Saturday, February 17, 2024

Avicenna, Aquinas, and Leibniz on the argument from contingency

Avicenna, Aquinas, and Leibniz all present versions of what would today be called the argument from contingency for the existence of a divine necessary being.  Their versions are interestingly different, despite Aquinas’s having been deeply influenced by Avicenna and Leibniz’s having been familiar with Aquinas.  I think all three of them are good arguments, though I won’t defend them here.  I discussed Avicenna’s argument in an earlier post.  I defend Aquinas’s in my book Aquinas, at pp. 90-99.  I defend Leibniz’s in chapter 5 of my book Five Proofs of the Existence of God.  Here I merely want to compare and contrast the arguments.

Because I want to focus on what I take to be the main thrust of each of the arguments rather get bogged down in exegetical details, I will offer my own paraphrases of the arguments rather than quote directly from any of these thinkers’ texts.

Here are the three arguments.  Avicenna’s is from the Najāt, and can be paraphrased as follows:

At least one thing exists.  It has to be either necessary or contingent.  If it’s necessary, then there’s a necessary being, and our conclusion is established.  But suppose it is contingent.  Then it requires a cause.  Suppose that cause is a further contingent thing, and that that further contingent thing has yet another contingent thing as its own cause, and so on to infinity.  Then we have a collection of contingent things.  That collection will itself be either necessary or contingent.  But it can’t be necessary, since its existence is contingent on the existence of its members.  So, the collection must be contingent, and in that case it too must have a cause.  That cause is either itself a part of the collection, or it is outside it.  But it can’t be part of the collection, because if it were, then as cause of the whole collection, it would be the cause of itself.  And nothing can cause itself.  So, the cause of the collection of contingent things must be outside the collection.  But if it is outside that collection, it must be necessary.  So, there is a necessary being.

Aquinas’s version is the third of his famous Five Ways in the Summa Theologiae.  It can be paraphrased as follows:

Some things are contingent in nature, as is evident from the fact that they come into existence and go out of existence.  Such things can’t exist forever, since whatever is contingent, and thus is capable of failing to exist, will in fact at some point fail to exist.  So, if everything was contingent, then at some point nothing would have existed.  But if there was ever a time when nothing existed, then nothing would exist now, since there would in that case be nothing around to cause new things to come into existence.  But things do exist now.  So, not everything can be contingent, and there must be a necessary being.  Now, such a thing might derive its necessity from another thing, or it might have its necessity of its own nature.  But there couldn’t be a regress of things deriving their necessity from something else unless it terminates in something having its necessity of its own nature.  So, there must be something which has its necessity of its own nature.

Leibniz’s version is in the Monadology.  It might be paraphrased as follows:

For anything that exists, there must be a sufficient reason for its existence.  In the case of the contingent things that make up the universe, this cannot be found by appealing merely to other contingent things, even if the series of contingent things being caused by other contingent things extended backward into the past without beginning.  For in that case, we would still need a sufficient reason why the series as a whole exists.  But the series as a whole is no less contingent than the things that make it up.  So, the explanation cannot lie in the series itself.  A complete explanation or sufficient reason can be found only if there is a necessary being that is the source of the world of contingent things.  So, there must be such a necessary being.

Each of these thinkers goes on to argue that, on analysis, it can be shown that the necessary being must have the key divine attributes, and therefore is God.  But I want to focus here just on the reasoning each argument gives for the existence of a necessary being.  And again, I won’t be defending the arguments here, but just comparing them.  So I won’t say anything about how the arguments might be fleshed out or the reasoning made tighter, how various objections would be answered, and so on.

What do the arguments have in common?  First, they all rest, of course, on the distinction between contingent beings and necessary beings, and argue that it cannot be that everything falls into the former class.  Second, they all reason causally to the necessary being as the source of everything other than itself.  Third, for that reason, they all have at least a minimal empirical component insofar as they appeal to the contingent things we know through experience and argue from their existence to that of a necessary being.

A fourth similarity is that all three thinkers cash out the nature of the necessary being’s necessity in terms of the distinction between essence and existence.  Though, as my paraphrases indicate, this is a distinct step that does not and need not be stated in the arguments themselves.  Furthermore, our three philosophers do not do this in quite the same way.  Avicenna, and Aquinas following him, hold that the cause of things in which essence and existence are distinct must be a necessary being in which they are not distinct.  Leibniz, however, does not say that God’s essence is his existence, but that his essence includes existence.  (This way of putting things has influenced much contemporary discussion – and not for the better, because it obscures the implications for divine simplicity that Avicenna's and Aquinas’s way of speaking makes clear.)

A fifth similarity is that none of the three arguments either presupposes or asserts that the universe had a beginning.  All of them hold that the existence of a necessary being can be established even if we were to suppose that the world of contingent things has always been here.

A sixth similarity is that each of the arguments moves from a claim about contingent things considered individually to a claim about contingent things considered collectively, albeit in different ways.  For Avicenna, just as an individual contingent thing requires a cause, so too does the totality of contingent things require a cause.  For Aquinas, just as individual contingent things must fail to exist at some point, so too must the collection of contingent things fail to exist at some point, at least if there were no necessary being.  For Leibniz, just as individual contingent things require an explanation outside them, so too does the collection of contingent things require an explanation outside it.

How do the arguments differ?  First some background.  Scholastic philosophers often distinguish physical from metaphysical arguments for God’s existence.  Physical arguments are those that proceed from facts about the concrete physical world as interpreted in light of the philosophy of nature.  For example, Aquinas’s First Way is commonly interpreted as a physical argument because it begins with the reality of motion, understood along Aristotelian lines.  Metaphysical arguments are those that proceed from more abstract considerations that are not tied to the physical world per se.  For example, Aquinas’s proof for God’s existence in De Ente et Essentia begins with the fact that there are things whose essence and existence are distinct and argues that such things require a cause whose essence just is subsistent existence itself.  Since there is an essence/existence distinction in angels no less than in material things, the argument does not depend on facts about the physical per se. 

Of the three arguments we’re considering here, Aquinas’s has a more physical cast than those of Avicenna and Leibniz.  For the observation that material things come into being and pass away, and the claim that material things individually and collectively would go out of existence given enough time, play a big role in the argument, and these are points about the physical qua physical. 

By contrast, Avicenna’s and Leibniz’s arguments have a more metaphysical cast.  Even if we take them at least to refer to physical things, what they focus on is the contingency of these things rather than anything specifically physical.  And angels, which are immaterial, are in a sense contingent too, insofar as there is in them an essence/existence distinction and thus the need for a cause which imparts existence to them.  (To be sure, there is for Aquinas also a sense in which angels are necessary beings, since once they exist, there is nothing in the created order that can destroy them.  Still, they have to be created by God, who could also annihilate them if he wills to.  Hence angels have only a derivative necessity rather than a strict necessity.  For that reason, they also have a kind of contingency.)

Hence, it seems that one could remove any reference to the physical as such from Avicenna’s and Leibniz’s arguments without altering their basic thrust.  Indeed, one could even remove any reference to any actual specific contingent things and argue simply that if there are contingent things (whether or not there really are any), they couldn’t be the only things that exist, for the reasons Avicenna and Leibniz give.  Aquinas’s Third Way, by contrast, would be a very different sort of argument if the physical claims it makes were removed from it.

A second difference is that the notion of explanation, and with it the Principle of Sufficient Reason (PSR), play an explicit role in Leibniz’s argument but not in Avicenna’s and Aquinas’s.  That is not to deny that Avicenna and Aquinas are at least implicitly committed to PSR, and that it lurks in the background of their arguments, which are of course offering explanations.  But this is not thematized in their arguments, the way it is in Leibniz.  This reflects Leibniz’s distinctively rationalist approach to metaphysics.

Here’s one way to understand the difference.  The Scholastics distinguish several “transcendentals,” attributes that apply to all things of whatever category – being, truth, goodness, and so on.  These are taken to be “convertible,” the same thing looked at from different points of view.  For example, truth is being considered as intelligible, and goodness is being considered as desirable.  (I say a lot more about the transcendentals in this article.) 

Avicenna’s and Aquinas’s arguments essentially consider reality under the guise of the transcendental attribute of being.  The being of contingent things, they argue, must derive causally from the being of something that exists in an absolutely necessary way.  Leibniz’s argument, by contrast, essentially considers reality under the transcendental attribute of truth.  The intelligibility of contingent things, he argues, presupposes a necessary being which is intelligible in itself rather than by reference to something else.

