QuestionsCan we be certain of anything?Cotton and fire: the cause is not necessary
Can we be certain of anything?We must doubt

Cotton and fire: the cause is not necessary

Al-Ghazali 1058 · 1111Reason has limits: certainty comes from experiencing God.
See profile
What it claims

Nothing proves that fire necessarily produces burning. It is possible for the fire to touch the cotton without burning it, if God does not create the burning: the link between cause and effect is not necessary.

Why

We see the fire, then the burning, never a link forcing one to follow the other; denying that link is not contradictory.

In short

"Cotton and fire: the cause is not necessary" is an argument by Al-Ghazali (The Incoherence of the Philosophers, discussion 17 (1095)) on the question "Can we be certain of anything?" (answer: we must doubt). Its claim: Nothing proves that fire necessarily produces burning. It is possible for the fire to touch the cotton without burning it, if God does not create the burning: the link between cause and effect is not necessary. According to AskPhilosophy's logic check (Z3 solver), the argument is valid only if one grants a premise the author leaves unstated (P2). AskPhilosophy presents 4 objections, 2 of them with a documented reply, and the original text behind each premise, checked word for word.

The Incoherence of the Philosophers, discussion 17 (1095)
Valid with an implicit premise

You bring a flame close to a piece of cotton: it burns. You say the fire caused the burning, and that it had to happen. Al-Ghazali asks: what did you really see? The fire, then the cotton burning. Never the link that would force one to follow the other. For him, this link is a habit that God keeps up, and can interrupt: that is what makes miracles possible, his real aim in this chapter. Six centuries later, Hume would make a very similar observation about causation, with no proven influence, and without drawing the same conclusion.

P1Cause and effect are two distinct things: affirming one does not force us to affirm the other, and denying one does not force us to deny the other. Saying "the fire touches the cotton, and the cotton does not burn" involves no contradiction.
P2 · implicit premiseA necessary link can be established in only two ways: either by showing that denying it would be contradictory, or by observing it. For the second, the philosophers have only one proof: we see the burning when the fire touches the cotton.
P3But observation only shows that the burning happens with the contact of fire, not that it happens by it, nor that it has no other cause.
P4Only what is contradictory (affirming and denying the same thing) is impossible. Everything that is not contradictory is possible, and God can do it.
Therefore nothing proves that fire necessarily produces burning. It is possible for the fire to touch the cotton without burning it, if God does not create the burning: the link between cause and effect is not necessary.

click a premise to reject it

If you accept every premise, logic forces you to accept the conclusion. So the debate is about the premises.
Checked by a logic solver: the premises cannot all be true while the conclusion is false.
Without the implicit premise: Without P2, the computer finds a case where the link is proved by a third way (for example a knowledge of the nature of fire), while neither logic alone nor mere observation proves it. The conclusion "nothing proves the link" no longer follows.
The argument is valid if we add P2: there are only two ways of proving a necessary link, logic or observation. Al-Ghazali only says the philosophers have no proof other than observation; he assumes there is no third way. This is where his opponents attack: for Avicenna and Averroes, reason can know the nature of things (fire is, by nature, what burns), without denying it being a mere verbal contradiction.
Analogy. Every time you enter the kitchen, the light comes on. You think your presence turns it on. One day you discover it is your sister, hidden, pressing the switch as soon as she hears you. You had seen the light come on with your entering, never by it. Al-Ghazali gives a similar example: a blind man who is cured thinks that opening his eyes is what makes him see colours, until the sun sets.

Objections

4

Each objection is a conversation: click to see who replies and where the debate stands.

Objection anticipated by al-Ghazali attacks the conclusion

If effects do not necessarily follow their causes, then anything becomes possible and we can no longer know anything. A book left at home might have turned into a young man, or an animal. We could no longer be sure of anything around us.

Al-Ghazali replies
Al-Ghazali
We never claimed that these things are necessary: they are only possible, they may happen or not. But when a possible thing is repeated again and again, this habit is fixed in our minds so firmly that we cannot shake it off. And God creates in us the knowledge that he will not do those things.
The debate goes on (our summary) The reply saves our confidence in the world, but grounds it in a habit and in God's will, not in the nature of things. His opponents judge this too high a price for science. Hume, later, would also ground our expectations in habit, but without God.
Avicenna and the philosophers (as reported by al-Ghazali) 1027 · attacks the conclusion

Suppose fire is not itself the agent: events come from higher principles (the celestial intelligences). But these principles act by necessity, as the sun gives light, without choosing. Then two identical pieces of cotton, touched in the same way by fire, will both necessarily burn. The miracle of Abraham thrown into the fire without burning is impossible.

Al-Ghazali replies
Al-Ghazali
We do not grant that the principles act without choosing, nor that God does not act through his will: we refuted this on the question of the world's beginning. And if it is established that the agent creates the burning by his will when the cotton touches the fire, reason admits that he may not create it, while the contact takes place.
Al-Ghazali (second answer)
We may even grant that fire is created in such a way that it burns two identical pieces of cotton without distinction. The prophet thrown into the fire may still not burn: God can change a property of the fire, or of the prophet's body.
The debate goes on (our summary) The first answer depends on discussion 1: it must have been shown that God acts through a will that chooses. The second keeps a nature of things, but subject to God. It is closer to the philosophers, but it no longer says the link of cause to effect is never necessary: only that it depends on God.
Averroes 1180 · attacks premise 2

To deny causes is to deny reason. To know a thing is to know its causes; and each thing has its own nature and actions: a fire that did not burn would no longer be fire. If this link is removed, nothing can truly be known, and science becomes impossible.

no reply
No reply possible: objection raised in 1180, after the author's death (1111).
The debate goes on (our summary) Averroes, born after al-Ghazali's death, wrote The Incoherence of the Incoherence to answer him point by point. Al-Ghazali had already partly answered: our ordinary knowledge stays secure through habit (first objection), and his second answer accepts that fire has a nature. The basic disagreement is the question: does the nature of things bind God, or does it depend on him?
Hume 1748 · attacks the conclusion

Hume starts from the same observation: we only see that things go together, never the link between them. But he turns the argument against those who conclude that God is the only true cause (he is aiming at Malebranche, not al-Ghazali). We are just as ignorant of how a will, even God's, produces an effect. If our ignorance were enough to deny a power to fire, it would force us to deny it to God as well.

no reply
No reply possible: objection raised in 1748, after the author's death (1111).
The debate goes on (our summary) For al-Ghazali, God's causation does not rest on observation, but on the proof that the world began and that a will chose it (see his argument on the world's beginning). Hume's objection thus mainly targets those who would infer God from the mere absence of a visible link. Historians stress that the resemblance between the two is real, but that no influence is proven.