The Table Video

Richard Swinburne

What Kind of Necessary Being Could God Be?

Emeritus Nolloth
 Professor of the Philosophy of the Christian Religion, University of 
Oxford
September 13, 2012

Visiting Scholar Richard Swinburne gives a lecture to the Biola University student body asking the question, “What kind of necessary being could God be?”

Transcript:

I’m a little nervous about giving this talk and the reason I’m a little nervous about giving this talk is it is a rather complicated talk. So why am I giving it? Well most of the things I would give a lecture on at Biola are things which on the whole, Biola faculty might agree with. And I thought it for a change I’d give something that they probably wouldn’t agree with, some of them, and it concerns a particular view of God’s nature with which I disagree and I suspect quite a number of the faculty who are present agree with.

And this is the view that God is a metaphysically necessary being. Those of you who were at Tim O’Connor’s lecture last night I understand that he claims that God is a metaphysically necessary being and it’s necessary to postulate that in order to have an explanation of anything. My view is that doesn’t make sense and that therefore we can’t say this about God. Now how does this issue arise? Well an issue arises because… people wanna say it can’t just be a lucky accident that there’s a God.

God can’t be saying to himself, well I’m rather lucky I exist. I might not have existed. And therefore, people want to say well, it must be a… it couldn’t in some instances couldn’t possibly be the case that there isn’t a God and how are we to spell out this couldn’t possibly? And the sort of really hard necessity that eventually strikes people is well, it’s like one and one make two. One and one couldn’t not make two.

And so, there is a God, must be like that. It must be logically necessary for that sort of reason. Otherwise there’s going to be some chance, some accident about there being a God and therefore, a lot of people for the past thousand years have wanted to say God is in this sense, something that such that in the end there would be some of contradiction in denying that there’s a God. I say to the past thousand years it never struck, these questions never struck anybody for a thousand years of Christianity and then Anselm was really the first person to want to say this sort of thing about God.

And some people did and some people didn’t say this sort of thing for the next few hundred years, but it somehow crept into the orthodoxy of Christian thought. It’s not a doctrine of any of the churches that we had to say this about God and Leibniz in particular made a lot of it. But it’s in no way necessary to be a Christian to hold this and I don’t think it makes sense and therefore I would like to see people realizing it doesn’t make sense and not backing up a view of God which is open to, obvious to my mind, objections.

So that’s why we’re on where we’re on and here it goes. And what the first part of this lecture, indeed much of this lecture is going to be about is explaining the kind of necessity which in the end I wish to claim God doesn’t have. So you’ve got a handout and you’ll need it to follow through. I should understand by metaphysical necessity the strongest kind of necessity there is and by metaphysical impossibility the strongest kind of impossibility there is and so by metaphysical possibility the weakest kind of possibility there is.

My concern in this paper is whether it’s metaphysically possible that God be a metaphysically necessary being in these senses. Now a substance or an event is metaphysically necessary or possible or impossible if and only if it’s metaphysically necessary, or possible or impossible, that it exists. And since we have no discussable knowledge of whether it is metaphysically necessary or whatever, that the substance or event exists, except in part by reflecting on the sentence which asserts this, it will be more convenient to speak of the necessity or whatever as belonging to the sentence.

So I’m going to investigate the metaphysical necessity of the sentence of English or its translation into another language, God exists. I shall come back later to the issue of whether these modal properties and necessity possibility, impossibility of modal properties belong primarily to entities of some other kind such as propositions and consider the consequences which would follow if they did.

I begin with general considerations about what determines the meanings in the sense of truth conditions of the sentences of a human language. That is what determines the conditions in which sentences are true and the conditions in which they’re false and so which other sentences they entail or are entailed by. Sentences of a language mean what it’s speakers or in the case of technical terms, some group of experts mean by them.

Each of us learns the meaning of certain sentences by being shown many observable conditions under which those sentences are regarded as true or as false and by being told of other sentences to which a speaker is regarded as committed by uttering the first sentences and other sentences which are such that someone who utters them is regarded as committed to the former sentences. We learned the meaning of a word by being taught the difference to the meaning of a sentence made by that word playing a certain role in the sentence.

By being taught the meanings of individual words and of sentences of various forms, that is subject predicate sentences, conditional sentences and so on, by being taught the meaning of individual words and of sentences of various forms, we may then come to an understanding of the meaning of a sentence in which those words are arranged in a certain way even if we have not been shown observable conditions under which that sentence is regarded as true or false.

Showing observable conditions may involve pointing to them or describing them by terms already introduced. For example, we learn the meaning of: there is a cat over there, by being shown observable circumstances under which this sentence is regarded as true and observable circumstances under which it’s regarded as false. And by being told that someone who utters this sentence is committed to: there is an animal over there, and someone who utters: there are two cats over there, is regarded as committed to the original sentence, there is a cat over there.

In these ways we come to know the meanings of cat and dog and so the kind of meaning possessed by sentences of the form, there is a so and so over there. We need to observe many different examples of observable conditions under which a sentence containing a certain word in various roles is regarded as true or false and many examples of the commitments that speakers who use sentences containing that word in various roles are regarded as having.

And this allows us to acquire an understanding of the conditions under which some new sentence containing that word would be regarded as true or false. Examples of different observable conditions under which some sentence is true or false and of sentences to which we are not committed by a given sentence, also illustrate which conditions do not rule out the sentence being true.

We extrapolate, that is, from a stock of supposedly paradigm examples of observable conditions and relations of commitment to an understanding that the sentence would be regarded as true or false as the case may be under conditions sufficiently similar in certain respects to most of the paradigm examples. These points may seem obvious. You learn particular words by being pointed to examples. You learn what it means to say there is a so and so over there by being given particular examples and so you learn what the form of the sentence means and then you learned to slot in a new so and so into it.

