DescriptionMorphology and syllable weight have both been shown to affect stress patterns, but these effects are analyzed in different ways. The theoretical goal of this dissertation is to propose a Optimality Theoretic model to account for how morphology influences stress, and to do this in a way that parallels the influence of weight upon stress. Prince (1990) lays out the WEIGHT-TO-STRESS PRINCIPLE, formalizing the principle by which heavy syllables attract stress in quantity-sensitive systems. I argue for the MORPHEME-TO-STRESS PRINCIPLE, a constraint that forces morphemes to attract stress in morphological stress systems. The WEIGHT-TO- STRESS PRINCIPLE has a counterpart, the STRESS-TO-WEIGHT PRINCIPLE, which forces stressed syllables to be heavy. The counterpart of the MORPHEME-TO-STRESS PRINCIPLE is the STRESS-TO-MORPHEME PRINCIPLE, which forces stressed syllables to belong to morphemes. This accounts for systems where epenthetic vowels resist stress assignment.The model proposed here has the following consequences. First, the MORPHEME- TO-STRESS PRINCIPLE can be invoked to account for the prosodic rooting constraint (as in Hammond 1984; or as in the LXWD=PRWD constraint of McCarthy and Prince 1993). This one constraint handles word minimality as a morphological effect, just as it accounts for the assignment of stress on morphological grounds in nonminimal contexts. Second, the formalization of the MORPHEME-TO-STRESS and the STRESS- TO-MORPHEME PRINCIPLES treats MORPHEME and STRESS as variables. I claim that all logical possible relationships are attested for these three variables: MORPHEME, WEIGHT, and STRESS. Preliminary results (in Chapter Five) suggest that each possible ordering appears to be attested.The foundation of the theoretical work is a description of the secondary stress patterns in Tohono O'odham, a Uto-Aztecan language formerly known as Papago. This description reveals that the primary way to predict the stress pattern of a word is the morphology. Words may surface with varying stress patterns depending on the number of morphemes, the presence of epenthetic vowels, or whether a word has been morphologically truncated. The descriptive work is the result of my fieldwork on Tohono O'odham.