A Novel Prosodic Morphology Account for Verlan

This paper continues the discussion of the French word game Verlan. Since the 1980s when Verlan first began being played (Fagyal 2003), there have been many different theoretical descriptions of the game. Some descriptions include a prosodic account of metathesis (Lefkowitz 1991), templatic systems (Azra and Cheneau 1993, Plénat 1992, 1995), and Optimality Theory analyses with ranked constraints (Gotowski 2019). I continue the discussion with a novel analysis that unifies all types of movement seen in Verlan and that applies the ideas of prosodic circumscription of the base algorithmically (McCarthy and Prince 1990). I propose an algorithm in three steps. The first step is PARSE, where the input is demarcated into its constituents. If the input is one or two syllables in length and the final syllable is closed, the historic /œ/ vowel surfaces in the derivation and PARSE is run again with the larger input. The second step is TARGET, where the largest, self-contained prosodic constituent on the right edge of the prosodic word is selected. The final step is MOVE, where the target is moved to the left edge of the prosodic word. This step is synonymous with “copy and delete” as presented in Bagemihl (1988). This algorithm applies to all instances of Verlan: open and closed syllables ranging in size from one to three syllables. Specifying the parameters for movement disambiguates the treatment of consonant clusters, possible and impossible derivations in three-syllable words, generates reverlanized forms through reapplication of the algorithm, and clarifies examples that were previously analyzed as exceptional or ungrammatical.