The “s”-ence of Pima Eric Jackson Workshop on American Indigenous Languages 27 April 2002 Goals: — Show that the grammaticality of the s- morpheme in Pima correlates best with stative lexical aspect — Show how stativity in Pima compares cross-linguistically. — Discuss some of the difficulties in formally analyzing the s- within a non- Lexicalist framework (Distributed Morphology). 1. Introduction • Pima is a language of the Tepiman branch of Southern Uto-Aztecan, spoken in central and southern Arizona. 1 Pima and its close relative Papago (or Tohono O'odham) have been described as dialects of a single language, O'odham (e.g. Saxton 1982). • Although Papago has been previously documented, very little work has been done to characterize Pima specifically. The data presented here, however, is taken entirely from Pima. • One interesting morpheme shared by Pima and Papago has the phonological form s-, which occurs in certain syntactic environments on many adjectives and a small set of verbs. • The best description of the set of stems which license this morpheme appeals to the lexical aspect of the stems—specifically, the factor which seems to license the s- is stativity.  Zepeda (1983) has claimed that the s- in Papago indicates stativity, though Saxton (1982) describes it as a positive polarity item indicating the affirmative for an arbitrary class of stems. Although Zepeda’s correlation with stativity does hold for many stems, the grammaticality of the s- may be arbitrarily determined for a small number of stems. • The set of stems which license the s- in Pima, however, is distinct from the sets of stems in other languages which are claimed to be stative; for Pima, factors of volitionality and controllability seem to be less important than lexical aspect. • An analysis within Distributed Morphology (a non-Lexicalist theory of 1 All data in this presentation, unless otherwise noted, has been graciously provided through the patience and good humor of Mr. Virgil Lewis, who is a native speaker of Pima from the Gila River Indian Community. I would also like to thank Pamela Munro, Carson Schütze, Tim Stowell, and Marcus Smith, whose comments and ideas have made this presentation more than it could have been otherwise. I cannot also fail to thank the other members of the 2000-01 UCLA Field Methods course for sharing their data and insightful comments. All errors remain my own. 1 morphology) which treats the s- as a verb class prefix requires that the relevant feature of the stem is able to percolate higher than the projection which mediates object agreement. 2. Pima clause structure • Word order in Pima is relatively free; all six possible orders of subject, object, and verb are judged acceptable in simple transitive sentences. (Munro 1984) • Most indicative sentences include a second-position auxiliary which indicates subject agreement, aspect, and optionally modality and evidentiality. • Object agreement is indicated by a marker on the main verb. (Agreement has no phonological expression for third person singular subjects or objects.) (1) First position Auxiliary Ha-pad:-c 'a-n-t heg heñ-gook kakalit. 3p:OBJ-bad-CAUS:PFV A-1s:SUB-PFV DET 1s:POSS-two RED:car 2 ‘I wrecked my two cars.’ • Elements which may occur in first position include argument DPs, VPs without objects (verb, floated quantifier + verb, adverb + verb, postposition + verb), PPs, complementizers, and certain grammatical particles ( e.g., focus particle ge, clausal negation pi, intensifier si). Pima, unlike Papago, does not allow object agreement markers alone to occur in first position. 3. Where the s- is preferred • The s- morpheme is preferred on a certain set of verbs and adjectives in simple, indicative, non-negated sentences. (2) Hega'i gogs 'o ge 3 s-'oam. that dog IMP FOC ST-brown ‘That dog is brown.’ 2 Data is presented in the orthography currently used by the UCLA Pima group. All Pima examples taken from Avelino et al (2001). A colon is part of a digraph in the orthography, but is also used in glosses where morphemes are not readily segmentable. Where glosses are not one-to-one, multiple words are joined with ‘.’. Abbreviations: 1s = first person singular, 2p = second person plural, etc., ¬1 = non-first person; A = filler vowel in auxiliary, ADV = adverbial, CAUS = causative, CMD = imperative, COMP = complementizer, COP = copula, DET = determiner, DEF = definite, DSD = desiderative, DXS = deictic particle, FOC = focus, GA = alienable possession, HSY = hearsay, IMP = imperfective, INCEP = inceptive, INCH = inchoative, INDEF = indefinite, IO = indirect object, IRR = irrealis, NEG = negative, NOM = nominalizer, NT = non-transitory, OBJ = direct object, PFV = perfective, POS = possessive, PROG = progressive, RED = reduplicated, REF = reflexive, SHD = shared knowledge, ST = stative (hypothetical—this gloss is always used for the s- prefix), SUB = subject, T = true characteristic, VB = verbalizer. 3 Here I follow Shademan (2001) in glossing ge as FOCUS. 2