«Тахиййат»: Сборник статей в честь Н. Н. Дьякова

Does Aten Live On in Kawa (K ó wwa) m 47 n An obstacle to this rule is raised by the seated man in Meroitic inscription REM 0704. The traditional interpretation of this Meroitic character would pre- sumably result in the reading Qaw-Ata, not Qaw-Yata or Qaw-Wata . Both / y / and / w/ were available as letters in Meroitic and they were not used. (As for / aa / versus / a /, vowel length was not indicated in Meroitic.) Another phonological process might have resulted in / ww /. It is the length- ening of a / w/ to fill a vacuum. The symbol / Ø / below indicates the position vacated by the lost consonant / y/ . The position may possibly be filled by a transitional sound in Meroitic (such as a non-distinctive glottal stop before the vowel represented by the seated man [?]). Rule 3: Compensatory lengthening (/w + y/ > /w + Ø/ > /w + w/) Gw-Yaata > Gw-Øaata > Gw-Waata The /w / is lengthened as / ww / to compensate for the loss of a following consonant / y /. This / ww / would be consistent with the doubling of the medial consonant in Kówwa . It would strengthen the argument that Meroitic / qaw- ata / may be linked to the modern Kówwa . Recently, Claude Rilly (1999) considered an argument by G. Böhm and demonstrated that the Meroitic letter Q was phonetically a labiovelar sound / k w /. Thus, / qaw-ata / would be represented as / k w aw-ata /. He also demon- strated that it coloured the following vowel. Thus, the Meroitic Qes found in lines 3 and 4 of the stela of Akinidad (REW 1003 in Leclant et al. 2000, 1380) would be read not as Ke:sh, but as Ku:sh or Ko:sh, the name of the country. Similarly, the impact of /k w / upon a following / a / would probably colour it as an / o /, e.g. / k w aw-ata / > / k w ow-ata /. Influenced by the Meroitic Q , this vowel is consistent with the / o / in the first syllable of Kówwa . Rilly’s demonstration can therefore be used to strengthen the argument that the Meroitic / qaw-ata / may be linked to the modern Kówwa . With the application of Rule 3 above, / k w ow-ata / would become / k w ow- wata /. Since the Nile Nubian languages do not have labiovelar consonants, this name would be reduced to / kow-wata / in Nubian. The Nubians would probably by that time have lost the awareness of an original word division ( - ) between the ancient elements of the name ‘Finding’ and ‘Aten’. Therefore, / kow-wata / should be represented as / kowwata /. Priese suggested that / ta /, the final syllable of the name, was lost. The loss of a / ta / might arguably have been reinforced among Arabic-Nubian bilinguals by a similar reduction of the word-final tā marbū ṭ a in formal Arabic (/ t/ > / h / when not supported by following elements such as case suffixes). Whether or not Arabic had this impact on Nubian, the loss of a final syllable is a familiar process of sound change. Change or loss of a final / t / in names written in