This paper investigates the origin of the Old Latin amāssō-type formations (e.g., amāssō, intrāssis, cenāssit, etc.) and their associated subjunctives (e.g., amāssim, cūrāssīs, seruāssīt, etc.), which correspond in meaning and function to sigmatic future indicatives such as faxō, respexis, rupsit, etc. and subjunctives such as faxim, dīxīs, surrepsīt, etc. Although there is general agreement that these forms are of analogical origin, most previous accounts fail to cogently explain the occurrence of geminate -ss- following a long vowel. Re-evaluating the distribution of intervocalic sibilants in pre-Classical Latin, including the treatment and nativization of Greek loanwords with intervocalic s, this paper argues against deriving amāssō from the present stem amāplus the -s- of faxō, respexis, rupsit, with some exceptional treatment of -s-. Instead, it proposes that amāssō reflects a morphological reformation based on the ‘t-stem’ (amāt-) of the perfect passive participle. By analogy with forms such as faxō (interpreted as underlying /fact-sō/), speakers created /amāt-sō/, which developed regularly into amāssō. This analysis not only resolves the formal anomaly of the amāssō type but also underscores the synchronic productivity of the participial t-stem in Latin verbal and nominal morphology.
Stefan Höfler (Fri,) studied this question.