledge.
Martin's argument, I think, is ultimately based on the idea of contradictions existing in the attributes we give to God. God's non-physicality, his existence as a spirit, contradicts the idea that he can experience swimming, since a spirit cannot swim. In the same way, God's perfection contradicts with the idea God can experience envy, since envy is a morally bad emotion. As with lust, because God can't have sex, he cannot, it seems, experience lust. So Martin is showing that no matter what, God cannot really experience things that we can experience, through such contradictions. And through these contradictions we can see that God can't experience things humans can, he can't have, as Michael says all kinds of 'procedural knowledge'.
To make my first argument, before getting on Michael Martin's objection to God specifically, (I will get on to that in a second) I would like to point out how dangerous it is to argue that because two attributes of God are contradictory, then God doesn't exist. This point applies to all God-contradiction arguments generally. If someone points out a contradiction in two supposed attributes of God, then we should defend the idea of God as much as possible, but if two supposed attributes really are contradictory then it does not mean that God doesn't exist, but that we misidentified and mis-analysed the attributes he does have. The idea of 'the greatest possible being', which is a philosophical definition of God, certainly makes sense. So it may be that God can't really be omnipotent and omniscient in the way some people understand it, but that just means he has the next best thing, whatever that may be. And if that means God can't make two and two equal five, then he has the next best thing, which is to do anything logically possible (so, everything apart from making two and two equal five.) So, the greatest possible being can't make two and two equal five, but he can do everything else. And if tomorrow we discovered that God can't for example, make more than a certain number of planets in a second, then that will mean that the greatest possible being can't make two and two equal five, and can't make more than a certain number of planets. So we would define omnipotence differently, to cope with our new understanding, just as we would define omnipotence differently if we understood something new about omniscience that contradicted our understanding of omnipotence. But that doesn't mean the greatest possible being can't exist, only that he actually has fewer abilities that we originally thought. Philosophers need to be careful lest they fall into definitional confusions, and mistake those definitional confusions for real discoveries about God, or the non-existence of God.
The question from this in my mind is whether a clear, well-thought out, comprehensive analysis of the greatest possible being would show the greatest possible being can't have the attributes of the kind of God described in the Bible, because it turns out the Biblical God really can't have attributes we think he should have. So, the real question is, not whether two attributes of God are contradictory, since ultimately that can't disprove the greatest possible being, it is whether on closer philosophical analysis, the greatest possible being is not only limited in making two and two equal five, but in, for example, how much love he can give his creatures at any one time. If he was limited in such a way, that would undermine Christianity.
I think specifically in regards to Michael Martin, that he is putting the cart before the horse in this aspect. Martin is arguing that a personal God must necessarily have experiential knowledge, but that he can't have experiential knowledge of how to swim, for example, so he can't have as much knowledge as humans, and hence can't be all-knowing. But, remember, omniscience is only the best omniscience any being can have. Michael is putting the cart before the horse. Our idea of omniscience shouldn't define how how we think of the greatest possible being, rather the idea of the greatest possible being should define how we think of omniscience. It may be that the best kind of omniscience God can have, and that is actually possible, is one in which a supreme being doesn't have all kinds of experiential knowledge, just like the best kind of omnipotence doesn't include the ability to make two and two equal five. To be sure, I'll grant to Michael Martin that God doesn't know some things that we know, like lust or envy. But then again, you'd think that the problem would be that omniscience is nonsensical, NOT that God can't experience lust or envy. In other words, I see how it undermines Christianity that God can't be omniscience, I don't see how it damages Christianity to say that God can't experience lust. But I think Martin's argument only applies on the weaker level.
That said, I would like to make a second argument that Michael Martin's objection isn't that strong, supporting my one above - based on the idea that we can make analogies with Michael Martin's argument where it doesn't seem fair that he should have to know some things to be omniscient. And I would argue that this is because the concept of omniscience, properly understood, automatically excludes experiences God can't have by definition. And this can be shown in examples where we feel such knowledge can be excluded from God without undermining his omniscience.
In some sense, I can know propositionally how to eat, and I can know how to eat because I have experienced eating firsthand. In the same way I can know propositionally that a person called Will G exists (myself), and I can know propositionally that I am Will G. But the propositional knowledge that 'I', and only I am Will G, is a kind of propositional knowledge that no one else can have. God can't know 'I am Will G' as a statement about himself, nor can you, nor can anyone else who reads this. Only I can. But does this mean that God isn't omniscience in a propositional sense? Of course not. That would be very unfair on God; to say he can't know all facts just because He can't know He is Will G. Actually no atheist philosopher I know of has said such a thing, because it is so ridiculous to demand that of God. This is because, by definition, God can't know 'I am Will G', so it's unfair to suppose that he should know it in order to qualify for omniscience. But, as I will argue, experiential knowledge like how to swim can be contradictory in much the same way.
Is it possible to have experiential knowledge of how to be Will G? For example, I have swum before, so I know how to swim, as an experience that I've had. But could you argue that I also have experience of being Will G, as I am Will G? Is being yourself something you can have experience of? In other words, can I know 'how' to be Will G as a kind of procedural knowledge? I would say there is no good reason to think just like swimming, or playing tennis, that you can't experience and get experiential knowledge of being yourself. But if this is true, then Michael Martin's argument can be reduced to absurdity. Demanding that God must have all kinds of experiential knowledge, including the experiential knowledge of how to be Will G, is like saying that because God hasn't experienced being George W. Bush, he can't know everything. That sounds like some kind of joke on atheistic arguments. But it's really just the experiential version of the propositional argument that since God can't know propositionally 'I am Will G' or 'I am George Bush', then he doesn't have the right amount of knowledge to be qualified as omniscient.
