137 - God Willing: the Ash'arites

Al-Ash'arī puts his stamp on the future of Islamic theology by emphasizing God’s untrammeled power and freedom.

Press 'play' to hear the podcast: 

You are missing some Flash content that should appear here! Perhaps your browser cannot display it, or maybe it did not initialize correctly.

Further Reading: 

• M. Allard, Le problème des attributs divins dans la doctrine d’al-Ashʿarī et de ses premiers grands disciples (Beirut: 1965).

• R.M. Frank, “Bodies and Atoms: the Ash'arite Analysis,” in Medieval Islamic Thought, ed. M.E. Marmura (Toronto: 1984), 39-53.

• R.M. Frank, Early Islamic Theology: the Muʿtazilites and al-Ashʿarī (Aldershot: 2007).

• R. McCarthy (trans.), The Theology of al-Ashʿarī (Beirut: 1953).

• P.E. Walker (trans.), Imām al-aramayn al-Juwaynī: a Guide to Conclusive Proofs for the Principles of Belief (Reading: 2000).

• H.A. Wolfson, The Philosophy of the Kalām (Cambridge MA: 1976).

Dave Martin's picture

A question about the Ash'arites

Hi Peter,
Thanks for another very interesting episode. There was one thing I didn't quite get.
At one point, you say that these guys denied the possibility of something having two causes, but you don't give an explanation of why they believed this. Surely, they do believe in actions that can have both an enabling component and an executive component, where both components have to be present for the thing to happen and where the two components can themselves have separate causes. Their whole theory of acquisition seems to depend on this (i.e. God provides the power, man provides the power to use the power, therefore both must be present to make the thing happen. As I see it, you either say that both are causes because both are needed to make the thing happen, or that neither are causes because taking the other one away always prevents the thing from happening, but if you look at it in this latter fashion, then you will have things happening that have no cause at all.)
Yours slightly baffled,
Dave M

Peter Adamson's picture

Two causes?

Dear Dave,

Well, you're putting your finger on the crucial issue there. This is quite difficult I think, if you really get into the details, but what I was trying to say in the episode is that if you think of a cause as something that is sufficient for the effect to be produced, they want to deny that there can be more than one cause. (Because then God could be trying to cause something to happen but He would need our cooperation to make this occur.) That's actually quite plausible - contemporary philosophers often speak of wanting to avoid "overdetermination" which is when you have more than one sufficient cause for the same event.

So, what they want to say is that God is the sole (sufficient, and in fact also necessary) cause of everything, including human actions. The mysterious part is how it happens that we "acquire" the action if we're not causing it, but I understand that to mean that we are carrying out the action which God causes to occur. So to use the analogy I mentioned in the episode, if God makes the sky blue it is the sky that is visible, not God, and likewise if He causes me to sin I am morally responsible, not Him.

What you are envisioning is a different situation, which might be held by other theologians or philosophers (perhaps some Mu'tazilites) in which God and the human agent are _jointly_ sufficient, i.e. together they constitute the cause but if either refuses to play ball, the effect will not occur. The Ash'arites can't live with that because they think it compromises God's omnipotence, and you can see why -- He needs my help in order to create my action.

Does that help?


Dave Martin's picture

Two Causes?

Dear Peter,
Thanks for the reply and yes, it does help to understand the Ash'arite position, but I must say, I don't buy into it. They seem to be using verbal gymnastics to fudge around the real problem, and not doing a particularly convincing job of it either.
As a physicist, I guess I can just about swallow the one-event-one cause premise. Most events in nature appear to be interactions between things which have at least two elements that have to co-operate by being in the right place at the right time, but - as long as we believe there is a universal equation somewhere that drives everything - I guess we could believe that all those elements eventually trace back to the one 'cause'.
I think my real problem with the Ash'rite position comes if I look closely at your analogy. OK, God makes the sky blue and it is the sky that shows the colour, but does that make the sky 'morally responsible' for its own blueness? The human concept of 'morality' seems to me to require that 'responsibility' is assigned to the sufficient cause, not to the mechanism by which an event becomes manifest. To think otherwise would require that we make a bullet (or perhaps even the victims inner organs) morally responsible for a murder rather than the person who pulled the trigger.
In fact, the more I think about it, I can't see how I would define 'responsibility' as anything other than 'being part or all of the sufficient cause'.
Thanks and Regards,
Dave M

