A framework used to analyze language that posits that different features of speech sounds are represented on separate tiers.