LCA Point 5: The Necessary Object is God

This is our third post in a series on Leibniz’ Cosmological Argument. You can find the first post here and the second here. The full argument is:

(1)               Every contingent fact has an explanation.

(2)               There is a contingent fact that includes all other contingent facts.

(3)               Therefore, there is an explanation of this fact.

(4)               This explanation must involve a necessary object.

(5)               This necessary object is God.

Today we will be examining…

(5)               This necessary object is God.

This is much like the first point, in that it’s not obvious and requires some argumentation. First, let’s talk about what it means to be a necessary object. A necessary object is an object that explains its own existence. It does not “depend” on any other object to exist. Using the possible world terminology we introduced last post, a necessary object is an object that exists in every possible world. 

So what properties can we deduce about the necessary object? Well it’s clearly not material, as material objects depend on matter to exist and are contingent. In fact, it’s not composite at all, since composite objects depend on their components to exist. So the necessary object must be entirely simple.

Whatever necessarily exists must always exist, since the fact that it necessarily exists is true no matter what time period you are in. So the necessary object is everlasting.

The necessary object must be unchangeable. Suppose that the necessary object were to change. That change must either be caused by the necessary object itself, or something else. It cannot be caused by something else, because that would make the state of the necessary object contingent on that thing that caused it. So if it changes, it must cause that change itself. If something changes itself, either part of it changes another part, or the whole changes the whole. Since the object in question has no parts, the whole necessary object must change the whole necessary object. But if this were the case, it would have already been the thing it was going to change into. So it cannot change.

The necessary object must also be omnipresent. The necessary object is free from any contingent properties, or any arbitrary properties, in virtue of necessity and simplicity. However existing at a particular physical location is a contingent property, since the existence of that physical location is contingent. Therefore the necessary object has does not exist at any particular physical location. This is omnipresence: present in equal measure at all places and at all times.

Now Leibniz presents an argument that the necessary object has understanding, a will, and power. He argues that it has understanding, as it can survey all possible worlds it can create. It has a will, as it selected one possible world. And it has power, as it actually created that world.

Is this argument reasonable? Consider first the understanding argument. Can the necessary object survey all possible worlds? A possible world is a world that could exist. If such a world contains any contingent objects, then the necessary object must exist. And if it doesn’t contain any contingent objects, and only the necessary object, then the necessary object must still exist.

Our necessary object explains all contingent objects. That means it must in some way contain information about those objects. And it contains information not only about contingent objects that exist, but also all possible contingent objects, since it would also explain them. And clearly it contains information of itself, since everything does. This kind of containing information is understanding.

Does it have a will? Of all the possible worlds, it chose one. Is simply choosing one possibility out of many having a will? Probably not, since my computer also does this and does not have a will. Instead, it seems more reasonable to say that having a will, or maybe a free will, is choosing an option apart from any outside compulsion. My computer chooses what it was programmed to choose, and so doesn’t have a will in this sense. But the necessary object is not compelled by anything external to it, and its choice comes entirely from within. So it seems it has a will.

And power? This one is more clear: it did in some sense create the world. It chose this possible world, and actualised it over other possible worlds. That is power.


Since the understanding, will, and power are directed towards all possibilities, they are infinite. The necessary being understands all possibilities, can will to create any of them, and has the power to carry out that will, whatever it may be.

And since choosing is to judge that the thing you choose is the best thing among the alternatives, and since the necessary object has full knowledge (knowledge of all possible worlds), it judges which is actually the best among infinitely many alternatives. This means that it is in a sense infinitely good.

Another argument can be made for a more general agency of the necessary object. As far as we understand explanations, there are three types: scientific, conceptual, and agential. The necessary object is clearly not a scientific explanation, since it cannot be material, due to simplicity. Further, science only explains things in terms of contingent things, or at least it has in every case we know of.

Nor is it a conceptual explanation. Among the contingent things we find substances, the stuff that other things are made from. Dualists will argue that the soul is a different kind of substance to matter, for example. Matter/energy is the substance that physical objects are made of. But the activity of a substance cannot be conceptually explained by the activity of something other than that substance. A conceptual explanation can tell us why things behave the way they do, in terms of what they are. But a conceptual explanation cannot give an account of why they are in the first place. So a conceptual explanation is insufficient for the big contingent fact.

So the only option remains is an agential explanation: an explanation in terms of some agent. So the necessary thing must be an agent in some sense.

