Burns E M, Campbell S L
Department of Speech and Hearing Sciences, University of Washington, Seattle 98195.
J Acoust Soc Am. 1994 Nov;96(5 Pt 1):2704-19. doi: 10.1121/1.411447.
The methodology derived from the trace-context theory of intensity resolution (Durlach and Braida, 1969; Macmillan et al., 1988) was applied to resolution over an octave range along two continua: a sequential-frequency-ratio continuum for possessors of relative pitch (RP), and a pure-tone frequency continuum for possessors of absolute pitch (AP). The performance of both RP and AP possessors was exceptional in that total identification sensitivity along both continua was much greater than identification sensitivity along unidimensional psychophysical continua characterized by the 7 +/- 2 rule. In addition, the performance of RP possessors was exceptional in that, on average, total sensitivity for identification resolution was greater than sensitivity for resolution in discrimination. Finally, identification sensitivity between category prototypes (chromatic semits) along both continua was approximately the same as identification sensitivity between phonemic category prototypes along speech continua, despite the fact that both the discrimination ranges and the total number of categories are much larger for the two pitch continua.
源自强度分辨率的轨迹-上下文理论(杜拉克和布拉伊达,1969年;麦克米伦等人,1988年)的方法被应用于沿两个连续体的一个倍频程范围内的分辨率:一个是针对相对音高(RP)拥有者的顺序频率比连续体,另一个是针对绝对音高(AP)拥有者的纯音频率连续体。RP和AP拥有者的表现都很出色,因为沿两个连续体的总识别灵敏度远高于沿以7±2规则为特征的单维心理物理学连续体的识别灵敏度。此外,RP拥有者的表现也很出色,因为平均而言,识别分辨率的总灵敏度高于辨别分辨率的灵敏度。最后,尽管两个音高连续体的辨别范围和类别总数都大得多,但沿两个连续体的类别原型(半音)之间的识别灵敏度与沿语音连续体的音位类别原型之间的识别灵敏度大致相同。