A third difference is that the impossibility of an infinite regress of a certain kind plays a role in Aquinas’s Third Way that finds no parallel in Avicenna’s and Leibniz’s arguments.  To be sure, and as I have said, none of the three arguments rules out the possibility of an infinite temporal regress – a regress of what Aquinas would call “accidentally ordered” causes extending backward into the past.  None of them supposes or tries to establish that the world had a beginning.  But Aquinas’s argument does include the premise that a series of beings that derive their necessity from something else would have to terminate in something that has its necessity of its own nature or built into it.  And here he is appealing to the impossibly of an infinite series of causes of what he calls an “essentially ordered” kind, also known as a hierarchical causal series.  (I discuss the difference between these two kinds of causal series in many places, including Aquinas and Five Proofs.)

Any further differences between the three arguments seem to me to reflect these three fundamental differences.  And the differences are important, both because they capture different aspects of reality, and because they entail that some objections that might seem to have force against one version of the argument from contingency will not necessarily apply to other versions.  (Though, as I have indicated, I think each version can successfully be defended against objections.)


  1. WCB

    Down at the bottom of this Universe is a basic brute fact. Whether it is physics at the bottom, or God, or Gods or something we cannot imagine, there is a brute fact, or a set of brute facts.

    All else is that exists because of this or these brute facts is contingent. but to state it must be a divine being, one and only one divine being needs evidence to be acceptable as an answer as to what that brute fact or facts, is or are.

    As David Hume pointed out, this is an analogy. An artisan creates something. God as an artisan. but large projects are created my numerous people, so maybe there are many Gods to create this complex Universe. Or perhaps, the Universe is an organic thing that just grows, like a carrot.

    The problem with claiming that brute fact is God is that this God, carefully examined is mired down in numerous logical difficulties. Free will vs foreknowledge, omnipotence and perfect goodness of God and the existence of great and horrendous evils and more.
    And no, these problems have NOT been solved, despite the usual claims that they have been resolved. Let the name calling begin.


    1. I wonder if you are functionally literate. The contigency argument leads to a conclusion that is not a brute fact. In addition if you truly "carefully examined " the arguments you will fine there is no logical inconsistencies or problems. Well informed atheist philosophers acknowledge that much (their qualms are more obvious the premises)

      Frankly, it's a pity this blog gets polluted these "anonymous" ignorant comments that have nothing relevant nor intelligent to say.

    2. What you did was incredible, and will be studied by philosophers for years to come. In four short paragraphs you demolished the arguments of Avicenna, Aquinas and Leibniz. Ed needs to close this topic and move on to something else, because really, what else is there to say?

    3. WCB signs his posts consistently as far as I know, so I wouldn't call him anonymous.

      He is wrong about the logical problems remaining unresolved, however. They most certainly have been, for a very long time. I'm frankly surprised they keep being asserted as strong arguments against God.

    4. @WCB

      "All else is that exists because of this or these brute facts is contingent. but to state it must be a divine being, one and only one divine being needs evidence to be acceptable as an answer as to what that brute fact or facts, is or are."

      Avicenna, St. Thomas and Leibniz* did explicit adress your concerns in the texts that Ed cited and others(at least on St. Thomas case), so yea, take a look someday.

      *besides most of the defenders of the cosmological argument

    5. Agree with FM, the low-quality comments by WCB pollute what could be a very interesting comments section.

      If I went onto an atheist blog and responded to a post that purported to give three or four reasons for God's non-existence, and simply ignored them and said "down at the bottom of everything is God, it cannot be otherwise", I would be mercilessly mocked and soon kicked out, and that would be a fair response.

  2. WCB

    Down at the bottom of everything is a brute fact or facts. It cannot be otherwise.


    1. I am not sure that material things individually and collectively would go out of existence given enough time. Individually, perhaps, but I do not see how every material object van go out of existence.
      There serems to be no physical support for that claim.

    2. I'm afraid that just repeating your thesis again is not a very persuasive way to argue for it. Have you considered repeating it more times? Perhaps about ten?

    3. I know Feser addresses this in Aquinas, but I gifted the book, so I cannot cite it.

      Although I know Aquinas was aware that change requires two terms. That is, corruption always requires generation (either absolutely or in a qualified sense). For example, Aquinas thought it was impossible for one creature to annihilate another creature. So I’m not certain that the third way is read physically as opposed to metaphysically. I’d have to hear arguments for and against.

    4. Our host regarding common complaints about the Third Way:

    5. I don't think the Necessary Being would be a brute fact, since its existence would be self-explanatory given its Necessity. But if you wanna insist on saying that it'd be a brute fact, be my guest. In that case you'd still have to make a distinction between Contingency vs Necessity, and then consider the fact that Brute Contingency would be absurd (i.e. you'll have to actually interact with the arguments of Avicenna, Aquinas and Leibniz, which you've been ignoring for whatever reason).

      So if you wanna say there MUST be a brute fact at the bottom, okay. If you insist and are so sure of it. But not all brute facts are equal. Saying that something that could have never existed, could have never been the case, just happens to be the case rather than not with no explanation whatsoever is crazy. And in addition, if you accept contingent brute facts in that manner, science is in some real trouble.

    6. @Walter, you write:

      I am not sure that material things individually and collectively would go out of existence given enough time. Individually, perhaps, but I do not see how every material object van go out of existence. There serems to be no physical support for that claim.

      Since Aquinas and the others here assume the eternality of the universe arguendo, your comment makes no sense. As usual, you comment on something you don’t understand.

      When you are repeatedly told that you don’t understand something, why don’t you take the time to understand it before offering a criticism? You’re skewering a scarecrow.

  3. WCB, is there a difference between something that has no possible explanation and something that explains itself?

  4. @WCB- Do you think we can say anything about the nature of these brute facts? Do you agree with Avicenna, Aquinas, and Leibniz that they are necessary? or are they contingent?

  5. WCB

    The problem is that when it comes to this idea, God, God is defined as having no underlying cause or causes, God is not in any way contingent. And God has always existed.

    Summa Theologiae I, q. 2, a. 3:
    Therefore it is necessary to admit a first efficient cause, to which everyone gives the name of God.

    Cutting through the empty verbiage, God has always existed, relies on no efficient causes for God's existence and is thus definable as a brute fact.

    Isaiah 40
    28 Hast thou not known? hast thou not heard, that the everlasting God, the Lord, the Creator of the ends of the earth, fainteth not, neither is weary? there is no searching of his understanding.


    1. Aseity for God is not a brute fact because a brute fact is one without a reason for obtaining. As God is Absolute Being or "Existence (in) Itself", to say "Existence Itself exists" is seen to be non-brute, as it has an internal explanatory rationale, that "This particular set of concrete entities 'obeying' this particular set of mathematical and qualitative laws exists" clearly does not have. Now, one can still argue about whether "Existence Itself" is meaningful, or is a coherent concept, or corresponds to theism's God, or can exist as such. But one cannot argue that "Existence Itself exists" proposes a brute fact. If it is a factual proposition, it is a rationally explicable one.

      Similarly, God can be defined as "Ultimate Reality". "Ultimate Reality must be real" is clearly not brute either. Again, that proposition might have other objections made to it, but it is clearly not the same kind of ambit claim as "This keyboard could not fail to exist".

    2. @Fr Matthew Kirby re your "a brute fact is one without a reason for obtaining."
      This claim is contentious. Another, less maximal view of brute fact: a state of affairs, the obtaining of which is not explained by the obtaining of another state of affairs. "... bruteness is an ontological rather than epistemological concept; that a given fact is brute does not imply that it cannot be proven or inferred from other things one knows. Also, that a fact is brute is consistent with its being explained by itself" (Erik Wielenberg). Swinburne wrote, "No other agent or natural law or principle or necessity is responsible for
      the existence of God. His existence is an ultimate brute fact." (Of course Swinburne became Orthodox not Catholic, but his confessional affiliation doesn't determine all his philosophical views.)

    3. Also, that a fact is brute is consistent with its being explained by itself" (Erik Wielenberg)

      ficino, I would take issue with this characterization. It seems to me that we are dealing with 3 logically possible categories: (1) Those things whose existence or nature has something else to account for them; (2) those things whose existence and nature needs nothing else to account for them because it is explained by themselves; and (3) those things whose existence or nature does not have something else to account for them but also do not account for themselves either. i.e. things that do not have an account for them.

      In one taxonomy, it is possible to lump (2) and (3) together by naming a category for "those things that do not have something else account for them". But in another taxonomy it is possible to lump (1) and (3) together by naming a category for "those things that do not account for themselves".

      It is my sense of the "state of the question" on brute facts that it is not appropriate (at least at this stage of the game) to assume the former taxonomy and then to simply identify brute facts as "those things that do not have something else account for them," because there is a grave epistemological gray area between, on the one hand, (1)(a) those things which have something else account for them and we know that because we know the something else that accounts for them and (1)(b) those things which are of such a nature that we think they have (or probably have, or ought to have?) something else account for them, but we don't yet know what that is, and (1)(c) those things that do have something else accounts for them but that WE are not exactly confident that they have something else that accounts for them; and on the other hand group (3) those things for which nothing accounts for them. On what basis could we know that an item is in category (3) rather than (1)(c)? Is there any sort of proof whose result can be of the form "therefore, X fact cannot possibly have an account for it?