We can also demonstrate uniqueness: every property the necessary object has must be a property it necessarily has, by a similar argument to the simplicity argument above. So if there are two necessary objects, they have the same properties. But two objects with identical properties are actually just one object. So there is only one necessary object.

At this point, we have proven that there exists a necessary object, which has infinite knowledge, infinite power, is infinitely good, is simple, eternal, unique, immutable etc. I know no name we could give this object other than “God”. And of course we have not yet arrived at Christianity. But I claim that after proving that this God exists, we may go looking for it in the world.

We find this: there exists a religious tradition spanning thousands of years who claim to worship something with all of these properties. They record centuries of interactions with this thing, who behaves exactly as we’d expect God to behave, including giving and fulfilling prophecy. And in the end we see the man Jesus who comes claiming to be God come in the flesh, who behaves as we’d expect God to behave, with a supremely good moral character, who fulfills prophecy given over the last few thousand years. I believe Him.

16 Replies to “LCA Point 5: The Necessary Object is God”

  1. I’m confused. How can something already be what it was going to change into? Why was it not one thing and then the other?

    1. It can’t be what it was going to change into. That’s my point, this is a reductio argument.

      It can’t be one thing and then another, because of the arguments I have just given.

      A fuller exploration of this can be found in chapter 4 of Aquinas’ Compendium of Theology.

  2. >Well it’s clearly not material, as material objects depend on matter to exist and are contingent. In fact, it’s not composite at all, since composite objects depend on their components to exist. So the necessary object must be entirely simple.

    I think we can say that the necessary being is matter or has parts because if it turns out to be matter or composed of parts, in this case, it would be dependent on matter and its parts meaning it is dependent on itself and so is still necessary since its explanation is not something exeternal.

    1. Being dependent on your parts is not the same as being dependent on yourself, because you are not equal to your parts, you are perhaps equal to the sum of your parts. This is not a good argument.

      1. Ok with that logic, what if the necessary being just is matter itself like a quark or another subatomic particle that is further indivisible. For example, Strings in M Theory are the most fundemantal building blocks of matter. How do we know that the necessary being is not only not material but how do we know that the necessary being is not matter itself?

        1. Well by “matter itself”, do you mean an individual particle like a single quark? That is clearly mutable, since quarks change position or colour, so it cannot be a quark.

          If you mean more the category of “matter”, then that has components, since there are things that are material. And God has no components.

  3. >Our necessary object explains all contingent objects. That means it must in some way contain information about those objects. And it contains information not only about contingent objects that exist, but also all possible contingent objects, since it would also explain them. And clearly it contains information of itself, since everything does. This kind of containing information is understanding.

    I don’t find this convincing at all. The necessary object could explain the existence of a said fact without containing explanation of that fact, much like the sun can give a partial explanation of the photosynthesis of a plant, although the sun does not contain partial information about that plant. I don’t see why it would contain information or understanding of itself either to be quite honest. Even if we take for granted that it has consciousness, how do we know that it knows everything about itself? A bay certainly has consciousness but it does not contain all the info about itself. Please explain this.

    >Another argument can be made for a more general agency of the necessary object. As far as we understand explanations, there are three types: scientific, conceptual, and agential. The necessary object is clearly not a scientific explanation, since it cannot be material, due to simplicity. Further, science only explains things in terms of contingent things, or at least it has in every case we know of.

    >Nor is it a conceptual explanation. Among the contingent things we find substances, the stuff that other things are made from. Dualists will argue that the soul is a different kind of substance to matter, for example. Matter/energy is the substance that physical objects are made of. But the activity of a substance cannot be conceptually explained by the activity of something other than that substance. A conceptual explanation can tell us why things behave the way they do, in terms of what they are. But a conceptual explanation cannot give an account of why they are in the first place. So a conceptual explanation is insufficient for the big contingent fact.

    >So the only option remains is an agential explanation: an explanation in terms of some agent. So the necessary thing must be an agent in some sense.

    Ok, this seems like non-sequitur to me. Why can’t the explanation be an immaterial substance? It’s as if you’re presupposing Avicenna’s metaphysics where he says that the properties of something immaterial must and can only be an intellect. I don’t see why this would be the case. Perhaps there are immaterial substances or properties, who knows? But we’re not in the position to say that something immaterial can only be an agent or intellect.

    I would really appreciate if you have the time to address my questions. I’m also a Theist, but I’m trying to be as objective as I can. God bless!

