Marathi Tense marking: A window into the lexical encoding of temporal meaning

Ashwini Deo (Ohio State University)

Partee (1973) first observed that natural language tense expressions are analogous to pronouns in that they can be interpreted indexically, anaphorically, and like bound variables. These referential (i.e. indexical+anaphoric) and non-referential interpretations of tense marking have not been yet shown to have distinct reflexes in natural language temporal expressions – i.e. no language has been claimed to lexicalize these distinctly. I argue that Marathi [mar, 71,700,000 speakers], an Indo-Aryan language of the Southern subgroup, morphosyntactically distinguishes between referential and non-referential temporal meanings. This lexicalization pattern, observed also in several other New Indo-Aryan languages, points the way to a more nuanced understanding of distinctions with respect to temporal reference in natural languages.

The second part of the talk traces the diachronic emergence of this encoding pattern in the  New Indo-Aryan languages, comparing it to the systems observed in Old and Middle Indo-Aryan. I tentatively suggest  that the referential—non-referential contrast in the temporal domain arises  due to the emergence of new present and past tense markers (auxiliaries) in a tenseless aspectually-based system.