140 - By All Means Necessary: Avicenna on God

Posted on

Avicenna’s proof of the Necessary Existent is ingenious and influential; but does it amount to a proof of God’s existence?

download-icon

Themes:

Further Reading

• P. Adamson, “On Knowledge of Particulars,” Proceedings of the Aristotelian Society 105 (2005), 273-294.

• P. Adamson, “From the Necessary Existent to God,” in P. Adamson (ed.), Interpreting Avicenna (Cambridge: 2013).

• A. Bertolacci, “Avicenna and Averroes on the Proof of God’s Existence and the Subject-Matter of Metaphysics,” Medioevo 32 (2007), 61-98.

• M.E. Marmura, “Some Aspects of Avicenna’s Theory of God’s Knowledge of Particulars,” Journal of the Americal Oriental Society 82 (1962) 299-312.

• M.E. Marmura, “Avicenna’s Proof from Contingency for God’s Existence in the Metaphysics of the Shifā’,” Medieval Studies 42 (1980), 337-52.

• T. Mayer, “Avicenna’s Burhān al-Siddiqīn,” Journal of Islamic Studies 12 (2001), 18-39.

• M.S. Zarepour, Necessary Existence and Monotheism: an Avicennian Account of the Islamic Conception of Divine Unity (Cambridge: 2022).

Comments

Khalil Habib on 4 August 2013

What a great podcast!

Peter, this is one of the best podcasts thus far. They are all excellent, but this one is really impressive, as explaining Avicenna's "proof" of God's existence is no easy task. I am just stunned at how much you manage to pack in a short podcast. Question: do you plan to connect future podcasts on Averroes to this particular aspect of Avicenna? Thanks and enjoy your holiday!

khalil

In reply to by Khalil Habib

Peter Adamson on 4 August 2013

Averroes on Avicenna

Hi Khalil,

Thanks, glad you enjoyed the episode! I think the Averroes episodes will not get much into his response to Avicenna; I am still writing them but I think it will be one on the Decisive Treatise, one on his theory of unity of the intellect, and then an interview (Richard C. Taylor). But I'll talk in general terms about his views on Avicenna and I guess I do need to discuss the fact that he thought God should be proved via physics rather than metaphysics, as Avicenna attempts to do.

Peter

Davlat on 4 August 2013

regarding an objection

hi prof. Adamson, enjoyed the episode. thanks!

i have a couple of comments to make. regarding the objection you raised, it seems to me that Avicenna's proof *itself* rules out the possibility of the universe being the necessary existent, which i gathered you don't seem to think, suggesting that it may commit the fallacy of composition here. i also don't think that when Avicenna says "in a way this is what was sought" he is referring to the universe being the necessary existent, as you suggest in the episode, but rather to the possibility that a certain part of it is so, which he then rules out. but of course this depends on what one takes the term 'universe' to mean in this context. in my view, it would be more reasonable, and consistent with the text of the Najat, to interpret 'the universe' as meaning the aggregate or sum of all contingent things and not just some contingent thing within that aggregate.

the fallacy of composition objection is i think easily answered: the purportedly necessary through itself aggregate, Avicenna says, subsists through its parts, all of which are possible in themselves. but that's a contradiction. therefore, etc. at this point, Avicenna continues, the cause of the aggregate will be either (a) external or (b) internal to the totality/aggregate (i noticed that you glossed over this distinction and what follows on it). if (b), it will either be (b.1) a necessarily existing internal cause or (b.2) a possibly existing internal cause. he then rules out (b.1), and focuses on (b.2), which is where the phrase "in a way this is what was sought" appears. but surely this internal possibly existent cause cannot be considered as 'the universe' at this stage of the proof (which you seemed to suggest in the episode). he only suggests that 'this is in a way what he seeks' only insofar as on this option this purported possible internal cause is sufficient to necessitate, to cause, not only the existence of every other contingent within the aggregate, but also its own existence. but this Avicenna rules out as impossible for two reasons: first, no contingent can cause it's own existence; and second, even if we assume it can do so and hence be necessary, that will contradict the initial assumption of it being merely possible. Avicenna had earlier ruled out disjunct (b.1) above on the same grounds.

In reply to by Davlat

Peter Adamson on 4 August 2013

The proof

Hi Davlat,

Thanks for that very detailed comment! I'd have to double-check where that "in a way..." remark comes, you might be right. But in a sense it doesn't matter: whether the opponent proposes that the whole set of contingents (that's all I mean by "the universe") or a subpart of the set is necessary, Avicenna can declare victory. Of course you're right that he has more than one way to defeat the idea that the universe, or for that matter a part of it, is necessary - for instance he will later argue that a necessary existent has to be immaterial, precisely because it can have no parts. So, both answers will probably work, as far as he's concerned. But I thought it was important to stress that at this stage of the argument he is only trying to show there is an NE, without yet having to argue about its nature or characteristics.

And you're also right that I deliberately skipped some of the details about internal causation wihtin the set, and so on, as being too complicated to explain in a podcast; I actually started to write it with fuller detail and realized it was going to take too long and be too involved. (There's not only what you mentioned, but also issues about circular causation for instance.) It helps if you can draw on a chalkboard while explaining it, too!

Thanks,

Peter

In reply to by Peter Adamson

Davlat on 4 August 2013

the proof

right right, i agree. i just wanted to make sure that it's clear that he refutes the idea that the universe is the necessary existent within, or in the course of, the proof itself and not by some other additional proof (in the way e.g., he establishes the Divine Attributes). i also definitely agree about making use of a chalkboard when explaining it, at least as long as we're embodied! ;)

wish you well on your time off.

Rhys W. Roark on 4 August 2013

By All Means Necessary

Dear Peter,

Great show! With its interspersed and wonderful humor, a show that upholds that Horatian maxim of both delighting and instructing!

First I was intrigued by the argument for a necessary existent that must be singular (i.e., thus absolute) through the co-eternal examples of Charles and Buster. This got me thinking of Christian Trinitarianism and how the 1st person and 2nd person of the Trinity both must ultimately express a unity. But I am curious, since I can also see the argument go in Avicenna’s way, did later Muslim polemicists use this line of thinking in attacking the Trinity doctrine? And who, by chance, (by necessity?) did so?