Peter Adamson's picture

Verbal gymnastics

Hi Dave,

Well, your verdict is shared by lots (maybe even most) people who look at the Ash'arite position. Even within Ash'arism there are some who accept the possibility of "secondary causation" which is what you suggested in your original post (i.e., I can cause something to happen, but God has a veto power). Some think this was Ghazali's view. My hope in the podcast was to make the view intelligible, not to defend it! Of course the key point, as you rightly say, is whether there is any rationale for retaining a notion of moral responsibility that can belong to the human agent if this agent is not the genuine cause. Here it might be helpful to think about why exactly the responsibility is not like the visibility in the sky case; remember that in general moral responsibility is not the same as causal responsibility (since it has to do with things like "being fittingly rewarded/punished") and in most cases causal responsibility doesn't even bring up issues of morality (e.g. if a rock randomly falls off a cliff onto a car and dents it, issues of reward and punishment don't arise). So there is more to moral responsibility than just being causally responsible; however I agree that it is plausible to think causal responsibility is a requirement for moral responsibility.

Remember too by the way that as far as the Ash'arites are concerned, God could in any case punish the good and reward the bad without being unjust, so this issue of moral responsibility is just an initial skirmish before that further battle.


Dave Martin's picture

Verbal Gymnastics

Hi Peter,
OK, let me thank you again. I think I do now understand their position, even though I find it irrational to define 'responsibility' such that a sentient being can be 'responsible' for something in which their body partakes but their mind and will does not. It seems a better term would be 'allocated to' or 'attached to'. Never mind.
Keep up the great work. Really looking forward to more on Avicenna, about whom I knew nothing before last week.
Dave M

Felix's picture

Certainly plenty of ideas to disagree with!


In general, if you were trying to put the best case possible for these theories then, allowing for your brilliance, the theories really suck!

The "ingenious" argument about pulling the trigger and dying is pathetic. It does not seem in the slightest bit absurd to suggest that the hunter is causing Hiawatha's death even though 'she [the hunter] does not exist at the point that she [the giraffe] expires'.
The act of pulling the trigger sets in motion a series of events which result in the death of the girafe.
The continuing existence of the hunter at the time the girafee dies is irrelevant.

For example, the hunter could put in orbit a weapon designed to re-enter the atmosphere in 1,000 years time and kill all humans and/or giraffes on the planet. The fact that they are dead long before this does not change their intent.

If there can be no unrealized power, and God puts the power to do evil in to me and, furthermore, via "occasionalism", he is responsible for the continuing existence of every atom at every moment along with the properties of that atom, then I am no more than a robot, that is a computer program that has been written to believe that it is experiencing pain but for which the sense of self-awareness is merely illusory.

God is definitely responsible for all the 'evil' in the world under this view.

However, it could be argued that shooting a 'robot' giraffe or torturing for enternity a 'robot' 'person' is not evil and therefore god is acquited.

Regarding your example: "if a rock randomly falls off a cliff onto a car and dents it, issues of reward and punishment don't arise".

However, if the rock is large and kills the occupants of the car then if there is an agent who is responsible for the falling of the rock (rather than just blind luck - which in the Asharite universe doesn't exist) then the agent, i.e. god, is morally responsible.

Are these models of ethics/morality still current in Islam today or are they of purely historical interest?


Peter Adamson's picture

Ash'arite arguments

I think the arguments are better than that, though I wouldn't go so far as to say they work in the end.

Regarding the chain of causation, consider the following argument:

1. Causation is a relation between a cause and effect

2. The existence of a relation requires the existence of its relata (=the things that are related)

3. Therefore, there can be no causation unless both cause and effect exist

Premise 1 looks hard to deny and premise 2 is at least plausible. In fact I think it is false but the fact that it is false is surprising. (To see that it's false consider this: if a star millions of light years away exploded and stopped existing 100 years ago, when we see it now we are seeing something that no longer exists, and seeing is a relation, in fact a causal relation.) I think the Ash'arites are presupposing an argument like this, so although they are wrong it is clever and convincing at first, and rejecting it is going to require giving an account of causation that doesn't commit us to premise 2; not easy.

