Graham Oppy examines contemporary arguments for and against the existence of God. He shows that none of these arguments are persuasive enough to change the minds of those participants on the question of the existence of God. His conclusion is supported by detailed analyses of contemporary arguments, as well as by the development of a theory about the purpose of arguments, and the criteria that should be used in judging whether or not an argument is successful. Oppy discusses the work of a wide array of philosophers, including Anselm, Aquinas, Descartes, Locke, Leibniz, Kant and Hume, and more recently, Plantinga, Dembski, White, Dawkins, Bergman, Gale, and Pruss. ~ Product Description
Table of Contents
- 1. Some preliminary questions addressed
- 2. Ontological arguments
- 3. Cosmological arguments
- 4. Teleological arguments
- 5. Pascal’s wager
- 6. Arguments from evil
- 7. Other arguments
- 8. Concluding remarks.
1 Preliminary Considerations
There are four preliminary topics that I wish to take up in this first chapter. The first topic is the question of how best to provide a taxonomy of arguments about the existence of orthodoxly conceived monotheistic gods. Here, I am happy to follow the more or less ad hoc system of classification that has grown up around Kant’s classification of theoretical arguments into the ontological, the cosmological, and the teleological. Since I think that each argument should be treated on its merits, I don’t care how the arguments are grouped together; what really matters is that no arguments should be neglected.
The second topic is the question of how best to think about the virtues of arguments. When should we say that an argument for a given conclusion is a successful argument? I defend the view that, in circumstances in which it is well known that there has been perennial controversy about a given claim, a successful argument on behalf of that claim has to be one that ought to persuade all of those who have hitherto failed to accept that claim to change their minds. While this view sets the bar very high, there are, I think, good reasons for preferring it to views that would have one saying that there are successful arguments for conclusions that one does not oneself accept.
The third topic is the question of thetenability of agnosticism. There are many people who have supposed that sensible, thoughtful, reflective, well-informed people cannot be agnostics: the only alternatives are (some kind of) theism and atheism. I think that this supposition is mistaken: if one supposes – as I do – that there can be sensible, thoughtful, reflective, well-informed people who are theists, and that there can be sensible, thoughtful, reflective, well-informed people who are atheists, then it is perhaps not very surprising that one can also maintain that there are sensible, thoughtful, reflective, well-informed people who are agnostics, that is, who suspend belief on the question of whether there is an orthodoxly conceived monotheistic god.
The fourth preliminary topic is the question of the alleged existence of cases for alternatives to orthodoxly conceived monotheistic gods that are no less strong than the case for orthodoxly conceived monotheistic gods themselves, for example, the alleged existence of a case for belief in an omnipotent, omniscient, and perfectly evil creator of the world that relies on ‘parallels’ to the various arguments that are standardly offered on behalf of belief in an orthodoxly conceived monotheistic god. Here, I argue that the existence of these ‘parallel cases’ does not provide the materials for a successful argument against the existence of orthodoxly conceived monotheistic gods, even though those who do not accept that there is an orthodoxly conceived monotheistic god may be perfectly justified in maintaining that the cases are, indeed, entirely parallel.
1.1. ARGUMENTS ABOUT THE EXISTENCE OF MONOTHEISTIC GODS
There are many different kinds of arguments that have been offered for and against the existence of monotheistic gods – and, indeed, for the existence of non-monotheistic gods as well, though we shall not be concerned here with any of these arguments. Attempts to classify – exhaustively and exclusively – all of these arguments for and against the existence of monotheistic gods are apt to end in frustration. Nonetheless, I shall attempt to provide a rough survey here, using traditional – Kantian – labels for classes of arguments that, in some sense, seem to be naturally collected together. Part of the point of the exercise is to test the Kantian system of labelling when it is extended to cover both argument for, and arguments against, the existence of monotheistic gods.
