Products & Services

A 'perfect' evidential: The functions of -shka in Imbabura Quichua

Description
A 'perfect' evidential: The functions of -shka in Imbabura Quichua
Published
of 21
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.
Related Documents
Share
Transcript
  A ‘Perfect’ Evidential: The Functions of   -shka  in Imbabura Quichua ∗ J ESSICA  C LEARY -K EMP University of California, Berkeley 1 Introduction Thispaperinvestigatesthefunctionsoftheverbalsuffix -shka inImbaburaQuichua(IQ),aQuechuaIIB language spoken in the northern highlands of Ecuador. Cole (1982:148) analyzes  -shka  as —among other functions — a perfect aspect inflection on main verbs, noting that it indicates inaddition “a degree of surprise.” Cognate forms in other Quechuan languages have been labelled‘sudden discovery tense’ and translated as ‘it turned out that...’, although they are reportedly notrestricted to past tense (see, e.g., Adelaar and Muysken 2004:223). These characterizations sug-gest that  -shka  is a perfect aspect marker, with additional semantics of mirativity. I will argue inthis paper that, although  -shka  does mark perfect aspect in periphrastic constructions (where theverb in  -shka  combines with copula  ka-  ‘be’), as a verbal inflection it marks ‘non-eyewitness pasttense’, contrasting in this function with  -rka  ‘eyewitness past tense’. In addition, I will argue thatthe constellation of functions modern-day IQ  -shka  fulfils — including deriving nominals and pas-sive participles — is explicable by assuming a series of typologically common diachronic changes,starting from a ‘resultative nominalization’ function.ThefindingsinthispaperarebasedondataobtainedfromtwospeakersofIQ—MC,a50-year-old woman from the area of Mariano Acosta, and ACO, a man from Otavalo. Both consultants arenative speakers of IQ, having learnt Spanish as a second language, and English as a third language.MC has been living in the United States for approximately 25 years, and ACO for 14 years. Thedata, collected by members of the 2009-2010 UC Berkeley Field Methods class, is of two maintypes: recorded monologues and dialogues, and targeted elicitation. The data presented in thispaper is from targeted elicitation, unless labelled otherwise. Some data is also taken from Cole(1982).The paper is structured as follows: §1.1 briefly describes the linguistic and social context of theImbabura Quichua language; §1.2 outlines the analytic framework adopted for describing tenseand aspect; §2 identifies the major categories of tense and aspect in IQ; §3 describes and exem-plifies the inflectional and derivational functions of   -shka  in IQ, and considers the uses of   -shka  in ∗ I would like to express my heartfelt gratitude to Mariana and Augusto for patiently and enthusiastically sharingtheir language and culture with us. I would also like to thank the members of the 2009-2010 UC Berkeley field methodsclass for very valuable comments and discussion relating to this topic. Thanks are also due to the audience at the 2012winter meeting of the Society for the Study of Indigenous Languages of the Americas (SSILA), who gave valuablefeedback on an earlier version of this paper. Finally, Lev Michael made numerous comments and observations thatsignificantly improved this paper. My thanks go to him, but as usual, any errors remain my own.  J ESSICA  C LEARY -K EMP personal and traditional narratives. §4 presents a diachronic proposal to account for the synchronicpolyfunctionality of   -shka ; and §5 briefly summarizes the findings of this paper. 1.1 The Imbabura Quichua Language and its Speakers Imbabura Quichua (IQ) is a Quechua IIB language spoken by approximately 30,000–50,000 peo-ple in the northern highlands of Ecuador (Cole 1982:3). The Quechuan languages have been incontact with Spanish for centuries, and the majority of IQ speakers are also fluent in EcuadoreanSpanish. The Quechuan languages of Ecuador (and of Colombia) have lost much of the complexhead-marking morphology exhibited in other Quechuan languages. Subject marking on the verbis retained, but only the 1 st person object marker remains, and possessive suffixes have been lostfrom nouns, possession now being marked on the possessor by genitive suffix  -pa . As such, IQis largely dependent-marking, especially compared to other varieties of Quechua (Adelaar andMuysken 2004:208). Argument alignment is nominative–accusative; case-markers attach to theNP phrase-finally, nominative being unmarked, and accusative marked by  -ta . Constituent order isrelatively free in main clauses, but SOV predominates, and is more strictly required in subordinateclauses. This correlates with general head-final order in constituents. Major word classes in thelanguage are ‘noun’ and ‘verb’, with small closed classes of demonstratives and quantifiers, plusa set of spatial terms that may be classed either as postpositions or relational nouns. No separateclass of adjectives is apparent (although close analysis may reveal the need for a separate adjectiveclass), and property-denoting terms essentially have the same distribution as referential nominals,being able to fill argument position and take case-marking, as well as functioning as modifiers andcopula complements. A pervasive phenomenon in IQ discourse is the use of a set of so-called ‘val-idators’—cliticsthatcombineevidential,modalandfocus-markingfunctions.Themainvalidatorsare  =mi  ‘direct evidential’,  =shi  ‘indirect evidential ∼ uncertainty’,  =chu  ‘polar question ∼ negativemarker’, and perhaps  =ta  ‘content question marker’. 