This model explains variations in sounds by tracing changes in phonemes or speech sounds over time and across languages.