There are some arguments for and against the existence of monotheistic gods that proceed more or less a priori. In particular, there are ontological arguments that start from definitions, or claims about the contents of conceptions or ideas, or claims about what is conceivable or logically possible, or allegedly analytic claims about the concept of existence, or the like, and – without the addition of further premises that are not claimed to be knowable a priori – draw conclusions about the existence or non-existence of monotheistic gods. On this characterisation, arguments to the conclusion that a postulated divine attribute is incoherent, or that a postulated tuple of divine attributes are jointly inconsistent, are classified as ontological arguments. If this consequence is deemed unhappy, then one ought to distinguish between ontological arguments for and against the existence of monotheistic gods and non-ontological a priori arguments against the existence of monotheistic gods. When thinking about ontological arguments, it is important not to lose sight of the fact that, for more or less any ontological argument in favour of the existence of a given monotheistic god, there are typically closely related ontological arguments against the existence of that god. One of the most important questions to be raised in the context of discussions of ontological arguments concerns the significance of the existence of these ‘parodies’ of ontological arguments for the existence of monotheistic gods.
There are many kinds of arguments for and against the existence of monotheistic gods that draw upon claims that are allegedly supported by a posteriori evidence. In some cases, the claims in question are alleged to be synthetic a priori, and hence distinct in kind from the key analytic a priori premises in ontological arguments. However – even setting aside Quinean scruples about the alleged distinction between claims that are synthetic a priori and claims that are analytic a priori – it seems doubtful that we shall go too far wrong if we suppose that the various kinds of arguments that we are about to describe are properly said to be a posteriori (evidential) arguments for and against the existence of monotheistic gods.
I shall suppose that cosmological arguments for and against the existence of monotheistic gods are arguments with key premises that advert to very general structural features of the universe and/or our ways of theorising about the universe – temporal structure, modal structure, causal structure, explanatory structure, intelligible structure, axiological structure, and the like – of which it is not plausible to claim that we have an exhaustive (analytic) a priori knowledge. Since I think that mereology is (analytic) a priori, I suppose that any argument for or against the existence of monotheistic gods that appeals only to very general mereological principles is properly classified as an ontological argument. Some might suppose that modality is similarly (analytic) a priori, and hence that arguments for or against the existence of monotheistic gods that appeal only to very general modal principles ought also be classified as ontological arguments. However, since I know of no such arguments for or against the existence of monotheistic gods, I think that nothing hangs on this decision. One consideration that will be important in our discussion of cosmological arguments for the existence of monotheistic gods is that these arguments very often involve claims about the impossibility of certain kinds of infinite regresses (in the general structural features under consideration). Nonetheless, I doubt that it should be built into the analysis of cosmological arguments that all arguments of this kind must have a premise of this sort.
I shall suppose that teleological arguments for and against the existence of monotheistic gods are arguments with key premises that appeal to particular (contingent) features of the world that are alleged to be prima facie plausible instances of intelligent design. Typical teleological arguments for or against the existence of monotheistic gods advert to particular cosmogonic, physical, chemical, biological, psychological, or social features of the world. Many arguments that are classified as teleological arguments under the above proposal are classified as minor evidential arguments by other philosophers. While it may be plausible to claim that the strongest teleological arguments are those that appeal to certain cosmogonic or biological features of the universe, it is not clear why this consideration should lead us to suppose that arguments from consciousness, or the existence of language, or the existence of politically structured societies, and the like, should not also be classified as teleological arguments. Since our discussion shall focus on arguments about cosmogonic and biological features of the universe, it won’t really matter to us how this issue is resolved.
Even if we suppose that the class of teleological arguments is fairly broadly defined, it seems that there are many evidential arguments left over – and it also seems reasonable to say that these are minor evidential arguments. Among the many different kinds of arguments that can be assigned to this category, we might include arguments from various kinds of authorities, such as scripture, bodies of religious believers, common consent, religious leaders, and the like; arguments from religious experience and revelation; arguments from alleged miracles; and so forth. Since the distinction between cosmological arguments, teleological arguments, and minor evidential arguments is not very clear, the most important point on which to insist is that any evidential argument that is not classified as either cosmological or teleological must be classified as a minor evidential argument, so that no argument is overlooked.
