Abstract This article will argue that the Indo-European s-aorist is an ancient verbal category, and thus can be reconstructed for the earliest reconstructable stage of Proto‑Indo‑European. Moreover, it will argue that evidence from Vedic Sanskrit indicates that its active paradigm originally had radical *ē‑grade in its 2sg. and 3sg. forms only, whereas its other forms (including 1sg.) showed *e‑grade. Furthermore, it will argue that s-aorists formed to roots of the structure *CeCH‑, which are the ancestors of the Sanskrit iṣ‑aorists, originally had *e‑grade throughout their active paradigm, and did not contain any *ē‑grade forms. The article will furthermore discuss the possibility that this specific distribution of *ē‑ and *e‑grade forms in the s‑aorist is the result of a pre‑PIE monosyllabic lengthening rule. It will argue that additional evidence in favor of this sound law can be found in other domains of the Vedic verbal system as well as in certain distributional patterns in the Tocharian verbal system.
Alwin Kloekhorst (Wed,) studied this question.