Regarding moral responsibility, that of course is the key issue: whether God can have all the causal responsibility and none of the moral responsibility. Again I tend to agree with you but the point that God can in general cause creatures to "acquire" properties He lacks does seem to me a strong one (e.g. He can cause something to be visible without being visible). So again the onus is on us to explain why moral responsibility can't work that way.

Regarding your last question I suppose that this sort of Asharism is more a historical than current phenomenon but certainly questions of divine power and fatalism are still relevant to Islam and indeed all the Abrahamic faiths.

Dave M's picture

Ash'arite arguments

It seems to me that, on the question of causation, it is the second part of the argument that needs qualification: 'the existence of a relation requires the existence of its relata at some time'. (In the words of your most recent podcast, they must be either necessary or contingent, they can't be impossible).

I think you're wrong that seeing is a relation between the person seeing and the thing seen. It's a relation between the person seeing and light emitted by the thing they are seeing. It's only causal in the sense you describe between star and viewer if you are going to allow that causality can be a chain of cause and effect, rather than one discrete interaction, in which case where are you going to stop in an infinite regress back to the creation of the universe?

Further, it seems to me, two things related by causation only need to exist simultaneously if we believe that causation must be an instantaneous effect.

However, modern physics will surely provide us with a neat proof that causation can never be instantaneous between two material things: material things can never be in absolutely the same place and interaction between them is ruled by the limitations of relativity, and therefore can never take no time. Thus, there must be a finite time in any causation between the action and its effect. Thus there is no reason why the cause must still exist when the effect occurs. (This argument could also be used to show that the cause must in fact exist before the effect.)

Dave M

Peter Adamson's picture


Thanks, that's really interesting. I agree that physics is throwing out some pretty surprising things about causation for instance that two things can act on each other at a distance. Your last point is particularly fascinating; unfortunately as I have admitted before I stopped paying attention to physics in the 13th century, so I can't venture to say anything about what contemporary physics is establishing about these questions.

I'm not sure I do agree about seeing, though; obviously you're right that the light coming off the visible object is an instrument by which the seen object is seen. Still, even if we are being careful to be precise, don't we still want to say that when I see an apple, I am seeing the apple and not the light waves coming off the apple? At least, I find that intuitive. It would be like person A hitting person B with a stick: person A is doing the hitting, not the stick, even though without the intermediary of the stick no hitting would be going on.

Dave M's picture


When it comes to the 'seeing' thing, think about it in terms of what happens if you put a distorting lens in the way of the light coming from the apple. Now you can see the apple as bigger, smaller, upside down, or even squashed in the middle. Are you seeing the apple, or an image of the apple caused by light hitting the back of your eye? It might have left the apple as a perfect representation of the apple, but, by the time it's reached you, what it causes is influenced by the intermediate cause-effect chain (i.e. a lot of factors about the refractive index of glass and air). The fact that normally (without the distorting lens) the intermediate chain doesn't do a lot of distorting and it still looks like the apple when it reaches you doesn't mean you're 'seeing' the apple.

Isn't this a bit like Plato's cave allegory? Even if the shadows seen in the light of the fire didn't distort the original image, you still wouldn't be seeing the 'real world', but an image of it.

Another way to think about it, I guess, is like a page of writing and the copy I get from a photocopier. I can look at the copy and know exactly what the original document looks like, but that doesn't mean I'm looking at the original. It also doesn't mean that the original still exists (it might have been shredded by now.)

Dave M
P.S I've been thinking about Avicenna and God as well, so I'm going to send a comment about that episode separately.

Peter Adamson's picture

Copies and causation

Right, the more we talk about this the more obvious it is that this is a tricky issue. I think I would want to distinguish between the case where I am looking at an apple through a distorting lens, and one where I am seeing the apple's shadow or, say, a photo of the apple. I would be more tempted to say that I am seeing the apple in the lens case than the shadow case, the photo case is rather less clear to me but I guess I lean towards thinking you don't see the thing in the photo, but rather the photo. Certainly I wouldn't say that I am looking at a piece of paper if I am looking at a photocopy of it, so I agree with you there. In cases where the object I am "seeing" no longer exists I guess that would push me towards thinking I am not seeing it (e.g. a photo of a dead person) but I don't think it's a decisive consideration, since as I say I would tend to think that we are indeed seeing far-away, now dead stars.