Many people have supposed that there is a category of moral arguments for and against the existence of monotheistic deities. The evidential arguments that fall into this alleged category can also be assigned to one of the categories that we have already distinguished. If we focus on the alleged fact that our world has a fundamental moral structure, then we might take ourselves to be concerned with cosmological moral arguments. If we focus, instead, on facts about the allegedly providential distribution of goods and rights in human societies, or on alleged facts about the existence of conscience and a sense of right and wrong in human agents, then we might suppose that we are concerned with teleological moral arguments. If we focus on considerations about the mismatch between happiness and desert in this life, or on the allegedly greater happiness of believers, or the like, then we might suppose that we are dealing with minor evidential arguments. Of course, if there are non-evidential moral arguments, then these must be assigned to different categories.
The classification of arguments from evil is also problematic given the categories that we have established thus far. Perhaps the most satisfying suggestion is to assign all of the arguments from evil to the class of teleological moral arguments: these arguments are, after all, concerned with the distribution of goods and rights in our universe. Since some – but not all – arguments from evil are standardly said to be evidential arguments from evil, it might be thought that there is a prima facie difficulty that arises for this proposed classification. However, the distinction between logical arguments from evil and evidential arguments from evil turns merely on the question of whether the claim that evil exists – or that such and such kinds of evil exist, or that such and such quantities of evil exist, or that such and such particular evils have occurred, or the like – is logically inconsistent with the claim that given monotheistic gods exist, or whether that claim merely lowers the probability – or makes it implausible to believe, or somehow undercuts the justification for believing – that given monotheistic gods exist. Of course, even if we are happy to classify arguments from evil as teleological moral arguments, we may nonetheless want to give a quite separate treatment of particular cosmogonic or biological arguments for the existence of given monotheistic gods, and particular arguments from evil against the existence of given monotheistic gods. Moreover, if we suppose that arguments from evil are teleological moral arguments, then we shall doubtless suppose that there are other arguments against the existence of given monotheistic gods – for example, arguments from divine hiddenness, arguments from non-belief, and the like – that should also be classified as teleological moral arguments.
Apart from a priori and evidential arguments for and against the existence of monotheistic gods, there are also various kinds of non-evidential arguments for and against the existence of monotheistic gods. Arguments that we might include in this general category include the various versions of Pascal’s wager, James’s arguments on behalf of the will to believe, the many variants of Kant’s prudential moral argument, and so forth.
It is sometimes said that, apart from the various kinds of arguments that we have mentioned so far, there are also cumulative arguments that somehow combine these arguments into more powerful meta-arguments (or cases) for and against the existence of monotheistic gods. While this is doubtless so, there are pitfalls that we must be careful to avoid.
If we have two valid arguments, each of which entails the conclusion that a particular monotheistic god exists, then we can form a disjunctive argument that also entails the same conclusion. More generally, if we have a large collection of valid arguments, each of which entails the conclusion that a particular monotheistic god exists, then we can form a multiply disjunctive argument that also entails that same conclusion. However, it should not be supposed that a ‘cumulative’ argument that is formed in this way is guaranteed to be a better argument than the individual arguments with which we began (even if we are properly entitled to the claim that the arguments with which we are working are all valid). For, on the one hand, if all of the arguments are defective on grounds other than those of validity – for example, because they have false premises, or because they are question-begging – then the cumulative argument will also be defective. But, on the other hand, if even one of the arguments with which we began is not defective on any other grounds, then it is a cogent argument for its conclusion, and the cumulative argument is plainly worse (since longer and more convoluted). So, at the very least, we have good reason to be suspicious of talk about a cumulative case for the claim that a given monotheistic god does – or does not – exist that is based upon a collection of (allegedly) valid arguments for the claim that the god in question does – or does not – exist.