      That isn't simply begging the question, (and that isn't the self-explanatory kind in group (2))?

      Russell's "the universe needs no explanation" appears to be more of the begging the question kind of comment in this vein. Given that every being that we (so far) have come across in the universe of a physical / energetic nature is contingent, it is not metaphysically obvious that the whole universe is not contingent. If it is possible to conceive of the universe as a whole as contingent, (which it is) then it is of such a nature that it MAY have an account, and simply stating that it needs no explanation is a self-serving and question-begging assertion.

      Indeed, if there are indeed brute facts in the sense that nothing accounts for them, it would be more expected that we would be unable to identify them as such precisely unless we could precisely and fully account for WHY they have no account - which would be a real trick, wouldn't it? It is harder still if we grant that absent the hypotheses that the universe is non-created and has no account,, the account of the theists that God is self-explanatory being and all others are accounted for by God, IS AN ACCOUNT of any otherwise brute-fact-seeming thing - including the universe. That is, the brute-factist can compete with the theist by simply adding to his primary hypothesis (that there are brute facts) that (a) there is no God, and (b) there is no account for the universe, but these are two additional hypotheses beyond the mere supposition that there are brute facts: he has to assume 3 things, whereas the theist only claims a single thing - God is his own self-explanatory kind of thing. Theists can account for the universe by their one premise, but the brute-factist requires 3 premises to get off the ground.

    4. @Tony: thanks for your reply. A reply to everything you wrote would become rather long!

      I did a quick search for "brute fact" on Philosopher's Index. A definition different from Wielenberg's: "Brute facts are those that require no human institution for their existence. To state a brute fact requires naturally the institution of language, but the fact stated is not the same as the statement of it. For example, regardless of any human institution or opinion, the presence of an extra chromosome is a brute fact, and despite of people's constructions or deconstructions,
      this fact remains." ~ S. Vehmas, P. Mäkelä

      Maybe a problem is, what sort of account do we seek when we ask for an account of a fact, or of a state of affairs' obtaining? I am thinking that we don't always get an account of how the constituent thing/s work. I get the impression that at the end of a series of explanatory accounts we get something like "that's its nature," or perhaps a less theory-laden response like "Well, that's the way it is." Isn't such a response a way of treating the state of affairs as providing its own primitive explanation?

      I don't know enough about Russell's notion of explanations/accounts to say whether Russell would allow that the universe provides its own explanation.

    5. @ficino4ml

      That's just a bad definition for brute facts because, if for no other reason, it introduces semantic drift in the middle of the discussion.

      If you want to insist on that definition for brute facts, fine, but then we need to agree on a different word that exclusively applies to the set of things Fr Kirby is referring to.

      You're muddying the waters with semantics.

    6. @Anonymous Feb 20 a 4:19 AM (BTW it becomes confusing when more than one posts as Anonymous):

      Which definition of brute fact do you think introduces semantic drift, and why? I proposed two, each of them maintained in the professional literature. To note that a term is used differently by different theoreticians is not to "muddy the waters." It is to push for clarification, as you yourself urge.

      As far as I can see, Fr. Kirby would want the set of brute facts as he defines brute fact -- "a fact without a reason for obtaining" -- to consist of no members. So I am confused about what "set of things" you say Fr. Kirby is referring to. Perhaps he can elaborate if he comes back on here.

      I do not know that there is any state of affairs that has NO explanation for its obtaining. It seems to me that what matters for the Contingency argument is whether there can be a collective or the like that provides or is its own explanation. I don't know enough about Russell to know whether he allowed that the universe's obtaining can be explained by the universe itself; of course he denied that it needs to be explained from an external explanans.

    7. Ficino4ml, I don't see anything wrong with defining "brute fact" that way for purposes of whatever your purposes are, and maybe I'm missing something, but I don't see the relevance to the discussion at hand. The fact that something's (or the universe's) existence doesn't depend on human institutions doesn't mean that it has no explanation or even that it isn't dependent on God.

    8. @WCB I dont understand your response (sorry if it was not directed at me). I wasn't asking about God, I was just wondering what you think we can say about the Brute facts you believe in, and whether you think they are necessary. Thanks\

    9. @Anonymous who wrote at Feb 20 11:36 AM:

      1. would you and others who post as "Anonymous" PLEASE create a profile? I don't know how many Anonymi are participating in this thread, or which one has replied what.

      2. I originally starting talking about brute facts because WCB said that God and/or God's existence is brute fact, and then Fr. Kirby said that WCB was wrong because a brute fact has no explanation for its obtaining, but God does have an explanation: His nature. This I take it is what Fr. Kirby pointed to when he said that the fact that Existence Itself exists is non-brute because "it has an internal explanatory rationale."

      My point was that some philosophers, unlike those followed by Fr. Kirby, do not define brute fact as a state of affairs that has NO explanation for its obtaining. Like Erik Wielenberg and others, they define brute fact as a state of affairs that merely has no EXTERNAL explanation for its obtaining, while they allow that it may have an internal explanation.

      On the latter conception of "brute fact," WCB might want to say that facts about both the universe and, ex hyopothesi, God are brute in that they have some internal explanation for their obtaining. The religious skeptic need not deny the PSR while being a skeptic about God Some skeptics do deny the PSR, but I submit that their skepticism about God itself need not require them to deny the PSR.


  6. To be honest, I have touble with these abstract concepts ( necessity and contingency) and their correct application.

    Is there any entity or thing in the world we live in to which we can attribute necessary existence?

    I know that oxygen needs to exist in order for their to be a fire. But does that mean it is necessary for oxygen to exist?

    1. Oxygen is clearly not necessary, for it can change to another substance, something that is necessary can't cease to be.

      In fact, thinkers like St. Thomas, Plotinus etc did argue that nothing material can be necessary*, needing something beyond the phisical world to explain why is there anything at all.

      *even Plato itself!

    2. Thanks for the explanation, Talmid.
      That helps to clarify the issue.

    3. That is cool to know! It is a bit of a hard area to understand at first, but it does gets interesting.

    4. @Hal, you write:

      Is there any entity or thing in the world we live in to which we can attribute necessary existence?

      In the context of these arguments, there are two types of necessity—absolute and hypothetical (suppositional). The former refers to God and the latter refers to creatures which do not have the capacity for generation and corruption (transitioning from one substance to another). Aquinas thought that things like planets had hypothetical or contingent necessity, but we now know that was mistaken. There may be things like quarks that have suppositional necessity, but the only created thing I can think of with hypothetical necessity is an angel.

    5. Thanks for the clarification, Bill. I didn't know about the two types of necessity. Also, hadn't considered the possibility you mentioned regarding quarks having suuppositional necessity.

    6. @Hal, I should add that on the supposition that God creates what He creates, He creates necessarily. Considering that, creation is not absolutely necessary, but it has a suppositional necessity. We can thus say that creation is a suppositional necessity. I was previously thinking only of beings which were considered necessary in the sense that they did not undergo generation and corruption.

  7. Paul Edwards, " A Modern Introduction to Philosophy. " This textbook (an anthology of readings)has been around for decades. In his chapter titled "Why," analyzing the views of numerous philosophers , Edwards explains that the question of the existence of universe, and "why there is something rather than nothing," is a question that cannot be answered and is meaningless.
    P. 796, free on the Internet Archive.

    1. There are plenty of questions that cannot be answered that are not meaningless questions. Math poses lots of them. So, hopefully, Edwards provides separate accounts for why that question cannot be answered and is meaningless. But I doubt it, I suspect that he is a typically sloppy writer who is unaware of the assumptions he is bringing to the issue.

      One might suggest that if the question "why is there something rather than nothing" is meaningless, then that casts a pall over all possible "why" questions: for, if the very existence of everything exists cannot be questioned, then how could any aspect of everything that exists be given a true and final account? The answer given for any "why" at all would need a prior why, on into a regression to ultimates, i.e. ultimately getting back to "why does anything exist" which is meaningless.

    2. Wow, this Edwards guy simply fails to understand Aquinas.

    3. Hey ficino4ml.

      Just because you've failed to understand Aquinas doesn't mean others have also.

      You're beginning to sound like a broken record. Someone makes a straw man argument and when he's called out you complain about the person calling out the fallacy rather than the person making the straw man argument.

    4. Did anyone actually read the entire article by Prof. Edwards? If so, rebut what he says.

  8. WCB

    The universe needs no explanation; it is “just there, and that’s all”
    - Bertrand Russell 1948


    1. Russell use of this in the Copleston debate was pretty lame. If Copleston wished to push him further i doubt he would have much to defend this going by what he had to say.

      "Just because every human has a mother it does not means that humanity has a mother".

      Well, a concept truly has no need of a mother, but the universe is kinda diferent, no?