The thing is, I'm not saying that experiential knowledge of how to swim is not knowledge, just as knowing how to lust is not knowledge. Just as I'm not saying that propositional knowledge that 'I am Will G' is not a kind of knowledge. But it doesn't matter to our idea of God or our definition of omniscience if God doesn't know some things. Just as we intuitively know that for some reason, it doesn't matter to God's omniscience that he can't have all propositional knowledge, like 'I am Will G', so it doesn't matter to God's omniscience that he can't have all experiential knowledge, like how to be Will G. There is something absurd about the example that since God can't know I am Will G, he can't know everything. And this absurdity that makes us reject that argument should make us reject Michael Martin's. The difference I think that makes it more absurd to say this than what Michael Martin was saying, is that many people can know how to swim, but not many people can know how to be Will G or George Bush, just one, and it is in fact logically impossible for all those other people to know.
So, it seems that if it's unfair in a propositional sense to declare that if God can't know 'I am Will G', or else he gets kicked out of the omniscience club, then it must be unfair in the experiential sense to demand God experience lust, swimming, or building a house with his hands or he gets kicked out of the experiential knowledge club. God can't know 'I am Will G' by definition, he can't experience being Will G by definition, he can't experience lust or swimming by definition. And if we say God is omniscient even though he can't know he is Will G, because God can't know that by definition, then it also must be unfair to demand God experience lust, blind hatred, swimming, and building a house if he cannot by definition.
The key point is what you can't know by definition. And what you can't know by definition is automatically excluded from the concept of omniscience as it can apply to you.
So, if an atheist does not wish to make the point that, because God can't experience being Will G, therefore he can't experience everything and is hence not omniscient, he cannot make the same point with regard to other experiential knowledge, like how to build a house, how to feel lust or how to swim. So whatever we may have thought at first, God gets off the philosophical hook in some way. And so it really doesn't undermine His existence at all to say He can't experience lust, envy, or how to swim.
1/26/07
Post Note
This is part of an email discussion I had with someone where I summarised some relevant points relating to this issue.
XXZ wrote,
Hey!
Have you seen this argument I found on a website of a philosopher?
[Link]
I wrote in reply
I had a very quick look at the article, just skimming it. It's interesting, because I think in a way it's similar to Michael Martin's disproof of God that I wrote about. No one that I know of has been convinced by Michael Martin's disproof of God, I personally cannot imagine anyone being convinced by it, because to me at least, it has so little force. Yet for me, actually figuring out why I probably thought this took a long time.
Ultimately I think a lot of atheistic arguments that attempt to disprove God are based on definitional confusions. For example, God is conceived of as being omnipotent, which means he can do anything, and omniscient, which means he knows everything. But he can't know how to lust, so he isn't omnipotent or omniscient. But the problem here is that omnipotence has been defined wrongly, as has omniscience. The argument arises because someone has made a *plausible* definition, that means God should know how to lust, but this plausible definition is just that: plausible, it is not the truth.
So in relation to this philosopher's argument I think he is using plausible definitions of spatial unlocatedness, space, spatial points etc. that have served him and philosophy well. But I think he is making definitions that are just plausible and do not encompass the whole truth of the concepts he is using. And thus he is getting caught in a definitional confusion, because he's using those 'only plausible' definitions to argue God doesn't exist, just like those who make up attributes of God and argue that since those attributes are contradictory, God can't exist.
Exactly how are his definitions just plausible and not the truth, you might ask? Well, first of all, philosophers have a really bad track record at defining things properly (we still don't agree on what 'mind' is, what 'person' is, what 'good' and 'evil' are, etc). Secondly, I think that his argument is only as strong as the plausibility not only of the definitions he is using, but of the precise delineation of those definitions, which I find implausible for any philosophical definition, and not only philosophical but just about any definition of anything generally except something really simple like, 'what is a tent.' And thirdly I don't find his argument at all persuasive I have the strong suspicion that his definitions do not encompass the whole truth of the concepts he is using.
I think these points should make us very cautious about making strong conclusions regarding ANY arguments that primarily rely on definitions about stuff. Whether it's an atheological argument for the nonexistence of God, or whether it's some elaborate theological argument regarding God's nature or something like that.
Labels: apologetics, michael martin, philosophy, philosophy of religion
posted by Will G | Thursday, January 25, 2007
2 Comments:
Hey there Will G! I have a question that has been driving me bananas. As a philosophy student, you might be able to help me with it. There was a book edited by Michael Martin which includes all sorts of "proofs" that God does not exist by redefining His attributes and showing that the redefined attributes are incompatible. Since these arguments are straw man arguments we, as theists, don't need to be worried about them. But there is one particular essay in this book (i.e. 'The Impossibility of God') which I am not quite sure how to react to. the author starts off discussion the modal ontological argument for God's existence, then says one of the steps was a logical fallacy, and with the "correct" version of the fallacious premise (which is also a lemma) derives that God necessarily does not exist. Why this argument against the existence of God is troubling for me, I suspect, is simply because I do not know very much about modal logic. Perhaps you could help me out here. The author of the paper (in 'The Impossibility of God') is John L. Polluck. And what I can understand of his ontological disproof seems fairly clearly to be obviously true. But we know God exists by faith (and reason). Anywho, if you could take a look at the paper for me-if you are willig and able to-and offer any clarification on what he is trying to say, that would be greatly appreciated. Thank you in advance for any help you may give. Shalom out!
Hi evangelical, I don't have Michael Martin's book and I haven't interacted with ontological disproofs of God that much. One thing to remember is that these are really abstract theories and unless they can be related to something more concrete then I wouldn't worry about them. Maybe there's some good stuff on wikipedia or William Lane Craig's site... sorry I can't be of more help.
Post a Comment
<< Home