Because humans have very similar cognitive mechanisms determining how they learn meanings and because members of a language group are exposed to very similar paradigm examples of observable conditions and rules of commitment, members of the same language group normally acquire a very similar understanding of the meaning of words and sentences. We all go through similar processes, one would expect us to as it were, extrapolate from them in the same way to get the same sense of what a word meant if we are presented with many similar examples.

This common understanding may be reinforced by dictionary compilers and philosophers who tidy up language by laying down rules for correct usage usually by codifying most people’s actual usage. The rules give general descriptions of the observable conditions under which various sentences of the language are true and of observable conditions under which various sentences of the language are false and of the kinds of other sentences to which a sentence of the given kind commits the speaker and by which sentences of other kinds a speaker is committed to a sentence of the former kind.

We get such rules, for example, as the rules of the syllogism. That all A’s are B and all B’s are C commits us to all A’s are C, but we can only understand such rules by means of further examples of observable conditions and kinds of sentences. One couldn’t understand the rule of the syllogism, which I’ve just quoted, without being given an example of what it is for an A to be B. All humans are mortal, Socrates is immortal, and we begin to understand what the rule means but only by being given examples and the examples in the end depend on things you can point to. This program of tidying up language aims to secure uniformity of use to the extent to which is successful in a language group there is a correct use of language and it is an objective matter to what one is and what one is not committed to by some sentences.

Words and sentence forms may however be ambiguous and new words and sentence forms enter language, but I shall count the language as having a correct use so long as speakers can be got to recognize the ambiguity or novel meaning. This can often be achieved by philosophical discussion forcing some speaker to admit that in one sense of a word, S is W is true, whereas in another sense, S is W is false. And because of a lack of sensory or cognitive apparatus some speakers do not have the capacity to extrapolate from any paradigm examples or inferential rules to the applicability of sentences in new situations.

Some people are colorblind and so unable to understand the sense of green to which they have been introduced by examples of green objects and so apply it to new instances. Other people do not have the cognitive apparatus to recognize some philosophical or mathematical concept such as tensor project or internal negation to which they have been introduced and so apply it to new instances.

But so long as those who purport to be able to extrapolate from paradigm examples can be got to agree as to how to do it, I shall count the expressions as having an objective meaning in the language and I shall call the assumption that all sentences of a language would have, in consequence of these procedures an objective meaning, the common language assumption. Assuming that these procedures give us a common language, a common way of understanding the words to which we have been introduced by these processes. And among these rules are rules about one is committed to by a given sentence and I shall call that a rule of mini-entailment. I think it’s on your handout.

A sentence S1 mini-entails a sentence S2 if and only if anyone who asserts S1 is thereby in virtue of the rules for the correct use of language committed to S2. S1 entails as the first dominion entails, entails the sentence Sn if and only if S1 and Sn can be joined by a chain of mini-entailments, such that S1 mini-entails some S2, S2 mini-entails some S3, and so on, until we reach a sentence which mini-entails Sn.

And I shall call a rule for what one is objectively not committed to by a sentence or its negation a compatibility rule. S1 is compatible with S2 if and only if S1 does not entail not S2. If a sentence S1 is compatible with S2, it is of course compatible with all the entailments of S2. Some of this may seem pretty obvious to start with, but just hold on, there’s some unobvious things will start to emerge. All right.

So we build up a language, later on correct rules, and among these rules are what you’re committed to by a given sentence and we get the notion of an entailment. Some sentence eventually entails, that is someone who says it, that sentence is such that if it’s true some other sentence must be true, but it may need a lot of processes of inference from S1 to another to another to get to the other sentence. Now so much general points about language.

Now I’ve used this word metaphysical necessity. Among metaphysic sentences that… if anything is are metaphysically necessary or metaphysically impossible or metaphysically possible. Some metaphysical necessities or whatever are ones discoverable a priori, that is discoverable by mere reflection on what is involved in the claim made by the sentence. That is in it’s entailments. I’ll call these necessities that are discoverable by mere a priori reasoning logical necessities. They include both logical necessities in a narrow sense, the sort of necessities that could be captured by a system of predicate calculus or some system that logicians have got hold of. And conceptual necessities like all squares have four sides would be logically necessary and also of course, the rules of the syllogism.
If A’s are B and all B’s are C then all is As are C. So these are logical necessities. The obvious examples by which we learn the meaning of logically impossible sentence, are self contradictory sentences and ones which entail self-contradictions. A self contradictory sentence claims both that something is so and also that it is not so. For example, he is taller than six foot and it’s not the case that he is taller than six foot. For such a sentence could only be true if that something was so and the sentence asserts that it’s not so.

No sentence could be more obviously or more strongly impossible than such a sentence and any sentence which entails a self-contradiction is as strongly impossible as a self-contradiction. And the natural understanding which most of us get from these examples is that a logically impossible sentence just is one which entails a self-contradiction and so any logically necessary sentence is one whose negation, that is the sentence that says the former one is not true, whose negation entails the self-contradiction and any logically possible sentence is one which doesn’t entail a self-contradiction.

Okay, I’m starting to spell out what this deep kind of necessity or impossibility and so on involves. And my first class of those sentences is the logically necessary impossible or possible and what makes it logically necessary or whatever is that it’s logical status can be determined by mere sitting down and thinking about it, a priori that is. So here is a suggestion of what it consists in, a logically impossible sentence is such that you can in the end entail a contradiction from it, a logically necessary one is such that in the end a contradiction follows from denying it. A logically possible one is in the middle.

Neither it, logically possible one is one which is such that it’s negation. There is such that it doesn’t entail a contradiction. Purported examples, somebody might say well yes, of course a logically impossible sentences, sentences which entail a contradiction are logically impossible, but perhaps there are logically impossible sentences of other sorts which are logically impossible because you can determine the impossibility a priori, but which don’t entail a contradiction. And likewise, logically necessary sentences whose negation doesn’t entail a contradiction, but I don’t think that there are such sentences.