1 The major previous work on IQ is Cole (1982), a grammatical description based on fieldwork undertaken with speakers of the Rinconada (including Mariano Acosta), Otavalo and San Roquedialects. Further references are listed therein. 1.2 A Neo-Reichenbachian Framework for Analyzing Tense and Aspect Klein (1994), building on work by Reichenbach (1947), outlines a framework for describing andanalyzing tense and aspectual distinctions cross-linguistically. The main elements of this approachare presented here, in order to lay the groundwork for the analysis presented in §3. 1.2.1 The Building Blocks The ‘lexical content’ (LC) of an utterance is the situation described in the clause, divorced from itsfinite temporal information. For instance, the LC of the English clause  John was studying Quichua is  { John study Quichua } . This is also the LC for the clauses  John will study Quichua ,  John hasstudied Quichua ,  John studies Quichua , and so forth. In other words, the LC is atemporal. But itis “timeable” — that is, it can be linked to a temporal structure. This is done by means of tenseand aspect marking. The temporal structure to which the LC is linked comprises three primaryelements: ‘utterance time’ (TU), ‘topic time’ (TT), and ‘situation time’ (TSit). The first two of these are relatively simple to define, while the third is somewhat more complex. 1 See Kwon (2010) and Cleary-Kemp (2010) for a more precise characterization of these elements and their functions.  The Functions of   -shka  in Imbabura Quichua TU (Reichenbach’s S) is the time of utterance. In conversations this is easy to calculate —it is the moment of speaking. In non-immediate communicative situations, such as letters, email,books, graffiti, voicemail messages, pre-recorded television or radio programs, etc., the TU is lessfixed, and more open to interpretation. The calculation of TU can be complex, but it is not crucialto the thesis of this paper, so I will not discuss it further here. TT (Reichenbach’s R) is the timeabout which a claim is being made. This can be overtly specified, as in  [At 6:40 in the morning, on August 24, 2003] TT   , Max was born , or it can be left to context. Occasionally, as in this example,TT is very precisely specified, but far more commonly the exact span of the TT is left open to beinferred from discourse context and world knowledge. Again, the details of how TT is determinedare not vital to the arguments presented here, and so are set aside.TSit (Reichenbach’s E) is the time for which the situation described by the LC holds. Forinstance, in the examples above, TSit is the temporal span during which it is the case that Johnis studying Quichua.  { John study Quichua }  is a one-state LC, because there is just one lexically-specified situation. In contrast, the LC  { John leave the house }  is two-state, since it encodes alexically-specified change of state: from John’s being inside the house to John’s being outside. Insuch complex LCs, the initial state is referred to as the ‘source state’ (SS) and the final state as the‘target state’ (TS). The TS in the case of   { John leave the house }  is clearly  { John not be inside thehouse } , but the SS, in contrast, is not simply  { John be inside the house } , rather it involves John’sbeing active in bringing about the TS. In other words, it is only felicitous to say  John is leaving thehouse ifJohnisintheprocessofbringingaboutthestateofbeingoutofthehouse,not,forexample,if he is simply sitting inside the house reading a book. Klein (1994:105) notes that, for purposesof aspect marking, languages tend to choose one of the two states to treat as TSit. In English it isSS. This is evident when we consider how two-state LCs behave in the progressive construction,whose function is to indicate that TT is fully included in TSit (see Table 1). The sentence in (1) isfelicitously uttered only if John is currently in the house, and is active in achieving the state of notbeing in the house.(1)  John is leaving the house In other words, this utterance denotes the structure in (2).(2)  { —–[—] TT — } SS ++++++ TS The sentence in (1) is not true if John is in the TS, having already left the house and being halfwaydown the street. 2 Given that the construction in (1) locates TT within SS, and not within TS, thisexample clearly shows that English treats the SS of two-state predicates as TSit. With respect tothis parameter, IQ behaves the same as English. 1.2.2 Defining Tenses and Aspects Utilizing the concepts outlined above, it is possible to precisely formulate the functions of themajor categories of tense and aspect in the world’s languages, in terms of how they situate thetemporal elements with respect to each other. This is shown in Table 1. Prior to Klein (1994), defi-nitions of tense and aspect had tended to be impressionistic, and therefore analytically inadequate. 