Abstract
The researcher of most varieties of Mixtec (Oto- Manguean) is faced with a high
level of surface alternations whereby the tones of some words vary according to their
context. Early researchers, such as Kenneth L Pike, accounted for these differences
by assigning morphemes to different classes according to the effect morphemes have
on the following morpheme. However a much more satisfactory explanation can be
achieved by positing the presence of floating tones which are the result of one of
three processes: right-ward shift of underlying tones, loss of CV segments, or the
delinking of Low tones.
The main focus of this thesis is to account for the tonal association patterns of Southeastern
Nochixtlán Mixtec (MXY). As background we present a brief summary of
the work of earlier researchers, including claims about the relationship between
stress and High tone, (Chapters 2 and 3), and then in Chapter 4 we show how
autosegmental phonology provides a more satisfactory account for these published
data. In Chapter 5 we show that in spite of surface differences, when the surface
tones of morphemes are compared across varieties, morphemes can be shown to
belong to tonal categories which reflect a previous stage of Mixtec. Chapters 6 to 10
present unpublished data from MXY. In Chapter 6 we show that underlying tones of
disyllabic morphemes usually align at the right edge of their sponsoring morpheme.
We also demonstrate how tones are provided for the unspecified initial syllables. In
this chapter we also present acoustic data to show that underlying Mid tones
participate in phonological processes which default Mid tones do not. Chapter 7
presents more data to show MXY tonal association patterns, including the behaviour
of floating High tones such as their tendency to align at the right edge of prosodic
words. Chapter 8 describes the complex tonal association of floating High tones
sponsored by four verbal prefixes. We show that the resulting surface forms depend
not only on the underlying tones sponsored by the verb root, but the form of the verb
stem with which it associates. In Chapter 9 we turn to examine whether there is any
relationship between stressed syllables and High tone. We analyse acoustic data to
show that initial syllables of roots show statistically significant increased duration.
By examining the tonal association patterns, we conclude that in MXY, the
association of High tones is governed by alignment rather than the locus of stress.
Based on the findings of Chapter 9, in Chapter 10 we look at the locus of stress in
two different contexts: one, stress found in compound words; and two, stress in verb
roots which co -occur with prefixes. In all these contexts we see no predilection for
High tones to associate with the stressed syllable. In this chapter we also look at
contexts in which Low tone spreads. Chapter 11 pulls together the data presented in
Chapters 6 to 10 and presents them against a theoretical background of the
interaction between prosody and syntax.
