Varieeruva vältega sõnade taju ja akustika

Acoustically as well as perceptually, Estonian quantity degrees are a complex phenomenon, depending on durational differences, pitch contour, intensity, sentence context, etc. The aim of the article is to find out how much variation there is in the perception of quantity degrees, what causes the variation, and how perception correlates with the acoustic features of quantity degrees. To achieve this goal, we conducted a perception test with 216 participants in which the participants were asked to evaluate which quantity degree (second or third) they heard in the word. Each word was pronounced with four different human voices and five different syntactic voices, and the same pronunciation case was asked to evaluate with and without the sentence context. Quantity degree perception varies from person to person, and listeners themselves often change their determination depending on whether they heard the word in or out of context. The listener’s own quantity degree preference has no more influence on the assignment of a pronunciation heard in the context of a sentence than on the assignment of a pronunciation heard out of context. A word heard in context is perceived more in the third quantity degree than the same word heard out of context. The listener’s dialectal background may in some cases be important for quantity degree discrimination. In human voiced speech, the values of the acoustic features varied widely, with the mean values of the basic quantity degree duration ratio being more characteristic of the third quantity degree. In synthetic speech, quantity degree ratings were more in line with acoustic feature values. The quantity degree classification ability on the basis of the acoustic features was moderate.