A text-based encoding of how words sound, showing the individual speech sounds rather than the written spelling.