Melanesian Languages on the Edge of Asia: Challenges for the 21st Century
edited by Nicholas Evans and Marian Klamer
Language Documentation & Conservation Special Publication No. 5 PUBLISHED AS A SPECIAL PUBLICATION OF LANGUAGE DOCUMENTATION & CONSERVATION
LANGUAGE DOCUMENTATION & CONSERVATION Department of Linguistics, UHM Moore Hall 569 1890 East-West Road +RQROXOX+DZDLދL USA
KWWS QÀUFKDZDLLHGXOGF
UNIVERSITY OF HAWAIދI PRESS .RORZDOX6WUHHW +RQROXOX+DZDLދL USA
© $OOWH[WVDQGLPDJHVDUHFRS\ULJKWWRWKHUHVSHFWLYHDXWKRUV All chapters are licensed under Creative Commons Licenses
Cover design by Susan Ford incorporating a photograph by Darja Hoenigman
Library of Congress Cataloging in Publication data ,6%1
KWWS KGOKDQGOHQHW Contents
Contributors iv
Introduction 1 Nicholas Evans and Marian Klamer The languages of Melanesia: Quantifying the level of coverage 13 Harald Hammarström and Sebastian Nordhoff Systematic typological comparison as a tool for investigating language history Ger Reesink and Michael Dunn Papuan-Austronesian language contact: Alorese from an areal perspective Marian Klamer Even more diverse than we had thought: The multiplicity of 109 Trans-Fly languages Nicholas Evans Projecting morphology and agreement in Marori, an isolate of 150 southern New Guinea I Wayan Arka ‘Realis’ and ‘irrealis’ in Wogeo: A valid category? Mats Exter From mountain talk to hidden talk: Continuity and change in 191 Awiakay registers Darja Hoenigman Cross-cultural differences in representations and routines for exact number Michael C. Frank .HHSLQJUHFRUGVRIODQJXDJHGLYHUVLW\LQ0HODQHVLD 7KH3DFL¿F and Regional Archive for Digital Sources in Endangered Cultures (PARADISEC) Nicholas Thieberger and Linda Barwick iv
Contributors
I WAYAN ARKA LV DI¿OLDWHG ZLWK WKH $XVWUDOLDQ 1DWLRQDO 8QLYHUVLW\ DV D )HOORZ LQ /LQJXLVWLFVDW6FKRRORI&XOWXUH+LVWRU\DQG/DQJXDJH&ROOHJHRI$VLDDQGWKH3DFL¿F DQG8GD\DQD8QLYHUVLW\%DOL (QJOLVK'HSDUWPHQWDQG*UDGXDWH3URJUDPLQ/LQJXLVWLFV His interests are in descriptive, theoretical and typological aspects of Austronesian and 3DSXDQODQJXDJHVRI,QGRQHVLD:D\DQLVFXUUHQWO\ZRUNLQJRQDQXPEHURISURMHFWV 16) funded research on voice in the Austronesian languages of eastern Indonesia ARC-funded projects for the development of computational grammar for Indonesian DQGWKH/DQJXDJHVRI6RXWKHUQ1HZ*XLQHD