Of course, the argument of the preceding paragraph is not meant to cast doubt on the obviously correct claim that one can set out a derivation of a conclusion in which there are lemmas, that is, derivations of sub-conclusions that are appealed to in the derivation of the final conclusion. But – despite the appearance of occasional claims to the contrary – it should not be supposed that separate (valid) arguments for or against the existence of a given monotheistic god ever stand in this kind of relationship to one another. The premises in a valid argument may well stand in need of further support before it is plausible to claim that one has a cogent argument for the conclusion in question; but whatever form those supporting arguments take, they cannot be valid arguments that have the conclusion of the original argument as their conclusion.
Talk about a cumulative case makes much more sense if we suppose that we are dealing with ‘probabilistic’ – or ‘inductive’, or ‘evidential’ – arguments, in which the premises provide ‘probabilistic’ – or ‘inductive’, or ‘ evidential’ – support for their conclusions. A proposition can be more probable given p & q, than it is given either p alone or q alone. Given that the existence of a given monotheistic god is made more probable – to degree D1 – by evidence E1, and that the existence of that monotheistic god is made more probable – to degree D2– by evidence E2, it may be that the existence of that monotheistic god is made more probable – to degree D3, where D3 > D2 and D3 > D1– by evidence (E1& E2). However, once we start talking about accumulating evidence in this sense, it seems to me that the only interesting question to consider is how a given proposition stands in the light of all of the relevant available evidence. That a given proposition is probable given a carefully selected part of the total relevant evidence is not an interesting result. But – at the very least – this makes it very hard to be sure that one has succeeded in setting out a good probabilistic argument for any hotly disputed conclusion. We shall return to these grounds for scepticism about probabilistic arguments for perennially controversial doctrines in subsequent chapters.
In the following parts of this section, I shall make some preliminary comments about the nature of arguments and argumentation, and about the connection that obtains between argumentation and reasonable believing. It seems to me that these topics need much more careful consideration than they are typically afforded in discussions of arguments for and against the existence of monotheistic gods. Even so, I am conscious that the following discussion is capable of improvement in many different ways.
It seems to me that a full account of what makes for a successful argument requires at least the following: (1) an account of rationality and rational belief revision; (2) an account of arguments; (3) an account of rational argumentation amongst rational agents; and (4) an account of the difficulties that arise as a result of the fact that we are not perfectly rational agents. Plainly, this is no small task; and I cannot pretend that I shall do justice to this topic here. Nonetheless, I propose to sketch my answer to the question of what we should suppose is required of a successful argument; I shall make use of this answer when we come to consider various arguments for belief in monotheistic gods in subsequent chapters.
1.2.1 Rationality and Rational Belief Revision
It seems to me to be plausible to think that reasonable people can disagree. Indeed, it seems to me to be more or less platitudinous that there are propositions that p such that some reasonable people believe that p, some reasonable people believe that not p, and other reasonable people are agnostic or indifferent in one way or another. Moreover, it would be wrong to think that, where there is disagreement about the truth of some proposition that p, all but one of the parties to the disagreement must be manifesting irrationality with respect to the subject matter at hand. That is, it seems to me to be more or less platitudinous that there are propositions that p such that some reasonable people act reasonably in believing that p, other reasonable people act reasonably in believing that not p, and other reasonable people act reasonably in being agnostic or indifferent in one way or another.
