and this is what he found problematic in his other book. The question then becomes: why cannot the combination have its existence determined by the individuals combined, and each individual by another? As explained above, the totality is not the successive individual members, but their collective configuration.

This answer can, however, be formulated if the collective configuration is treated as one of the totality’s parts. This is a matter of convention: in a whole composed of parts, the combination itself may be regarded as either not constituting a part of the whole or constituting one. If the combination is treated as part of the whole, the question can be formulated as follows: this part is caused by the remaining parts, and each part has its existence determined by another. A contingent part having its existence determined by another contingent part is like a contingent part having its existence determined by several contingent parts.

In that case, his answer—“The aggregate of the individual members is the very totality posited. This entails a thing determining its own existence”—does not address the question. They did not claim that the whole has its existence determined by the whole, but that the combination has its existence determined by each and every successive part. Although the combination is itself a part, it is not one of the successive parts.

Yet this involves some parts having their existence determined by others, and is therefore like some contingent things being caused by others. The issue thus returns to the same question: is the totality necessary in itself or contingent? Is it caused by itself, by some part of itself, or by something external to it? This follows Avicenna’s formulation of his argument, presented earlier.

It has already been explained that the totality either has no cause, but is necessary