Also, I was interested in the empirical argument for God’s existence by noting the features of the natural world. I am familiar with this line of thought by Aquinas, for example, who rejects the ontological argument in favor of this one, as the latter has a scriptural warrant in Romans 1:20 (“Ever since the creation of the world his eternal power and divine nature, invisible though they are, have been understood and seen through the things he has made.” NRSV trans.).

But I was not aware that it equally had Aristotle’s warrant as well—I know, in a general way, he talks about how things moved need movers, but the way this was described in the show, it had a more expansive sense, similar in resonance with Rom. 1:20 (i.e., there are other things we can also observe that points to an ultimate Unmoved mover / Divine Mind).

Is that expansive sense correct and where in Aristotle do we find this? I presume something in the (Meta)Physics, at least?

Thanks again—have a happy break, with happy viewings of Hiawatha and Buster!

Rhys
OKC, OK

In reply to by Rhys W. Roark

Peter Adamson on 4 August 2013

Proofs for God

Hi Rhys,

I don't know of a text that applies Avicenna's reasoning to the Trinity specifically, but I may come across it as I research the later Eastern tradition for future episodes. Of course Avicenna himself is probably well aware that he's ruled out the Trinity here.

Aristotle and design arguments: a contentious question since there are debates about the extent to which God intends for the universe to be the way it is. But my own view would be that Aristotle simply thinks that the good order of the world depends on there being a single principle -- as he says at the end of Metaphysics XII, quoting Homer ("the rule of many is not good"). That doesn't mean he thinks that god actually designed the world, or anything at all, though.

Thanks!

Peter

Parzival on 5 August 2013

Thank you!

Thank you so much for this wonderful discussion! I've been waiting for this episode the most since you started your podcast. I love this so much.

roman prychidko on 9 August 2013

Avicenna on God

Hi Peter
A couple of points that occur to me.
1. God as a intellect. Can we ever know what this means. In a world were the only real mening is through personifications. You only need to ask Buster or Charlie
2. God as a necessary existance through universally and not as a particular. If a killed a wasp which was in my wine glass God would not know that I have killed that wasp ( too particular). But would he know a wasp is dead because the universal sum total of a wasps has decreased by one. Or is he just flowing though my veins no matter what i do.

Roman

In reply to by roman prychidko

Peter Adamson on 9 August 2013

God's knowledge

I think Avicenna would say that universal knowledge does not include statistical information about how many times the universal in question is instantiated. That is, it is no part of the scientific knowledge of wasps to know that N wasps currently exist. Also N would change from moment to moment, so there is good reason to think He'd better not know this: if He did, He would change in what He knows, as the number of wasps increases and decreases.

Your first question is harder to answer, but I guess Avicenna would just insist that there is knowledge of things, not only via representations, and that even humans have this never mind God.

Thanks,

Peter

ajmal on 9 August 2013

Question

Hello.

Great podcasts, and thank you for the informative and intellectual inputs.

I had a question regarding avicenna's concept of eternity of the universe: if the universe (according to him and aristotle) is eternal, how does the Necessary Existent, if for him is the cause of the universe, causes it when it is supposed to be eternal?

One can object and say (as aristotle might say) the ingredients of the universe was already present, and the Necessary Existent simply caused them to be (the universe), but then who caused those already present ingredients from which the universe was caused?

Regards.

In reply to by ajmal

Peter Adamson on 9 August 2013

Avicenna's God

Spoken like al-Ghazali! Avicenna does though think he can explain how something could be caused without starting to exist. In effect what he wants to say is that the universe depends on God constantly for its existence: God is permanently making it necessary, whereas in itself it is only contingent (and thus neither "deserves" to exist, or not to exist, as Avicenna would put it). So actually, you could argue that Avicenna gives God more causal power than someone who just says that God makes the universe start existing, and then sits back and does nothing thereafter. For Avicenna, God is causing the universe to exist all the time. Of course, he can say this whether or not the universe is eternal; his reason for thinking it is indeed eternal is that everything about God is necessary. Thus if God causes the universe then He necessarily causes the universe, so it can't be that at some point in the past He was not causing the universe.

Does that help?

Peter

In reply to by Peter Adamson

ajmal on 12 August 2013

Question

"Thus if God causes the universe then He necessarily causes the universe, so it can't be that at some point in the past He was not causing the universe."

Hello and thank you for the input.

ALong with those lines, Avicenna must deduce and infer that time (zamaan) is also eternal, which would put God ontologically prior to everything else (in line with quran and Avicvenna as he says the only Necessary Existent is God).

If God is prior to everything else, in atemporal realm, It (God) being the source of necessitating everything necessarily, does not that again make It (God) the Cause of everything (sounds a bit aristotelian I guess)?

Regards, and thank you for your time.

Vogelweide on 12 August 2013

Transcript?

Is there a transcript of this? I have a friend who can't get the audio to play.

In reply to by Vogelweide

Peter Adamson on 14 August 2013

Transcripts

I don't make transcripts available, for a variety of reasons but one is that the scripts will all appear later as a series of books (just finishing off vol.1 which goes up through Aristotle). If your friend is having technical difficulty we might be able to help, though.

Vogelweide on 14 August 2013

Avicenna's Proof of God

Dear Peter,

Can you tell me which texts and passages within have Avicenna's proof of God's existence (esp. ones in English)? I'd greatly appreciate it.

WvdV

In reply to by Vogelweide

Peter Adamson on 14 August 2013

Texts for the proof

There is a version in the Hackett reader on Classical Islamic Philosophy which from memory I think is from the Najat. Also check out the second Marmura article on the podcast page for episode 140, “Avicenna’s Proof from Contingency..."; I think it has translations of relevant texts. The Mayer article has translations from the Isharat (again, this is from memory so might be wrong).

And most importantly there is Marmura's translation of the Metaphysics of the Healing... though the proof famously doesn't really appear there, the classic version is elsewhere for instance the Isharat.

Felix on 15 August 2013

Impressive

Peter,

you have certainly convinced me that Avicenna was a philosopher of the highest order.

Looking forward to future episodes!

Felix

In reply to by Felix

Peter Adamson on 16 August 2013

Mission accomplished

Thanks! Glad you liked him better than the Asharites.