LINDA BARWICK is Associate Professor in the School of Letters, Art and Media at the 8QLYHUVLW\RI6\GQH\DQG'LUHFWRURIWKH3DFL¿FDQG5HJLRQDO$UFKLYHIRU'LJLWDO6RXUFHV LQ(QGDQJHUHG&XOWXUHV 3$5$',6(& 6KHWUDLQHGLQ,WDOLDQODQJXDJHDQGGLDOHFWRORJ\ with Antonio Comin at Adelaide’s Flinders University, and in ethnomusicology with WKH ODWH &DWKHULQH (OOLV DW WKH 8QLYHUVLW\ RI $GHODLGH 6LQFH WKHQ VKH KDV XQGHUWDNHQ HWKQRPXVLFRORJLFDO ¿HOGZRUN LQ ,WDO\$XVWUDOLD DQG WKH 3KLOLSSLQHV DQG KDV SXEOLVKHG widely on Australian Indigenous Music as well as continuing her engagement with Italian WUDGLWLRQDOPXVLF
MICHAEL DUNN is an evolutionary linguist with a background in linguistic typology and ODQJXDJH GHVFULSWLRQ +LV FXUUHQW UHVHDUFK WDNHV D TXDQWLWDWLYH SK\ORJHQHWLF DSSURDFK WR ODQJXDJH FKDQJH DQG OLQJXLVWLF GLYHUVLW\ 5HFHQW SURMHFWV KDYH DGGUHVVHG TXHVWLRQV of coevolution of typological parameters, as well as the ecological and social factors LQÀXHQFLQJODQJXDJHFKDQJH+HKDVDOVRZRUNHGRQWKHFODVVL¿FDWLRQDQGSUHKLVWRU\RI 3DSXDQ ODQJXDJHV DQG WKH SK\ORJHRJUDSK\ RI ,QGR(XURSHDQ DQG$VOLDQ ODQJXDJH +H leads the Max Planck Research Group “Evolutionary Processes in Language and Culture” DWWKH0D[3ODQFN,QVWLWXWHIRU3V\FKROLQJXLVWLFVLQ1LMPHJHQWKH1HWKHUODQGV
NICHOLAS EVANSLV'LVWLQJXLVKHG3URIHVVRURI/LQJXLVWLFVLQWKH&ROOHJHRI$VLD3DFL¿F $XVWUDOLDQ1DWLRQDO8QLYHUVLW\+HKDVFDUULHGRXWZLGHUDQJLQJ¿HOGZRUNRQWUDGLWLRQDO ODQJXDJHV RI QRUWKHUQ$XVWUDOLD DQG VRXWKHUQ 3DSXD 1HZ *XLQHD 7KH GULYLQJ LQWHUHVW of his work is the interplay between documenting endangered languages and the many VFLHQWL¿F DQG KXPDQLVWLF TXHVWLRQV WKH\ FDQ KHOS XV DQVZHU ,Q DGGLWLRQ WR JUDPPDUV of two Aboriginal languages, Kayardild and Bininj Gun-wok, dictionaries of Dalabon and .D\DUGLOG HGLWHG FROOHFWLRQV RQ D QXPEHU RI OLQJXLVWLF WRSLFV DQG RYHU VFLHQWL¿F papers, he recently published the widely-acclaimed crossover book Dying Words: Endangered Languages and What They Have to Tell Us which sets out a broad program for HQJDJLQJZLWKWKHZRUOG¶VGZLQGOLQJOLQJXLVWLFGLYHUVLW\+HKDVDOVRZRUNHGDVDOLQJXLVW interpreter and anthropologist in two Native Title claims in northern Australia, and as a SURPRWRURI$ERULJLQDODUWE\WKH%HQWLQFN,VODQGZRPHQ¶VDUWLVWV v
MATS EXTER studied General Linguistics, Historical-Comparative Linguistics, Phonetics DQG )LQQLVK 6WXGLHV DW WKH 8QLYHUVLW\ RI &RORJQH +H KDV KHOG SRVWGRFWRUDO UHVHDUFK DQGWHDFKLQJSRVLWLRQVDWWKH8QLYHUVLWLHVRI%RQQDQG'VVHOGRUI+LVZRUNIRFXVHVRQ language description and documentation, experimental phonetics, laboratory phonology, DQGPRUSKRV\QWDFWLFW\SRORJ\+HKDVFRQGXFWHG¿HOGZRUNRQ:RJHRDQ$XVWURQHVLDQ language of Papua New Guinea, where he has done descriptive research on the phonological DQGPRUSKRV\QWDFWLFVWUXFWXUHDVZHOODVSURGXFLQJDFROOHFWLRQRIWUDGLWLRQDOWH[WV0RUH IRUPHUO\µ6RXWKHUQ.KRLVDQ¶ ODQJXDJH XXD7XXۄUHFHQWO\KHKDVFRQGXFWHG¿HOGZRUNRQ1 RI6RXWK$IULFDIRFXVLQJRQWKHSKRQHWLFDQGSKRQRORJLFDOVWUXFWXUHRIWKHODQJXDJH