    2. @Talmid: perhaps a better analogy would be to say that if every person in the tribe depends on at least one other person in the tribe, it does not follow that the whole tribe depends on at least one person outside the tribe. ??

    3. @ficino4ml

      That one is way better! Even a genious like Russell can fail to state his points well, i guess. Your analogy works well to the case.

      Of course, one can make the case on a certain tribe that the economy inside it depends on trading with a rich man with lands nearly, so one would need to look at the case well to know who is right. This is the problem with Russell and WCB spirit, though: these questions deserve our energy and time.

    4. The universe needs no explanation; it is “just there, and that’s all

      This has always seemed to me like pointing to a pile of rocks on the ground and saying: "Every individual rock in this pile needs an explanation...but the 'pile' doesn't!"

    5. @DrYogami: I've proposed a sense of "brute fact" used by some philosophers, according to which a fact is "brute" just in case there is no EXTERNAL explanans of its obtaining. It may have an INTERNAL explanans. I am wondering whether the person who is skeptical about the God of classical theism can still adhere to the PSR and accord some internal explanation of the universe's existing, as theists accord an internal explanation to God's existing.

    6. Hello, ficino. What I was trying to get at with my analogy was that Russell (and by extension you) seems to think of the 'universe' as a kind of container or 'box', rather than an abstract aggregate. But a pile of rocks is not two distinct entities 'pile' and 'rocks' and so there must be an 'internal explanans' of 'pile'. The pile just IS the rocks...

    7. @DrYogami, if I understand your last correctly, it sounds as though we are in agreement. I am supposing that the universe is just constituted by all that there is. I'm positing that it can have an internal "explanans" and thus be "brute" only insofar as it does not need an external "explanans." That soungs like what you said about the pile of rocks.

    8. @ficino4ml

      " I am wondering whether the person who is skeptical about the God of classical theism can still adhere to the PSR and accord some internal explanation of the universe's existing, as theists accord an internal explanation to God's existing."

      One could argue for that, that the universe is not contingent but necessary.

      The thing is that folks like St. Thomas, Avicenna etc do argue for a necessary being need to have certain characteristics that the universe lacks and that pretty much only God has such as immateriality, eternity, ominiscience etc.

      The cosmological argument has 2 stages, one can say:

      1. That there needs to exist a being whose explanation of its existence is internal.

      2. That this being is what we would call God.

      Oponents of the cosmological argument normally tend to focus on attacking stage 1: we do not need a being with a internal explanation at all.

      Can we say that you are okay with stage 1 but just disagree when we get to stage 2?

    9. @ficino4ml:

      Of course this gets to what sort of properties a thing has to have in order to have an internal explanans. A pile of rocks certainly doesn't.

      I sometimes think the problem people have with classical theism is that they don't like the 'God' language and think that it carries unnecessary baggage. So don't call it 'God'. Call it 'X'. What sort of properties would X need in order to be internally explanatory?

    10. @DrYogami: I don't know enough to enumerate in advance all the sorts of properties a state of affairs would need in order to be internally explanatory. What I have read so far suggests that if p explains q, then p and q must both hold, and that a sufficient explanans must be either self-explanatory or necessary.
      An example of an internally explained state of affairs I've seen in the literature is that pain is intrinsically bad. That doesn't require explanation in terms of some other state of affairs that obtains. Another is "7 + 5 = 12." The thought is that no explanation external to the sentence is required. Perhaps one could say that any external ground to which one might appeal would be less perspicuous than the truth of the equation itself.

      That's the best I can do so far!

  9. Can I ask a question here? I might consider becoming a theist again if someone can answer it. I accept that there is Necessary Existence, but I don't accept that it is "God" in any way.

    The question is this: The action of creation from a Necessarily Existing "God" would have to take place somewhere externally, thus for His action to occur, it is dependent on the existence of something external to Himself, if only logically or as "space" (since He is unchanging in His "essence" and so cannot have an internally-directed action or an "inbetween" action in any way, shape or form). Hence, if He creates anything, He is contingent for otherwise, He cannot necessarily do that act and actualize Himself (be purely actual). Therefore, you arrive at a contradiction if the NE God creates anything. But if He doesn't create anything, then its not a creator God at least. Moreover, thus since knowledge is not actually necessary in any way, the NE God cannot have any knowledge. And also, if He cannot create, then He is not all-powerful in reality, and thus, its not God, and certainly not a God of any scriptures. Therefore, the Necessary Existence cannot be God, albeit it could be perhaps something like the Buddhist nibbana depending on the school of thought.

    Does anyone have a reply to this?

    1. JMM

      Maybe you can begin by explaining why you believe that "The action of creation from a Necessarily Existing "God" would have to take place somewhere externally, thus for His action to occur, it is dependent on the existence of something external to Himself, if only logically or as "space" "

      It sounds like you are trying to say that God depends on there being some kind of "space" that is "outside" of himself in which he can make things. Is this what you mean? That there's nowhere for God to put the things he's made unless there's some place that exists for him to put them?


    2. In line with what I posted below, I'd suggest that one line of thought leads to the question, of what is "necessary" a property? AFAIK Russell expressed a standard, analytic philosophy view when he said to Copleston in their famous debate that he regards "necessary" as a property of propositions not beings (or, as I summarized, things). Questions like that may lead you further to consider whether "the Necessary Existence" about which you wrote is a coherent concept. Russell wanted to restrict its application to analytic propositions. There is also the problem, is existence a first-order predicate.

    3. JMM: What I'm trying to say is that God is unchanging in His essence and one. So he has no internal potency, so the only way He can actualize and create anything is external to Himself. But I think my comment about needing something external is actually a tangent, my main point was this:

      If God is necessary, and He creates *ANYTHING*, then it means that this contingent existent that He created is actually necessary as well, meaning that NOT creating it would result in a logical contradiction, and thus God HAS TO create it. And if He HAS TO create a contingent existent, then it means that He depends on something external to Himself in order to be fully actualized, meaning that He is imperfect, contingent, etc. But we already said that He is absolutely necessary (not merely necessary via another), so we get into an issue here with contradiction.

      ficino4ml: I would say that most Classical Theists would say that God is not a "being", but transcends such a description and dichotomy. Indeed, being itself is grounded within a Necessarily Existing God. I think what I said was standard in this type of philosophy, but correct me if I'm wrong.

      I hope this clarifies what I meant.

    4. @Anonymous who posted at 2/19 7:01 PM:

      I think we get bollixed up easily when we start using the verb "to be." It's contentious, for example, whether ancient Greek has an "existential" sense of "einai" or whether "einai" always means "to be F," whatever value we give to F in a given sentence.

      Anyway, as far as the Russell-Copleston debate goes, Copleston himself started off by defining God as "a supreme personal being." Later on Copleston invokes Leibniz's argument from Contingency, about which Prof. Feser wrote just recently. Copleston describes that argument as working from "beings in the world which do not contain in themselves the reason for their existence" to "a being which contains within itself the reason for its own existence."

      It was in response to that argument that Russell maintained that "the word necessary ... can only be applied significantly to propositions."

      I don't know whether Ed or others would want to reword what Copleston said about the argument from Contingency.

    5. Can you rephrase this second part a different way "Hence, if He creates anything, He is contingent for otherwise, He cannot necessarily do that act and actualize Himself (be purely actual)."?

      Your first statement is not controversial for a classical theist, that creation must take place externally to God/the necessary existence, but I don't quite understand what you're saying with this second part.

      Maybe this will help, I don't see why the argument as presented is a problem unique to a necessary existence that looks like God. While it seems like you have ruled out certain kinds of necessary existence by deduction, your supposed contradiction doesn't seem dependent on any particular divine attribute. It seems like if that contradiction is a problem, it's a problem for any kind of necesary existence that creates contingent things, and since we agree that contingent things exist, that's the kind of necessary existence that must exist,

    6. @Original anon

      On your clarification, it seems like in your objection, you have independently come to what is known in the literature as the "modal collapse" argument. Feser has written on this topic before on this blog, so that would be a good place to start.

    7. Anonymous at 4:35AM: I guess its a type of modal collapse, but the issue here isn't that all contingents get created as with my understanding of traditional formulation of modal collapse, but rather that if an NE God creates *ANY* contingent existent, even a singular one, it would implicate dependency, and thus, contradiction. So an NE God would have no capacity to create, which means its not a creator God, nor is it all-powerful, and therefore, the Necessary Existent is not God.

      Am I understanding this correctly?

    8. ficino4ml: Is "God exists" not a proposition though? Can it not apply to it? Russell-Copleston already begin by defining God as something Classical Theists would object to, so I'm not sure how much their criticisms would apply. What you wrote about Lebiniz contingency argument has been discussed I think in some Ismaili criticisms of the contemporary Islamic contingency arguments, and also by some later Arab scholars, but I'd have to dig into the literature to find them again. I don't think its anything crushing for the contingency argument though.