Dave Martin on 17 August 2013

Avicenna on the Necessary Existent

Hi again Peter,

I've been thinking about your last podcast all week, trying to get the arguments straight.

It seems to me that Avicenna’s argument for the necessary existent does not work if you combine it with his argument that rejects there being two such existents (your Buster and Charlie example) on the grounds that it causes absurd consequences. I think, if you run the same argument with only one necessary existent, you still get the same kind of absurdity from the same logic chain.

Forgive my clumsy paraphrasing, but Avicenna’s argument for the necessary existent is that there has to be such an existent because if all things were caused then – following back a long succession of prior things causing subsequent things - there would be nothing to cause the first thing. Thus there must be something that is completely necessary and uncaused, and we’re going to call this the ‘necessary existent’. The only alternative to this is non-existence. But that can’t be, because I think, therefore I am, therefore there is some existent, therefore there is a necessary existent.

However, if I now apply the Charlie and Buster argument, I must be able to show some difference between non existence and the necessary existent. Using Avicenna’s own argument: there has to be some distinguishing or individuating feature that prevents the necessary existent from being identical to non-existence, and if both are uncaused this cannot happen. Non existence, by definition has no cause, and a necessary existent has no cause. So now there are only three possible conclusions:
1) The necessary existent doesn’t exist
2) The necessary existent is identical to non-existence.
3) Even non existence has a cause, in which case the original argument for the necessary existent doesn’t work any more because the chain of prior things causing subsequent things would have to stretch beyond existence itself and then I don’t have a necessary existent as it goes ‘poof!’ in a cloud of non-existent smoke.

As I said, absurd consequences!

I still loved the podcast though. Happy holidays
Dave M

In reply to by Dave Martin

Peter Adamson on 17 August 2013

Non-existence

That's very clever, thanks. So, just to recapitulate to make sure I understand, your objection is that non-existence has no cause, and neither does the NE, thus there is a threat that they be identical. And no further factor can be brought in to explain the difference because of the individuation argument.

I think Avicenna could respond in several ways. Firstly, he could (in fact would) say that non-existence does have a cause: in each case where a possible essence is preponderated not to exist, God is ultimately the cause of its non-existence. So he would block your argument at the first step. I guess this is the response you were anticipating in your third option, but I don't quite understand what you say there. Avicenna, I believe, would say that you can prove God's existence even from the non-existence of a contingent entity: since it neither "deserves" to exist nor not to exist, an NE is needed to explain non-existence just as much as existence.

Another reply he could give, if you don't like that, is that God's "guaranteed existence" already differentiates Him from non-existence. So there is no further individuating feature needed: by virtue of being the NE, He exists, whereas the non-existent or non-existence itself does not exist.

In reply to by Peter Adamson

Dave Martin on 17 August 2013

Non Existence

Peter,
Yes, you have correctly summarised my argument.

However, my 'non existence' is not the non-existence of one contingent thing that has not been preponderated. Rather, what i was proposing was that we should look for a way to differentiate the NE from a complete non existence. Such a non existence could not have a cause because then the cause would have to exist. Therefore in order to be different from non existence, the NE would have to have a feature that differentiates it, and as soon as you get to that point, you get hung on the same argument that makes Charlie and Buster impossible.

If you say, at this point, ah well, existence is the differentiating feature and existence doesn't need a cause, then why do any other possible differentiating features (eg a moustache) need a cause?

The second argument you are putting forward seems to be circular. It seems to be saying that God exists because he has existence that diffentiates him from non existence, which may be true, but isn't a proof. If existence is the feature of difference that you are going to rely on, then following the Buster and Charlie argument, it must have a cause.

Dave M

In reply to by Dave Martin

Peter Adamson on 17 August 2013

More on causation

Hi Dave,

I think you are presupposing that there is no more to the notion of a necessary existent than "something uncaused." If this were the case, you would have a point: absolute non-existence is also uncaused. So, what would make them different?

But (and this was the point of my second response) this is not what Avicenna would understand by the NE. Rather the NE is something that _exists_ without a cause (it is a necessary _existent_). So, unless you can tell me why absolute non-existence exists without a cause, I don't think he has a problem here: it's quite obvious that the two are not identical.

I think you are suggesting in your post that this line of thought is somehow circular or question-begging, but bear in mind that the mere definition of the NE as "that which exists without a cause" does not, according to Avicenna, immediately imply that there is such a thing. He proves that the NE exists with the proof, not as an analytic truth about the NE (the way Anselm does with the ontological argument).

Does that clarify things?

Peter

 

In reply to by Peter Adamson

Dave M on 18 August 2013

More on Causation

Peter,

Ok, I have had a long think about it. You object to my proof on the basis that I have not proved that non-existence exists without a cause, but I think I could run the same argument as I was trying to run yesterday using slightly different terminology if I compared the 'universe with an NE in it' to an 'empty universe', and I think I can prove that an empty universe has no cause. Apologies for the clumsy scientist's terminology, but here goes:

Avicenna uses the grouping of all the contingent things in the universe as the basis of his argument, so let me consider the problem in the same terms (i.e. using groupings).

Let me first take a group of all the things that exist in an empty universe (i.e. the null set).

Does the anything about the set have a cause? Well, there is nothing outside the set as I've included everything in the universe so I can’t find a cause there. There is nothing inside the set, so no luck there either. Therefore, an empty universe has no cause.

Ok, so now I’ve proved that an empty universe has no cause, I can go back and ask Avicenna to show a difference between the empty universe and a universe with an NE, which - as I argued yesterday - leads to the same absurdities as the Charlie and Buster argument (i.e. the cause must be some feature of the NE and that can't be, etc. etc.)

As far as I can see, Avicenna now has two objections he could raise to my proof of nullness having no cause.

1) He could say that the fact that I form a ‘group’ creates ‘something’ in my universe and therefore that my null group is not in fact an empty universe. He might say that the idea of 'universe' has an existence even if its empty. I wouldn’t agree that the act of grouping creates anything which has existence, or that an empty universe has existence. But if he insists, I would respond that in that case, when he makes his own argument about God, Avicenna is making the same mistake. He introduces an uncaused thing by grouping all the contingent things in the universe (either the 'group' or the 'universe' are uncaused) in order to prove that an uncaused thing (i.e. the NE) exists. This cannot be allowed either.