I think all logically necessary sentences are ones whose negation entails a contradiction. Purported examples of logically necessary sentences whose negation do not entail a contradiction, turn out, I suggest, on examination, either to be such that their negations do entail a contradiction or not to be nearly as strongly necessary as ones whose negations do entail a self-contradiction. And there is a general reason for denying that there are any logically impossible sentences other than ones which entail a self-contradiction.

And that is this, that any sentence whatever must have the form of a declarative sentence in which the component words already have a sense in the language. It will be, for example, a subject predicate sentence, an existential generalization, or some other one of many recognized forms of declarative sentence. It will, to put the point loosely, assert about some substance or property or event or whatever that it does or does not have some property or relation to some other substance property and so on, or that there are not certain substances, properties or whatever. Any sentence must be a certain general form and it will contain words and the words have a sense insofar as it’s clear what is the criteria for an object property or whatever to be that object property or whatever.

And therefore, as it were, sentences have a certain form and the words in them only have a meaning if there are rules for what counts as an instance of that thing or what doesn’t count as an instance of that thing. And so it will be inconsistent to affirm that an object picked out by some expression is of a kind ruled out by the very criteria for being that object. And the form of a sentence S1 will exclude some alternative S2 and so it will be inconsistent to affirm S1 and S2. That is the very use of any word: chair, table, planet, means that it can only apply to certain sorts of things and not do other sorts of things.

Red, green, square, we’ve got rules for use of these words such that only certain kinds of things count as red or green or square. And you can’t have a prime number which is red, you can’t have inflation which you square, the very words themselves, there are rules for the sorts of things they can apply to and the sorts of things they can’t. And likewise for sentences, this A is a B there, are rules for what can count as a case of that, it could only apply to something which has a certain property being an A and it can only apply to us in virtue of being B or not B. And it therefore follows, the very form of a sentence excludes some alternative and the very use of words exclude certain alternatives to it therefore follows that sentences exemplifying what used to be called category mistakes, doesn’t matter if you haven’t heard of this, such as Caesar is a prime number or this memory is violet, are in my sense logically impossible sentences.

They’re logically impossible because if you look rather carefully at them, assuming Caesar is used as a name for a person and a prime number is a number and to say that something is Caesar is to say that it’s a person, a substance, and to say that it’s a prime number is to say that it’s a number and number is an abstract object and substance is a non-abstract object and to say the object is both abstract and not abstract is a contradiction. So in the end even if it doesn’t look as if there’s any very obvious contradiction in some sentence which obviously is impossible, if you look at the rules for the use of the words in it and use the form in it, it turns out to be, it too entails a contradiction.

Any rate, I’m going to act on this assumption that if a sentence is logically necessary, it is such that it’s negation entails the contradiction. If it’s logically impossible, it must be such that it entails a contradiction. If it’s logically possible, it must be such that it doesn’t entail a contradiction. And given the common language assumption that we all know what we mean by the same words, we should be able in that case within a finite time to agree about many sentences that they entail self-contradictions and are so logically impossible.

And about many sentences such that their negations entail self contradictions and so are logically necessary. Compatibility rules also allow us to recognize many logically possible sentences and so since any sentence entailed by a logically possible sentence is itself logically possible, we should be able to recognize many more logically possible sentences. So once we’ve got hold of the rules for the use of sentences we should be able to see what they involve and to sort out what are the entailments of sentences.

Of course a philosophical discussion often begins with disagreement about what entails what or what is compatible with what. And the way to resolve a disagreement about whether a sentence P entails the sentence Q is to find a root of many entailments from P to Q. And if somebody says: all men are mortal doesn’t entail there are any men, and somebody says: all men are mortal does entail there are men, and how do we resolve the disagreement? Well you resolve or somebody may say there are two objects which erect two substances which are in no distance in no direction from each other. And somebody may say that is logically impossible and somebody else may say it’s logically possible.

How do we resolve the disagreement? Well we resolve the disagreement about whether P entails Q by finding a root of many entailments from P to Q, that is to say, we find P entails S1 and S1 in fact entails S2, so on till you get to Q. And so the way to resolve disagreements about whether P is logically possible is to find some sentence, normally a long conjunction describing some circumstance, which disputants agree to be logically possible which entails Q or to find some self-contradiction entailed by P. That is philosophers disagree about whether such and such state of affairs is logically possible or impossible or necessary and the way to resolve this is to, for the person who says this sort of thing is logically impossible is to find a route from the sentence by a number of steps of many entailments to a final entailment which is a contradiction.

And the way to show that it’s logically necessary is to show that it’s negation leads to such a route, to a contradiction. The way to show that it’s logically possible is a little more complicated, but it in fact it’s this. You get some sort of conjunction of sentences which is more obviously logically possible and show that, that does entail the sentence you’re interested in. So there are ways of resolving disagreements about whether something is logically possible or logically impossible or logicless, their ought to be if our language is common, we share it, we know the rules for the words, we’ve been taught them, we ought to be able to settle these issues straight away. Why not? And of course philosophers haven’t been settling these issues for many centuries.

And so what’s happened? Well prolonged failure to resolve disagreements about which sentences entail which other sentences or which are logically possible is evidence of a failure in the common language assumption and the failure would mean that the examples and rules by which a word or sentence form has been given a sense has led to different un-shareable concepts, different in communicable understandings of that word by the disputants and this may happen either because disputants differ in their sensory life or in their cognitive abilities. Some of us may simply lack the ability to recognize the influence that somebody’s pointing out or alternatively we may have a different sensory apparatus so although we’ve been taught the meaning of these words by the same examples, the example are a lot different to us, and they’ve given us a different understanding of the word.

But alternatively of course, we may just not be very clever. But where this doesn’t happen there ought to be agreement about what’s logically impossible and so on, given the common language assumption. Thus to modify an example used by Kripke to illustrate this class of sentences. Suppose that in the days long before people knew the geography of the Himalayas, explorers named a mountain of a certain visual shape seen from Tibet “Everest.” And a mountain of a certain different shape seen from Nepal “Gaurisanker.”

