van Lieshout P H, Starkweather C W, Hulstijn W, Peters H F
Nijmegen Institute of Cognition and Information, The Netherlands.
J Speech Hear Res. 1995 Apr;38(2):360-72. doi: 10.1044/jshr.3802.360.
In this study changes in upper lip and lower lip integrated electromyographic (IEMG) amplitude and temporal measures related to linguistic factors known for their influence on stuttering were investigated. Nonstuttering subjects first read and then verbalized sentences of varying length (sentence length factor), in which meaningless but phonologically appropriate character strings were varied in their position within the sentence (word position factor) and their size (word size factor). It was hypothesized that the production of stressed, vowel-rounding gestures of words in initial position, longer words, and words in longer sentences would be characterized by specific changes in IEMG amplitude that would reflect an increase in speech motor demands, intuitively defined as articulatory effort. Basically, the findings corroborated our assumptions, showing that words in sentence initial position have shorter word and vowel durations in combination with an increase in IEMG activity. Similarly, we found shorter vowel durations for longer words, and in sentence final position an increase in IEMG activity. For longer sentences we found a clear increase in speech rate, but contrary to our expectations a decrease in IEMG activity. It was speculated that this might relate to the use of a movement reduction strategy to allow higher speech rates with increased coarticulation. These findings were discussed both for their implications in normal speech production, as well as for their possible implications for explaining stuttering behavior. To this end our data can illustrate both why stutterers might run a higher risk of stuttering at these linguistic loci of stuttering, and why they might come up with a strategic solution to decrease the motor demands in speech production. The basic outcome of this study shows that higher order (linguistic) specifications can have clear effects on speech motor production.
在本研究中，调查了上唇和下唇的综合肌电图（IEMG）幅度变化以及与已知对口吃有影响的语言因素相关的时间测量指标。非口吃受试者首先阅读然后说出不同长度的句子（句子长度因素），其中无意义但语音合适的字符串在句子中的位置（单词位置因素）及其大小（单词大小因素）有所变化。据推测，句首位置、较长单词以及较长句子中的单词产生的重音、元音圆唇手势，其IEMG幅度会有特定变化，这将反映出言语运动需求的增加，直观地定义为发音努力。基本上，研究结果证实了我们的假设，表明句首位置的单词其单词和元音持续时间较短，同时IEMG活动增加。同样，我们发现较长单词的元音持续时间较短，而在句末位置IEMG活动增加。对于较长的句子，我们发现语速明显增加，但与我们的预期相反，IEMG活动减少。据推测，这可能与使用减少运动策略以允许更高语速并增加协同发音有关。这些发现既讨论了它们在正常言语产生中的意义，也讨论了它们对解释口吃行为的可能意义。为此，我们的数据可以说明为什么口吃者在这些口吃的语言位点可能有更高的口吃风险，以及为什么他们可能会想出一种策略性解决方案来降低言语产生中的运动需求。这项研究的基本结果表明，高阶（语言）规范会对言语运动产生有明显影响。
