Introduction
Al-Ghazali is one of the most influential philosophers the world has ever seen, and has been called the most original philosopher of Islam. His demolition of Greek philosophy in Tahafutal-Falsafa (The Incoherence of the Philosophers) is considered a turning point in the history of philosophical thought. It is fascinating that al-Ghazali adopts the methods of logic and reason (i.e. philosophy itself) to demonstrate that the conclusions of Neoplatonic thinkers are invalid.

Necessary Causality
Although in modern western philosophy Hume is well known to have denied necessary causality, al-Ghazali had done the same several centuries earlier. Necessary causality is the idea that the relation between a cause and its effect is necessary and always true. Al-Ghazali gives us an example: when a piece of cotton is brought near a flame, the cotton burns. In fact, whenever any similar piece of cotton is brought sufficiently close to a similar flame, it also burns. So there appears to be a causal link between the flame and the burning of the cotton. That is, we would be inclined to say that the flame caused the cotton to burn. Furthermore, we think that this link is necessary, i.e. it always happens.
We can think of this in terms of physics. For example, whenever I hit a glass window with a certain tennis ball, at a certain velocity, from a certain angle, under certain defined circumstances, the glass window breaks. Furthermore, I can set parameters such as speed, acceleration, angle, curvature of the ball, and characteristics of the ball, that would be necessary for this to occur. In other words, we scientifically define that, for example, a tennis ball at 100 mph at an angle of 0-20 degrees in a horizontal plane and in a vacuum will always break the window. So we say that under these circumstances, the ball causes the window to break.
Al-Ghazali, however, points out that what we are really observing is simply a quick succession of events, not any sort of causation. All we see is a tennis ball hit the window, and then we see the window break. All we see is the cotton a few centimeters away from the flame, and then the cotton burning. We are not observing any direct causation, only an association of events.  
He gives the example of a blind man who is unable to open his eyes. One day, through a miraculous event, he is able to open his eyelids and see the world. In his joy he attributes his ability to see to the opening of his eyelids.  (He thinks the opening of the eyelids causes him to see). But he fails to realize that it is not the eyelids, but rather the rays of light refracting into his eyes that give him the ability to see. The blind man mistakes this association of events for causation.
Hume puts forth the same criticism. In talking about the cause (flame) and its effect (cotton burning), he asks, where is the causal glue that holds the events together? In other words, all we see are two events occurring one right after another, there is no reason to think that one event causes the other. There is no causal glue that holds the cause and effect together. Why think that the association would always be true?
Hume points out that the only reason we have for thinking that it will persist in the future is an argument from induction of the following form:
Premise 1: All As observed before time t have been followed by B
______________________________________________________________
Conclusion: The first A observed after time t will be followed by B
That is, all series of events A (cotton near flame) observed before time t have been followed by B (cotton burning). Therefore, the conclusion, that the next time we observe a series of events A, it will also be followed by B. Hume rejects this argument. What reason, he asks, do we have to think that this is so? Why think that simply because something has held true in the past, it will it necessarily continue to hold true in the future? To accept the above argument would be to accept the idea of necessary causation between A and B, something which al-Ghazali rejects. 

This is precisely the problem with causation.  
To illustrate the point further, let us consider the following example. Suppose we meet a person named Ahmad. It so happens that whenever Ahmad sneezes, it rains in outer Mongolia. We have observed this over a period of several years, and it has been confirmed by the national weather service. In this situation, we do not think that Ahmad’s sneezing causes it to rain in Mongolia! We think this is absurd, if it were true, we would consider it an extraordinary coincidence. In the same way, the notion of necessary causation seems to be flawed. All we observe is:
1- Flame
2- Cotton near flame
3- Cotton burns
To say that the flame necessarily causes the cotton to burn is not valid, according to the above argument. Note that this has wide ranging consequences, from experimental scientific research to simple things, like pulling a trigger and the firing of a bullet! If there were no necessary causation, then anything would be possible! There is no guarantee of what will happen next.

Resolution of the Problem
So why do we as humans adopt this notion of cause and effect if we are not justified in holding such a position? Al-Ghazali states, and Hume agrees, that this is due to the habitual course of events. That is, nature tends to follow certain rules and laws, and this happens to be the way nature works. Things have a nature, and by their nature, they tend to act in a certain way. As Hume would say, nature is uniform. According to al-Ghazali, God created the universe to act in a certain way, and God can suspend this this natural course of events.
Al-Ghazali thought that accepting necessary causation would deny God the power to do what he wills. That is, if necessary causation is true, then God would not be able to perform miracles, since miracles do not follow causal laws. Hence, when God wishes to perform a miracle, he suspends the habitual course of nature and allows such an aberration of natural laws to take place.
So, we see that the reason we are led to a conclusion such as causation, is  the remarkable consistency with which such events occur. Hume states that “nature is too strong for principle.” That is, although we clearly realize that there is no logical basis for accepting causation, we continue to believe that, for example, the next time I clap my hands (cause) it will make a noise (effect). The only time when such causation is not seen, is during the performance of a miracle by a prophet, in which case the natural and habitual course of the universe is suspended.
Students of philosophy will realize that although both al-Ghazali and Hume agree on this point, they have different agendas, one is a theist, the other a skeptic. In Part Two, we will consider Ibn Rushd’s and Hume’s objections to al-Ghazali’s proposed solution. Of the series, we will see how the problem of induction is superimposed on this apparent resolution of the problem of necessary causation. We also discuss objections raised by Ibn Rushd.

