Acquiring Acquiring Ergativity: Ergativity: Cognitive Cognitive

steps towards steps towards syntactic syntactic

organizationorganization

Helen Charters, Helen Charters, Dept Applied Language Studies and Linguistics , Dept Applied Language Studies and Linguistics ,

University of Auckland.University of Auckland.H.Charters@auckland.ac.nzH.Charters@auckland.ac.nz

Structure of TalkStructure of Talk 1. Syntactic Organization 1. Syntactic Organization

–Syntactic Pivots: SAO Syntactic Pivots: SAO –Accusativity: S/A PivotAccusativity: S/A Pivot–Ergativity: S/O PivotErgativity: S/O Pivot

2. Tests for Subjecthood2. Tests for Subjecthood

–deceptive casedeceptive case–syntactic structuressyntactic structures

3. Semantic vs syntactic Subjects3. Semantic vs syntactic Subjects

–Manning (1996)Manning (1996) 4. Evidence of Ergativity in (TL) Samoan4. Evidence of Ergativity in (TL) Samoan

–Cook, 1991:Cook, 1991:–Mosel and Hovdhaugen, 1992Mosel and Hovdhaugen, 1992

5. Processing the role-function interface (TL)5. Processing the role-function interface (TL)

–Feature UnificationFeature Unification–Encoding GFsEncoding GFs

Configurational vs non-configurational languagesConfigurational vs non-configurational languages–Linking roles to GFsLinking roles to GFs

Lexical mapping theory Bresnan (1982, 2001)Lexical mapping theory Bresnan (1982, 2001)–Completeness and coherenceCompleteness and coherence

6. GFs and Stages in SLA (Pienemann (2005)6. GFs and Stages in SLA (Pienemann (2005)

–Structural simplicityStructural simplicity–Topic HypothesisTopic Hypothesis–Direct Alignment HypothesisDirect Alignment Hypothesis

7. Acquiring Ergativity: processes and words7. Acquiring Ergativity: processes and words

–Direct mapping: from role to discourse functionDirect mapping: from role to discourse function–Mediated mapping: from role to GF (in f-structure)Mediated mapping: from role to GF (in f-structure)–Licensed mapping: roles and rulesLicensed mapping: roles and rules

C-structure rules C-structure rules Grammaticalized predicatesGrammaticalized predicates

–Constructive mapping: Constructive mapping: 'Free' word order'Free' word order Constructive Case-markersConstructive Case-markers

–Variable mappingVariable mapping GF linking: functional and Anaphoric ControlGF linking: functional and Anaphoric Control

–MarkedMarked mapping: An ergative lexicon? mapping: An ergative lexicon? A-structure and the divided hierarchyA-structure and the divided hierarchy

88 conclusions conclusions

1. Theory and Linguistic 1. Theory and Linguistic developmentdevelopment

Samoan is an ergative language Samoan is an ergative language spoken in 1,000s spoken in 1,000s of Auckland homesof Auckland homes

Little is known about developmental milestones in Little is known about developmental milestones in Samoan Samoan

Processability Theory links linguistic development Processability Theory links linguistic development to processing demands of syntactic structuresto processing demands of syntactic structures

Contributors to processing include Contributors to processing include – agreementagreement– assignment of case and Grammatical Functions (GFs) assignment of case and Grammatical Functions (GFs) – marked mapping of semantic roles to GFsmarked mapping of semantic roles to GFs

Bresnan's Lexical Mapping Theory (LMT) suggests Bresnan's Lexical Mapping Theory (LMT) suggests that Ergative Systems are 'marked' systems. that Ergative Systems are 'marked' systems.