And they use these names as rigid designators of the mountains. A rigid designator is a word which picks out the same object, however the object may change in respect of its non-essential properties. These mountains Everest and Gaurisanker are in fact the same mountain and being the same mountain they are by the necessity of identity necessarily the same mountain. And so it seems Everest is Gaurisanker is necessarily true with as hard of necessity as any logically necessary sentence. However, we may suppose the explorers did not know that.

That’s to say they just saw a mountain with a certain shape in the distance and their own from Tibet and when they went to Nepal they saw a mountain with a certain shape in the distance but they didn’t know it was the same one, but it was the same one, and clearly if it was the same one it couldn’t not be the same one. If something is the same it couldn’t not be the same, and so Everest is Gaurisanker is necessarily true with as hard as necessity as any logically necessary sentence. However, we may suppose the explorers did not know this and clearly they would not have been able to discover its truth by mere a priori means. Hence it is not a logically necessary truth. Or consider Putnam’s example of water is H2O.

Water being understood, as Putnam supposes that it was in the early 19th century, as a rigid designator of the transparent drinkable stuff in our rivers and seas. What makes the stuff that stuff is its chemical essence, being H2O. Having that essence it could not not have that essence so water is H2O is metaphysically necessary but again, not so discoverable a priori. Hence it must be an a posteriori metaphysical necessity.

What has made these necessary sentences a posteriori is that the sentence contains at least one rigid designator of which we learn the meaning by being told that it applies to certain particular things, especially substances, mountains, and kinds of substances, water, having certain superficial properties, but where we are told what makes a thing that thing, what makes the thing that substance or substance of that kind is the essence of which we may be ignorant underlying these properties. In ignorance of the latter, we do not fully understand what we are saying about a substance when we say that it’s that substance or a substance of that kind.

Hence I shall call such designators uninformative designators. I define a rigid designator of a thing as an informative designator informative, if and only of someone who knows what the designator means, that is has the linguistic knowledge of how to use it, knows a certain set of conditions necessary and sufficient for a thing to be that thing whether or not he can state those conditions in words. And a rigid designator is uninformative if and only if those conditions are not satisfied.

To know these conditions for the application of a designator, as I shall understand this expression, just is to be able, when favorably positioned with faculties in working order and not subject to illusion, to recognize where the designator, or if it’s defined in words, the words by which is defined, applies and where it does not and to know the mint-entailments of the sentences in which it occurs. Having the ability to recognize something when favorably positioned with faculties in working order and not subject to illusion, involves knowing what that thing is. In the case of technical terms, it’s experts whose knowledge is relevant.

Many of the words, for example, red, square, or has a length of one meter, by which we pick out properties are informative designators. They are such that if we know what the words mean, we can recognize subject to the stated restriction where they do or do not apply and can make the requisite inferences. Other words by which we pick out properties can be defined by words for which those conditions hold.

For example, has a length of 10 minus 15 meters can be defined in terms of the informatively designated property has a length of 10 meters and the informatively designated relation being shorter by 1/10 than, used 15 times. So the reason why the claims originally made by the sentences “Everest is Gaurisanker” and “water is H2O” are necessary with as hard as necessity as logical necessity is that they are logically necessary, but the use of uninformative designators has the consequences that speakers did not know fully what these claims were until they had done some a posteriori investigation.

When we know fully what we are talking about, mere a priori considerations can show whether some sentence is metaphysically necessary or impossible. Hence there is available a more general definition of a sentence being metaphysically necessary, impossible or possible, if and only if it’s logically necessary impossible or possible, when we substitute co-referring informative designators for uninformative designators. That’s water is H2O is metaphysically necessary, although it’s not logically necessary.

You can’t work it out by pure reason, but if you find out what the essence of the thing you’re referring to by water is, and the essence is H2O, it reduces to H2O is H2O and you can know by pure reason that that’s true. And Everest is Gaurisanker, if you know that Everest in fact refers to a certain chunk of rocky matter and that Gaurisanker refers to the same chunk of rocky matter and you call the chunk rock then you can know that rock is rock and you can therefore, you can substitute that word for each of the Everest and Gaurisanker and so know that it’s true.

So it would seem that no sentence could be as strongly impossible metaphysically as one which is in reality logically impossible, and so there can be no metaphysically impossible necessary or possible sentences apart from the ones of the kind analyzed in this section. The kinds of examples which Putnam and Kripke gave suggests that there are necessities other than these logical necessities. I’ve wanted to suggest to you that when we, the reason why this is so is we don’t fully know what we’re talking about when we refer to something as Everest or something as water because we define the terms as applying to something if they have the essence underlying the visible in ignorance of what that essence is, we don’t know fully what we’re talking about. But when we found out what the dissonance is, then we do what we’re talking about.

So a metaphysically necessary sentence is a sentence which is logically necessary when we substitute informative for uninformative designators and similarly for possible and necessary. So now I’ve introduced this term metaphysically impossible as the strongest kind of impossibility which a sentence could have and I’ve defined it in terms of logically impossible. So we’ve got the examples by which philosophers introduced this kind of a posterior metaphysical necessity, can I suggest be reduced to logical necessity and I can’t see that any recent philosophical discussions have suggested that there are any other plausible examples of something which is as strongly necessary as either something that’s logically necessary or something that can be reduced to it in the way that I’ve suggested and ditto for possibility and impossibility.

With one exception and the one exception is that of course a lot of philosophers who do philosophy of religion want to say there is a God is metaphysically necessary and this is a suggestion which is put forward and the question is would it count? What reason what we have to suppose that it’s metaphysically necessary if my definitions of the matter hold? And they’re going to seem to have to be forced to suggest there’s a special sort or special kind or wider sort of metaphysical necessity which would capture it.