    9. @Anonymous writes:

      If God is necessary, and He creates *ANYTHING*, then it means that this contingent existent that He created is actually necessary as well, meaning that NOT creating it would result in a logical contradiction, and thus God HAS TO create it. And if He HAS TO create a contingent existent, then it means that He depends on something external to Himself in order to be fully actualized, meaning that He is imperfect, contingent, etc.

      This doesn’t appear coherent. On Thomism, there are different types of necessity—absolute and suppositional (or hypothetical). God exists in Himself and everything else exists through God. These are clearly different modes, so they cannot be conflated. Creation is necessary on the supposition that God wills it. It is therefore a suppositional necessity and never an absolute necessity. Since God is infinitely perfect, His act is infinitely perfect. Everything He does is infallible, so creation follows from God’s infinite perfection. But since creation is always reliant upon God’s conserving act, it can never be, by definition, an absolute necessity. God is thus “reliant” on Himself and never upon anything ad extra. There is no relation from God to creature, only from creature to God.

      To say that God’s having to do anything implies dependence implies that the concept of necessity is itself incoherent. Does God “have to” tell the truth? Does God “have to” stay in existence? As a necessary existence, there are things He cannot do (e.g., commit suicide). To argue that this kind of necessity is contradictory is to presuppose a definition your interlocutors do not assert. Creation exists and is sustained by the act of God, and everything that follows therein is pursuant to His infinite decree. Creation is thus dependent on God, not the other way around.

    10. @Bill

      Sir, I don't think you're correct. If God has to necessarily will something, then He depends on the act of willing it. And since He is unchanging, He cannot will something within His essence, nor can He will anything in time because it would implicate that He is not timeless. So He must always be willing something. But if He is willing it, then it must be willed external to Himself, thus making Him dependent on the existence of externality in order for that thing to exist. So in a way, He becomes contingent on it.

      Also, yes, God HAS TO tell the truth, because according to classical theism, He is identical to truth for He literally grounds the existence of all, and He cannot go against His essence. And God has to stay in existence if He is beyond time and unchanging.

      Yes, creation is dependent on God, but then if God is creating anything, then He is also dependent. So you either get a contradiction, or you get that He cannot create anything, and thus is neither a creator, nor is omnipotent. And if He is absolutely Divinely Simple, and unchanging, then He cannot perceive us or anything here that is changing, and thus, He is not all-knowing, etc. After you exhaust all the attributes, you realize that there is just no way that the Necessary Existence can be God. It is closer to Buddhist nibbana than a Judeo-Christian-Islamic-Samaritan-etc. "God".

      The issue ultimately is that there is no way to prove via contradiction that it is NECESSARY for the Necessary Existent to posses attributes such as being all-powerful, all-knowing, creator, etc. Feser's views are unsatisfactory IMO and don't deal with emergent effects properly. You can perfectly have a Necessary Existence, which has no thoughts, knowledge, power or anything like that, just a lifeless innanimate "thing" which grounds the existence of everything else, and the universe has always been cycling over and over, ad infinitum. There is no contradiction or difficulty here, and no need for a God of Classical Theism, and as a matter of fact, even Neo-Classical Theism (Theistic Personalism) as well as Open Theism (e.g. all-good God) and Process Theology and Panentheism. All of them fail ultimately. The only one that might work is some type of pantheism or process pantheism/panentheism but I'm getting besides the point here now I think.

      I hope you understand what I'm trying to say. I feel like I'm repeating myself and that people are not listening to me. I wish I could post a drawing of what I have in my head or something like that. I made distinctions in my replies between different kinds of necessity (e.g. necessary via another to borrow the Avicennist terminology), but it still doesn't work. Ultimately, there has to exist some externality for God to create anything, and thus, it has to be explained, not to mention that no matter how much you spin it, its neither necessary (causing contradiction if its otherwise) nor non-dependent.

      Cheers. If you disagree, please clarify.

    11. AnonymousFebruary 21, 2024 at 6:53 PM,

      If God has to necessarily will something, then He depends on the act of willing it.

      Not necessarily.
      ST:First Part:Question 19:Article 3.

      But you should read the entire section of Question 19 since it addresses other things such as the question of whether God has free will or not.

    12. @Anonymous who posted Feb 20 at 9:39 AM:

      PLEASE help us out by creating a uniquely named profile! (:

      I am sorry to confess that I don't understand the import of this of yours: "Is "God exists" not a proposition though? Can it not apply to it?"

      Of course "God exists" is a proposition. Can what not apply to it? So far I am supposing that on his theory of denotation, Russell would say that this proposition is false. On that count, he would deny that it is a necessary truth, since he'd deny it is a truth. An example of a necessary truth that Russell would accept might be "an even number is a number that is divisible by two without remainder." I am not sure what we're debating at this point, to be honest.

      As to applying the term "a being" to God, as I pointed out in another comment, even Prof. Feser uses "a being" in his OP, viz. "existence of a divine necessary being."

      Russell would want to deny that we legitimately apply "necessary" to beings.

    13. AnonymousFebruary 21, 2024 at 6:53 PM,

      The Summa Contra Gentiles also addresses your concern with an explanation about the relationship of absolute necessity (God wills goodness) and suppostional necessity (creation is not necessary for the goodness God wills) in Chapter 83:

      Chapter 84 is worth reading too since it explains why God cannot do things contrary to His being such as your concept of God's own will making God a contingent being.

    14. @ficino4ml

      Okay, I will make an account later (not right now). Refer to me as J.G. for now, though my account name won't be J.G. If you reply to me the next time, I will make a proper account.

      Now, I'm confused about what Russell is saying here, is his example of a necessary truth literally not just a Kantian A priori Analytic statement? What is supposed to be innovative here, the idea that necessity can only apply to apriori analytic statements which are in a way, tautological-definitional?

      I think the argument here is that I would contest the claim that "a being" cannot be necessary. Necessity can apply to the proposition "God 'exists'". Prof. Feser as far as I understand uses "being" in an Aristotelian-Thomistic analogical sense, not in the concrete sense. Its just a term we use for simplicity and to "approach" the proper view of God, sort of like in calculus if you get what I mean. Its an approximation.

      Anyway, if we understand that Unconditioned Reality, which isn't a proper "being" is necessary for otherwise it would be self refuting, then obviously it becomes necessary. This would I guess be true in both epistemic and ontological sense. As a matter of fact, I don't think you've articulated as to why Russell thinks we can't apply necessity to the Unconditioned Reality when there is no way it could not not exist. You've just stated that Russell thinks we can't apply necessity to it, which is sort of self-refuting IMO.

      I wanted to say something else, but I forgot now. Doesn't matter, I think this gets to the gist of it.

      Please let me know if you think I'm wrong. Thx.

    15. @bmiller

      I don't really get it though. How does it not follow that God willing is not necessary according to Aquinas, if God is:

      a) Unitary/Simple/Without distinction
      b) Will has to come internally from Him
      c) He is timeless, and hence, He must've always been unchangingly willing something, for otherwise no creation is possible and necessary since it would all be contingent and He lacks nothing
      d) But evidently, there is contingent creation
      e) Not to mention that if He is willing something he needs two things: externality to Himself, which is almost never explained how it arises or exists, since if He cannot change anything in His eternal and perfect and unchanging essence, no any creation must take place outside of Himself (though keep in mind that if the universe is not the first creation, then creation is occurring without time, which needs elaboration in and of itself), but He is the only that exists so its a bit difficult to conceptualize this. I forgot what the second one was as I was writing this comment but I hope you get what I'm saying

    16. ficino4ml

      Russell would want to deny that we legitimately apply "necessary" to beings.

      Maybe it would help if you explained why anyone should accept Russell's argument. Just telling us Russell says so doesn't shed any light on why he said so.

    17. AnonymousFebruary 22, 2024 at 9:01 PM

      Aquinas does not say that God's willing is not necessary. He says there are 2 types of necessity, absolute and suppositional. He gives the example that Socrates is a man absolutely necessarily because it is definitionally so. Socrates can either sit or not sit so:

      In this way it is not necessary that Socrates sits: wherefore it is not necessary absolutely, though it may be so by supposition; for, granted that he is sitting, he must necessarily sit, as long as he is sitting.

      So God wills creation necessarily as long as He is willing it (suppositionally, not absolutely) like Socrates necessarily sits as long as he is sitting. If Socrates sits eternally it is still only suppositionally necessary that he sits.

      I think a Thomist could agree with most of your bullets* but I don't understand your e)

      Would you like a reference from the ST or SCG regarding how God is different from His creation?

      *b) seems muddled since God and His will are not distinct as your a) points out.