MICHAEL C. FRANK LV$VVLVWDQW 3URIHVVRU RI 3V\FKRORJ\ DW 6WDQIRUG 8QLYHUVLW\ +H UHFHLYHGKLV3K'IURPWKH'HSDUWPHQWRI%UDLQDQG&RJQLWLYH6FLHQFHVDW0,7LQ and now heads the Language and Cognition Lab, which uses experimental, observational, DQGFRPSXWDWLRQDOPHWKRGVWRVWXG\ODQJXDJHDFTXLVLWLRQDQGODQJXDJHXVH+HLVEURDGO\ interested in the reciprocal interactions of language and cognition: both how languages affect the thoughts of their users and how the structure of cognition (especially social FRJQLWLRQ IDFLOLWDWHVWKHDFTXLVLWLRQRIODQJXDJHLQLQIDQWVDQGFKLOGUHQ
HARALD HAMMARSTRÖM studies linguistics and computer science at Uppsala University (Sweden) and went on to do a PhD in computational linguistics at Chalmers University 6ZHGHQ +LV LQWHUHVWV DUH OLQJXLVWLF W\SRORJ\ HVSHFLDOO\ QXPHUDO V\VWHPV 3DSXDQ ODQJXDJHVODQJXDJHFODVVL¿FDWLRQDQGFRPSXWDWLRQDOWHFKQLTXHVIRUPRGHOLQJWKHGLYHUVLW\ RIKXPDQODQJXDJHV+HLVFXUUHQWO\ZRUNLQJDVD3RVW'RFDWWKH03,(9$ *HUPDQ\ DQG Radboud University (the Netherlands) on the documentation of the Papuan language Mor DQGRQDUHDOOLQJXLVWLFVLQ6RXWK$PHULFD
DARJA HOENIGMAN is a PhD candidate in Anthropology at The Australian National University, working among the Awiakay, a community of 300 people living in Kanjimei YLOODJH LQ (DVW 6HSLN 3URYLQFH RI 3DSXD 1HZ *XLQHD ,Q KHU FXUUHQW SURMHFW VKH LV investigating socio-cultural continuity and change in Kanjimei and its relation to linguistic UHJLVWHUV,QVWXG\LQJWKHVHVSHHFKYDULHWLHVDQGWKHLUUHODWLRQWRWKHRYHUDOOVRFLDOVFHQH VKHEULQJVWRJHWKHUOLQJXLVWLFDQWKURSRORJ\DQGHWKQRJUDSKLF¿OPPDNLQJ
MARIAN KLAMER WHDFKHV DW /HLGHQ 8QLYHUVLW\ DQG KDV GRQH SULPDU\ ¿HOGZRUN RQ D GR]HQ$XVWURQHVLDQ DQG 3DSXDQ ODQJXDJHV LQ ,QGRQHVLD RYHU WKH ODVW WZR GHFDGHV +HU research centres on language description and documentation, typology, and historical and FRQWDFWLQGXFHGODQJXDJHFKDQJH+HUSXEOLFDWLRQVLQFOXGH$JUDPPDURI.DPEHUD $JUDPPDURI7HLZD $VKRUWJUDPPDURI$ORUHVH DQGRYHUDUWLFOHVRQ DYDULHW\RIWRSLFV.ODPHUKDVFRRUGLQDWHGQXPHURXVUHVHDUFKSURMHFWVRQODQJXDJHVRI ,QGRQHVLD LQFOXGLQJ WKH 12:9,', SURMHFW µ/LQJXLVWLF YDULDWLRQ LQ (DVWHUQ ,QGRQHVLD¶ ± DQGWKH(XUR%$%(/SURMHFWµ$ORU3DQWDUODQJXDJHV 2ULJLQVDQGWKHRUHWLFDO LPSDFW¶ ± IXQGHGE\WKH(XURSHDQ6FLHQFH)RXQGDWLRQ vi
SEBASTIAN NORDHOFF is a postdoctoral researcher at the Max Planck Institute for (YROXWLRQDU\$QWKURSRORJ\LQ/HLS]LJ+HVSHFLDOL]HVLQODQJXDJHFRQWDFWDQGODQJXDJH change and the interface of language description and documentation on the one hand and HOHFWURQLFSXEOLFDWLRQRQWKHRWKHU+HLVDPHPEHURIWKHZRUNLQJJURXSRQ2SHQ'DWDLQ Linguistics of the Open Knowledge Foundation, where he works on integrating typological GDWDLQWRWKH/LQJXLVWLF/LQNHG2SHQ'DWD&ORXG
