Lung Volumes and Linguistic Units as Predictors of Respiratory Pattern in Speech

M Denny


Investigations of lung volume control during speech have shown considerable unexplaiined variability.  Likely predictors of expired volume for a breath group include volume at inspiratory onset, inspired volume, and end-inspiratory volume.  Predictors of inspired volume may include expired volume for the preceding breath group and an estimate of the number of linguistic units to be produced. This investigation used multiple regression to evaluate the relative influence of these variables on speech breathing. Twenty normal subjects, instrumented with a respiratory inductance plethysmograph, described a simple map and told a story from a picture book.  Inspired and expired volumes were measured and expressed as percentages of vital capacity; transcriptions of speech were made and divided into clauses and higher-order, multi-clause linguistic units, called thought units and conversation units.  Analyses to date indicate no differences in mean volumes or clauses per breath group for the two tasks.  Multiple regression revealed that inspiratory onset volume and inspired volume predict expired volume for the following breath group; end-inspiratory volume does not.  Final regression analyses will add counts of linguistic units per breath group to the predictor variables.