But I'm only reporting on my own intuitions here, really what is needed is a precise account of what it means to see something, and then an analysis of each of these cases in terms of that account. Philosophy is hard!

Dave Martin's picture

Copies and Causation

Physics is hard too, and the answer here isn't perfectly clear because we don't exactly understand the nature of light. Your podcast no 132 showed pretty accurately that light has to pass from object to eye, rather than the eye in some way directly reaching out and interacting with the apple. The verb 'to see' has to be describing something that works by a somewhat different and more complex interaction than 'to hit'.

However, I think the question we're debating really comes down to what you think 'I see Object A' actually means. Is it sufficient that Object A emits or reflects light that is turned into an accurate understanding of its form and colour by a combination of my eye and brain? If yes, then it doesn't matter how long the chain of cause and effect is between the apple and your eye as long as the information is received and reconstructed as knowledge accurately by the brain.

Of course, if this is our accepted definition of seeing, then it's going to be harder to say that seeing a photograph, or perhaps a live TV broadcast, isn't seeing the thing depicted. To take it further, is, for example, seeing an X-ray plate 'seeing' the bones that are shown there, even though the X-ray radiation that produced the images isn't directly visible by the eye, and despite the fact that the bones we'd be seeing are actually the shadows where the radiation didn't get through?

I think we can agree, though, that - however you define seeing - cause and effect don't have to be coexistent. Seeing long dead stars or images caused by long dead stars, if you prefer to think about it that way, seems to prove that point either way, which I think is where we started this discussion.

Dave M

Peter Adamson's picture


Yes, I think I agree with all that: in general, a cause can be connected to an effect by some sort of intermediary chain, and this needn't take away the cause's status as a cause. It may depend on the nature of the chain, of course, but it seems like there are some such chains anyway (the hunter shooting the giraffe for instance: here the hunter is pretty clearly the cause of the death). And in those cases we probably want to say that the cause can no longer exist when the effect obtains. That, I think, is why the Ash'arite argument is wrong; but I still think it isn't obviously wrong, because this point about causes no longer existing yet still being causes is quite subtle and, at first blush, rather paradoxical.

David Taylor's picture

Sources of Islamic Atomism?


I've been loving these podcasts since I started listening to them a few weeks back.

I was wondering about the popularity of atomism among these early Islamic thinkers (both the Mu'tazilites and the Ash'arites). I take it that what they have in mind is not the "geometrical atomism" of the Timaeus, but then where exactly are they getting it from? Is it Epicurean-influenced? If so, how was it transmitted?

Just something that I was curious about while listening to these episodes.



Peter Adamson's picture

Islamic atomism

That's an excellent question and as it happens I have done some research on the issue. I could go on about this at some length but here are the main points:

1. Plato's Timaeus was known in the Arabic world but does not seem to be the source for kalam atomism. As you say they aren't using the idea of geometrical solid "molecules." However in my forthcoming book on al-Razi I argue that his main inspiration for his atomism is the Timaeus.

2. Kalam atomism seems to be closely related to developments in mathematics; they often conceive of atoms as "points." On this check out Marwan Rashed's piece about physics in the Cambridge Companion to Arabic Philosophy.

3. Other possible Greek sources are Aristotle (since he attacks atomism in the Physics) and Galen (also attacked atomism in works known in Arabic). There seems to be little or no transmission of Epicurean atomism, so that is not a likely influence.

4. One shouldn't rule out atomist ideas coming from India but I've never seen an argument for influence from that quarter.

So summing up it may be that the kalam atomism is a (nearly) autonomous development applying mathematical ideas of the period to physics; in any case it doesn't respond directly to the kind of Eleatic puzzles that inspired Democritus. Incidentally the best study of the whole issue is probably the monograph by Alnoor Dhanani, "The Physical Theory of Kalam."