PT not previously applied to any ergative PT not previously applied to any ergative languagelanguage

Lexical Mapping TheoryLexical Mapping Theory Thematic hierarchy: Thematic hierarchy: agent > beneficiary > experiencer> instrument > patient/theme > agent > beneficiary > experiencer> instrument > patient/theme >

locativelocative GF Hierarchy:GF Hierarchy:

SUBJ > OBJ / OBLSUBJ > OBJ / OBL > OBJ > OBJ[-o -r]>[+o -r] [-o +r] > [+o +r][-o -r]>[+o -r] [-o +r] > [+o +r]

Highest role mapped to highest GFHighest role mapped to highest GF Other roles mapped by reference to featuresOther roles mapped by reference to features

– Patients [-r] Patients [-r] 'Secondary patient-like roles' [+o]'Secondary patient-like roles' [+o]– All other roles are [-o]All other roles are [-o]

No verb can have two arguments with identical No verb can have two arguments with identical featuresfeatures

Default system is Accusative Default system is Accusative – Agent ...instrument => SUBJ // Patient => OBJAgent ...instrument => SUBJ // Patient => OBJ

Research QsResearch Qs

How does Ergativity impact on How does Ergativity impact on Acquisition order?Acquisition order?

What are the processing demands of What are the processing demands of Ergative structures?Ergative structures?– Skin-deep ErgativitySkin-deep Ergativity– Ergatives as Fossilized PassivesErgatives as Fossilized Passives– Alternative mapping algorithmAlternative mapping algorithm

2. Ergativity in Samoan2. Ergativity in Samoan

Identification of Samoan SubjectIdentification of Samoan SubjectNo syntactic Subject - No syntactic Subject - Mosel & Hovdhaugen, Mosel & Hovdhaugen,

19921992

S/A (Nominative) pivot despite case-S/A (Nominative) pivot despite case-marking marking - Cook, 1991- Cook, 1991

S/O (Absolutive) syntactic pivotS/O (Absolutive) syntactic pivot– Dukes, 1998– Dukes, 1998

Tests for SubjecthoodTests for Subjecthood– Keenan (1975)Keenan (1975)– Manning (1994)Manning (1994)

Samoan Verb classesSamoan Verb classes

Ergative VerbErgative VerbNaNa fufulefufule ((e e le tama)le tama) le ta’avalele ta’avalePAST wash PAST wash ((ERG the boy)ERG the boy) the carthe carAbsolAbsol

The boyThe boy washed washed the carthe car

Non-Ergative VerbNon-Ergative Verb'Ua'Ua alu alu le tamale tama (‘i (‘i Samoa)Samoa)PERF go PERF go the boythe boyAbsolAbsol (DIR Samoa)(DIR Samoa)The boyThe boy has gone (to Samoa) has gone (to Samoa)

- Cook, 1991:78Cook, 1991:78

Labile Case alternationsLabile Case alternations

Labile VerbLabile VerbSa Sa 'ai 'ai e e le teinele teine le i'ale i'a

past eat past eat ERG the girlERG the girl the fishthe fishAbsolAbsol

The girl ate the fishThe girl ate the fish

Sa Sa 'ai 'ai le teinele teine

past eat past eat the girlthe girlAbsolAbsol

The girl ate / was eatenThe girl ate / was eaten

Sa Sa 'ai 'ai le teine le teine i i le i'ale i'a

past eat past eat the girlthe girlAbsolAbsol Dir Dir the fishthe fishThe girl ate from the fishThe girl ate from the fish

Mosel & Hovdhaugen Mosel & Hovdhaugen 1992:1081992:108

Tests for SubjecthoodTests for Subjecthood- Keenan (1976)- Keenan (1976)

A Subject is:A Subject is: indispensable indispensable omitted in omitted in

imperativesimperatives causer in causer in

applicativesapplicatives target of raising target of raising relativised firstrelativised first

A Subject can:A Subject can: control equi-NP & control equi-NP &

deletion in deletion in conjuncts conjuncts

launch a Floating Qlaunch a Floating Q be cliticisedbe cliticised trigger Agreementtrigger Agreement

Cook's applicationCook's application

Agreement (Ergative)Agreement (Ergative)‘‘UaUa ōō tamaititamaiti ‘i Samoa‘i Samoa

