The lexical prosody of Digor has long been a subject of disagreement among researchers. While some have reported Digor primary stress to fall on the last strong vowel of the word, others have reported the Iron and Digor systems to be identical (with the exception of a larger stress window in the latter), implying that the accent is placed on the first strong vowel. A previous acoustic study of Digor words containing the same vowels in isolation and in utterance-final position showed vowel duration to increase gradually towards the end of the word. This study aims to add nuance to the existing knowledge of Digor prosodic patterns by considering more types of word structure and the influence of isolated vs utterance-medial environments.

The phonological "strength" of a vowel was established as the deciding factor in its duration: strong vowels are consistently longer than weak ones regardless of their position. When vowels of the same phonological status are considered, the aforementioned lengthening towards the end of the word is visible in both strong and weak vowels. In sentences containing words with three strong vowels in utterance-medial position, this tendency is shown by the second and third vowels, but not by the first and second. Given that the same vowel phonemes in Digor have been reported to have shorter realisations before strong vowels than before weak ones, it can be assumed that the position before a strong vowel causes a durational decrease in the preceding vowel, but the position before an already shortened strong vowel does not have the same effect. Such quantitative marking of the first syllable appears to be completely neutralised by the utterance-final (but not word-final) increase in duration, which might explain the stark differences in previous descriptions of Digor stress: although vowel duration generally tends to increase towards the right edge of the word, a minority of words with non-final strong vowels (followed by either a weak or a shortened strong vowel) have what could be perceived as word-initial stress.

Varvara Petrova is a graduate student at the Institute of Phonetics and Speech Processing, LMU Munich.