Of course, some of the actual disagreements among reasonable people can be traced to irrationalities: sometimes reasonable people do have ‘blind spots’ where irrationality creeps in. Moreover, there is a substantial body of psychological research that suggests that our ‘reasonableness’ is actually quite imperfect – that is, even at the best of times, we are prone to all kinds of lapses from ideal rationality (especially when it comes to statistical and probabilistic reasoning). However, there are at least two other sources of disagreements among reasonable people that are equally significant. One is that we all have different bodies of evidence – we draw on different bodies of information – that we obtain in all manner of different ways. Even if – perhaps per impossibile – we were perfectly rational, it would still be possible for us to disagree provided only that each of us had different partial bodies of evidence. Moreover, even if we were perfectly rational, and had accessed the same full body of evidence, it might still be possible for us to disagree provided that we accessed the evidence in differing orders (and provided that our finite capacities ensured that we could not ‘store’ – or access – the full body of evidence all at once). The other source of disagreement among reasonable people is that there is no one set of ‘priors’ that any reasonable person must have. Again – though I admit that this is slightly more controversial – it seems to me that, even if we were perfectly rational, it would be possible for us to disagree simply because we have differing ‘priors’ (and this would remain true even if we accessed the same full body of evidence in the same order). Of course, in saying this, I am not committing myself to the claim that there are no substantive constraints on ‘priors’ – it may be that there are quite severe constraints on reasonable sets of ‘priors’; however, I am claiming that I can see no reason at all for thinking that there is a unique set of ‘priors’ that any reasonable person must have on pain of conviction of irrationality.
The above remarks seem to me to fit naturally into a neo-Quinean picture of the web of belief.1 At any time, a person has a network of beliefs that are connected together in various ways. Under the impact of evidence, the person will be disposed to revise his or her beliefs in various ways. If rational, then he or she will be disposed to revise his or her beliefs in accordance with the canons of belief revision (whatever those happen to be – it is not part of my present brief to elaborate any substantial account of the content of those canons). Perhaps, for any given reasonable person in any given state, there is a unique rational revision that he or she ought to make to his or her beliefs under the impact of a given piece of evidence; perhaps not. Even if there is a unique rational revision to be made (for any person in any state given any evidence), there is no reason to think that there is bound to be convergence of belief given similar (or identical) evidential inputs. (If there is no unique rational revision to be made in the envisaged circumstances, then the prospects for convergence are even dimmer. However, I am prepared to suppose that there are unique rational revisions.) I see no reason at all why it could not be that a single piece of evidence leads you to believe that p and me to believe that not p, even though we both act with perfect rationality. And even if that claim is too strong, it seems pretty clear that what one ought to come to believe under the impact of any given evidence depends upon what one already believes.
My earlier talk about ‘priors’ was meant to suggest a Bayesian conception of belief revision. However, it was also intended to be deliberately ambiguous between ‘prior probability’ and ‘prior belief’. It is a crucial part of the picture that I am sketching that all assessment of evidence takes place against an already existing background of beliefs, so that there can be no question of one’s ‘examining’ all of one’s beliefs at once. The crucial questions about reasonable belief revision thus turn out to be questions about coherence with beliefs that one already has. Moreover it seems psychologically reasonable to conjecture that there are bound to be ‘environmental’ influences on the network of beliefs that one comes to hold (one’s belief system is in part caused by one’s upbringing, etc.). However, there is no room for the thought that one might ‘make over’ one’s system of belief, throwing out those beliefs that have ‘mere causes’ and preserving those for which one has ‘reasons’ – for every belief stands or falls by coherence with the rest, and the environmental influences have an impact on the network as a whole. Of course, it could be that considerations of coherence lead you to revise families of belief – for example, belief in Santa Claus ceases to cohere with the rest of a child’s body of beliefs, and is replaced (in part) by beliefs about Santa Claus stories and the like – and that this process of revision leads one to hold that certain beliefs were merely ‘caused’ – ‘I only believed in Santa Claus because my parents inculcated the belief in me’. But this process of labelling can only proceed ex post facto: in order to judge that some of one’s beliefs are ‘merely caused’, one must already have reached a state in which one is giving them up. (There is a first-person/third-person asymmetry here: I may be perfectly well entitled, or even obliged, to judge that some of your beliefs are ‘merely caused’, particularly if they manifest a sufficiently deeply rooted disagreement between us.)