2) He could say that the universe never contains only the NE. Indeed, he could say that the only set that I'm allowed to compare against is the NE plus all the contingent things that he causes. But, it cannot be that the NE needs the contingent things to distinguish it from the null set, because then the difference between NE and the null set would be caused by the contingent things.

Does that work for you?
Dave M

In reply to by Dave M

Peter Adamson on 18 August 2013

Empty universe

Hi Dave,

Thanks, that's a nice next move. So, again to recapitulate, you are shifting ground slightly (or maybe just specifying what you mean) by demanding from Avicenna a way of distinguishing the NE from non-existence, by now understanding non-existence as an empty universe. And then you want to say that the empty universe both exists and has no cause; since I said earlier that the notion of the NE is that it is something that exists and has no cause, you will have him where you wanted him which is that he needs to invoke an additional factor that individuates God from non-existence (i.e. from the empty universe).

Again, very clever! Again, taking up Avicenna's side of the story I think he could respond in two ways, which are not quite the responses you envision I think:

1. The empty universe does not exist. Remember that for him "existence" is something that comes to an essence, either from itself or from another. I don't think the empty universe has an essence, or at least, I guess Avicenna wouldn't even be able to make sense of the proposal that it does. Therefore we have our distinction: God exists, the empty universe doesn't, in fact to talk about it "existing" or not is probably just a category mistake.

2. But suppose that this is wrong and that the empty universe could exist. In that case, it would not be uncaused. Why not? Well, it seems clear that the empty universe is not necessary -- in fact, it is not even actual (since in fact the universe is not empty), so it is either impossible or contingently non-existing. Rather, if counterfactually the empty universe "existed" then it would exist by God making it exist (which He would do by preventing any essence other than Himself from receiving existence, with the result that the universe is empty). Therefore, we have our distinction: God is necessary/uncaused, the empty universe isn't.

It may be that you are imagining the empty universe as something like an independent vessel or container into which God can, if He wishes, place existent entities -- if He creates anything it will no longer be "empty". And then you want to say that this vessel itself exists necessarily. That would be a lot like Razi's notion of absolute place. Obviously though Avicenna would have lots of reasons for rejecting this conception, for instance if you imagine it as spatially extended it would have parts and nothing necessary can have parts.

Your turn!

Peter

In reply to by Peter Adamson

Dave M on 19 August 2013

The Empty Universe Cont.

Hi Peter,
Yes, I think you summarise my point correctly. I am saying there is no possible distinction between the NE and an uncaused nothingness, without destroying the very features that make the NE into the NE. I am essentially trying to force Avicenna into proving that the NE is different from nothing. At the moment, I don't think he can do this without invoking a feature that would be available to use against him in the argument he uses to dismiss the possible existence of two NEs.

The first objection that you put forward doesn't seem to me to work. If the question is 'show me something that proves the NE has existence by showing me it's different from non existence?', I don't think it's sufficient to say 'Ah, well, it exists therefore it's different, therefore it proves that it exists.'

The second objection is much more interesting. It's essentially challenging whether it is possible to conceive of an uncaused nothingness (which I am currently arguing is provided by an empty universe). Indeed the idea that an empty universe might need a cause took me back to the Ash'arite questions about causality. Does Avicenna ever give a comprehensive explanation of what he believes it means for something to cause something else. This seems to be critical to the argument. For example, his system of classifying things seems to involve a 2x2 matrix defining whether they have essence and/or cause. In order to be the NE, you have to be uncaused, so presumably the NE is pure essence, and indistinguishable from that essence. Therefore, the existence of the NE (which is pure essence)suggests that an essence can cause something without anything causing (preponderating) the essence to exist. This suggests a cause doesn't have to have 'existence' in the sense that Avicenna is defining it, only essence, which seems to me to have serious consequences for Avicenna's proof of God. (In fact, it suggests to me that things could be self-preponderating, which is why I think you have to go back and take a very close look at what we mean by causality, before you can decide whether Avicenna's argument works or not.)
Dave M

In reply to by Dave M

Peter Adamson on 19 August 2013

More on the empty universe

Hi Dave,

You're underestimating the first objection, I think. Don't confuse the following two tasks:

1. Explaining what an NE is

2. Showing that the NE exists

The first task is obviously prior to the second since before we can prove the NE exists we need to know what we are talking about. Now, sticking only with the first task, by definition the NE would (if there were such a thing, but we haven't decided that yet) be something that (a) exists and (b) is necessary (which at least means uncaused). As I say, I don't think the empty universe or non-existence qualifies on either score (it is something that is caused not to exist) so I really don't think there is any danger that Avicenna will not be able to distinguish the two. Indeed, there is nothing that is in less danger of being confused with the NE than non-existence, since it shares neither of the NE's two defining characteristics! So to sum up, you can easily show that the NE is not the empty universe without being committed to, or even raising the question of, the existence of the NE.

It is then a further question whether such a thing as the NE (i.e. something that exists, and is necessary) actually exists. This is the second task, and his famous proof (as described in the podcast) is intended to carry out the task. The proof is not what you seem to be assuming it is, namely that we say the NE by definition exists. Again, that would be more like Anselm's proof, it is not what Avicenna is doing.

Regarding the second issue you discuss in your post, there are two ways you can go with an NE: you can either say that it is self-causing, or you can say that it has no cause. Avicenna says the latter (at least I don't know of a passage where he suggests it is self-causing; you do find this sometimes in Neoplatonism and in Latin medieval thought). I think he would back that up by saying that questions of causation only arise for the contingent things: in their case, you need an external cause to explain either existence or non-existence. Thus a definition of causation at least in this context could be: preponderating a contingent essence either to exist, or not to exist.

Peter

In reply to by Peter Adamson

Dave M on 20 August 2013

More on the empty universe

Hi Peter,
OK, I think I'm understanding why you object to my empty universe being uncaused. While I'm not sure about it, I started to look elsewhere for something uncaused I could hang my argument on. Thus, I have been thinking harder about the effect of splitting the two tasks in the way you describe.

