Since this is my first post on the matter I'll briefly explain the purpose of the project which I am titling 'Perception of Prominence Correlates'. The goal is to see what acoustic values exactly drive perception of meaning-based acoustically marked words in an utterance. Specifically, I am looking at words in subject position in contrastive focus vs. broad focus. E.g. the response "Did Mary see the movie" and response "No, *John* saw the movie" vs. "What happened yesterday" "*John* saw the movie". Where in the first case John is acoustically marked--the word may be uttered with more acoustic intensity, longer, and with a sharper or higher pitch accent than in the second situation.
To do this task, I collect these two responses and iteratively resynthesize the contrastive condition to be like the broad condition. I am able to do this with word duration and the pitch contour (both of which, I am still fine-tuning) however, I don't have a good way to do this for intensity. I may try some things if I have time but otherwise the experiment will be run without those. I am using LMEDS to run this experiment.
From this experiment, broadly speaking, I hope to find out what triggers the perception of contrastive focus. What drives perception? Are the cue thresholds (largely) consistent within a person? Across people?
To do this task, I collect these two responses and iteratively resynthesize the contrastive condition to be like the broad condition. I am able to do this with word duration and the pitch contour (both of which, I am still fine-tuning) however, I don't have a good way to do this for intensity. I may try some things if I have time but otherwise the experiment will be run without those. I am using LMEDS to run this experiment.
From this experiment, broadly speaking, I hope to find out what triggers the perception of contrastive focus. What drives perception? Are the cue thresholds (largely) consistent within a person? Across people?
No comments:
Post a Comment