PastPast go.go.PLPL children DIR Samoachildren DIR Samoa

The children went to SamoaThe children went to Samoa

‘‘UaUa nunutinunuti ee le tamale tama fuāmoafuāmoa

PastPast crush.crush.PLPL the boythe boy egg.PLegg.PL

The boy crushed the eggsThe boy crushed the eggs

Two Two Relativisation Relativisation strategiesstrategies

Gapping (Ergative)Gapping (Ergative)le teine sā moe le teine sā moe ___ ___ i lo’u falei lo’u fale

the girl [PROG sleep the girl [PROG sleep GAPGAPabsabs LOC 1S.GEN house]LOC 1S.GEN house]

‘‘The girl [The girl [____ sleeping in my house].’ sleeping in my house].’– Chung & Seiter 1980:632Chung & Seiter 1980:632

le teine sā fasi ____ le teine sā fasi ____ e le tamae le tama

the girl [PST hit the girl [PST hit GAPGAPABSABS ERG the boy]ERG the boy]

‘‘the girl [the boy hitthe girl [the boy hit____ ]’ ]’– Chung 1978:235Chung 1978:235

Resumptive pro-formsResumptive pro-formsle tagata sale tagata sa (ia) ‘aveina la‘u ta‘avale‘aveina la‘u ta‘avale

the person[ PST the person[ PST 3S 3S drive 1S.GEN car]drive 1S.GEN car]

‘‘the person [the person [he drove my car]’drove my car]’- Chung & Seiter 1980:633- Chung & Seiter 1980:633

le fale‘oloa [sā ‘ou maua le fale‘oloa [sā ‘ou maua ai Ioane]Ioane]

the house PST 1S live the house PST 1S live PRN JohnJohn

‘‘the store [I found John in the store [I found John in it]’]’- Chung 1978:236- Chung 1978:236

Indispensability (Ergative)Indispensability (Ergative)

NaNa fufulefufulele ta’avalele ta’avale

PAST wash PAST wash the carthe carAbsolAbsol

[Someone] washed [Someone] washed the carthe car

'Ua'Ua alu alu ‘i ‘i SamoaSamoaPERF go PERF go DIR SamoaDIR Samoa[He] has gone (to Samoa)[He] has gone (to Samoa)

Sa Sa 'ai 'ai le teinele teine

past past eat eat the girlthe girlAbsolAbsol

The girl ate / was eatenThe girl ate / was eaten

Cook's Results in Manning's Cook's Results in Manning's frameworkframework

SemanticSemanticAccusative Pattern control in equi control in equi ImperativeImperative ?Raising ?Raising Causer in ApplicativeCauser in Applicative

Ergative PatternErgative Pattern binding reflexive binding reflexive ?Agreement?Agreement

SyntacticSyntacticAccusative Pattern Cliticisation Cliticisation

Ergative PatternErgative Pattern obligatorinessobligatoriness co-referent deletionco-referent deletion relativisationrelativisation launching a Floating Qlaunching a Floating Q Agreement Agreement

3. L1 Study3. L1 Study

200+ utterances from published 200+ utterances from published sources sources – Elinor Ochs (1982 a,b,c; 1986), Ochs and Elinor Ochs (1982 a,b,c; 1986), Ochs and

Duranti (1996), Ochs and Schieffelin Duranti (1996), Ochs and Schieffelin (1982, 1986); Kernan (1969), Schieffelin (1982, 1986); Kernan (1969), Schieffelin and Perry (1986); Platt (1986); Podmore and Perry (1986); Platt (1986); Podmore (2004)(2004)

11 children, 1;7 to 5;0 years11 children, 1;7 to 5;0 years– Most from traditional villages on Upolu. Most from traditional villages on Upolu.

AnalysisAnalysis

Utterances parsed and assigned to a Utterances parsed and assigned to a structural class. structural class.

