Detecting word endings in an unknown script

Date: Around 2200 BC.Location: Western, southern and eastern Iran.Type: Syllabic Script.Text Corpus: 22 (known a long time), plus 15 (known since 2015).Sign Corpus: 110 sign type, 1340 sign tokens.Status: Principally undeciphered, except the sound values for in, šu, uš, ši, na, and k, drawn from the divine name Inšušinak found in the only bilingual inscription. Several further sound values were proposed. In our paper, some of them are being corroborated, and a new one is presented.Language behind the signs: Based on graphotactical patterns found in the texts, this paper claims that it must be Elamite or a language closely related to it