If an NE is to be defined as something that a) exists and b) is necessary, and I think I remember things being necessary when they are always true, then surely 'time' must qualify as an NE. It always exists.

This proposition would make a good deal of sense to me, because if I look at all the things in the universe at a point in time (as Avicenna does at the beginning of his argument by asking us to consider them as a group), it seems that I need 'time' to exist to be able to distinguish the necessary from the contingent. Not only would there be no difference between necessary and mere contingent things in a scenario where time was not both existent and flowing, but I think the concept of 'cause' also requires time, because without it nothing can change.

Of course, if time is an NE, I need no other, and Avicenna's subsequent argument (the Buster and Charlie argument) would tell me that I can't have more than one even if I wanted it.

Dave M

In reply to by Dave M

Peter Adamson on 20 August 2013

Time as necessary

Hi Dave,

Firstly I just want to say to anyone else who is following this thread, that you provide a model for how to do philosophy -- keep thinking it through from different angles!

Secondly, I like that a lot. If you don't mind, it again makes me want to mention my man Abu Bakr al-Razi, who would likewise say that time is necessary (well, he said it is "eternal," he doesn't talk about necessity -- but uncaused anyway). I think that's very plausible, and in fact I would say you are getting here to what is for me the fundamental problem with Avicenna's philosophical theology. This is something I also mentioned in the episode: there might be LOADS of necessary things, apart from God. Avicenna wants to try to show that there can be only one necessary thing and that it has all the traditional divine attributes. But if admit that there are other uncaused, existing things -- and time may be a good example, perhaps also things like numbers, logical laws, moral laws, whatever -- then he loses. The difficulty will be, precisely as you said, that the mere concept of "necessary existent" is not sufficient to imply all the divine features, so some other factor will have to be brought in to explain why God is different from time, or the number 7. So I would agree more with you now, than when you were going for the non-existence example.

But perhaps someone else would like to jump in to defend the Shaykh.

Cheerio,

Peter

In reply to by Peter Adamson

Dave M on 20 August 2013

On to the Flying Man

Dear Peter,

Thank you for the compliment. I am waiting now for Podcast 141, because I think the flying man thought experiment (which I know nothing about) must have some profound implications for what we think, or could possibly think, about time and our experience of it.

Enjoy the beer gardens until then. There are many in Bavaria (especially around Munich) which are old and traditional enough to feel both eternal and necessary.
Dave M

In reply to by Peter Adamson

Davlat on 20 August 2013

time as necessary

hi prof. Adamson,

i'll jump in to defend the Shaykh. first, we have to keep in mind the distinction between (a) necessary in itself and (b) necessary through another but possible in itself. time cannot be a necessary existent - that is to say, a necessary existent in itself (which is what God is) - because, according to the Shaykh, time not only has internal constituents (muqawwimat) which constitute its quiddity, but it is also externally (i.e., causally) dependent on motion and ultimately on God. so even though time may be eternal, it is not necessary in itself. as for things like logical laws and other absracta, none of these things, according to the Shaykh, are substances in themselves. as such, they have no independent existence in themselves. that is to say, qua intelligibilia, they all depend on an intellect for their existence, i.e., that of God ultimately but also the separate intellects as their necessary concomitants (lawazim). (of course all these claims require proof and a refutation of a kind of Platonism (which i believe the Shaykh provides) but the point is that in his system, there's no inconsistency between these propositions).

In reply to by Davlat

Peter Adamson on 21 August 2013

Avicenna on time

Hi Davlat,

Yes, sorry -- I have actually been writing about Fakhr al-Din's response to Avicenna on time and will discuss that later in a podcast episode, so I should have said something about this too! To make a long story short, Avicenna would most definitely say that time is not only non-necessary, but in fact supervenes on contingent (changing) things. To that extent he agrees with Aristotle about time, though in the details his view is rather different. Still I think Dave's contention that time might be a necessary feature of the world is plausible and, as I say, the other earlier Razi would presumably agree.

Thanks!

Peter

In reply to by Peter Adamson

Davlat on 21 August 2013

on time

hi prof. Adamson,

i'm looking forward to reading your article when it comes out!

"Still I think Dave's contention that time might be a necessary feature of the world is plausible and, as I say, the other earlier Razi would presumably agree."

sure, but i was claiming that this isn't a point against Avicenna. in fact, he endorses it, just as long as time's necessity is understood as a 'necessity through another'. otherwise, we need an explicit proof (from someone) that time is necessary in itself (in the way God is) in Avicenna's sense.

In reply to by Davlat

Dave M on 22 August 2013

On Time and Cause & Effect

Hi Peter and Davlat,
Thank you all for your replies on ‘time’. As you can probably tell, I’m having a problem with ‘time’, as the philosophical conception of it does not, at the moment, seem to accord with the empirical evidence that modern physics would put forward. I can see that I have to do some background reading because I have no idea what a temporal frequency account modality would be, or how time can be dependent on motion, given that any kind of ‘motion’ needs time to be flowing in order to have any meaning (which seems to make it logically prior).

Neither do I see how Avicenna’s proof shows an NE that exists ‘in itself’ (whatever that means); rather his proof (at least as summarized in Peter’s podcast, and I accept that he was time-constrained and summarizing) is centred around proving the necessity of its existence as a pre-condition of any capability to preponderate anything else. I think this all that it does, or is capable of doing, at least in its summarized form. It doesn’t prove that it exists in itself. In order to exist ‘in itself’, I think you have to provide a proof whose line of logic does not involve the existence or essence of anything else. Without the existence of contingent things, the line of reasoning given in the podcast would not lead you to a conclusion.

Indeed, Avicenna’s line of reasoning also needs to invoke causality and I think he’s on shaky ground if he is trying to prove that causality does not presuppose the existence of time. The proposal that time ‘supervenes’ on contingent things seems to me to deny their nature. Their properties demand that they can change from non-existence to existence when they are preponderated, or at least that they can experience preponderance as the creation of their existence. Either way ‘time’ seems to be logically prior to their ability to exist, rather than supervening it.