At any rate, let’s see if it could be fitted in, see if we could get it to satisfy this definition of metaphysical necessity. Given my understanding of these concepts, together with the assumption that modal properties are properties of sentences, and given the common language assumption that we all mean the same by the same words, it seems fairly implausible to suppose that a positive existential sentence, that is the sentence which says there is a so-and-so or there isn’t a thing of this kind, seems fairly implausible to suppose that a positive existential sentence can be logically necessary.

For to be so it would need to be such that it’s negation entails a self-contradiction. The negation of an existential sentence, an existential sentence is there is a set of something which has the properties so and so. The negation says there is no object which has the property so and so. It claims that a certain property or conjunction of properties is not instantiated. A self-contradictory sentence claims that the actual word has a contradictory quality and that will be so either because some object within it has such a quality or because it both does and does not contain an object of a certain kind.

That is a self contradictory sentence will either be there is an X such that X is phi and there is not an X such that X is phi. Or it will be there is an object X such that X is phi and X is not phi. It doesn’t look very plausible to suppose that a sentence which says there is a so and so, whatever the so and so may be, could be logically necessary because if it was then it’s negation, there is no object which is so so, would have to entail a contradiction. But how can the mere denial of the existence of anything be contradictory? You can just merely saying there is no so-and-so. How can that can’t contradict anything? One can simply say the world doesn’t contain an item of a certain sort. That would would seem… contradictions arise from saying something both is of a certain sort and isn’t of a certain sort.

But merely to just say there isn’t an object of a certain sort, doesn’t seem to entail any contradiction. It just says the world is empty of a certain kind of thing. Plausibly the mere non-existence of anything of some kind cannot entail the existence of anything. That is to say the sentence which says there is no object of a certain kind, will have typically the form, if it did entail a self-contradiction it would say there is a phi and there isn’t a phi. That’ll be a self-contradiction, but how can the mere non-existence of something entail the existence of something? Or it might have the alternative form.

There is something which is phi and at the same time there isn’t something that’s phi, but again it has an existential claim. There is something that’s phi and how can the mere non-existence of something entail the existence of something? At least it doesn’t look as if that is in the least possible. This plausible suggestion that the mere non-existence of anything of some kind cannot entail a contradiction and so no positive existential sentence can be a necessary truth is of course due to Hume. It will hold whether that the thing is of a concrete or abstract kind. So the supposed necessary existential truths of arithmetic do not constitute as an exception. It might be said look, there are exceptions like this.

There is a prime number greater than three. Here is a an existential claim, normally supposed to be a necessary claim, and yet I am denying that there can be any necessary existential truths. But there is a prime number greater than three does not entail a contradiction by itself, it only entails a contradiction when joined with the axioms of arithmetic. And two of those axioms, one which arithmetic is based, are themselves existential claims. Naught is a number, for each number there is a successor. That is to say there is a prime number greater than three entails nothing unless you already join it with existential claims and the contradiction of the negation involves joining it with an existential claim, but by itself the mere denial does not entail any contradiction.

However of course, as we know, and that’s what the lecture is about some people claim that a particular negative existential sentence, there is no God or there is no omnipotent, omniscient being, does entail a contradiction and some of them claim to have demonstrated this. If they say that this is metaphysically necessary then the claim must be that it’s logically necessary when you fill in something about the nature of God and although we may not know very much about the nature of God, clearly we do wish to say that he’s omnipotent, omniscient and so on being, he might go a lot else as well, but we want to say this.

But in that case if there is a God is necessary there is no God must entail a contradiction and if that’s the case then the description of God including those predicates must entail a contradiction. There is no being or such, there is no being who is omnipotent, omniscience and whatever the divine properties there are would have to entail a contradiction. But if it’s just a negative existential claim, my suggestion is it’s difficult to see how it contains a contradiction. It merely says there isn’t a being of a certain kind. Well of course some people hold that nevertheless this is the one exception.

That this there is God is the one exception, that it’s negation does entail a contradiction. But if you say that then you learnt in this difficulty. There are a lot of other sentences which it rather looks as if they are logically possible, by any normal standards they seem logically possible, and they may too must entail contradictions. For example, once upon a time there were no rational beings or no one knows everything or no one is perfectly good or among sentences evidently describing non-actual worlds “the only substances are for mutually repelling steel balls” or “no one knows what is happening outside the sphere “of one mile diameter surrounding their body.” All of these which we might as it were, given our philosophy class as examples of logically possible sentences or so on.

They must all entail contradictions if there is a God entails contradictions because sorry, if there is no God entails contradiction because there is a God does entail each of these sentences. If there is a God then that entails, sorry, if there is no God then that entails… Sorry. If someone holds that there is a God is logically necessary and therefore there is no God entails a contradiction, they will also have to hold that once upon a time there were no rational beings also entails a contradiction because if there is no God is necessarily true then it’s necessarily false that once upon a time there were no rational beings because there is a God entails there was always a rational being and therefore since it’s entail by, since there is a God is supposed to be logically necessary then once upon a time there were no rational beings must in fact be logically impossible and it must entail a contradiction.

But I can give numerable examples of other sentences as it were which have this property. There is a God entails all sorts of sentences of this kind and they are sentences which in any other context the normal rules for understanding the meaning of sentences would tell you that these are logical possibilities, could be true, could be false, but it turns out that if there is a God is logically necessary all of these are going to be logically impossible. And that seems very strange. Someone could only derive a contradiction from these sentences if they understood an enormous number of predicates, not just one or two technical philosophical terms in different senses from the rest of us. That is to say if you really want to say no one knows everything is really logically impossible then it will have to be the case that that entails a contradiction.

And if you want to say no one is perfectly good is really logically impossible then you’ll have to say that entails a contradiction. And you try and get it, it doesn’t look as if these would generate a contradiction and no one is to my mind ever shown they would. And I think you’d have to understand the words in very different senses from what the rest of us would have to understand. Or you’d have to understand the formal terms like exist or not or all in different senses if you’re to get a contradiction out of these. But in view of the similarity of the respective cognitive mechanisms and the process of language acquisition between humans who believe in ontological arguments and those who don’t, I don’t find it very plausible to suppose that the former understand all these predicates in different senses from the rest of us.