Composite Longitudinal Observations Composite Longitudinal Observations from two girls between ages of 2,3 and from two girls between ages of 2,3 and 3,63,6

– Pesio sampled at 2;3 2;4, 2;9 2;10 Pesio sampled at 2;3 2;4, 2;9 2;10 – Naomi sampled at 2;11, 3;0, 3;1, 3;6 Naomi sampled at 2;11, 3;0, 3;1, 3;6

Implicational hierarchy based on data Implicational hierarchy based on data from 11 childrenfrom 11 children

Pesio Pesio

2; 3 1-2 2; 3 1-2 wordword ai Alesagaai AlesagaLook. AlesanaLook. Alesana

La ia okeLa ia okeLOC. 3PS hibiscusLOC. 3PS hibiscusThere it hibiscusThere it hibiscus

UmaUma ai ai a'ua'uFinish eatFinish eat me/myme/myAlgone me Algone me

2;4 Aspect-V2;4 Aspect-V *ua *ua pa pa pipiPRT burst pee?PRT burst pee?

2;10 2;10 Tense, Tense, Agentive Poss'rAgentive Poss'r

sa sa fai fai makou mea'aimakou mea'ai

Tense make 1PL.excl. foodTense make 1PL.excl. food

We made our foodWe made our food

Masae Masae le (ofu)vae (o) Fineasole (ofu)vae (o) Fineaso

Ripped art pants (of) FineasoRipped art pants (of) Fineaso

Fineaso ripped his pantsFineaso ripped his pants

NaomiNaomi

2.11 2.11 Agentive Poss'r

Uma Uma ai a'uFinish eat me/myFinish eat me/mymy eating is finishedmy eating is finished

Uma Uma mago aufinish mango mefinish mango memy mango is my mango is

finished finished I finished my mangoI finished my mango

3;03;0 CaseCase

Ai Ai ee oeoeEat Erg. YouEat Erg. YouYou ate it. You ate it.

Moe Moe 'oe'oeSleep youSleep youYou go to sleepYou go to sleep

4. Processing Ergativity4. Processing Ergativity Direct arguments are separate from and Direct arguments are separate from and

higher than obliques in a-structure (Manning, higher than obliques in a-structure (Manning, 1996)1996)<DIRECT || OBLIQUE ><DIRECT || OBLIQUE >

Modified hierarchy:Modified hierarchy:Direct: experiencer/goal > instrument > Direct: experiencer/goal > instrument >

patient/themepatient/themeOblique: Oblique: agent/ stimulus >beneficiary > locative agent/ stimulus >beneficiary > locative

In syntactically ergative languages, agentive arguments are specified as obliques in a-structure< paient || agent >< paient || agent >< experiencer || locative >< experiencer || locative >

Ergative and non-Ergative Ergative and non-Ergative verbsverbs

-r-r +r+r-o-o SUBJSUBJ OBLOBL+o+o OBJOBJ OBJOBJ

ergative Vs:ergative Vs: sasasasa 'hit' 'hit' < patient | < patient | agent>agent>

togitogi 'throw' 'throw' < theme | < theme | agent>agent> [-r][-r] [-O][-O]

– [-r] patient maps to SUBJ [-r[-r] patient maps to SUBJ [-r,-o,-o]]– [-o] agent maps to OBL[-o] agent maps to OBL [ [+r+r,-o] ,-o]

Non-ergative V: Non-ergative V: ataata 'laugh' 'laugh' <experiencer | (stimulus) ><experiencer | (stimulus) >[-o][-o] [-o] [-o]

– [-o] experiencer maps to SUBJ [[-o] experiencer maps to SUBJ [-r,--r,-o]o]– [-o] stimulus maps to OBL[-o] stimulus maps to OBL [ [+r+r,-o] ,-o]

Omission of stimulus leaves ranking, GF and case assignment Omission of stimulus leaves ranking, GF and case assignment unaffected.unaffected.