Being a beginner at this, I had not realized until Davlat’s earlier post that Avicenna thought cause and effect were always simultaneous (Peter, I think this puts a different spin on what we were previously discussing in comments about the Ash’artite episode when you said you could ‘see’ the stars), but as I said, I’m now off to read some more so I have the technical terms at my fingertips the next time. I'll post again on this issue, but the word limit's about to cut this post off in its prime.
Dave M

In reply to by Dave M

Dave M on 22 August 2013

More on Cause and Effect

Just a bit more on cause and effect. I would say there is strong empirical evidence that cause and effect are not always simultaneous.

When we look into the sky, we are ‘seeing’ stars from the past, and whilst you can say that this is in fact a chain of causes and effects, rather than one cause and effect, something (at least, one cause and effect) in that chain appears to be temporally extended (i.e effect at time=t is caused by a cause existing at time less than t.)

You could propose that the process is not simply a chain of causes and effects, but something that looked like a necklace in which there were beads of simultaneous causes and their effects, followed by periods of unchanging being; the whole process looks temporally extended because of the periods without change. For example, I could propose the star emits light (simultaneous cause and effect), the light travels (taking time but not involving a change therefore not requiring a new cause), then the light strikes the eye and causes seeing (again simultaneous at the point of impact). If you want to consider it that way, you have to deal with 'something' (perhaps a photon in the case of light) changing in position, so you would have to deny that position is a proper feature of something which requires a cause to change it. That’s OK, I think. You are in effect saying that momentum/velocity is the proper feature and is unchanging during the ‘time’ that the light is in flight, therefore there is no cause-effect extending over this period.

That’s a neat move, but then empirical experiment would give you a problem. Think of interference patterns that are seen when two beams of coherent light come together. Basically, the idea here is that I split a coherent beam of light into two, send the light on different paths and then redirect it so the two halves hit the same surface. I get a pattern of light and dark patches in which the effect of the light on any point it lands on is dependent on the relative time taken for the two journeys that are coming together. (If I alter the length and therefore time taken for either of the two beam halves even slightly, I see a change in the light pattern.) The simplest way to look at it is that the pattern is caused by light from two different times, but that’s probably too easy a way to claim victory in the cause-effect simultaneity debate.

A more complete way is to say that the experiment shows empirically that the light is not in a constant state during its journey; it in fact holds a real property (which scientists call ‘phase’) which is constantly changing with time, and thus requiring a cause. Thus there is a real problem in denying that there is a temporal extension in the cause and effect chain, particularly because, if I made the change in the path length of my second beam at a point close to the source end of the light, I can measure a time delay before it affects the pattern at the other end.

I suppose we could propose, on Avicenna’s behalf, that the modern measurement of ‘time’ is simply wrong- headed, and that in fact our measurement of time in the light source’s location should be shifted so that its emission of light is simultaneous with its reception at the other end. But this doesn’t work either. If I put a mirror at the reception end and reflect the light back, the light arrives back at its starting point after the time taken for two journeys, rather than instantaneously.

Dave M

In reply to by Dave M

Peter Adamson on 22 August 2013

More on the NE

Just a quick response on your worry that "In order to exist ‘in itself’, I think you have to provide a proof whose line of logic does not involve the existence or essence of anything else." I think that Avicenna would say this objection confuses the manner in which God exists, with the way we know He exists. Limited creatures as we are, we might need to invoke the existence of contingent things (or the fact that they are possible) to prove that God exists. But precisely what we are proving there is the existence of an NE, i.e. something that does not depend on the existence of anything else, i.e. something that exists "in itself." Another way of of putting this is that if, like God, we could just grasp God's essence, we would see that He cannot fail to exist. But epistemically we need to approach Him indirectly, from His effects.

In reply to by Peter Adamson

Dave M on 23 August 2013

Stll Don't Get the Nature of NEs

Dear Peter,
Thanks for the response. When I made that comment, I was trying to understand Davlat's comment that God 'necessarily exists *in itself* (not just necessarily exists)'. I was questioning a) what is the difference and b) if there is a difference, Avicenna seems to be proving the former rather than the latter statement. You seem to be saying that something that necessarily exists does 'exist in itself'(i.e. there is no difference) which I agree with. I can't see what the distinction would be, but if there is a difference, I'd like to understand what it is and see it proved.

On a separate issue, given the fact that I've clearly got a problem with 'time' as a concept, I think I'd try to insist that saying we've proved the existence of 'something that does not depend on the existence of anything else' is imprecise. I think we've proved it only in as far as we can prove we're right about our axioms concerning causality and time. For example, I struggle to believe you can establish what causes what (i.e. which is cause and which is effect) if you are going to insist that cause and effect are simultaneous, since neither has logical priority. Secondly, you have to establish that 'forward' is a special direction in time, rather that just the way this particular branch of contingent things experience it. And when you've done those two things, I think you come back to my problem that you have to conceive of 'time' as something which has an uncaused existence.
Dave M

In reply to by Dave M

Davlat on 25 August 2013

time, cause and effect

Hi Dave,

first, i think it's key that you get a grasp of some of the technical terms involved in this issues. otherweise, will be talking past each other all day. second, i'm curious as to what is this so-called ‘empirical evidence’ is? third, regarding the dependency of motion on time, i think that in fact just the reverse is the case, i.e., it is time that depends of motion/change for its existence - at least according to Avicenna. if you think otherwise, then it seems to me that you think time is some sort of a substance, i.e., something that exists in its own right not dependent on the existence of anything else. if so, i’d like to see an argument for your claim.

regarding the "in itself necessary existence" of the NE, all it means is that the NE does not depend, either externally or internally, on the existence of anything else for its existence. another way to state the matter is to say that necessity (of existence) according to Avicenna is two-fold; it’s either conditional or absolute. conditional necessity is the kind of necessity which holds e.g., between a cause and its effect. that is, if the cause exists, the effect necessarily exists. in that sense, the effect does not necessarily exist ‘in itself’ - it’s necessity is dependent, i.e., on that of its cause. on the other hand, something absolutely necessary (like the NE) lacks this notion of dependency, both externally (in terms of existence) and internally (in terms of quiddity), on something other than itself.

In reply to by Davlat

Dave M on 25 August 2013

Time, Cause and Effect

Hi Davlat,

Yes, you're right that I lack the technical terms.

