| This dissertation presents a constraint-based theory of consonant-tone interaction called the 'Extended Tone Bearing Unit Theory (xTBU theory)'. In this presentation, I provide synchronic evidence from a variety of languages that shows how laryngeal features can influence tone. These laryngeal features include voicing, aspiration, glottalization, and voicelessness.;There are three parts to xTBU theory: representation, markedness constraints, and faithfulness constraints. I propose that while tone prefers to associate to moras, tone can also be associated to non-moraic root nodes. I argue that tone directly associated to root nodes acts differently from tone associated to moras due to the interaction of constraints.;The presence and type of directly associated tone is restricted by markedness constraints only. Constraints such as *[+SPREAD GLOTTIS]/L can prevent H tone from spreading over spread glottis consonants; it can also force L tone to change to H, and even force consonants to change their laryngeal features. A significant result is the ability of the constraints to account for apparently unnatural classes of tone-affecting consonants.;An important aspect of the constraint formalism in xTBU theory is that faithfulness constraints only target tone associated to moras; they never target tone associated to non-moraic root-nodes. The lack of faithfulness to directly associated tone means that tone is not contrastive when associated to non-moraic root nodes. Consequently, tone on non-moraic segments is only regulated by markedness constraints. |
