Abstract This paper argues that soil mineralization should be seen as an emergent process. While it is effecting mineralization the soil population shows many of the characteristics of recognised emergent systems such as ant colonies and slime moulds. In particular, the whole is more than the sum of the parts, and the system shows clear evidence of bottom-up organization. Furthermore, the ‘quorum sensing’ recently discovered in soil and other bacteria strongly suggests that chemical signalling plays a part among the soil population, as it does in other biological emergent systems. Emergent behaviour has consequences for measuring and modelling mineralization. The idea that the whole is more than the sum of the parts suggests that measurements must be made at a scale that includes all the parts, while the bottom-up organization implies that there is no controlling parameter on which a model can be centred. Other emergent systems have been simulated using simple sets of rules for organisms that correspond to the local information they use in practice, and this approach could be tried with mineralization. Candidate rules are suggested.