The parallel evolution of the construction of perfect and modal future from Sanskrit to early Indo-Aryan, in the West as in the East, shows that both originally represented "views on action" rather than actions (Benveniste, Kurylowicz), structured as locational predications (by/to-me X done ; by/to-me X to-be-done). This view is confirmed by the locative origin of modern case markers for ergative, a construction directly inherited from the Sanskrit nominal sentence. Further these markers happen to also mark the "subject" of modal sentences in Marathi/Gujarati, wether transitive or intransitive. The history of ergativitiy in Indo-Aryan then shows not a "reversed" image of the active construction of transitive verbs (Dixon) but rather an 'inversion' of subject marking for stative locational prodicatios, due to the absence of 'have' verb. This phenomene, which occured both in the West (modern ergative languages) and in the East (modern non ergative languages), shoud be considered in parallel with the development of the so-called "dative subject".