But if they do understand the predicates in the same senses as the rest of us, they’ll have to admit that there is no God does not entail a contradiction and so there is a God cannot be logically necessary in the sense in which I have spelt it out. But if there is no logically necessary sentence of the form there is an X such that phi X, where phi designates the definite description of God, no substitution for God in God exists of a co-referring designator will yield a logically necessary sentence.

And so God exists cannot be metaphysically necessary in the sense in which I have spelt it out. Let’s just see if you’ve got the last state of this argument. Suggestion is that there is a God is metaphysically necessary. Is it logically necessary? If it is logically necessary, well then it’s negation, there is no God, is going to be logically impossible. If it is logically possible then all sorts of other things are going to be logically impossible, which don’t look as if they are logically impossible at all. If there is a God is logically necessary, then it must be logically impossible that once upon a time there were no rational beings and no one knows everything and no one is perfectly good because if it’s necessary there’s a God then it’s necessarily impossible that no one is perfectly good because it’s necessary that one person is perfectly good, it is God. Okay. So it rather looks as if there can’t be and he logically necessary sentence of the form there is an X such that phi X where phi describes a definite description of God.

That is to say whatever of the typical divine predicates you put in to substitute for God, there is no God, God might be a proper name defining a thing, you don’t know altogether what it is, but you know something about it and so it clearly refers to a being who has certain properties omnipotent, omniscient, and so on and a lot more, so if there is a God is to be logically necessary, it is to be metaphysically necessary, it must be logically necessary when you substitute for God, that sort of description. If that is logically necessary then it must be logically impossible, there is no being which has these such properties must be that that entails the contradiction.

But how can stage one, how can a mere denial of the existence of something entail a contradiction? Especially as contradictions themselves seem to say there is a so and so and there isn’t a so and so. In other words, they do involve an existential claim. But in any case apart from the general point, the point is that if you say there is a God is logically necessary then you have to say, well a lot of other things which don’t look logically impossible really are logically impossible. If it’s necessary there’s a God then it’s necessary that there is a rational being and in that case it’s impossible that once upon a time there were no rational beings and so on generally in innumerable examples.

If we’re using words in the same normal sense, well some of us might have got a bit of a different understanding of some of these words, but it’s very implausible to suppose that all of us could have got a totally different understanding of all of these words. They’re part of our common language, we know what they mean. And that being the case, if these these sentences do entail contradictions we ought to be able to derive them, but of course we can’t. No normal rule of inference, no normal understanding of what these words mean, and what we’re committed to by them will give you a contradiction. So I find it pretty implausible to suppose the sentence there is a God does entail a contradiction.

Now I think I’m going temporarily to stop at that point. I do have a little more to say, but what I’ve been assuming is that what we are interested in is the modal status, that is to say as necessary and possible or whatever, of the sentence that there is a God. And I’ve been arguing that that can’t be a metaphysically necessary sentence. If you suppose that the necessity might be a posterior, It might be a matter that God picks out something with a certain essence then of course we don’t know fully what their essence is, but we know enough about it to say that it involves… there is no being as omnipotent, omniscient, and so on has such and such properties. So if there is a God is to be metaphysically necessary, it must be logically necessary that there is no being with such and such properties.

If that is to be logically necessary, it must be sorry, there is a being with such and such properties must be logically necessary. If that’s to be logically necessary, must be logically impossible that there isn’t a being with such and such properties. In other words, that must entail a contradiction. But if that did entail a contradiction, all sorts of other sentences which don’t look as if they’re anything to do with God would also have to entail a contradiction and quite pretty fairly obviously they don’t.

And for that reason I think there is a God can’t be metaphysically necessary, but I do have a bit further to say which will arise if we want to say that well, it’s not a property, this this logical necessity isn’t a property of sentences, it’s a property of propositions. But as the clock is moving on I think I’m going to stop at that point for you to raise objections to what I’ve said so far and if you’re short of those I’ve got plenty more to say. [audience applauding]

Man: You say there is no god doesn’t entail contradictions?

That is the the claim of those who say that it’s logically necessary, I say it isn’t logically necessary. [man speaking off mic]

Couldn’t that be a string of?

Man: Some of the evidence is for the existence of God.

Yeah, what about it? Yes. I believe there’s a God, what I’m trying to argue is that we mustn’t burden him with this property of being metaphysically necessary. The design argument certainly and I am going on to add the point in due course, God is certainly going to be in some sense a necessary being and I think there is plenty of evidence for his existence. But I don’t think he’s metaphysically necessary in this sense whereas a number of, a quite a lot of modern philosophers of religion do think he’s metaphysically necessary in this sense and that’s what I wish to object you because I think they’re landing theism with an unnecessary burden. [man speaking off mic] Al Plantinga does have an ontological argument, an ontological argument is an argument for God being a metaphysically necessary being. I think that argument fails.

For the reasons I’ve just been giving, but that doesn’t mean to say that there aren’t plenty of other good arguments for the existence of God as an omnipotent, omniscient, perfectly good being who is necessary in a causal sense that that he is the ultimate explanation of all things and all things depend on him, but he can’t be necessary in this sense. That’s my argument. [man speaking off mic] Yes, I haven’t got on to that point I’m afraid yet, I’m only about two-thirds of the way through the notes and where is my sheet of the notes?

Sorry, just a minute. As far as your handout is concerned, I’ve got it, as far as that handout is concerned I’ve dealt with the first two pages, so don’t raise any questions on the last page. I’ll come to that in due course. [man speaking off mic] Wait a minute, wait a minute, yes. [man speaking off mic] I too agree that there is a perfectly good being, but the issue is whether it’s logically necessary that there is a perfectly good being and it doesn’t look as if the sentence “no one is perfectly good” entails a contradiction.