GER REESINK studied psychology at the University of Utrecht, after which he spent 15 \HDUVLQ3DSXD1HZ*XLQHDXQGHUWKHDXVSLFHVRI6,/)LQLVKLQJKLVDI¿OLDWLRQZLWK6,/ he spent more than 15 years at Leiden University, mostly doing research on the languages RIWKH%LUG¶V+HDGRI3DSXDSURYLQFH,QGRQHVLD6LQFHKHKDVEHHQDSRVWGRFWRUDO researcher at the Radboud University and the Max Planck Institute for Psycholinguistics at Nijmegen, involved in typological research of Papuan and Austronesian languages in order WRWUDFHWKHDQFLHQWKLVWRU\RIJHQHDORJLFDODQGFRQWDFWUHODWLRQV
NICK THIEBERGERLVDQ$5&4(,,)HOORZDWWKH8QLYHUVLW\RI0HOERXUQH+HUHFRUGHG Paakantyi (NSW) speakers in the early 1980s and then worked with Warnman speakers (Western Australia) when he was setting up the Wangka Maya language centre in Port +HGODQG+HEXLOWWKH$ERULJLQDO6WXGLHV(OHFWURQLF'DWD$UFKLYH $6('$ DW$,$76,6 LQWKHHDUO\ VDQGWKHQZDVDWWKH9DQXDWX&XOWXUDO&HQWUHIURP ± +HZURWH DJUDPPDURI6RXWK(IDWHDODQJXDJHIURPFHQWUDO9DQXDWXZKLFKZDVWKH¿UVWJUDPPDU WRFLWHDGLJLWDOFRUSXVRIUHFRUGLQJVLQDOOH[DPSOHVHQWHQFHVDQGWH[WV,QKHKHOSHG HVWDEOLVK WKH 3DFL¿F DQG 5HJLRQDO$UFKLYH IRU 'LJLWDO 6RXUFHV LQ (QGDQJHUHG &XOWXUHV 3$5$',6(& +HWDXJKWLQWKH'HSDUWPHQWRI/LQJXLVWLFVDWWKH8QLYHUVLW\RI+DZDLµL ± +HLVDFRGLUHFWRURIWKH5HVRXUFH1HWZRUNIRU/LQJXLVWLF'LYHUVLW\ 51/' DQGWKHHGLWRURIWKHMRXUQDO/DQJXDJH'RFXPHQWDWLRQ &RQVHUYDWLRQ+HLVGHYHORSLQJ PHWKRGV IRU WKH FUHDWLRQ RI UHXVDEOH GDWD IURP ¿HOGZRUN RQ SUHYLRXVO\ XQUHFRUGHG ODQJXDJHVDQGWUDLQLQJUHVHDUFKHUVLQWKRVHPHWKRGV /DQJXDJH'RFXPHQWDWLRQ &RQVHUYDWLRQ6SHFLDO3XEOLFDWLRQ1R 'HFHPEHU Melanesian Languages on the Edge of Asia: Challenges for the 21st Century, HGE\1LFKRODV(YDQVDQG0DULDQ.ODPHUSS± KWWS QÀUFKDZDLLHGXOGFVS 1 KWWS KGOKDQGOHQHW
Introduction: Linguistic challenges of the Papuan region
Nicholas Evans The Australian National University
Marian Klamer
Universiteit Leiden
7KHUHJLRQZKHUH3DSXDQODQJXDJHVDUHVSRNHQ±FHQWUHGRQWKH,VODQGRI1HZ*XLQHD with extensions westward into Timor and the islands of eastern Indonesia, and eastward LQWRWKH6RORPRQ,VODQGV±LVDWWKHVDPHWLPHWKHPRVWOLQJXLVWLFDOO\GLYHUVH]RQHRIWKH SODQHWDQGWKHSDUWRIWKHORJRVSKHUH1 ,WSDFNVDURXQGRIWKHZRUOG¶VODQJXDJHVLQWR OHVVWKDQRILWVVXUIDFHDUHDDQGOHVVWKDQRILWVSRSXODWLRQ7KHDEVROXWHOHYHO RIOLQJXLVWLFGLYHUVLW\±ZKHWKHUPHDVXUHGLQVKHHUQXPEHUVRIODQJXDJHVRULQWHUPVRI µPD[LPDOFODGHV¶RIXQUHODWDEOHXQLWV±LVFRPSDUDEOHWRWKHZKROHRI(XUDVLD Getting the right term to describe the region of interest in this collection is a famously GLI¿FXOWSUREOHP0HODQHVLDLVDOLWWOHWRREURDG±H[WHQGLQJRXWWR)LML9DQXDWXDQG1HZ Caledonia to the east, a little beyond the scope of the present collection, and on the other