    1. I think your sun example is an interesting one, because the sun is not only a partial explanation, but actually a very very small part of a partial explanation. But I think the sun does contain some information about the plant, it contains information about the source of some of the energy that the plant has, which the plant has used to become what it is. A very small part of the information about the plant, but it is there. What contains more information about the plant is the DNA of the plant, and the history of that individual plant, and all the plants that came before it, etc.

      A better question is one about the relationship between what we might call the “possession” of information and the “knowledge” of information. I possess all information about me in that all that information exists within me somewhere. But I don’t know myself perfectly, I don’t know all that information. So how do I bridge the gap between God’s possession of information, and God’s knowledge of information?

      If you’re willing to grant consciousness, I don’t find it all that difficult. If God is conscious, God knows at least one thing. Now under divine simplicity, all of God’s attributes are identical attributes. Internal possession (but not necessarily knowledge) of information is an attribute, and that attribute must be identical to God’s knowledge. But if the internal possession is about everything, then God’s knowledge is about everything. This should be fleshed out a bit more, but it’s easy to establish the kind of link we need between God’s possession of all information and God’s knowledge of all information.

      Now you might not be willing to grant consciousness, but I think a good argument can be made for consciousness without presupposing will and power and knowledge. In Rasmussen’s paper From a Necessary Being to God, he presents an argument for consciousness that doesn’t depend on the will or on knowledge. He argues that non-conscious explanations necessitate the thing they explain. But if God necessitates the thing God explains, then we are forced into modal collapse and everything true becomes necessarily true. That’s got some pretty bad consequences, such as knowledge of counterfactuals being impossible, so we don’t permit that. Therefore God has a volitional will, and is conscious.

      (Without consciousness, Aquinas would argue that all effects pre-exist within their cause. I don’t know if I can demonstrate that, but it would be enough.

      Let’s not take consciousness, since I think we would go the other way. We would prove that God has knowledge and a will and is therefore conscious. So let’s not presuppose that, and we will make a slightly trickier argument without it. We will still suppose that God possesses all information, but we won’t suppose that God is the kind of thing that has knowledge. One avenue that I won’t explore in depth, which is Rasmussen’s perfection argument. I talk about it a bit here, and I think it’s reasonably obvious that knowledge is a perfection.

      Now about different kinds of explanations, I believe this is an argument I originally lifted from Pruss. I think I already do demonstrate that the explanation cannot be an immaterial substance, I don’t need Avicenna here. Clearly we don’t have a scientific explanation, I think that’s obvious if the substance doing the explaining is immaterial. All we need to rule out is a conceptual explanation. But as I say above, the activity of a substance cannot be conceptually explained in terms of anything other than that substance. Supposing we have an immaterial substance doing the explaining. That immaterial substance is explaining things that involve material substances. This means that it’s not a conceptual explanation. Conceptual explanations are more like internal models of how a thing behaves.

      But if it’s not a conceptual explanation or a scientific explanation, then it is an agential explanation.

      Continue being objective, that is a good practice. And out of necessity I don’t have as much time as I’d like to go into the depth I want on all these issues. But thankfully, many people have already done so. I recommend reading Josh Rasmussen and Alexander Pruss on these when you can.

  4. >In fact, it’s not composite at all, since composite objects depend on their components to exist.

    Jon, this is not a good argument at all. The type of self-sufficiency that you proved God to have is such that “God has no external sufficient explanation” where external just means non-identical. Fair, parts are technically not identical to the whole and so are external. If God depends on his parts, however, his parts would not be a sufficient explanation of God (each essential part individually would perhaps be a necessary partial explanation but not a sufficient explanation by itself) and so that is not inconsistent with the conclusion you made in the previous steps of the argument. Moreover, some constituents ontologists like myself don’t consider contingent properties to be parts at all but rather further facts about the particular with the particular being identical only to its essential properties (+ maybe some substance). So even if I grant you that there are no essential parts, there would still be further contingent facts about God.

    1. God doesn’t have partial explanations either, so God’s parts being necessary partial explanations rather than constituting a sufficient explanation is fine I think.

      In terms of properties, I think you’re thinking along the right lines with accidental and essential properties. First, let’s distinguish between real and Cambridge properties. I’m only talking about real properties here, not Cambridge properties.

      In terms of real properties, God only has essential properties, only necessary essential properties, and only one of them, and that property is identical to God.