You might think “there is no God” entails a contradiction, but my argument suggests that are innumerable other sentences that you would have to hold, entail contradictions. Which as it were, just look like ordinary sentences you might find in a logical textbook to illustrate logical possibility. And we’ve all learned that these words good and so on, that no one is all of these words. We’ve had the same instruction in what these words mean so one would expect in general, us to have the same understanding of what sentences containing them mean. If that’s right… if no one can produce a contradiction from that then it rather looks as if there isn’t a contradiction being produced from it.

And we might disagree about God, but suppose that these rather ordinary life sentences, which just contain words like no and is and so on, suggest that they all entail contradictions suggests that we don’t quite understand. The only way one could think that some group of people might hold that is if they had… understood a lot of these words, including the words like is and no in different senses from the rest of us. That seems very implausible. We’ve all learned these words in the same way, although we might have learned the word God in a different way from anyone else, it’s not very plausible to suppose that we’ve learned good and manners and rational 10 years ago and so on in different ways.

All right well I will in that case just move on to a little further, it’s only a few more pages, speculation. Now there are some hard lined defenders of divine metaphysical necessity around. And what they are not likely to like about what I have said is that it starts from the supposition that logical necessity is a property of sentences. They would say it’s a property of propositions. Now what’s the difference? Well sentences of different languages can say the same thing, can mean the same thing. Rex mortuus est, the king is dead, Le Roi Est Mort, these all mean the same thing.

So it’s natural to, in virtue of them meaning the same thing that there are certain similarities in what they entail and what is entailed by them. So it’s natural to say well, there’s a proposition which they express and this isn’t confined to any particular language. This is a timeless entity which as is it were, these sentences of the English language express. Well that seems a harmless way of going about things and I am happy with the idea that we can can talk about propositions. But this is just a useful fiction, that’s to say there really aren’t such things as propositions around, but it’s useful to talk as if there were.

Why do I say there are not such things as propositions around? Well the things of which this world contain or the normal things we suppose this world contains are things that we interact with. Propositions are supposed to be timeless entities, which exist eternally, individual sentences might express them. We don’t need to postulate propositions in order to explain what happens with sentences, we can explain why humans utter the sentences they do without bringing in this notion of proposition. We don’t interact with propositions. They don’t cause things, we don’t cause them, they exist eternally. They’re just a useful way of talking about the properties of summarizing the properties of sentences. If they were these timeless things, then the picture which is given by the proponent of them is that they exist eternally, there is an eternal proposition of the form two plus two is four, which is eternally true.

There’s another proposition of the form all squares are round, which is eternally false and they’re all up there in the heavens and Plato would have been happy with that, but human sentences just try and imitate them as it were. And necessity belongs to these timeless things or impossibility belongs to these timeless things and therefore all that I’ve shown about sentences is that the real truths might be necessary or whatever, but our human sentences don’t quite capture them.

If that was true and all their real relations of logic and a necessity belong to propositions and it’s their entailments that are the true entailments as it were and human sentences just imperfectly catch of that, why should we believe anything about, how are we in a position to know which propositions entail which other propositions, if they don’t have much to do with human sentences? What gives us this intuition into the platonic realm? Why I’m able to tell you about what entails what is because I’m dealing with human sentences and why I know what they entail is because I’ve been taught the meaning of words in the same way as you have and therefore because of that it might be expected that I might know what a sentence entails or not.

I can know what a sentence entails, but these timeless entities are, why should I know what they entail? And I think that all talk about propositions which has any significant meaning can be analyzed in terms of human sentences. That is to say it’s quite true that. I mean you might want to say the proposition “All men are mortal” and the proposition “Socrates is a man” entails the proposition “Socrates is mortal” but I suggest that all that is meant by that is these sentences have the relation of entailment described and also any other sentences of another language which have the same entailments. That exhausts the meaning of the statement about the propositions, what else is there? And in that case we’re back to ordinary language.

But now suppose we did go along with this way of thinking. There are propositions, real timeless entities which have a modal status quite independently of any human sentences which might express them. In that case there might be some plausibility in saying there is a God might be a metaphysically necessary truth because it might be the case that the proposition “There is no God” does entail a contradiction only it does so by means of all sorts of other propositions that we don’t have any access to because they are quite inexpressible in human language. It’s wild, but that’s the way the argument would go.

But suppose you did go along with them, then there are other problems. And other problems which are unwelcome consequences for theism. And they are that… there is a God is one of these necessary propositions which we can’t show by ordinary language to be necessary but is necessary because it’s a proposition expressed by human words and it’s necessary because “There is no God” in fact entails by means of propositions no human is able to express eventually a contradiction. If that were the case, well it’s not the only necessary proposition. There are lots of other necessary propositions like there are no round squares or if all A’S are B and all B’s are C then all A’S are C. Or two plus two equals four, lots and lots and lots of other necessary propositions. So what’s the relation of these necessary propositions, all these other ones to the proposition, there is a God?

And among the necessarily true propositions is no one can make me exist or not exist at the same time at least if you care to count anything as a proposition. I exist and don’t exist at the same time does look like a necessarily false proposition and so no one can bring it about would seem to be a necessarily true proposition. So there would seem to be a lot of other… These other necessary truths which are really real propositions at a timeless heaven, which could constrain the way God can act. There is God up there and he exists necessarily, but there’s all sorts of other necessary propositions over which God has no control or at least so it would seem and all sorts of things that God can’t do.

Now if all we’re dealing with is sentences, the reason why God can’t make me exist and not exist at the same time is not any weakness on the part of God, but simply the sentence ultimately doesn’t itself make any sense. The weakness is in the sentence. I put together human words in such a way that nothing would count as satisfying me existing or not existing at the same time. But if there are these real timeless propositions and these relations between what can be and what can’t be aren’t just relations about human sentences, but are timeless realities in a platonic world then they would seem to constrain what God can do, not merely what I can do with a human sentence, but what God can in fact do. And that’s beginning to look a bit difficult for theism and the way out of this for those who believe in propositions is to say well, these other propositions are unto themselves made necessary by God.