This is no more 'marked' overall than a Subj/OBJ systemThis is no more 'marked' overall than a Subj/OBJ system

Application of split hierarchyApplication of split hierarchyErgative VerbErgative Verb

fufule fufule < patient | agent>< patient | agent> [-r] [-r] [-O][-O]

Na Na fufule fufule ((e e le tama)le tama) le ta’avalele ta’avalePAST wash PAST wash ((ERG the boy)ERG the boy) the carthe carAbsolAbsol

The boyThe boy washed washed the carthe car

Non-Ergative VerbNon-Ergative Verb alu < alu < themetheme >>

[-r][-r]'Ua'Ua alu alu le tamale tama (‘i (‘i Samoa)Samoa)PERF go PERF go the boythe boyAbsolAbsol(DIR Samoa)(DIR Samoa)The boyThe boy has gone (to Samoa) has gone (to Samoa)

- Cook, 1991:78Cook, 1991:78

-r-r

+r+r

-o-o SUBJSUBJOBLOBL

+o+o OBJOBJOBJOBJ

-r-r

+r+r

-o-o SUBJSUBJOBL

+o+o OBJOBJOBJOBJ

Labile VerbsLabile Verbs

'ai 'ai 'eat'<(patient)|'eat'<(patient)| >> [-r] [-r] [-O][-O]

Sa Sa 'ai 'ai le i'ale i'apast eat past eat the fishthe fishAbsolAbsol

ate the fishate the fish

'ai 'ai 'eat'<|(agent)>'eat'<|(agent)> [-O][-O]

Sa Sa 'ai 'ai le teine le teine i i le i'ale i'apast eat past eat the girlthe girlAbsolAbsol Dir Dir the fishthe fishThe girl ate from the fishThe girl ate from the fish

-r-r

+r+r

-o-o SUBJSUBJOBLOBL

+o+o OBJOBJOBJOBJ

-r-r+r+r

-o-o SUBJSUBJOBLOBL

+o+o OBJOBJOBJOBJ

(agent)(agent)

e e le teinele teineERG the girlERG the girl

The girlThe girl

5. Conclusions5. Conclusions Agency is expressed through lexical (possessive) Agency is expressed through lexical (possessive)

case before syntactic (Ergative) casecase before syntactic (Ergative) case Ergative case may be initially semanticErgative case may be initially semantic Specification of agent as Oblique is typologically Specification of agent as Oblique is typologically

marked, but...marked, but... Morphological marking provides direct evidence of Morphological marking provides direct evidence of

marked rankingmarked ranking The split hierarchy is systemic not lexically The split hierarchy is systemic not lexically

specifiedspecified Application of LMT is NOT exceptionalApplication of LMT is NOT exceptional Acquisitional tasks involveAcquisitional tasks involve

– acquisition of split hierarchy acquisition of split hierarchy – Assignment of Subj GF to unmarked argument. Assignment of Subj GF to unmarked argument.

LimitationsLimitations

The data is opportunistic The data is opportunistic – No controlled elicitation techniques were No controlled elicitation techniques were

employedemployed– Linguistic context not controlledLinguistic context not controlled– The child's identity not always knownThe child's identity not always known– data was not phonetically transcribeddata was not phonetically transcribed– orthography not standardizedorthography not standardized

The results of this survey are necessarily The results of this survey are necessarily tentative / indicative onlytentative / indicative only

Future researchFuture research Adult L2 learners at University of AucklandAdult L2 learners at University of Auckland Bi-lingual children in AucklandBi-lingual children in Auckland

– 4- year olds; 5 year olds4- year olds; 5 year olds– Younger age groupYounger age group– English dominant vs Samoan dominant ?English dominant vs Samoan dominant ?

Controlled elicitation targetting key structuresControlled elicitation targetting key structures Assessment of phonological development and Assessment of phonological development and

syntactic developmentsyntactic development Ethnography of language socialisation in Auckland Ethnography of language socialisation in Auckland

environmentenvironment Psycho-linguistic techniques for exploring Psycho-linguistic techniques for exploring

relationships between phonology, morphology and relationships between phonology, morphology and syntaxsyntax

ReferencesReferencesBresnan, Joan. 2001. Lexical Functional Syntax. Chung, S. 1978. Case Marking and Grammatical Relations in Polynesian. Austin: University of Texas Press.Cook, Kenneth W. 1991. The search for Subject in Samoan. In Robert Blust (ed) Currents in Pacific Linguistics: Pacific Linguistics,

Series C, no 117. Cancerra: ANU Duranti, A. & Ochs, E. 1996 "Use and acquisition of genitive constructions in Samoan" in Social interaction, social context and

language: Essays in honor of Susan Ervin-Tripp, ed. by D.Slobin, J. Gerhardt, A. Kyratzis, & Guo Jiansheng. Lawrence Erlbaum Associates, pp.175-190.

