There are remarkable individual differences in the ability to recognise individuals by the sound of their voice. Theoretically, this ability is thought to depend on the coding accuracy of voices in a low-dimensional “voice-space”. Here we were interested in how adaptive coding of voice identity relates to this variability in skill. In two adaptation experiments we explored first whether the aftereffect size to two familiar vocal identities can predict voice perception ability and second, whether this effect stems from general auditory skill (e.g. discrimination ability for tuning and tempo). Experiment 1 demonstrated that contrastive aftereffect sizes for voice identity predicted voice perception ability. In Experiment 2, we replicated this finding and further established that this effect is unrelated to general auditory abilities or general adaptability of listeners. Our results highlight the important functional role of adaptive coding in voice expertise and suggest that human voice perception is a highly specialised and distinct auditory ability.

在通过声音识别个人的能力上存在明显的个人差异。从理论上讲，这种能力取决于低维“语音空间”中语音的编码精度。在这里，我们对语音身份的自适应编码如何与技能的这种可变性相关感兴趣。在两个适应性实验中，我们首先探讨了对两个熟悉的声音身份的后效应大小能否预测语音感知能力，其次，该效应是否源于一般听觉技巧（例如，对音调和节奏的辨别能力）。实验1证明，语音识别的对比后效应大小预测了语音感知能力。在实验2中，我们复制了这一发现，并进一步确定该效果与一般听觉能力或听众的一般适应能力无关。我们的结果突出了自适应编码在语音专业知识中的重要功能，并表明人类语音感知是一种高度专业化和独特的听觉能力。