    18. @bmiller: I would say that in the debate with Copleston, Copleston was the one developing arguments, because he was the one making the assertive claim that God exists: sc. the Argument from Contingency (Metaphysical Argument) and the Moral Argument. Russell did not defend the thesis, God Does Not Exist. His replies to Copleston were to give reasons why he was not convinced by either of Copleston's arguments.

      In his replies, Russell makes explicit and also tacit appeal to positions he had argued for elsewhere. For example, when he says a subject named can never be said significantly to exist, only a subject described, this trades on Russell's theory of definite description. He doesn't argue for it in this debate. Similarly when he refers to the ontological argument. As I quoted earlier, when R says that necessary applies to propositions not beings, again he speaks from work in logic that he did previously.

      A reason to start inquiring might be that there are big theory gains made since what is now standard predicate logic (I understand that there is more than one "logic" since the Principia Mathematica) has been used. In that endeavor, existence is not a first-order predicate. That position in turn was argued by Kant in other terminology. If you're going to maintain, on the other hand, that it is meaningful to talk of beings that do or do not make an act of existence, etc., beings, some of which have existence and others of which do not have existence, then it's not obvious how you're going to standardize claims about them w/ logical notation. Then it will not be obvious how to establish that they're true. Something like "There exists an x such that x is F and x exists" is, as far as I know, nonsensical.

      But perhaps there is a branch of modal logic today that avoids the limitations of the Aristotelian/Porphyrian syllogistic AND provides an accepted way of applying "existence," "necessity," and such to what we've been calling beings. If there is I'd like to know about it.

      I am not an expert on Russell, and I'm not a professional philosopher. I am a Classicist. So I have much to learn about various branches of philosophy.

    19. ficino4ml,

      I'm not a professional philosopher either but that doesn't mean we can't understand and discuss their ideas.

      If Russell didn't explain his reasoning in his debate with Copleston and no one is willing to explain why we should reasonably accept it then I don't see the point of bringing it up other than an historical one.

      We can't evaluate the claim one way or the other unless reasons are given can we? If all you want to do is to point out that different philosophers have different opinions on different matters then I'm thinking pretty much everyone knows that. What we don't know is why.

    20. @J. G. and bmiller:

      As far as I can see, the closest Russell came to giving Copleston an argument for why "necessary" applies to propositions/propositional functions and not to beings/things is this:
      "I will say that what you have been saying brings us back, it seems to me, to the ontological argument that there is a being whose essence involves existence, so that his existence is analytic. That seems to me to be impossible, and it raises, of course, the question what one means by existence, and as to this, I think a subject named can never be significantly said to exist but only a subject described. And that existence, in fact, quite definitely is not a predicate."

      transcript of debate:

      As I said in an earlier post, when Russell says that an object named cannot be significantly said to exist, only an object described, he relied on his earlier theory of definite descriptions. A little more on that below.

      As I understand Russell about "necessary," he is understanding "necessary being" as a thing/being whose mode of existing is logically necessary. It can't be the case under any conditions that this being would not exist. But as I understand Russell, he thinks that Copleston, following long tradition, is basing logical necessity on ontological necessity: the necessary being cannot not exist under any conditions because its very essence includes its existence. Put another way, it will exist in all possible worlds. (Russell already had a doctrine of necessity = "in all possible worlds" in 1912 in The Problems of Philosophy ... so maybe even earlier.) And for Russell, this brings back the Ontological Argument. So the argument against applying "necessary" to beings/things rides on or collapses into an argument against the Ontological Argument. I think. This is just my take as I read the transcript of the debate.

      I don't have any writings of Russell where he makes a long argument against the Ontological Arg. In his History of Western Philosophy, he says that Leibniz' formulation of it was the best in history. I take the Ont. Arg. as read. Against it R said:
      1. it relies on a distinction between essence and existence. "Kant countered the argument by maintaining that 'existence' is not a predicate." I take Kant's position as read. I don't think we can prove, in a strict sense, that existence is not a predicate, but the whole of Frege/Russell standard modern predicate logic requires that existence not be a first-order predicate like "horse" or "in a field." And that's implicit in Kant, I think, who urges that if we say we have 630 (I think it was) Pfennigs, we don't get any more if we add that they exist. Once you enumerate all the coin bundle's properties, to say it exists just is to say that all those properties are instantiated at least once.

      So if Kant was right that existence is not a predicate, or "perfection" as he called it following Leibniz, then it is senseless to talk about things' having existence, whether necessary or contingent existence. It's senseless to try to add existence into an essence or make them identical.

      If one wants to say, no, it's not senseless, then I think with Russell that such a person will be falling back on Aristotelian subject-predicate logic. I learned in college that the Aristotle/Porphyry-based syllogistic is inadequate. A colleague says that it can ground absurdities that follow from the Square of Opposition. Dumb stuff can come out of fourth-figure modes of the syllogism, like "some mammals are animals". You have to say that the syllogistic assumes that there are no empty classes, but that limits the old subject-predicate logic. As far as I know. Brandon may want to make corrections.

    21. more @J. G. and bmiller:
      As for Russell's theory of definite descriptions, it's very complex. I have waded through his paper "On Denoting" several times. His goal was to get around the problem of standardizing sentences with names in logical notation, because predicate logic takes assertive and negative existential sentences as saying that the conjunction of n predicates is instantiated or is not instantiated at least once. But you can't do that with a name. You have to replace it with a description so that you get predicates to quantify, like "author of Waverly" instead of Sir Walter Scott.

      The result is that certain puzzles, esp. about refernce to nonexistents, can be solved. R's famous example was "the present king of France is bald." Some said that this sentence is meaningless because its terms don't refer - there is nothing in existence that they pick out. But if you reword it Russell style, you see that it's simply false to say that "there exists an x such that it is king of France and bald..." A logic that can show a sentence as false rather than stop at declaring it meaningless is a more powerful logic.

      Russell wanted to do the same thing with names.

      I think his reason for saying that the theory of definite descriptions helps refute the Ontological Argument may have been that he thought the Ont Arg requires us to use a name, that God is a name. But I can't establish from the text of Russell's History that he had that motive.

      Sorry, can't write any more. And this is way too long already. Note: Russell of course says that Aquinas does not accept the Ont Arg. But he does disagree with Aquinas' tendency to treat existence as a perfection. Anthony Kenny does the same in his book on Aquinas on Being. I can't go into Kenny's arguments, have to be upstairs!

    22. Adding re "author of Waverley": I don't see Russell in "On Denoting" confronting the question whether "Waverley" too must be translated into a definite description. In some comments on that paper in later literature it looks as though that's a move Russell could make; it's not clear to me whether he needs it.

      Here is a comment by Russell from his My Philosophical Development on some things he thinks he argued successfully in "On Denoting":

      "The theory also threw light upon what is meant by ‘existence’. ‘The author of Waverley exists’ means ‘there is a value of c for which the propositional function “x wrote Waverley” is always equivalent to “x is c” is true’. Existence in this sense can only be asserted of a description and, when analysed, is found to be a case of a propositional function being true of at least one value of the variable. We can say ‘the author of Waverley exists’ and we can say ‘Scott is the author of Waverley’, but ‘Scott exists’ is bad grammar. It can, at best, be interpreted as meaning, ‘the person named “Scott” exists’, but ‘the person named “Scott”’ is a description, not a name. Whenever a name is properly used as a name it is bad grammar to say ‘that exists’"

      OK enough about Russell.

    23. ficino4ml,

      I can't tell. Is there some argument you wish to put forward and defend?

      BTW, have you ever read any of Professor Feser's books? Or even searched this blog to see if he has addressed some of the topics you've read about?

    24. @bmiller: I predicted your response almost to a word.

      Earlier you wrote: "I'm not a professional philosopher either but that doesn't mean we can't understand and discuss their ideas.

      If Russell didn't explain his reasoning in his debate with Copleston and no one is willing to explain why we should reasonably accept it then I don't see the point of bringing it up other than an historical one."

      So I "discuss their ideas" and offer explanations of Russell's reasoning.

      But now that's not good enough. Rather than discussion, as you invited, you consider my long explanations no longer to the point. Rather, you shift the point and say you want something else. I am going to consider you on "ignore."

    25. ficino4ml,

      I read your entire piece. As far as I can tell you did not explain his reasoning in the debate. You made some guesses but perhaps you actually did intend to give us a definite answer and I missed it. If so please tell me what it is. Then I assumed you would want to defend it. Did I get that wrong too?

      I asked you if you read any of Professor Feser's books or even did a search. I'm inclined to believe the answer is "no" since you decided not to answer. It seems that question hit a nerve.

    26. I have read and reread much of Feser's work, including the books you mention. Theory embedded.

      I violated my own statement, so no more, and nothing more to guesses about nerves.