Keenan, E.L. 1976. Towards a Universal Definition of “Subject”. In: Li, C. (ed.), Subject and Topic. New York: Academic Press. 305–32.

Kernan, K. T. 1969. The Acquisition of Language by Samoan Children. Working Paper of the Language Behavior Research Laboratory, No. 21. California Univ., Berkeley. Language and Behavior Research Lab.[CIQ11410].

Manning, Christopher D. 1996. Ergativity: Argument structure and Grammatical Relations . Stanford: CSLI PublicationsMosel, Ulrike & Hovdhaugen, Even. 1992. Samoan Reference Grammar. Oslo: Scandinavian University Press. Ochs, E. 1982a. Ergativity and Word Order in Samoan Child Language. Language, 58(3), 646-671.Ochs, E. 1982b. Talking to Children in Western Samoa. Language in Society, 11(1), 77-104Ochs, E. 1988. Culture and language development : language acquisition and language socialization in a Samoan village.

Cambridge ; New York: Cambridge University Press. Ochs, E., & Schieffelin, B. B. 1982. Language Acquisition and Socialization: Three Developmental Stories and Their Implications.

Sociolinguistic Working Paper Number 105 (Research/Technical): Southwest Educational Development Lab., Austin, TX.[BBB00950].

Pienemann, M. 1998. Language Processing and Second Language Development: Processability Theory. Amsterdam: BenjaminsPienemann, M. 2005. Cross-linguistic aspects of processability theory. Amsterdam: BenjaminsPlatt, M. L. 1980. The Acquisition of Deictic Contrasts by Samoan-Speaking Children. Stanford Child Language Research Forum.Platt, M. 1980b. The Acquisition of 'Come' and 'Bring' by Samoan Children. Papers and Reports on Child Language Development,

19, 60-69.Platt, M. 1980c. The Acquisition of "Come," "Give" and "Bring" by Samoan Children. Papers and Reports on Child Language

Development, Number 19 (Research/Technical No. NSF-53-482-2480): Stanford Univ., CA. Dept. of Linguistics.[BBB34247].Platt, M. L. 1982. Social and Semantic Dimensions of Deictic Verbs and Particles in Samoan Child Language. Unpublished PhD,

University of Southern California.Platt, M. 1983. Deictic Particles in Samoan Child Language. 35-38 ISSN 0110-6376.Podmore, V. S. L. a. i. M. L. 2001. Transition to school from Pacific early childhood centres. SET: Research Information for Teachers,

3 7-10 ISSN 0110-6376.Schieffelin, B. B., & Ochs, E. 1986. Language Socialization across Cultures. vii+274pp, New York, NY: Cambridge University Press.

Composite Emergence Composite Emergence Order Order

Speaker I1 K Kal An1 Sip Tof P1 I2 P2 Na1 Ma Na2 Ni1 Na3Ni2 Mas An2

AGE 2.1 1.8 2.12.6 2.6 2.2 2.4 2.9 2.11 3.0 -3.3 3.6 - 3.84Simple SSimple S

S V/ loc S V/ loc XX XX    XX   (V) VOA(V) VOA XXVP loc/ AdvVP loc/ Adv XX XXV/loc SV/loc S XX XX XXASPPASPPASP (ua) V ASP (ua) V ?? XX       XXVO (A =poss'r)VO (A =poss'r) XX XX XX XXVO (= N Adj)VO (= N Adj) XX XX       XXComplex DPComplex DPN Dem (lea)N Dem (lea) XX XXArt-Poss NArt-Poss N XX XX XX XXIPIPTense V …Tense V … XX       XX XXCaseCaseE-Ergative VO e A E-Ergative VO e A XX       XX    XXModified Modified SSEmph PartEmph Part XXBecause SBecause S XXAdv Asp V …Adv Asp V …    XX