Plainly, there is much more to be said about the nature of reasonable belief. However, the last remaining thing that will be important for what follows is to note that nothing in the account that I have given is inconsistent with the idea that the aim of belief is truth, and the further idea that truth is a non-epistemic, non-relative notion. Nor, indeed, is anything in this account inconsistent with the idea that justification and warrant can be external matters, that is, matters that have more to do with how an agent is connected to the world than with how things are inside the agent’s head. All that is being insisted on is that there is an important sense in which reasonable people must be amenable to reason, where “amenability to reason” is a matter of how things are inside the agent’s head. If you are going to argue with someone – that is, to present an argument to them, or to engage in genuine argument with them – then you need to suppose that they are reasonable in this sense; but it is perfectly consistent with this assumption of rationality that you suppose that many of that person’s beliefs are not warranted or justified, let alone true, because they are not appropriately connected to how things are outside that person’s head. (Of course, it is a controversial question whether we should be externalists about warrant and justification. I think that nothing in what follows turns on how we choose to answer that question; in any case, I do not propose to try to pursue that issue here.)
1.2.2. Arguments and Their Role in Belief Revision
I take it that the proper function of arguments is to bring about reasonable belief revision: the aim of my argument for the conclusion that p is to bring you to reasonable acceptance of that conclusion. Of course, there are all kinds of other things that can be done with arguments – I may seek to dazzle you with my brilliance, or entertain you with my logical facility, or…– and there are all kinds of other ways in which I may try to bring you to (reasonable) acceptance of the conclusion that p – I may tell you a story that illustrates its truth, or show you some evidence, or….However, the crucial point is that the telos of argumentation is bound up with reasonable belief revision.
Given this much, what shall we take to be the characteristics of a good (or successful) argument? Perhaps this seems easy: a good argument is one that succeeds – or perhaps would or ought to succeed – in bringing about reasonable belief revision in reasonable targets. The most successful argument would be one that succeeds – or perhaps would or ought to succeed – in persuading any reasonable person to accept its conclusion; good, but less successful arguments would be ones that succeed – or perhaps would or ought to succeed – in persuading a non-zero percentage of reasonable people to accept their conclusions. However, as we shall now see, there are here many difficulties that lie just below the surface.
Some arguments are deductively valid (or, in some cases, mistakenly supposed to be deductively valid). In this case, what the argument establishes – or purports to establish – is that it is a logical error to accept all of the premises of the argument, and yet to reject the conclusion of the argument: no reasonable person can accept all of the premises, and yet also reject the conclusion. Other arguments – including those that appear to rely upon induction, or inference to the best explanation, and the like – are typically not supposed to be deductively valid. In this case, what is typically supposed is that the argument establishes – or, at any rate, purports to establish – that it is reasonable – perhaps even most reasonable – to accept the conclusion of the argument on the basis of the premises. (If it is most reasonable to accept the conclusion on the basis of the premises, then it seems, again, that no reasonable person can accept all of the premises and yet also reject the conclusion.)
In both kinds of arguments, there are various kinds of things that can go wrong. For instance, in both kinds of cases, one can be mistaken about the kind of support that the premises actually lend to the conclusion. As we noted before, people are far from perfect performers of deductive reasoning, statistical reasoning, probabilistic reasoning, and the like. While it is perhaps dangerous to suppose that these kinds of errors are very widespread, it seems to me that it is highly plausible to suppose that at least some of the arguments that we shall go on to examine do indeed suffer from defects of this kind. However, given the telos of argumentation, there is another kind of difficulty that arises in the assessment of the merits of an argument, concerning the acceptability of the premises of the argument. If a reasonable person need not accept all of the premises of an argument, then that argument does not give all reasonable people a reason to accept its conclusion. If a reasonable person ought not to accept all of the premises of an argument, then that argument cannot give any reasonable people a reason to accept its conclusion. Hence, in determining the merit of an otherwise acceptable argument, we always need to ask whether reasonable people may, or must, accept all of the premises of the argument. If the argument has no premises – as may be the case, for example, in reductio arguments – then this is not a substantive requirement; otherwise, it is.