Perception of speech is effortless under most circumstances; this may explain why the complexity of this cognitive process is so often unappreciated. While most other sounds in the environment are acoustically very distinctive, the 30 or 40 speech sounds people make are discriminable only by rapid analysis of rather subtle acoustic cues. Given the importance of language, it seems plausible not only that our auditory system evolved a degree of specialization for handling such analyses, but also that this specialization should assimilate substantial neural resources. Recent work in comparative physiology and human brain imaging suggests, in fact, that much of the superior temporal gyrus in humans is devoted to this perceptual feat.