    27. ficino4ml,

      I wonder what you expected as a response to, these in order:

      God does not exist is a nonsense statement.
      Because Kant says its senseless?
      Square of Opposition means all syllogisms are nonsense so Kant is right?
      Russell agrees with Aquinas "the Ont Arg." is wrong (which "the Ont Arg."?). He doesn't like the Thomist term "perfection" and neither does Kenny.
      It comes down to Bad grammar.

      There doesn't seem to be any consistent theme or even consistency regarding which Ontological Argument you are talking about, so there is no where to start any evaluation. Is it just bad grammar?

      Dr. Feser's book Aquinas addresses Kenny's treatment of the Five Ways. That's why I asked if you'd read books like Aquinas. It also seems to me that there have been discussions in this blog regarding predictate logic and it's relationship to metaphysics. Recently in fact. So I wondered if that was somehow one of the arguments you thought needed to be addressed. If so, it has been and the search box at the upper left will help you find where it was discussed.

      Thanks for answering the question though.

    28. Anonymous writes:

      If God has to necessarily will something, then He depends on the act of willing it. And since He is unchanging, He cannot will something within His essence, nor can He will anything in time because it would implicate that He is not timeless. So He must always be willing something.

      Sorry for the delay. Other matters were pressing. Now, there’s nothing objectionable about your first sentence. God is the sufficient explanation of Himself. The act of God is God. There is no bifurcation between act and will (or between God and act, for that matter). Our saying that God “depends” on His necessity is a human convention. God simply is because he exists necessarily. Your next two sentences are incoherent. First, He cannot will, then He must always be willing something. Which is it? Since God is not affected by time, the fact that His will unfolds in time has no bearing on His transcendence. Since God exists eternally, His act of willing is eternal as well.

      But if He is willing it, then it must be willed external to Himself, thus making Him dependent on the existence of externality in order for that thing to exist. So in a way, He becomes contingent on it.

      Not at all. You are presupposing univocal causal agency. Mutual relations occur within the same order of being, but God is not correlative to the world. Many critics miss this because they unwittingly project finite causality onto God. Creation does not move God into a new state of existence because He is Pure Existence. Every perfection in creation is already in Him transcendently, so it is impossible that anything new could accrue to Him. God does not inhabit the same ontological order as the universe. If you’re familiar with the Thomistic causal argument, an acting cause qua acting cause does not receive anything (because it is already actual—not in potency to the immediate effect). The agent doesn’t receive anything qua agency but as a patient. One’s feet causing an impression on sand (while the sand causes granule impressions on the feet) shows a reciprocal causality or correlative change, which is true if the agent is also a patient. If something is Pure Act, there is no univocal causal agency, which means that the Agent (God) qua agency can produce an effect without change (because the principle of passivity is absent in act or cause). Thus, it is impossible for the act of creation to entail dependence given the Agent’s unqualified actuality.

      And if He is absolutely Divinely Simple, and unchanging, then He cannot perceive us or anything here that is changing, and thus, He is not all-knowing, etc.

      I think you need to study what being simpliciter (unqualified existence) means (and I’m not being snarky). If such a being exists, and arguments from motion and contingency (among others) show this to be the case, then there is no state of reality ad extra that isn’t in The Being eminently. Indeed, that would be impossible. Your statement is question-begging because you haven’t shown why God’s knowledge of changeable man affects His omniscience. God knows from eternity everything that man will do, so there’s nothing taking Him by surprise.

      I hope you understand what I'm trying to say. I feel like I'm repeating myself and that people are not listening to me.

      Part of the problem is your critiquing a system before fully understanding it. You’re bringing some presuppositions to the table which skews your understanding of the underlying metaphysics of Thomistic arguments and positions. Somebody said something to the effect that one should understand an opponent’s argument better than he does. Thomism is complex, and it’s easy to get tripped up.

  10. Has anyone refuted Russell's reply that necessity is a proper of propositions not things?

    1. It's the classic de re vs. de dicto debate. Not to treat it too flippantly, but the main issue with Russell's reply is that it doesn't actually engage with the theist's arguments. We do not merely stipulate God's necessity but derive, as a conclusion, from the various arguments employed (the most pertinent being cosmological arguments).

      In short, Russell's reply denies the conclusion without having to do the dirty work of denying one of the premises. Now, if he has some positive argument that de re modality is in fact incoherent, or something to that effect, then we at least have a deadlock we need to break. But if it's simply a matter of not having a reason to believe necessity applies to things in addition to propositions, then that is exactly what a successful cosmological argument will prove. Hence one must contest a premise rather than merely balk at the conclusion that God is necessary.

    2. @ccmnxc: "In short, Russell's reply denies the conclusion without having to do the dirty work of denying one of the premises."

      We'll probably disagree about the Russell - Copleston debate as well as about necessity and contingency, lol. I would say that Russell did mean to deny the premise that there are contingent beings. That premise is implicit in the title that Copleston gave to Leibniz's argument, which he summarized. Russell denies that there is any particular meaning in calling other beings (i.e. the ones said to be non-necessary) contingent. He does not try to prove that denial in the debate, but he appeals implicitly to all the work on logic that he and others had done in preceding decades.

    3. Adding to what ccmnxc says, it is obvious that that are contingent beings, since some things come to be and pass away. It would be silly to urge that contingency is not a property of things. Well, from there it is an obvious step to at least ask the question whether there are also non-contingent beings. If the question is one of those "that doesn't make sense" questions, it is not obvious - you would have to demonstrate or explain why it doesn't make sense. And, by the way, do so in a way that doesn't equally invalidate the observation that there are contingent beings.

      Note that it is not sufficient to demonstrate merely that the set of "necessary beings" happens to be an empty set: if the concept of "necessary beings" is a sensible concept but there just doesn't happen to actually be any of them, then Russell's comment is still wrong. (Not that I imagine it is any easier to actually prove that the set of necessary beings is an empty set.)

      A further issue would require getting into what qualifies as "a being" in the proper sense. Admittedly, a truth about a material being (or about a class of beings) doesn't have the same kind existence as the material being: it is not itself material. But it also isn't pure nothingness. Aristotle argued that forms don't properly "exist" except as in instantiated substantial beings, but Plato thought Forms (in some sense) "exist" separately. So, while I am not a platonist, I would demand of Russell an account of how he qualifies "a being" so as to claim his point.

    4. Russell changed his views of necessity and possibility multiple times, so it's always hard to say for sure what his view is at any given time, but on the view most associated with him, and the view he was probably assuming in the Copleston debate, necessity is not a property of propositions but of propositional functions and therefore only apply to propositions in a looser sense of 'apply' than 'being a property of'. The distinction is significant, since Russell is actually famous for arguing that the property-of-propositions view is wrong -- as he puts it somewhere, propositions are only true or false, not necessary or possible.

      In any case, Copleston's response at the time, that such a conclusion would depend on which logical system you were using and which metaphysical theory you were using to interpret it, is entirely adequate and has only become more cogent over the years. When the debate happened, Carnap had only recently been laying the first crude groundwork for what eventually became the explosion of advances in modal logic in the second half of the twentieth century. In the usual sense of 'property' used at the time (and the most common sense since), properties are associated with predicates, but in almost all modal logic since Russell's days, modalities are not associated with predicates but with logical operators. Nothing requires, however, that logical operators only apply to propositions; the rise of computing showed us that very clearly, because you can use the same logical operators whether you interpret them as applying to propositions or to bits. This doesn't cease to be true for modal logical operators, which is why you can find them applied to imperatives (modal imperatival logic), actions and action-types (logics of action and certain deontic logics), events, doxastic states, regions of space (certain representations of topology in modal logic), temporal intervals, ordered pairs, parts and wholes (mereology as a two-place modal logic), arguments (some modal representations of validity), and the like. Many of these are quite niche and not in general use, but they exist. Thus both Russell's semantic view involving propositional functions and the property-of-propositions view seem to be dated, based on views that have been behind the logical state of the art for at least forty years now. This of course is not a refutation; things go in and out of fashion even in logic. But there doesn't seem to be any obvious way to make either the view usually associated with Russell or the property-of-propositions view consistent with much of the actual work in modal logic that began to take off literally in the decade after the Russell-Copleston debate.

      Of course, one could take 'necessity is a property of propositions' much more loosely and just mean something like, 'The best or most appropriate logic for discussing metaphysical questions about necessity and possibility is modalized propositional logic'; but even if that's true, one can simply define other kinds of necessity in an extended sense by saying something like," 'X is a necessary being' just means BOX(There is an X) for such-and-such relevant interpretation of the Box operator."

    5. Fr Copleston's last book "Memoirs," was published almost 50 years after he debated Russell. Looking back on that debate, he said he "a formally valid argument for the existence of God is doubtless possible, given certain premises. But as the premises may be challenged , it is important to state and discuss them."

      He didn't think the debate went well for either him or Russell. He said he liked his debates with A.J. Ayer much better and considered Ayer ( "Freddy," as he called him) a friend.