Firstly, thanks for clarifying what 'existing in itself' means. I think maybe I misunderstood what you were getting at in your first post. In it, you seemed to be drawing a distinction between something that 'necessarily exists' and a 'thing that "necessarily exists *in itself*"', but God is the only NE and the only thing that necessarily exists in itself, am I right? Thus, there is no distinction between the two terms? Basically, I can't see a way of conceiving of an NE that doesn't 'exist in itself'.

As for the empirical evidence about time, I was thinking of certain natural laws that seem to presuppose that time does have independent reality. Time definitely can't have substance (no mass, no dimensional extension), so if it exists, it belongs to the same class of metaphysical things as the NE.

Some empirical evidence to consider: motion has hard physical boundaries in the observed universe. It is impossible to move from a position A to a position B in less than a predefined time T. It is hard to conceive therefore that position and time are merely numbers attributed to things, since attributions cannot constrict the nature of things.

If you try to run an argument that says it's the motion that's real and somehow bounded independently of time and dimension, leaving both time and dimension(position) as attributed numbers with no proper existence (which seems like a possible objection to the above), then you have the problem that 'force', which appears to rule all interactions in our universe, depends on the change in motion with time, which requires that both motion and time have a separate existence.

Avicenna is empirically correct, I think, in that the nature of time is - or feels like - a change in 'something', so maybe it's incorrect to think of time as loosely as we do. If it's a change in 'something', that something is not motion, because of the arguments above, and that 'something' must exist. It must be logically prior to time; it must exist eternally.

Therefore, to put a provocative argument forward, if Avicenna has shown that an NE exists and there can only be one NE, the above evidence suggests that time (or at least the 'something' that it is recording a change of) is the NE. More provocatively, since God is supposedly unchanging, perhaps the 'something' time is monitoring is God.

I have a bunch of problems with this argument (even though it's me proposing it), not least of which is that 'change' seems to be logically prior to time. Human conception of change is something that has one property at time t and a different property at time t2. Thus, if time is to be defined in terms of change, we need a different conception of change, since you can't define something in terms that depend on itself. (For example, you can't define 'dimension' in terms of its size.) Thus, although I might agree instinctively with Avicenna that time is a change in 'something', I struggle with knowing what I, or he, mean by 'change' in that sentence. I have to solve that before I even start to consider what the 'something' might actually be.

In reply to by Dave M

davlatdida@gmail.com on 29 August 2013

time

Hi Dave,

"Firstly, thanks for clarifying what 'existing in itself' means. I think maybe I misunderstood what you were getting at in your first post. In it, you seemed to be drawing a distinction between something that 'necessarily exists' and a 'thing that "necessarily exists *in itself*"', but God is the only NE and the only thing that necessarily exists in itself, am I right? Thus, there is no distinction between the two terms? Basically, I can't see a way of conceiving of an NE that doesn't 'exist in itself'."

The problem here is that ‘necessarily exists’ and the ‘Necessary Existent (NE)’ are not identical. The former is more general than the latter; that is, something that necessarily exists can either (a) necessarily exist in itself or (b) necessarily exist through another. Although both things ‘necessarily exist’, for Avicenna, the Necessary Existent is just short hand for ‘necessarily exists in itself’.

"As for the empirical evidence about time, I was thinking of certain natural laws that seem to presuppose that time does have independent reality. Time definitely can't have substance (no mass, no dimensional extension), so if it exists, it belongs to the same class of metaphysical things as the NE."

There’s some confusion here. First, in Avicenna’s system (and the Peripatetic (Aristotelian) system more generally) a thing does not “have substance”; it either is a substance or it is not. Second, on that system, things such as mass, dimensionality, etc. are not constitutive of a substance, although they may characterize it in various ways. Why? All such things presuppose the existence of substance; that is, there are no such things as just dimensionality by itself or mass by itself in the external world, but always something with a mass or something with dimensions, and etc. Third, regarding “natural laws”, according to the Aristotelian system, there are no such things as natural laws if by that you mean independently existing laws which somehow act on and constrain the activities of various independently existing objects in the external world. IOW, ‘natural laws’ are just abstractions, just shorthand descriptions for the activities of actual/real concrete objects vis-à-vis each other, not actual things themselves which exercise causal efficacy of some sort on those objects.

"Some empirical evidence to consider: motion has hard physical boundaries in the observed universe. It is impossible to move from a position A to a position B in less than a predefined time T. It is hard to conceive therefore that position and time are merely numbers attributed to things, since attributions cannot constrict the nature of things."

Well, I certainly didn’t and wouldn’t say that time and position “are mere numbers attributed to things” though. I said that time presupposes motion insofar as there is no time without motion. I don’t see how what you’ve said above contradicts that.

"If you try to run an argument that says it's the motion that's real and somehow bounded independently of time and dimension, leaving both time and dimension(position) as attributed numbers with no proper existence (which seems like a possible objection to the above), then you have the problem that 'force', which appears to rule all interactions in our universe, depends on the change in motion with time, which requires that both motion and time have a separate existence."

I think you need to clarify what you mean by “proper existence” and “separate existence.” You seem to mean an existence which isn’t like the existence of numbers. If so, I agree. Time is not just an “attributed number”. It has a ‘separate/proper existence’ if by that you mean “outside the mind”. But again, according to Avicenna, it’s extra-mental existence is bounded to motion/change.

"Avicenna is empirically correct, I think, in that the nature of time is - or feels like - a change in 'something', so maybe it's incorrect to think of time as loosely as we do. If it's a change in 'something', that something is not motion, because of the arguments above, and that 'something' must exist. It must be logically prior to time; it must exist eternally."

I really am unclear about the ‘argument above’ you’re referring to. Avicenna (nor I) never said that time is a ‘change in something’; rather, it results from a type of change in something (more on this below. Change, according to Avicenna, is related to motion as a genus to species. That is, motion is a type of change. Avicenna’s position would more accurately be stated as: time is the measure of the motion *of something* according to the motion’s before and after

"I have a bunch of problems with this argument (even though it's me proposing it), not least of which is that 'change' seems to be logically prior to time. Human conception of change is something that has one property at time t and a different property at time t2. Thus, if time is to be defined in terms of change, we need a different conception of change, since you can't define something in terms that depend on itself. (For example, you can't define 'dimension' in terms of its size.) Thus, although I might agree instinctively with Avicenna that time is a change in 'something', I struggle with knowing what I, or he, mean by 'change' in that sentence. I have to solve that before I even start to consider what the 'something' might actually be."

