Reduplicative allomorphy and language prehistory in Uto-Aztecan

Reduplicative allomorphy and language prehistory in Uto-Aztecan
of 37

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  To appear in Studies in Reduplication, Bernhard Hurch (ed.), Berlin:  Mouton de Gruyter Reduplicative Allomorphy and Language Prehistory in Uto-Aztecan  Jason D. Haugen 1. Introduction 1   Although Uto-Aztecan is one of the largest and most well-established lan-guage families of North America, the internal organization of sub-groups within Uto-Aztecan is still a matter of debate. At the center of discussion is whether or not the geographical split of Uto-Aztecan into “Northern” and “Southern” branches can also be considered to be a genetic split. Also at issue is the legitimacy of various proposed sub-groups within these two larger possible branches. The Northern branch is generally considered to  be a legitimate unit, based on shared morphological features (Heath 1977, 1978) and a sound law involving the lenition of intervocalic **-c- to *-y- (Manaster Ramer 1992). The unity of the Southern branch is more contro-versial but it does have its proponents, largely due to lexicostatistical evi-dence (i.e. percentage of shared cognates) (Miller 1984, Cortina-Borja and Valiñas 1989). In addition, for some sound correspondences between the two branches, such as NUA / N / ~ SUA /n/ and NUA /n/ ~ SUA /l, r/, it is difficult to determine which branch is the innovator, with some scholars (e.g. Hill 2001) arguing for innovation in NUA, and others (e.g. Silver and Miller 1997) for innovation in SUA. 2  Nevertheless, for the reconstruction of grammatical elements back to the level of Proto-Uto-Aztecan (PUA), it is necessary to provide corresponding elements from both  NUA and SUA as a minimal requirement for a likely PUA reconstruction. Following the call to find evidence beyond shared sound changes in the establishment of genetic relationships heralded by such scholars as Thoma-son and Kaufman (1989), Campbell (1997), and Dixon (1998), among many others, this paper began as an attempt to discover patterns of syn-chronic reduplicative allomorphy in an effort to identify shared allomorphy  patterns, and thus possible family-internal relationships, among the lan-guages within Uto-Aztecan. Barragan and Haugen (2002) identify four  2  Jason D. Haugen reduplication patterns in the Uto-Aztecan languages of Miller (1984)’s  proposed “Sonoran” branch of Southern-Uto-Aztecan, and in this paper I  present evidence that the same patterns appear across Uto-Aztecan, in both the Northern and Southern branches, and can thus be reconstructed for the  protolanguage. These reduplication patterns are a light syllable reduplicant; a disyllabic (CV(C)V-) reduplicant (or possibly a full root reduplicant); a  bare mora affix; and a fourth reduplication type which surfaces as a heavy syllable variation on the light syllable reduplicant, and will be called here “marked heavy syllable reduplication”. All of these patterns of reduplication are retained in at least one Uto-Aztecan language: Yaqui. 3  While it is true that the typology of reduplica-tion forms is limited cross-linguistically, the occurrence of all of these pat-terns in this language and others of the family is more likely to be the result of shared inheritance rather than convergent development in multiple lan-guages. These varying reduplication patterns would provide evidence for sub-groupings within the family if it could be demonstrated that certain  patterns were limited only to certain languages, but the actual distribution of the patterns in both the Northern and Southern branches does not allow for any particular sub-groupings. That is, the lack of evidence for shared innovation in the domain of reduplication in some sub-grouping of Uto-Aztecan suggests that each of these patterns must have existed in PUA. This paper is structured as follows. Section 2 will illustrate each of the reduplication patterns as they surface in one Uto-Aztecan language, Yaqui, and Section 3 will then provide examples of cognate reduplication patterns in both the Northern and Southern branches of Uto-Aztecan. Section 4 will discuss previous efforts at the historical reconstruction of Uto-Aztecan reduplication patterns and the reconstruction proposed here for PUA, and section 5 will discuss the issue of diachrony in reduplication more gener-ally. Section 6 will conclude. 2. Reduplicative allomorphy in Yaqui (SUA) Spaelti (1997) distinguishes between two different kinds of multi-pattern reduplication. The first type consists of “duplemes”, where different redu- plication patterns (i.e. different reduplicative morphemes) are used for dif-ferent semantic functions. An example would be the three patterns used in Sawai (Whistler 1992, cited by Spaelti 1997), where one reduplicant copies only the first consonant of the base for the durative (e.g.  g  E  lay  ‘to scream’     Language Prehistory in Uto-Aztecan  3      g    -   g  E  lay  ‘wailing’), a second copies both the first and second consonant for nominalization (e.g. l  E   s E  n  ‘to sweep’    l    s  -l  E   s E  n  ‘broom’), and a third copies only the second consonant for the reciprocal (e.g.  gali  ‘to help’     fa l  -gali  ‘to help one another’). The second kind of multi-pattern reduplica-tion is the “alloduple”, where different patterns of reduplication serve the same semantic function, but their shapes are phonologically conditioned. An example of this type is the Ponapean durative, where the reduplicant surfaces in a variety of shapes, depending on the phonological properties of the base: (1) Ponapean durative (Rehg 1981) a. tep    tepi -tep * ten -tep ‘beginning’  b. dod    don -dod * dodi -dod ‘frequenting’ c. pa    paa -pa ‘weaving’ Spaelti acknowledges the existence of mixed