Overview Summary eBook 53 Beal Soyaltepec Mazatec

297 CHAPTER 6 CONCLUSION

6.1 Overview

In this chapter, I conclude the work of this dissertation which has documented the sound system of Soyaltepec Mazatec. The complexity of the tone system of Soyaltepec Mazatec has provided ample opportunity to explore the geometry and features of tone suggested by RTT. After summarizing the main points discussed above, I discuss the advantages of adopting Register Tier Theory to describe the tone in Soyaltepec Mazatec. I also point out that while the processes discussed above generally support the adoption of the UAC in Soyaltepec Mazatec, there are some issues which preclude its strict application. Next, I comment on the placement of Soyaltepec Mazatec in traditional tonal typology. Finally, I propose topics for future research.

6.2 Summary

This dissertation set out to describe the basic phonology of Soyaltepec Mazatec by investigating the segmental and tonal elements. In Chapter 2, I detailed the primary literature available and phonological analyses that utilize Soyaltepec Mazatec data. Pike 1956 initially described Soyaltepec Mazatec tone, but did so without differentiating between phonological and phonetic processes. Her article demonstrates the incredible number of surface variations that one underlying tone can produce. Several of the tonal changes that Pike portrays appear to describe differences in the phonetic realization of tones rather than phonological processes. I include all of the alterations caused by single tones which were discussed in §2.2.1 in this category, as well as the sandhi that was initiated by changes in phrasal position which was discussed in §2.2.3.2. Pike also described behavior which is more likely phonological in nature, in particular the behavior surrounding the contour tones. The behavior she described mostly hinted at processes typical of contours composed of individual segments; however, some of the behavior of the Mid- 298 high to High contour as described by Pike is potentially explained if the contour tones are unitary. The composite behavior was explored in analysis by Biber 1981 and Goldsmith 1990. Pike’s data helped ferment and exemplify phonological thinking regarding the nature of contour tones. Building on these analyses, Pizer 1994 proposed an expanded tonal geometry to allow for a structure that was able to both reflect this sequential nature of tone, while also allowing unitary behavior. Pizer’s analysis is interesting and accounts for more of the processes described by Pike than her predecessors, but it does not solve all of the problems introduced by Pike’s data, specifically, concerning the behavior of the Mid-high to High contour. In order to understand this complex system, I collected fresh data. The basic non-tonal segments were introduced and used in Gudschinsky 1959, Pike 1956 and Kirk 1966. While the segments were previously introduced, they were never exemplified in a systematic manner. In Chapter 3, examples were given to demonstrate the co-occurrence of each possible onset with each nucleus. In so doing, an argument in favor of considering the complex onsets of Soyaltepec Mazatec to be consonant clusters rather than modified phonemes immerged; however, the argument is not definitive and the result does not fundamentally influence the tone of the syllable. This is an area for continuing research. Concerning the tones of Soyaltepec Mazatec, my data confirms that the tonal system is indeed complex, but I found differences from what Pike described. In Chapter 4 the four levels of tone of Soyaltepec Mazatec were described along with five rising contours and two falling contours that can occur on monomorphemic, monosyllabic words. I was not able to replicate all of Pike’s data or the processes she described but I discovered the tonal process of register lowering heretofore never described for a Mazatec language. This process is best described using Register Tier Theory. Soyaltepec Mazatec contour tones behave in a manner consistent with sequential tones, and processes that need to be explained using unitary tones do not occur. Specifically, no instances of Mid-high to High rising tones which reduced to Mid tones were found. The system does not require the geometry described by Pizer; however, the typical single tiered geometry employed by Autosegmental Geometry which offers only one level of tone specifications is inadequate to explain the behavior exhibited. 299 Register Tier Theory was demonstrated to both fully specify four levels of tone and allow enough flexibility to fully describe and explain the tonal phenomena present in an insightful manner. Specifically, the explanation of the lowering behavior of the floating Mid-high tone hinges on RTT features and the geometry which allows two tonal features to be specified separately and function either as a unit or separately. Examples such as those in 1 in which a floating M 1 causes the following syllable to assume its value through spreading both the tonal melody and the tonal register are not surprising; however, the resulting Low tone on the final syllable is surprising unless the M 1 is specified to contain a l-register which can spread independently. 1 The M 1 raises the adjacent syllable and lowers the following syllable ndɛ˧˦ + tia˧wa˧ → ndɛ˧ tia˦wa˩ ‘shoe’ ‘white’ ‘white shoe’ When the M 1 tone is specified to contain a l-register as prescribed by RTT, simple register spreading without tonal melody spread accounts for the surface tone. The example in 2 reconfirms that the l- register from the M 1 can spread completely independently of its tonal melody. 2 The lowering behavior of the M 1 tu˧˦ + si˧nɛ˥ → tu˧ si˩nɛ˦ ‘fruit’ ‘yellow’ ‘yellow fruit’ The features and geometry of RTT simply and elegantly explain this otherwise mysterious behavior.

6.3 Tonal Processes of Soyaltepec Mazatec and the UAC