I think that’s right, and Avicenna would I think agree. Change is prior to time insofar as it is related to motion - which time presupposes - as a genus to a species. This is because for Avicenna, change is defined as any passage from potentiality to actuality. Now, this passage can either be instantaneous or gradual. Strictly speaking, motion is only the latter; and time is just the measure of this gradual passage from potentiality to actuality. Hence, time then is not a change in something but results from a kind of change - i.e., motion - in something, say, the mobile.

In reply to by Peter Adamson

ajmal on 20 August 2013

Time as necessary

Hello.

To jump right in, what if a) time is not eternal at the first place but has a moment of origin (like during big bang), and 2) there is no reality of time but a concept (called time) to tarry, measure and make sense of our relation to the world in different moments or instants?

If we go by the aristotelian line of inquiry, a number cannot exist on its own since it is a predicate, not a subject. There is one specific buster as a person, not a number (one) to which buster is attributed. If by existence of numbers as necessary existents we mean (platonic) form of numbers, there is quite a lot stated by parminides to challenge the existence of forms at the first place, including the form of numbers.

About logical and moral laws, logic is pertaining to the world of generation and corruption (cosmos), and if we go by platonic line of inquiry, the cosmos has lots of opposing features (a tall tree is tall, only relative to x tree, and short relative to y tree, making the initial tree long AND short, which is logically impossible). So we cannot say that those laws are necessary because there is implicit contradiction embedded in the laws (again in platonic way of inquiry).
Morality again could be emanating from the Necessary existent to everyone, and it could be (again platonic) we all possess moral laws wired in us set up by the necessary existent, and the necessary existent is the source of that emanation. If there is no necessary existent, then there wouldn't be any laws at the first place. But we do have innate disposition or understanding of what is good and what is not good (if we think carefully), making the need for a necessary existent to exist ontologically prior to moral laws.

The very fact that time, and number 7 (as pointed by peter) can leave an impression on our intellect and make us think about it (metaphysically) tells us those things cannot be necessary. Time does not have to have an existential reality (as described above), making it contingent (if time does not exsist but it is just a relation), and number 7 can mean the combination of 7 objects which could easily have been 6 (not 7 objects), making it not necessary again. But to talk about time and number 7, we need a source through which the understandng of time and numbers can be understood, and if learning IS recollection as plato says, and if the forms of time and numnbers are nothing but concepts in the mind of the necessary existent, then there is the necessity of the necessaryexistent to exist, not necessarily the existence of time and/or numbers.

In reply to by Dave M

Davlat on 20 August 2013

time

David,

a NE is something that (a) exists an (b) necessarily exists *in itself* (not just necessarily exists). time according to Avicenna does not satisfy condition (b) because its quiddity has internal constituents and because it is dependent on motion for its existence.

as for your claims that (a) time is needed to distinguish between the necessary and the contingent/possible and (b) that causality always requires time, both of these according to Avicenna are not true. the first presupposes a temporal frequency account modality, which Avicenna rejects because (he has a proof for how) something can be eternal but still caused. moreover, for Avicenna, time is just a certain kind of possibility associated with motion. as such, it presupposes the notion of possibility and hence cannot be included in any explanatory account of the latter (as you seems to suggest in (a)). as for (b), this Avicenna would reject because he holds that every effect is simultaneous with its per se cause, especially with the per se cause of existence (as opposed to merely motion).

Muhammad on 16 March 2016

Greetings from Iran.

Greetings from Iran.

Dear Peter,

I have to admit, i feel a strong partaility towards the podcasts. Not only do i enjoy them, but they also have valuable statments within.

I read some of the stuff you have mentioned in original text, and i have to say, you are doing a great job.

 

Sincerely,

Muhammad

Hoom on 25 June 2016

Does this proof has a name?

One of the most interesting episodes so far!

Does this proof/argument has a name? I mean, as comparison, Anselm's proof is called the "(Anselm's) ontological argument", and Aquinas' proofs are often referred to the "Five Ways" and each of the five even has their own name. As you say, Avicenna's proof is no less intriguing or influential, do people have  a name for it? If it doesn't, what would you call it? :)

In reply to by Hoom

Peter Adamson on 25 June 2016

Avicenna's proof

Well in Arabic it was called burhan al-siddiqin, "the demonstration of the truthful/those who grasp the truth". I have seen it called a "modal argument" which gets it pretty well ("modality" means the notions of necessity, contingency and impossibility). But it has no real official title in English I guess.

In reply to by Peter Adamson

Hoom on 21 January 2017

burhan al-siddiqin

Where does the name burhan al-siddiqin come from? Did Avicenna himself call it that, and if yes, in which work? It's an impressive name

In reply to by Hoom

Peter Adamson on 21 January 2017

Burhan al-Siddiqin

Good question - couldn't tell you off the top of my head but it is, I believe, a title given the proof by later authors not by Avicenna himself.

Muhammad on 21 August 2016

Interpreting Avicenna

Hello,

Is there a chance to get a PDF version of "Interpreting Avicenna" -- I live in Iran, there is no access to Amazon from here.

 

Thank You

In reply to by Muhammad

Peter Adamson on 22 August 2016

Interpreting Avicenna

Not the whole book but if you are interested in one chapter I think I would be allowed to send you the proofs for that (which is all I have).

Jonathan on 5 November 2016

the proof and eternity

Hi Peter,

I've been studying Avicenna and turned to your podcast for some help (and great help I did find). I was hoping you could help clarify something for me about Avicenna's proof. Does it presuppose the eternity of the world? I was under the impression that it does. If everything (the aggregate) is possible, then it is possible for everything to pass into non-existence (every possibility has to become actual over the course of eternity, right?), and that has not happened, so there has to be cause that "preponderates" its existence. Does this idea that everything possible has to come about over the course of eternity factor into the proof? Thanks!  

Add new comment

The content of this field is kept private and will not be shown publicly.