ø. Introduction

The reflection on whether and on what grounds the absence of a sign turns out to be as just significant as the sign itself inevitably leads, in the linguistic field, to concentrating one's attention on the heterogeneous set of linguistic phenomena which modern linguistics generally refers to with the term ‘zero’. Examples are easily found in many languages, such as in Engl. sheep (sing.) vs. sheep (pl.); cut (present) vs. cut (past) or, even more interesting, as it involves a transcategorisation, cheat (noun) vs. cheat (verb). In all these and many other examples the ‘absence of an otherwise necessary sign’ to stay with Whitney's words may be recognised either through analogical reasoning (which allows one to postulate for example the necessity of an s to mark the morphological function of plural as in brook vs. brooks etc.) or through opposition which highlights the morphological role of some absences of sign like in Czech žena ‘woman’ (nom.sg.), ženy ‘women’ (nom. pl.) opposed to žen ‘of women’ (< asl. ženu).

Similar phenomena have also been taken into account in the first Sanskrit grammar ever written, the Aṣṭādhyāyī of Pāṇini (4th c. BC) in 239 rules involving the technical device of lopa lit. ‘suppression, erasing’. 212 of these rules are operative ones, of which 95 can be considered as teaching the zero of sounds (either a single sound or sequences of sounds), 117 the zero of morphs (either inflectional or derivative affixes), rule the zero of an inflected word (pada), and 1 rule the zero of its, i.e. of the so-called ‘markers’ used by Pāṇini to connect some rules to some specific units.