The accentual system
The general limiting rule
A binary accentual opposition in Attic was recognized from the time of Plato (Cratylus 399 A). A term oksUs ‘acute’ was applied to the positive, culminative feature occurring only once in each full word (= high pitch); and the term barus ‘grave’ was applied to the negative, non-culminative feature occurring in all the other syllables (= low pitch) with the exception of the syllable immediately following the high pitch. The post-tonic syllable carried a falling glide starting at a high pitch and finishing low. The accentuation rules and graphic notations for different types of accent were elaborated by Aristophanes of Byzantium (about 200 BCE) in order to teach foreigners the “correct” accent in pronouncing the Attic based Hellenistic Koine. His rule of recession, limiting the distance of the accent from the end of the word, is presented in various school grammars in lengthy and atomistic statements referring to the last three syllables. Jakobson ( 1971: 263) simplified considerably the rule of recession by introducing the concept of mora: “The span between the accented and the final mora cannot exceed one syllable”. Allen (1973: 234) further simplified the rule by introducing the new notion of “contona- tion” (= the combination of high pitch + falling glide): “Not more than one mora may follow the contonation”. In Attic (and apparently in other Ionic dialects) the phonological opposition between the falling melodic pattern (= circumflex) and the high pitch without falling glide (= acute) only existed in long vowels (or diphthongs) in final syllables as in the following minimal pairs: phos ‘light’ versus phos ‘man’; heis ‘one’ vs. heis ‘having sent’ (Part. Aor. of h^ёmi); theas ‘of the goddess’ vs. theas (Acc. Pl.). In the penultimate syllable this alternation was predictable since it was governed by the general limit?ing rule. For instance, the pair oikoi ‘houses’ vs. oikoi ‘in the house’ features the mono- moric /oj/ in the plural vs. the dimoric /o+i/ in the locative singular: /oikoj/ vs. /oiko+i/. In Aeolic the accent was not contrastive as in Attic since it recedes to its full limit which makes the mora irrelevant for the accentual description (cf. Garde 1968: 148). Thus the following pairs of words would be homophonous in Aeolic (pronounced as in the first of each pair): tomos ‘a cut’ and tomos ‘cutting’; ekhthra ‘hatred’ and ekhthra ‘hated’ (Fem); therotrophos ‘feeding on beasts’ and therotrophos ‘feeding wild beasts’.