2 Of course there is some leeway in the interpretation of this utterance, as with all natural language. If John has juststepped outside the house and is still on the front steps, perhaps the utterance in (1) will still be felicitous. But thisflexibility does not detract from the overall point of the example.  J ESSICA  C LEARY -K EMP Aspect:  situates TT with respect to TSitP ERFECTIVE  TT (partially) includes TSit  { ——[—– } TSit  ] TT  or [  { —— } TSit  ] TT I MPERFECTIVE  TT is fully included in TSit  { ——–[—] TT ——– } TSit P ERFECT  TT is after TSit  { ——– } TSit  [ ] TT P ROSPECTIVE  TT is before TSit [ ] TT  { ——– } TSit Tense : situates TT with respect to TUP AST  TT is before TU [ ] TT  ( ) TU P RESENT  TT includes TU [ ( ) TU  ] TT F UTURE  TT is after TU ( ) TU  [ ] TT Table 1: Characterization of the major categories of tense and aspect (after Klein 1994)For instance, Comrie (1976:3) defines ‘aspects’ as “different ways of viewing the internal temporalconstituency of a situation,” but later adds that “the perfect is rather different [...] since it tells usnothing about the situation in itself, but rather relates some state to a preceding situation” (p.52).He defines perfect aspect as referring to “a past situation which has present relevance” (p.12). Thisdefinition captures an important intuition about uses of the perfect cross-linguistically, but it failsto provide diagnostics for identifying perfect constructions across languages. In contrast, Klein’sframework allows one to devise tests that categorically differentiate between the major categories‘tense’ and ‘aspect’, and between different tenses and aspects in a language. The distinction be-tween tense and aspect under Klein’s approach is that tense locates TT in relation to TU, whileaspect relates TT to TSit. For instance, as noted above, an imperfective aspect, such as the Englishprogressive, locates TT  within  TSit. The utterance,  When I walked into the room, John was readinga book   indicates that the TSit of   { John read a book  }  extends either side of the TT (which is hereovertly specified with the adverbial clause  when I walked into the room ). In contrast, the use of past tense in this utterance indicates only that the TT precedes the TU, and says nothing directlyabout the TSit. This is demonstrated by the fact that it is perfectly felicitous to say  When I walked into the room, John was reading a book, and he is still reading it now . If past tense situated TSitprior to TU, then it should not be possible to use past tense when the TSit is still ongoing, as in thisexample. 2 Major Categories of Tense and Aspect in Imbabura Quichua As background to the analysis of   -shka , I briefly outline here the major tense and aspect distinctionsin IQ. The description here should by no means be viewed as exhaustive; numerous other aspectualnuances, such as durative, habitual and pluractional, can be expressed by verbal morphology in IQ,but these are not yet well-understood, and their analysis is beyond the scope of this paper. (SeeCole (1982) for an overview.)  The Functions of   -shka  in Imbabura Quichua SG PL 1  -ni -nchi 2  -ngi -ngichi 3  -n Table 2: IQ subject agreement suffixes in present tense 2.1 Present Tense Present tense in IQ is not marked by an overt morpheme. As shown in (3), it is indicated by theabsence of overt past or future tense marking. 3 (3)  Nyuka 1 SG ri-xu- /0 -ni go- IMPFV - PRES -1 SG . SBJ  yachachi-k  teach- NMLZ wasi-man house- ALL ‘I am going to school’The subject agreement markers used in present tense are given in Table 2. They distinguish threepersons, and two numbers in 1 st and 2 nd person. 2.2 Past Tense:  -rka The suffix  -rka  indicates simple past tense. It is compatible with imperfective and (null) perfectiveaspect marking.(4)  Kayna yesterday nyuka 1 SG ri-xu-  rka -ni go- IMPFV - PAST -1 SG . SBJ  yachachi-k  teach- NMLZ wasi-man house- ALL ‘Yesterday I was going to school’(5)  Kayna yesterday nyuka 1 SG ri- /0 -  rka -ni go- PFV - PAST -1 SG . SBJ  yachachi-k  teach- NMLZ wasi-man house- ALL ‘Yesterday I went to school’As shown in Table 3, person marking in the past tense is the same as in present tense, except that3 rd person subject agreement is null. This is illustrated in (6).(6)  Pay 3 SG ri-xu-rka- /0 go- IMPFV - PAST -3 SBJ  yachachi-k  teach- NMLZ wasi-man house- ALL ‘S/he was going to school’The past tense suffix  -rka  is described by Cole (1982:144) as a simple past tense, but I will arguein §3.5 that it in fact has an additional entailment of ‘eyewitness’ evidentiality. 3 Abbreviations:  ABL , ‘ablative’;  ACC , ‘accusative’;  AG , ‘agent’;  ALL , ‘allative’;  CAUS , ‘causative’;  CL , ‘clitic’; COM  /  INSTR , ‘comitative/instrumental’;  DIM , ‘diminutive’;  DIST , ‘distal’;  DS , ‘different subject subordinator’;  EYE ,‘eyewitness’;  FUT , ‘future’;  GEN , ‘genitive’;  HAB , ‘habitual’;  IMPFV , ‘imperfective’;  INDEF  ‘indefinite’;  INDIR ,‘indirect evidential’;  INF , ‘infinitive’;  LIM , ‘limitative’;  LOC , ‘locative’;  NMLZ , ‘nominalizer’;  NONEYE , ‘non-eyewitness’;  PERF , ‘perfect’;  PFV , ‘perfective’;  PL , ‘plural’;  PLUR , ‘pluractional’;  PRES , ‘present tense’;  PROX ,‘proximal’;  PURP , ‘purposive’;  SBJ , ‘subject’;  SG , ‘singular’;  SS , ‘same subject subordinator’;  TOP , ‘topic’.
Search
Tags
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