1 We gratefully acknowledge the support of the various bodies which supported the original confer- HQFHLQ0DQRNZDULIURP)HEUXDU\XQGHUWKHWLWOHµ0HODQHVLDQ/DQJXDJHVRQWKH(GJH of Asia’: the Australia-Netherlands Research Collaboration, the Australian National University 'HSDUWPHQWRI/LQJXLVWLFV&ROOHJHRI$VLDDQGWKH3DFL¿FDQGWKH6WHSKHQDQG+HOHQ:XUP Endowment), the Max Planck Institute for Evolutionary Anthropology, and the Universitas Negeri Papua (Unipa) and the Centre for Endangered Languages Documentation (CELD) for so gener- RXVO\KRVWLQJWKHFRQIHUHQFH:HZRXOGIXUWKHUOLNHWRWKDQNWKHYDULRXVFROOHDJXHVZKRDFWHGDV referees for the present collection (in alphabetical order): Sander Adelaar, Wayan Arka, Johan van der Auwera, Matthew Baerman, Rene van den Berg, Sonia Cristofaro, Mary Dalrymple, Philippe Grangé, Simon Greenhill, Harald Hammarström, Andy Pawley, Ger Reesink, Malcolm Ross, $ODQ5XPVH\5XWK6LQJHU/RXUHQVGH9ULHVDVZHOODVWR6XVDQ)RUGDQG$XQJ6LIRUWKHLU HGLWRULDODVVLVWDQFH (YDQV ZRXOG IXUWKHU OLNH WR WKDQN WKH$OH[DQGHU YRQ +XPEROGW 6WLIWXQJ WKURXJKDQ$QQHOLHVH0DLHU)RUVFKXQJVSUHLVIRU¿QDQFLDOVXSSRUWZKLFKDVVLVWHGLQVRPHRIWKH ¿QDOSURGXFWLRQRIWKLVFROOHFWLRQ
cc Creative Commons Attribution Non-Commercial No Derivatives Licence Introduction
KDQGQRWJHQHUDOO\LQFOXGLQJWKH/HVVHU6XQGDLVODQGVLQWKH,QGRQHVLDQDUFKLSHODJR 1RUDUHGH¿QLWLRQVLQWHUPVRIODQJXDJHIDPLOLHVHDV\WRPDNHFOHDQO\7KH$XVWURQHVLDQ languages have wrapped New Guinea and its surrounding islands in a three thousand year embrace that is still being played out in intimate language contact with all the other ODQJXDJHVRIWKHUHJLRQ6RPHRIWKHSDSHUVKHUHFRQFHUQHLWKHU$XVWURQHVLDQODQJXDJHV ZLWKVLJQL¿FDQWVWUXFWXUDOUHVHPEODQFHVWRQRQ$XVWURQHVLDQODQJXDJHVRIWKHUHJLRQ±VHH ([WHU¶VSDSHURQ:RJHR±RUYDULRXVW\SHVRIKLVWRULFDODQGW\SRORJLFDOLQWHUDFWLRQEHWZHHQ $XVWURQHVLDQDQGQRQ$XVWURQHVLDQODQJXDJHV±VHHWKHSDSHUVE\5HHVLQN 'XQQDQGE\ .ODPHU For the non-Austronesian languages of Melanesia and its surrounds (excluding Australia), the collective name ‘Papuan’ has been widely used and we continue that SUDFWLFHKHUH7KLVXVHEDVHGRQGH¿QLWLRQE\H[FOXVLRQKDVKXQJRQIRUZDQWRIDEHWWHU term long after comparable terms like ‘Palaeosiberian’ have been abandoned, but includes XSZDUGVRIIRUW\GLVWLQFWIDPLOLHVDQGLVRODWHV7RJHWDQLGHDRIKRZGLVWRUWLQJDWHUPOLNH this is, consider how unsatisfactory it would be to