Date of Award

Spring 5-20-2019

Document Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Stephen Anderson

Second Advisor

Claire Bowern


Mandan [ISO: mhq] is a Siouan language traditionally spoken in northwestern North Dakota on the Fort Berthold Indian Reservation. The language no longer has any L1 speakers, and fewer than a dozen L2 speakers remain. This dissertation provides a description of the phonological and morphological systems of the language, as well as contextualizes these systems within a formal framework. The data come from an assembled corpus of over five hundred pages of transcribed traditional narratives and twenty hours of recordings of those narratives done in the 1970s, which is supplemented by data from more recent fieldwork done in the early 2010s and by the author. Mandan features a system of templatic morphology, whereby affixes are ordered according to a proscribed order and whose linear order does not align with the semantic scope of those affixes. This dissertation proposes a morphologically-motivated reason for how these affixes come to appear in the order they do.

Chapter 1 provides an introduction to the Mandan people and their language. The ethnographic content of this chapter shows the relationship between the Mandan people and the Upper Missouri region of the Great Plains, and the unfortunate decline in their population due to their status as organizers of one of the largest hubs of trade near the center of the North American continent and the successive waves of smallpox epidemics brought to their villages due to this trade. After discussing previous research on the Mandan languages, the chapter concludes by posing three questions regarding whether affix order in Mandan is truly unmotivated and if not, why affixes appear where they do within the verbal template. Chapter 2 gives an overview of the phonological inventory of Mandan, describing its consonants and vowels, as well as providing a look at phonetic characteristics of Mandan consonants. This overview includes an analysis of several phonological processes that are blocked by some non-featural element in the vicinity of preverbs, and the chapter concludes by ruling out phonology as a motivating factor for affix order.

Chapter 3 identifies and describes individual affixes and clitics found on Mandan verbs. This description of verbal morphology includes reanalysis of elements previously considered to mark tense, but which actually mark aspect or evidentiality. Further analysis on postverbal elements reveals that Mandan has an extensive enclitic field rather than a large number of suffixes. Ordering of these enclitics can change depending on semantic scope, but the ordering of prefixes remains unchanged. The conclusion of this chapter states that a syntactic approach also cannot yield a convincing result for why the prefix field is ordered the way it is, so there must be a factor beyond phonology or syntax that is regulating the ordering of prefixes. Chapter 4 argues that since the prefixes in Mandan are true prefixes, there must be a motivating factor for their organization within the template. Previous accounts of clitic order has had little controversy over ad hoc ordering of clitic fields, so the question is raised: why the controversy over affix ordering? If clitics are the morphology of the phrase level, then affixation is the morphology of the word level. The argumentation in this chapter holds that Mandan has a templatic morphology because of the diachronic reanalysis of preverbal elements as being integral parts of the verb rather than independent elements, causing an expansion of brackets around that material while maintaining the old bracketing of the head word. These composite constructions allow for multiple word edges within a single morphological word and, as such, affixes are able to target a specific edge within the composite word. This accounts for the linear order of prefixes in Mandan, as well as in other Siouan languages using an OT framework to align prefixes to their target word edges.

The dissertation concludes in Chapter 5 that templatic morphology in Mandan is definitely motivated by some principles, but those principles are purely morphological. By virtue of templatic morphology being motivated, we can then say that all language with affixal morphology has some kind of motivation for the order of affixes. We can extend this composite analysis to templatic morphology in general By looking briefly at several languages from different families that have templatic morphology, revealing that languages with composite words are not uncommon, and that this composite structure may be the cause of unexpected phonological or morphological phenomena in languages around the world. As such, this dissertation serves not just as the first detailed account of the grammatical system of Mandan, but as an attempt to show the similarity between the behavior of clitics and affixes in that they both are able to target the edge of a specific domain: clitics target an edge within the domain of a phrase, while affixes target an edge within the domain of a word.

Open Access

This Article is Open Access

Included in

Morphology Commons