Actually, the correct definition of a "free market" is a market where the price of commodities isn't controlled by some entity, where free competition on equal terms is possible, but does NOT mean that any transaction is possible, such as a person wanting to sell a nuclear bomb to some terrorist
This still doesn't prove (what RB tries to claim) that the concept of free market is atomic, that it can't be further reduced to distinct interactions of individuals (with all their whimsies and fancies) by way of deduction, on the one hand, or that it can't be arranged from the same interactions by way of induction, on the other...
As I've said, free market is a particular form of individual economic interactions taken integrally, as a whole