It may come as a surprise that it is possible to present even a tentative phonology of Meroitic when there are so few certainties about the language. Paradoxically, the reconstruction of the pronunciation of Meroitic is more certain than the translation of minor phrases.
The sources for this phonological system are of three kinds. The first may compare Meroitic terms with those of other languages: either those that have been borrowed in different periods from Egyptian or those that have been transcribed in Egyptian, Greek, or even Latin. There is, for instance, a Meroitic parallel kdke “Candace [Queen]” in Greek transcription: κανδάκη. This method of comparison was used extensively by Griffith for the decipherment of the script. The second may additionally examine variations in orthographic conventions, which, because of the phonetic nature of the script, give information about the similarity of the phonemes. For instance, the variant mxe for the most frequently written word mhe “abundant” suggests an articulatory analogy between /x/ and /h/. Finally, the third may attempt to use the first two methods to create a linguistic system of sounds. For instance, the presence of only four written vowels in the Meroitic vocalic system makes it more likely that there is a /u/ among them.