Hume on al-Ghazali
Does Hume succeed in undermining al-Ghazali’s theory of causation with the problem of induction?
Al-Ghazali correctly points out many of the flaws in a theory of strict and necessary causality. However, the kind of solution offered by al-Ghazali seems unsatisfactory to Hume. Al-Ghazali does not accept the kind of induction mentioned earlier in this paper. That is, he rejects the argument of the following form:
In the past, cotton near a flame always resulted in the cotton burning
______________________________________________________
The next time cotton is placed near a flame, it will burn
Al-Ghazali does not think that this relationship is necessarily true, as he points out by his argument. Instead, al-Ghazali states that the only reason the above relationship holds true in general is that God has created things with certain natures, and these natures tend to act in certain ways. That is, the reason we can rely on an argument of the above form in our everyday lives is that nature is uniform.
Hume’s response is that the theory advanced to support the above argument from induction is itself subject to the same argument. In other words, no progress is made.  Namely, in defending the above position, al-Ghazali proposes the idea of the uniformity of nature. Since nature is uniform, it makes sense for us to accept an argument of the form above. But this defense itself relies on the same sort of argument. The problem being as follows:
In all observed cases, nature has been uniform
____________________________________________
In the next observed case, nature will be uniform
If an argument of this kind is rejected in the first place, why accept it now?  Why think that this will always be true? Is al-Ghazali not overlooking this problem in his argument? In fact, al-Ghazali would not accept this argument either. He would continue to state that there is no necessary link causing nature to continue to be uniform. In fact, the uniformity of nature can be suspended if God so wills.
Now a greater question arises, a question upon which much of modern day skepticism is based. If there is no way for us to determine truth from experience, then how can we be sure that we know anything at all (besides necessary truths)? For example, if a medicinal drug has been demonstrated to cure cancer in clinical trials, there would be no reason to think that it will necessarily work when used the next time. This would undermine vast portions of human knowledge. Bertrand Russell states,
It is therefore important to discover whether there is any answer to Hume within the framework of a philosophy that is wholly or mainly empirical.  If not, there is no intellectual difference between sanity and insanity....This is a desperate point of view, and it must be hoped that there is some way of escaping from it.(Russell 646)
This is precisely what worries Ibn Rushd, as we see in the following section.

Ibn Rushd on al-Ghazali's Theory of Causation
Ibn Rushd points out that if we accept al-Ghazali's theory of causality, this would preclude any possibility of yaqqini (definite) human knowledge. That is, if there is no necessary causality between events, then we cannot claim to have knowledge about them. At most, we can claim to have probable knowledge concerning science. We can say that since nature is uniform, A tends to cause B, but we will never have any definite knowledge. We will lose the ability to verify all knowledge by experience. There can be no conclusive scientific knowledge if al-Ghazali’s view is accepted.
Ibn Rushd accepts causality and rejects the driving force (the ability of God to perform miracles) behind al-Ghazali’s attack on it. He states that miracles are supernatural events, and hence by definition are not subject to human understanding. In other words, they are beyond reason and it is inappropriate to attempt to use logic and reason to justify events which are beyond reason. He argues that the cause and effect relationship is unalterable and is testament to the wisdom of God in giving us scientific knowledge. (Qur’an 35:43) But Ibn Rushd fails to give a reason why he thinks necessary causation is in fact necessary.
In Ibn Rushd’s argument against al-Ghazali, he is assuming that definite scientific knowledge is in fact possible. That is, he is being an epistemic optimist in the Aristotelian sense. However, al-Ghazali would, surprisingly, agree with Ibn Rushd. He would say that in fact, it is true that no scientific knowledge is possible by experience alone. Only probabilistic or working knowledge is possible. Since there is no necessary intrinsic causality, the cause and effect relationship observed is due to the nature of the things involved. The nature of those objects is created by God, and God has the ability to change them at his own will. God is the ultimate cause of all events (even if in the less occasionalistic sense). Therefore, definite knowledge can only be achieved from the ultimate cause, that being God. Only God, then, is the source of yaqqini (definitive) knowledge.
So we see that al-Ghazali's strategy to escape the problem of skepticism is through a direct relationship with God. Al-Ghazali accepts many of the skeptics’ arguments, realizing their power, before most western philosophers came to that realization. He continues to state that the source of true and definite knowledge is God alone, all other knowledge will in fact be, as Ibn Rushd points out, only probable.
                       ..........
Note:  Whether or not al-Ghazali is a strict occasionalist (i.e. one who thinks God is the ‘link’ between each cause and effect) is beyond the scope of this paper. It has been a matter of debate; see Professer Riker's paper in the Monist.  This paper assumes al-Ghazali took the more moderate position of God working through nature.

References:
Riker, Ghazali on Necessary Causation,”- Monist, July 1996
Al-Ghazali, The Incoherence of the Philosophers
Ibn Rushd, Incoherence of the Incoherence
Al-Ghazali on Science, Routledge (pg. 939)