    6. @ one of the Anonymi: I said that Copleston and Russell both used the word "being" when talking about God, but I should have noted also that Prof. Feser uses the word "being" (sc. "a divine necessary being") when discussing Contingency arguments in his OP.

      @Brandon: thanks for this about Russell. As you know, Russell did say "propositions" not "propositional functions" in his debate:

      "The word "necessary," I should maintain, can only be applied significantly to propositions. And, in fact, only to such as are analytic -- that is to say -- such as it is self-contradictory to deny."

      Do you have references to works that flesh out how Russell, as you say he did, really meant to say that "necessary" can only be applied significantly to propositional functions? At first glance I find that strange, since the variables in a propositional function aren't bound by values. Tx.

    7. I was the Anon of Feb. 20 @8:03 am, and it was an accident, I simply forgot to change my comment from Anonymous to a name. I am not the other anonymi.

      We'll probably disagree about the Russell - Copleston debate as well as about necessity and contingency, lol. I would say that Russell did mean to deny the premise that there are contingent beings. That premise is implicit in the title that Copleston gave to Leibniz's argument, which he summarized. Russell denies that there is any particular meaning in calling other beings (i.e. the ones said to be non-necessary) contingent... but he appeals implicitly to all the work on logic that he and others had done

      I can understand the reluctance to grant the premise that there are contingent beings, but I would offer that reluctance doesn't amount to a position without an argument, and without knowing which part of the "work done" on logic explicitly (rather than just the implicit of "the body of work") one cannot judge its validity. And one might well reject some of it as being unfounded, or at least resting on premises that are in dispute, so that his conclusion here would also be in dispute.

      For instance, while it was a commonality in the late 19th and 20th century to believe that physics had "established" determinism as a universal reality, so that the very possibility of "truly" contingent events (or beings) is denied, later physics is no longer quite so absolute about that. And while eliminative reductionism might reject the distinctness of apparent "beings" that are in the common experience of people, reducing them to mere appearances from the only "real" beings as quarks or strings (take your pick), such eliminative reductionism is hardly the prevailing philosophical view, and other views make room for the possibility of more complex "beings". As a consequence, it is not facile to require an account of whether there are distinct beings and whether any can be considered contingent, but it IS facile to simply dismiss the category of contingent beings without making the arguments AND accepting the limits (e.g. that some of the premises are in dispute).

    8. As you know, Russell did say "propositions" not "propositional functions" in his debate:

      Yes, and as I said, he almost certainly meant it in a looser sense than the claim that necessity is a property of propositions; notice that he says not that it is a property of propositions but that it 'is applied' to the latter, which is also true on the propositional function account, on which 'necessity' applies to propositions in a derivative and looser sense.

      Do you have references to works that flesh out how Russell, as you say he did, really meant to say that "necessary" can only be applied significantly to propositional functions? At first glance I find that strange, since the variables in a propositional function aren't bound by values. Tx.

      Almost any discussion of Russell on the subject of modality will discuss it; it's one of his more famous positions, and is likely a contributor to his resistance to early advances in modal logic. You might start with Jan Dejnozka's *Bertrand Russell on Modality and Logical Relevance*.

      It's precisely the fact that propositional functions involve variables that are, so to speak, not locked down -- which can be related to all values -- that I think makes them attractive to Russell when he is trying to give an account of modality. He thinks 'necessity' can't add anything to saying that a proposition is true, but that you can make sense of saying that a propositional function is necessary, when plugging in any value will still get you a true proposition.

    9. @Brandon: thank you for these explanations and for the reference to Dejnozka. I have started to poke through his book. So many deadlines in the way, aargh.
      I do see that in his 1912 paper, "On the Notion of Cause," Russell wrote that "A propositional function is necessary when all its values are true," and that "A proposition is necessary with respect to a given constituent when it is the value, with that constituent as argument, of a necessary propositional function..."

  11. I find Leibniz' formulation the most persuasive. Denying the PSR is just too high of a metaphysical cost for me. I don't think any rational person actually *can* deny the PSR because to do so is to deny rationality.

    It really seems to me that the people who insist on denying the PSR are doing so, not for any principled reason, but because they can't find any other way to deny the conclusion of the argument, because it's so obvious that denial of the PSR leads to a metaphysical picture that is literally ridiculous.

  12. I've heard it said that Aquinas actually gives another version of the contingency argument in the Summa Contra Gentiles. That version doesn't rely on the notions of time, at least that's what Thomists like White have claimed. Nevertheless, one could always re-formulate Aquinas' argument, regardless.

    Certainly, it would seem that Aquinas wouldn't have a problem with saying that there can't be an infinite essentially-ordered series of perishable/contingent things existing here and now. To my mind, that makes the argument simpler and similar to the 1st and 2nd ways.

    1. Yes, you're correct. In SCG 1.15, Aquinas offers a simpler version of the argument (see The Third Way: Another Take On The Argument). Since Aquinas assumes the eternality of the universe arguendo, he is using "possible" in the concurrent sense, so that if everything is possible, then nothing would be. The "at one time" in Latin refers not to progression or digression; it is merely a moment. If there is no ground for possible being, then there would be nothing in a moment (there would be a time where there is nothing). Hence, not everything can be possible.

  13. I really enjoyed the five proofs book and all the proofs, and this discussion. This post was helpful too. I was trying to compose a simple paragraph that integrated all of the five proofs (more as a memory aid than anything else, as I find keeping the five proofs in my head correctly organised hard). I came up with this, not sure it works but there you go, also it would be good to get the Augustinian proof from universals integrated into it also, but I could not find a way to do so as it seemed the most different from all the others. If anyone wishes to respond I would appreciate just a comment on the logic itself within the accepted Thomistic framework and the Five Proofs book, i.e. I do not want comments that state science disproved this whole enterprise and there are just brute facts and the void, or metaphysics itself is worthless. Just the pure logic within the framework and if I am integrating like with like correctly or not.

    1. Beings are everything that has existence (that they are) in actuality and every being has its own unique essence (what they are). Metaphysically there can only be two types of being (note this is metaphysical, neither being has been proven to exist yet).
    2. Contingent beings: Contingent beings might never have existed, which means their existence in actuality is not necessary, and therefore their existence in actuality is not their essence, which makes their existence and essence distinct. Thus they have at least two parts and are composite. Because their essence is not actuality, it is therefore only a potency. Thus their essence-as-potency needs to be actualised into existence, which can only be done by something already in actual existence, and therefore contingent beings are dependent on already actual beings for their existence.
    3. A Necessary being: A necessary being is a being which must exist, and therefore their essence is existence. Existence is actuality and essence is potency, thus their essence is actuality, thus they are purely actual without potency. The actualisation of potency is the measure of material objects, change, time, power (as active potency) and goodness. Thus something purely actual is immaterial, unchanging, timeless, all powerful, and perfectly good. This is what we call God. Thus the necessary being is God.
    4. Contingent beings exist (self-evident, assume various refutations of sense scepticism, evil demons, etc), and per (2) their existence can only be actualised by another being.
    5. If that Being were also Contingent, then it would require another Contingent Being to Actualise its own Existence, and so on, leading to an infinite regress, which is metaphysically impossible.
    6. There are only two types of Being. If a series of Contingent Beings leads to an Infinite Regress, then the Regress can only be broken by a Necessary Being. Thus the Necessary Being exists and therefore God Exists.

  14. You said: "they all have at least a minimal empirical component insofar as they appeal to the contingent things we know through experience and argue from their existence to that of a necessary being."

    But this isn't true of Avicenna's arg. - at least not in the way it is of the other two. For their arguments are based on there being contingent existents; they start from that fact. Avicenna's is based merely on their being some existent. And this latter is not an empirical premise (and so not an 'empirical component' of the argument).

  15. WCB

    Augustine in his confessions, Book 11, On Time, proposed the idea that God is outside of time. For God, God exists in an eternal now. For God, there is no past, no future, no now separate from these. All is now. m And this seems to have become a dogmatic claim about God and time.

    Things have always been as they are, and cannot be otherwise. It seems that if we take Augustine seriously about time, then nothing is contingent and everything is necessary. Always has been, always will be.


  16. Dr Feser,
    I saw your X about Prof. Gorman's forthcoming book on metaphysics and I am sure it will be an excellent scholarly book ( his C.V, is very impressive) ,but no one makes metaphysics understandable to the average reader as you do.

  17. Regarding PSR it’s interesting that in Aquinas and Leibniz the foundation of all contingent things (truths) is another contingent thing (truth), which is “All contingent things are created by the free act of God”. Now I think you could take this to be a brute fact (or truth or thing). Leibniz of course doesn’t take it as brute but nevertheless sees it as essential that contingent truths are grounded in a fundamental contingent truth, only that way can the whole be contingent. What makes this pertinent here is, despite having a grounding contingent fact it’s still required that something is necessary. A brute fact just doesn’t help in that regard