The Third Pillar of the Altaic Hypothesis
(45th Meeting of the PIAC, Budapest 2002)
Among the still debated propositions of the “classical”, i.e. Ramstedt-Poppe reconstruction of Altaic phonology, ‘zetacism’ / ‘sigmatism’, resp. ‘rhotacism’ / ‘lambdacism’ are most notorious, having been dubbed the “twin pillars” of the Altaic hypothesis either in irony, by those contesting the genetic relationship of the languages involved, or in earnest, by those still holding the Ramstedt-Poppe position. But their reconstruction has yet another “pillar”, hardly noted in the controversy about Altaic to date, i.e., Poppe’s postulation of a significantly variable position for a suprasegmental pitch- or stress-feature that would then account both for certain instances of vowel length and for a number of forms where Mo. –g- apparently corresponds to Tk –b-. The great of this third “pillar” has of course been the lack of any early evidence for suprasegmental phonemes in the languages involved. But in Korean we do have such evidence, unknown to Ramstedt, thanks to the elaborately precise “tone marks” of the Middle Korean written records. The paper will attempt to demonstrate how these resources, mostly ignored in the literature to date, may help to support this “third pillar” of Ramstedt-Poppe.