We human language users learn about speech all our lives. In fact if the beginning of our individual life is taken to be the moment of our birth, we learn for more than our whole lives, because even prior to birth we acquire much basic information about language sound structure. Then in early life, before we are capable of uttering any recognizable words, we have already learned a huge amount about the sounds and words of our language; furthermore, the efficiency with which we learn this is predictive of our later linguistic facility. We learn to process speech in the way best suited to the particular language (or languages) we acquire as children. This leads to a formidable efficiency and robustness in native-language processing, but it actually inhibits learning in the case of other languages we encounter only later — children are clearly very much better at learning a new language than adults are! Learning about speech nonetheless continues throughout life, in particular the everyday perceptual learning that enables us to adapt our speech processing to newly encountered talkers, and to adjust our own pronunciation in keeping with pronunciation changes in our speech community across time. This learning (at least in the native language) can draw on a wide variety of information sources, is fully in place in childhood, and is apparently unattenuated in older language users.