This study examined the relationship between lung capacity, speech volume and duration of utterances. 8 adult subjects, 4 males and 4 females (24-36 yrs), participated. By breathing into a digital spirometer, lung capacities such as Vital Capacity (VC) and Inspirational Capacity (IC) were estimated. Respiratory movements were measured using Respiratory Inductance Plethysmography (RIP), and these respiratory movements were calibrated in litres using the spirometer. The proportion of lung capacity used for speech was estimated from the calibrated RIP signals during reading of a nonsense text without punctuation marks. This proportion was defined as the median volume of exhaled air per utterance (SV) (in litres) during text reading, divided by the speaker’s VC (SV/VC). Utterance durations (UD) and Respiratory Rates (RR) were estimated from acoustic recordings of the text readings as well as the RIP signals, displayed in Praat. This allowed investigating the relationships between lung capacity, respiratory rate, utterance durations as well as the proportion of lung capacity used for speech. Put differently, the question was whether people with larger lungs speak in longer utterances and inhale less frequently, as well as whether people with smaller lungs use a relatively larger proportion of their lung capacity for speaking. Additionally, where SV initiated (SVIN) and terminated (SVTER) within VC was calculated based on the RIP signals. There were no significant relationships between VC and UD or RR. In addition, there was no significant relationship between SV/VC and VC. SVIN ranged from 43%-71% and SVTER ranged from 17%-55%. The results indicate no relationship between VC and UD or RR nor that people with a smaller VC use more of it for speech. The range of SV within VC suggest that people maintain a fairly relaxed state with regards to muscle pressure.

The present study is the first to examine the effect of first versus second language (L1 versus L2) speech production on specific temporal parameters of speech in bilingual speakers with neurogenic speech disorders. Three persons with apraxia of speech (AOS), three with phonemic paraphasia (PP) and five normal speaking participants were included as subjects in the study. Subjects were required to read phonemically similar L1 and L2 target utterances in a carrier phrase, five times each, at a normal and fast speaking rate, respectively. This rendered four speaking contexts that included speech production in L1 at either a normal (L1NR) or fast speaking rate (L1FR) and speech production in L2 at either a normal (L2NR) or fast speaking rate (L2FR). Acoustic analysis of on-target productions involved measurement of utterance onset duration, vowel duration, utterance duration and voice onset time. Results revealed that in normal speakers, speech production in L2 results in greater token-to-token variability than in L1. However, token-to-token variability in the experimental subjects did not tend to increase whilst speaking in L2, most probably because these subjects generally decreased their speaking rate in this context, resulting in more consistent production. The subjects with AOS and PP seemed to be influenced by the increased processing demands of speaking in L2 to a greater extent than the normal speakers, in that they more frequently experienced difficulty with durational adjustments (decreasing duration in the fast speaking rate) in L2 than in L1. Furthermore, the subjects with AOS or PP also exhibited a greater extent of durational adjustment in L1 than in L2. The durations of most of the subjects with either AOS or PP tended to differ from those of the normal group to a greater extent in L2FR that was hypothesized to be the most demanding speaking context for these subjects. The longer than normal durations and greater than normal token-to-token variability in the subjects with either AOS or PP imply the presence of a motor control deficit. The extent of the motor control deficit appears to be more severe in AOS than in PP as is evident from the finding that the subjects with AOS generally exhibited longer durations and greater token-to-token variability than the subjects with PP. The pattern of breakdown in respect of different parameters and utterance groups also differed between subjects with AOS and PP. The nature of the disorder in AOS and PP thus appears to be both quantitatively and qualitatively different. Regarding measurement of the different temporal parameters, voice onset time appears to be less subject to the influence of L2 than the other measured temporal parameters. The results of this study imply that bilingual AOS is as much a reality as bilingual aphasia. Furthermore, the results underscore the importance of taking contextual factors, specifically L1 versus L2, into account when compiling assessment and treatment procedures for persons with either AOS or PP, since speech production in L2 appears to be motorically more difficult than in L1 for persons with neurogenic involvement. The significance of the results is discussed with reference to the influence of speech production in L2 on temporal control and the underlying nature of AOS and PP with regard to theories of speech sensorimotor control. Copyright / Dissertation (DPhil (Communication Pathology))--University of Pretoria, 2004. / Speech-Language Pathology and Audiology / unrestricted