systems, of which the Southern Uto-Aztecan language Yaqui is apparently a case: different (not-entirely-prosodically-determined) patterns can express the same function (e.g. habitual action), and the same pattern can indicate different meanings. I use the term reduplicative allomorphy   in a theory-neutral way to encom- pass both kinds of multiple-pattern reduplication within a mixed system. I assume that the mixed system in Yaqui has come about via convergence, where what used to be different duplemes have merged through time, giv-ing the synchronic effect of suppletive allomorphy where multiple redupli-cant shapes serve the same semantic function (usually habitual action). The facts of Yaqui reduplication are discussed in more detail in Haugen (2003), which discusses the phonology of the allomorphy in Yaqui reduplication, as well as in Harley and Amarillas (2003), which gives a detailed account of the various semantic functions of reduplication in Yaqui. The Yaqui data in this paper come from Molina, Valenzuela, and Shaul (1999) or forms collected in the Yaqui Reduplication Survey at the Univer-sity of Arizona. Reduplication can serve several purposes within Yaqui grammar, and reduplication can co-occur with verbal tense/aspect/mood suffixes, nouns, numerals, adjectives, etc. The focus here (and throughout this paper) will be on prefixal reduplication, which usually expresses habit-ual action in Yaqui. However, as Harley and Amarillas show, each of the reduplication patterns in Yaqui can also be used to mean other things as well: iterative, continuative, imperative, etc. The semantics of the Yaqui reduplicant are often tied to the meaning of the verb root.  4  Jason D. Haugen Yaqui is a pitch accent language, and pitch accent does not determine the shape of the reduplicant. Thus, there are minimal pairs such as bwí.chi.a  ‘worm’ and bwi.chí.a  ‘smoke’, both of which reduplicate as bwi. bwi.chi.a , which can mean either ‘wormy’ or ‘smoky’. On the received view of Yaqui reduplication (e.g. Escalante 1985, Martínez 1995, Demers, Escalante, and Jelinek 1999, etc.), there are two kinds of reduplication in Yaqui: “primary reduplication” with a light sylla- ble reduplicant that copies the first syllable of the base (e.g. bwii.ka  ‘sing’    bwi  .bwi.ka  ‘usually sings’), and “secondary reduplication” with a heavy syllable reduplicant that triggers gemination of the onset of the base into coda position of the reduplicant (e.g. bwii.ka  ‘sing’    bwib  .bwi.ka  ‘sing from time to time’). We will see that this does not completely capture the facts of Yaqui re-duplication, since there are two more kinds of habitual reduplication: disyl-labic (or full root) reduplication and mora affixation (usually appearing as morphological gemination). In sum, Yaqui has three (largely unpredict-able) “alloduples” of a single dupleme, ‘habitual action’: a light syllable reduplicant, a disyllabic reduplicant, and a bare-mora affix. In addition, there is another dupleme which surfaces as a phonological twist on the light syllable reduplicant: the “marked heavy syllable reduplicant.” All four of these patterns have cognates across the Uto-Aztecan language family. 2.1. Yaqui light syllable reduplication The usual pattern of reduplication in Yaqui is light syllable reduplication (previously “primary reduplication”). This pattern exhibits a kind of redu- plication that has often been claimed not to exist: the pattern of “syllable copy” reduplication (cf. Moravcsik 1978, Marantz 1982, McCarthy and Prince 1986, etc.). This reduplicant usually appears as a copy of the entire first syllable of the base: (2) Yaqui monosyllabic habitual reduplication (Molina et al. 1999) a. ‘awaken’ vu  b. ‘comb one’s hair’ chi c. he.wi.te ‘agree’ he .he.wi.te d. ko.’a.rek ‘wear a skirt’ ko .ko.’a.rek e. cho.’ ‘lasso’ cho .cho’ila     Language Prehistory in Uto-Aztecan  5   (3) a. ‘hurry’ vam.  b. chep.ta ‘jump over’ chep. chep.ta c. chuk.ta ‘cut with a knife’ chuk  .chuk.ta d. hit.ta ‘make a fire’ hit .hit.ta e. bwal.ko.te ‘soften, smooth’ bwal .bwal.ko.te There are a few cases where a coda consonant does not copy (e.g. bwakta     bwa  .bwakta ‘take out of a container’), but no cases where we see copy into a second syllable: i.e. there are no forms like * vus.vusa . Because the reduplicant surfaces with a short vowel, Demers et al. (1999) and Haugen (2003) argue that the reduplicant is a light syllable, and coda con-sonants are not moraic, except when they are geminate consonants serving some morphological function. Also, Haugen (2003) argues that the base for reduplication in these cases is only the first syllable of the word. 2.2. Yaqui marked heavy syllable reduplication Previously referred to as “secondary reduplication”, the marked heavy syl-lable reduplicant surfaces as a heavy syllable reduplicant which triggers gemination from the onset of the base into the coda position of the redupli-cant: (4) Yaqui marked heavy syllable reduplication (Molina et al. 1999) a. bwii.ka ‘sing’ bwib .bwika  b. tee.ka ‘lay it across’  tet .teka c. ‘swim’  vav .vahume d. ‘smoke (tobacco)’  yey .yena e. ’om.te ‘get angry’ o’ .’om.te The marked heavy syllable reduplicant is always a different dupleme from habitual reduplication, and because it is always a different morpheme it is an example of what Urbanczyk (2002) has referred to as the “enhanc-ing of contrast” between morphemes. However, this process is apparently losing its productivity in the Arizona dialect, with speakers varying as to whether, and if so to what extent, this process is active. In many cases, the semantics of this reduplication pattern are idiosyncratic to the verb root. This loss of productivity is probably a concomitant of language shift in the Arizona Yaqui community.
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks