This study is an investigation of the expression of negation in the history of Greek, through quantitative data from representative texts from three major stages of vernacular Greek (Attic Greek, Koine, Late Medieval Greek), and qualitative data from Homeric Greek until Standard Modern. The contrast between two complementary negators, NEG1 and NEG2, is explained in terms of sensitivity of NEG2 μη to nonveridicality: NEG2 is a polarity item in all stages of the Greek language, an item licensed by nonveridicality. The asymmetry in the diachronic development of the Greek negator system (the replacement of NEG1 and the preservation of NEG2) is explained with reference to the particulars of the uses of NEG2, specifically the inertial forces drawn by the nonnegative uses of NEG2, which being nonnegative did not experience the renewal pressures predicted by the Jespersen’s Cycle. These are its complementizer uses: (i) as a question particle, and (ii) in introducing verbs of fear complements. A viewpoint for Jespersen’s Cycle is proposed that abstracts away from the morphosyntactic and phonological particulars of the phenomenon and explicitly places its regularities in the semantics, accommodating not only for Greek, but for numerous other languages that deviate in different ways from the traditional description of Jespersen’s Cycle. The developments observed in the history of the Greek negator system agree with current generative theories of syntactic change, regarding the notions of up-the-tree movement.
Inertial forces and the foundations of optical geometry