They are part of God’s nature that necessarily he brings it about the two and two make four rather than two and two make five, those our ideas in his mind. God exist is the supreme necessary proposition and other propositions are necessary in virtue of being brought about by God in this way. But then… you have to ask are they all part of God’s nature if they exist… if you want to say they are not independent of God, they’re all part of God’s nature. But then this is going to make God a remarkably complex thing. God has all these propositions in him.

God is traditionally supposed to be simple. God is good, perfectly good, perfectly knowledgeable, totally powerful, everlasting, you can get all the other divine properties from that very simple notion. But now we’ve got all these propositions built into God and that makes him a very complicated being and therefore since the simpler your explanation of the world, the more likely it is to be true, you are postulating a very complicated being responsible for the world therefore less likely to be true. So you might say well no, they’re not exactly ideas actually in the mind of God, but God has a nature such that necessarily he creates two plus two equals four and he creates seven plus five equals 12 and he creates nothing can exist and not exist at the same time.

Yes, well you might say yes, he does have these nature, but why should he be like that? Why should it be that God has a nature such that he makes it necessarily true that two plus two equal four rather then making it necessarily true that two plus two equals five? What’s specially good about the former ledger? And you’d have to say well, God always does what is good and it’s better that the world should be such that two plus two equals four rather than two and two equals five.

But if that were true, it could only be true in virtue of the fact that there is some other truth that a world with two plus two equals four is better than the world with two plus two equal five and that would be independent of God’s will and God’s action as it were. God would bring about such propositions as it’s good to bring about, but if there’s to be any meaning in saying God does what is good then it can’t be just God does what God does. It must be God does it because it has certain non moral properties. And so it must be there are awful lot of truths independent of God that two plus two equal four is good thing for a universe to have and two plus two equals five is a bad thing for a universe to have.

Now you’re thinking aren’t getting a bit wild and of course I am, that is what I’m trying to bring out. If you think that despite my simple operation with sentences and I want to say well if metaphysical necessity and so on are properties of sentences then it’s going to be extremely odd to say that an existential proposition there is a God is going to be metaphysically necessary. But as I say those who defend this notion are very keen on thinking of propositions and indeed other abstract things like properties and so on as being independent of the world and existing in a platonic realm.

And of course if they also believe in God, they must think the other things are dependent on God and so they’ve got to fill up the world with all these propositions dependent on God. So the question arises just what makes it the case that God has this very complicated nature? One might say well he has the best nature he can have, but what makes this the best nature he can have? Well because there’s some other proposition up there that says this is good and that is bad and generally you get into an awful mess if you start thinking along those lines.

I think you’ve no reason to suppose there are propositions in the first place, let alone, but if you think of God’s necessity as a matter of necessity in necessary relation between propositions which cannot be captured by human sentences, you’ve no reason to suppose that there is a God is or one of those propositions and if you confine metaphysical necessity to sentences then there is the sort of argument that I have given against it. If you put it in propositions not merely have you got no reason to suppose there are such that there’s a God is such, but if you do suppose that, it’s very difficult to give an account of how God is related to other necessary propositions, which make sense of God not being somehow constrained by them and that is the problem.

And as I say there really is no need for Christian theism to take this one on board. Theism certainly wants to say that God is in some sense unnecessary being and that is that he’s not causally contingent on anything. He is the ultimately supreme principle, everything depends on him, nothing exists apart from him. But there is no contradiction in saying that there is no God. The fool who the psalmist says, according to the psalmist says there is no God, he’s not contradicting himself, he’s just wrong.

But what about the original objection that I started with? Isn’t it going to be lucky that there is a God if there’s no metaphysical necessity about it because there could be a world without a God so what makes it the case that there is a world with a God? And of course and I would want to say in response to that well talk of luck is in place when as it were you think there are certain principles governing the universe and they don’t determine this rather than that so it’s a lucky thing that this turned up rather than that. I mean if the supreme laws of the world are those of quantum theory, but they don’t determine exactly what happens in a very small realm of the of the world.

And then this happens rather than that and you might say well it’s lucky that this happened rather than that, but luck is sort of… things are lucky or unlucky within the general scheme of things they turn out this way rather than that way, but the supreme principle governing all things doesn’t seem to me a proper subject for saying well it’s lucky that it’s that way or not that way. However if you think too much anthropomorphically and too much of God as a person, who like us something that makes me, me and you, you, you can imagine God saying to himself, well I’m lucky there’s not a different God in charge of the universe, fortunately it’s just me.

But if you think that although of course it’s right to talk of God as having being like ourselves in many respects and having beliefs and powers and purposes, if you think of him rather as not an omnipotent being rather than another one, but but omnipotence itself. That is to say you think of him as lacking a particular vastness. We are persons and we are this person and not that person and there’s something that makes us that and it’s something that individuates us, but God as the supreme power isn’t as it were this supreme power rather than that supreme power, he is just supreme power goodness in charge of the universe.

There wouldn’t be a difference between the universe governed by this God and one governed by that God if he had all the same properties. And I think if you begin to think in that way then it becomes you’re less inclined to pose the question isn’t it lucky that this God was in charge of the universe? Well there’s no difference between this God and that God being in charge of the universe and you’re less lucky to pose the question, isn’t it lucky that there is a God in charge of the universe because whatever explains anything just is God.

He exists in virtue of his own nature, but this is a causal relation. It isn’t a matter of logic, it’s a matter of that’s the way the world is and if that seems a bit weak compared with what it would be if logic determines it, well I’m sorry, it does. But what I think it reveals is that we humans are not fully in a position to understand the answer to why there must be a God and I think the answer is a lot deeper than I’ve suggested it is. But I want to suggest that it isn’t this answer and there’s no need for us to be burdened with that. [audience applauding]

About the Author