Date of Award

January 2017

Document Type


Degree Name

Master of Arts (MA)



First Advisor

Stephen A. Marlett


Murui is a Witotoan language spoken in Colombia and Peru. This thesis focuses on alternations of voiced and voiceless alveolar and velar stops in Murui that occur at some morpheme boundaries in verbs. The alternations of the voiced and voiceless alveolar stops occur in the active indicative suffix allomorphs [‐dɯ] ~ [‐tɯ] ~ [‐d] ~ [‐t] and the alternations of the voiced and voiceless velar stops occur in the passive indicative suffix allomorphs [‐ka] ~ [‐ɡa]. These stops may also become voiced or voiceless when other suffixes are present in the verb. I focus my analysis on the inflectional morphology of active and passive non‐causative and causative verb forms. I discuss the inflectional morphology of future, negative, andative, venitive, and desiderative verb forms, and I claim that my analysis applies to other verb forms in Murui as well.

The data for this thesis is found in a word list compiled by Shirley Burtch and found in S. Burtch (1978). I use a model of word‐based morphology for my analysis. In this model of morphology word forms are put into paradigms and the recurring morphological patterns of a language are visible through these paradigms. I use four‐part proportional analogies to deduce unknown verb forms from known verb forms.

I make two claims. First, I claim there are two verb classes in Murui. Class I verbs have the voiced allomorphs of the active indicative suffix [‐dɯ] ~ [‐d] and the voiceless allomorph of the passive indicative suffix [‐ka]. Class II verbs have the voiceless allomorphs of the active indicative suffix [‐tɯ] ~ [‐t] and the voiced allomorph of the passive indicative suffix [‐ɡa]. I claim that the principal parts of the verb classes are the third singular active indicative forms and I show that each verb form is deducible through proportional analogy to the principal parts of each verb class.

Secondly, I claim that when other suffixes precede the active and passive indicative suffixes, the suffix which immediately precedes the active or passive indicative suffixes in the verb structure determines the voicing of the stops in these suffixes. When a suffix precedes the active and passive indicative suffixes it may condition the voicing of the stops so that the two verb classes are no long distinguishable. I conclude that these alternations are not phonological but morphological and I contrast my analysis with a previous analysis of Murui done by Petersen de Pi񥲯s (1994). I claim that the morphological analysis is the better analysis for these alternations in Murui and I draw attention to the fact that this pattern of alternations is also seen in other Witotoan languages.