"The Concept of Phonetic Drift"

The idea of the drift goes back to Sapir and has often been used in diachronic studies.  "Drift" is a more substantive concept than "tendency," for "tendency" is simply a cover term for a group of related phenomena and has no explanatory value.  Yet "drift" appears to be a kind of linguistic perpetuum mobile (hopefully, not a kind of phlogiston) and should be resorted to with caution.    Major phonetic changes are usually caused by commands from the higher levels of language, but, once the stone begins rolling downhill, its movement becomes to a certain extent predictable.  Although the existence of a drift can never be proved, the possibility that we are dealing with a "rolling stone" should be considered, however long the process may take.  In my presentation, I will touch on the voicing of fricatives in the history of the Germanic languages, beginning with Verner's Law and up to the rise of voiced consonants in English words like /o*f*/, /*th*e/, and /i*s*/.  All those processes seem to be the consequence of a "command" given about 2000 years ago by the First Consonant Shift, a major event in the restructuring of the entire system of Old Germanic phonemes.  I am fully aware of the speculative nature of my reconstruction.