The evolution of human speech

As we have noted throughout this book, human speech appears to involve a number of innate, i.e. genetically transmitted, anatomical and neural mechanisms. The plurality of these mechanisms is consistent with the mosaic nature of evolution (Mayr, 1982), which basically states that living organisms are put together in genetically transmitted “bits and pieces.” Some of these mechanisms appear to be species-specific. Other aspects of human language, which are beyond the scope of this book, like rule-governed syntax, may also involve species-specific neural mechanisms. These species-specific anatomical and neural mechanisms probably evolved to enhance human linguistic ability. As we noted in Chapter 7, human speech allows us to transmit phonetic “segments” at a rate of up to 25 segments per second. In contrast, it is impossible to identify other nonspeech data at rates that exceed 7–9 items per second. A short two second long sentence can contain about 50 sound segments. The previous sentence, for example, which consists of approximately 50 phonetic segments, can be uttered in two seconds. If this sentence had to be transmitted at the rate of nonspeech, it would take so long that a human listener would forget the beginning of the sentence before hearing its end. The high data transmission rate of human speech thus is an integral part of human linguistic ability that allows complex thoughts to be transmitted within the constraints of short-term memory.