In traditional studies of historical phonology, there is frequently a division between sound change and analogy. Sound change is said to proceed gradually and inexorably under strictly phonetic conditions; “La règle générale est que les transformations phonétiques s’opèrent avec une constance absolue c’est-à-dire que les mêmes phonèmes placés dans les mêmes conditions, se développent d’une manière identique.” (Schwan-Behrens 1963: 12). Occasionally, when the conditions are not apparent, there may be some concern until further research uncovers the conditioning factors. Such was the case with Verner’s law, for example. Yet in any interesting situation, there remains a body of forms for which the proper phonetic environment cannot be found. The items just do not follow the normal phonetic “laws.” In these cases, analogy is often called on to explain the discrepancy, and also to account for the reintroduction of regularity into a system “ravaged” by the forces of phonological change; “L’analogie joue un rôle considérable en roman (et en français) qui s’est reconstitué, comme on l’a dit, sur les ruines du latin où les ravages d’une évolution phonétique brutale avaient entièrement boulversé le système des oppositions morphologiques et de leurs valeurs” (Guiraud 1965: 63).