      In terms of Cambridge properties, these are probably the further contingent facts about God you have in mind. Properties like “God communicated with Moses”. If Moses is contingent, clearly that Cambridge property is contingent.

      1. >God doesn’t have partial explanations either, so God’s parts being necessary partial explanations rather than constituting a sufficient explanation is fine I think.

        How would you prove this? The ‘God does not have any partial explanations’ does not seem to follow from anything previously said in the argument including the claim ‘God does not have any sufficient explanations. An argument has to be made that the latter entails the former.

        >In terms of Cambridge properties, these are probably the further contingent facts about God you have in mind

        If this is the case and God does have contingent properties in a sense then the arguments for omnipresence and uniqueness dont go through since they both depend on absence of contingent properties. If God has contingent properties (cambridge or real) then these are the properties that can be used to distinguish one God from another.

        I want to make a separate objection, how do you go from the explanation merely possessing the information about the BCF to actually knowing the information? A rock in a sense contains information about minerals without knowing that information.

        1. > How would you prove this?

          Nothing in my argument actually depends on the notion of sufficient explanations, so we can safely run the argument with only partial explanations, or explanations that do not entail. We end up still getting a necessary thing, but in this case a necessary thing is a thing that is self-explanatory and has no outside partial explanation.

          > If this is the case and God does have contingent properties in a sense then the arguments for omnipresence and uniqueness dont go through since they both depend on absence of contingent properties. If God has contingent properties (cambridge or real) then these are the properties that can be used to distinguish one God from another.

          Only if Cambridge properties are able to differentiate different beings. But it seems to be the case that if there is some difference in Cambridge properties, this must actually be grounded in a difference in real properties. If God 1 and God 2 have identical real properties, then that would entail that they have the same relationship with Moses, and so the same Cambridge properties. Clearly this needs to be fleshed out a bit more, but it is easy enough to do 🙂

          > I want to make a separate objection, how do you go from the explanation merely possessing the information about the BCF to actually knowing the information? A rock in a sense contains information about minerals without knowing that information.

          I want to flesh this out more in a blog post soon, but I am trying to do it in a way that is relatively agnostic of phil mind, since I don’t want to force my readers into a particular phil mind. But it’s hard to make it general enough.

          Either way, God probably doesn’t really actually have a mind or knowledge like we do, but only analogically. So that’s fine.

  5. >I want to flesh this out more in a blog post soon

    Sure, I will check it when it comes out.

    >Only if Cambridge properties are able to differentiate different beings. But it seems to be the case that if there is some difference in Cambridge properties, this must actually be grounded in a difference in real properties.

    This seems false. Lots of counterexamples come to head but would love to see you defend these. Btw, if you take a relative account of space, then your argument for omnipresence fails given contingent cambridge properties. However, yeah this type of argument would be valid with a controversial premise. If God’s cambridge properties can vary that would entail that God’s essential properties could be different among divine beings which can’t be so God’s cambridge properties can’t vary and so must be the same among different divine beings. Seems valid.

    >Nothing in my argument actually depends on the notion of sufficient explanations, so we can safely run the argument with only partial explanations, or explanations that do not entail.

    This seems iffy to me. If you use sufficient explanations then I grant steps 2-4 but you can’t derive many divine attributes because the conception of explanation is too specific and niche. If you use partial explanations, then I grant many of the divine attributes (except for your argument that God has no accidental properties which still would not give us simplicity) but then I would not grant steps 2-4, even if some contingent fact X explains the BCCF and is a conjunct of the BCCF and X is a partial explanation of X, that would not make X necessary since it could have some other external partial explanation Y (in which case X is still contingent despite being a partial explanation of itself), so your a reductio would fail. You need to give me a proper definition of “explanation” that reconciles these issues.

    I would be interested to see how you would also respond to the other constituent ontologists that believe contingent attributes are genuine parts of particulars. That ontology would seem to undermine your argument for simplicity (it would not apply to contingent parts since they would not be a partial explanation of God at all given that without them, God would still exist and that they are not a part of God’s identity).

    1. > would not grant steps 2-4, even if some contingent fact X explains the BCCF and is a conjunct of the BCCF and X is a partial explanation of X, that would not make X necessary since it could have some other external partial explanation Y

      I do not think it can be the case that contingent things are even partial explanations of themselves. This is the view that Pruss takes, perhaps some of his work would be useful to you.

      I don’t think there was much else in here that really needed a response. Was there something else that you thought was important?

Leave a Reply

Your email address will not be published. Required fields are marked *