are most worthy of being fulfilled are those by which you have made intimate relations lawful.”

Both reports are sound and are recorded by al-Bukhārī and Muslim. However, what is widely accepted among the legal theorists is that specific evidence prevails over general evidence; hence the conditions in question must be honoured. This is the apparent sense of what is found in al-ʿUtibiyyah, even though the position most often cited there is to the contrary.

With regard to conditions that are linked to some portion of the mahr, the Mālikī school displays considerable divergence—by which I mean a divergence over whether such stipulations are binding or not—and our present work is not intended to delve into every subsidiary detail.

As for the ruling when an invalid marriage has in fact been contracted:

• Some cases are unanimously subject to annulment both before and after consummation. These are marriages rendered invalid by omitting a condition whose presence is universally required for validity—such as a man marrying a woman who is permanently forbidden to him.

• Other cases are disputed, the disagreement turning on the strength or weakness of the cause of invalidity and on how far it involves neglecting essential conditions. In most of these instances Mālik annuls the contract before consummation but upholds it afterwards.

His basic principle is that no annulment is required, yet he adopts a cautious stance, much as he does with many corrupt sales that, in his view, become irrevocable once market prices fluctuate and similar factors intervene. It seems that he classifies such unions as merely reprehensible rather than null; otherwise there would be no rationale for distinguishing between consummation and non-consummation. The school’s rulings in this area are marked by considerable inconsistency, and the matter appears, for him, to hinge on the strength of the evidence demanding annulment: when the proof strikes him as strong, he annuls both before and after consummation; when it is weak, he annuls before consummation but not after—whether or not that strong proof enjoys consensus.

For the same reason the school has also differed over whether inheritance arises in invalid marriages when death occurs before annulment, and likewise over whether a divorce pronounced in such a union takes effect. At times they have looked to whether the underlying cause is agreed upon or disputed; at other times they have considered whether annulment follows consummation or not.

We deem it best to conclude the discussion here, for what we have set out suffices for the purpose at hand.