use a term like ‘Eurasian’ for the set RIODQJXDJHVLQFOXGLQJ%DVTXH)LQQLVK*HRUJLDQ,QJXVK&KLQHVH7DPLO&DPERGLDQ Japanese, Kurdish, Japanese, Hmong, Ket, Chukchi, Burushaski and all the other non-Indo- European languages of Eurasia (where we partition off Indo-European languages only, in WKHVDPHZD\WKDWZHSDUWLWLRQRII$XVWURQHVLDQODQJXDJHV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Introduction 3 DQDSSUR[LPDWHDQGUDWKHUXQTXDQWL¿HGZD\WKDW0HODQHVLDDQGLQSDUWLFXODUWKHLVODQGRI New Guinea, contains both the greatest concentration of linguistic diversity anywhere on HDUWKDQGWKHORZHVWOHYHORIGRFXPHQWDWLRQ7KH¿UVWSDSHULQWKLVYROXPHThe languages of Melanesia: Quantifying the level of coverage by Hammarström & Nordhoff, adds SUHFLVLRQWRWKLVVWDWHPHQWE\SUHVHQWLQJUHOHYDQW¿JXUHVIURPWKHLU/DQJ'RFGDWDEDVH This database aims to give comprehensive global listing of all materials existing on all ODQJXDJHVDORQJZLWKDQLQLWLDODSSUR[LPDWHPHWULFRIGHJUHHRIFRYHUDJH$VWKHDXWKRUV SRLQW RXW WKHUH DUH PDQ\ VKRUWFRPLQJV WR WKHLU PHWULF ,W LV UHODWLYHO\ XQDPELWLRXV D ODQJXDJHSRVVHVVLQJDORZTXDOLW\JUDPPDURISDJHVDQGQROH[LFRQRUWH[WFROOHFWLRQ would already be placed at the highest level—well short of the modern gold standard of a %RDVLDQWULORJ\VXSSOHPHQWHGE\DZLGHYDULHW\RIDQQRWDWHGPXOWLPHGLD¿OHV²DQGWKHUH LVQRPHDVXUHRITXDOLW\RIDQDO\VLV'HVSLWHWKHVHÀDZVLWKDVWKHJUHDWYLUWXHRIEHLQJ operationalisable and applicable to all the world’s languages in a relatively automatic way, and in their paper they outline their scheme in detail as well as comparing the level of FRYHUDJHIRU0HODQHVLDZLWKWKHUHVWRIWKHZRUOG First, regarding the total proportion of the world’s languages spoken in Melanesia, WKHLU¿JXUHVFRXQWODQJXDJHVWKDWDUHµ0HODQHVLDQ¶LQWKHLUVHQVH $XVWURQHVLDQ QRQ$XVWURQHVLDQ RIWKHVXEUHJLRQRI2FHDQLDH[WHQGLQJIURPWKH$UDIXUD6HDLQ WKHZHVWWR)LMLLQWKHHDVW VHH¿JXUHLQWKHLUFKDSWHU 7KLVLVMXVWRYHURIWKH ZRUOGWRWDORI OLYLQJ ODQJXDJHVRQWKHLUFRXQW VHHWKHLUWDEOH ZLWKVRFDOOHG µ3DSXDQODQJXDJHV¶WKHQPDNLQJRIWKHZRUOG¶VWRWDO,QDEVROXWHWHUPVWKHQXPEHU RI ODQJXDJHV LQ 0HODQHVLD LV DOPRVW LGHQWLFDO WR WKRVH LQ WKH ZKROH RI (XUDVLD WKHVHWZREHLQJVXUSDVVHGRQO\E\$IULFD Second, for their assessment of level of documentation, they lump together Papuan ZLWKDOO$XVWURQHVLDQVRWKHLU¿JXUHVDOVRLQFOXGHWKHUHVWRI,QGRQHVLDWKH3KLOLSSLQHV 0DOD\VLD DQG VR RQ 'UDZLQJ RQ WKHVH ¿JXUHV WKH\ GUDZ VRPH VWULNLQJ FRQFOXVLRQV First, in absolute terms, Papua + Austronesian has the largest number of languages with RQO\DZRUGOLVWWRWKHLUGRFXPHQWDWLRQ LHWKHORZHVWOHYHORIGRFXPHQWDWLRQZKLFKWKH\ UHFRJQLVH 7KHFRPSDULVRQZLWK$XVWUDOLDDQRWKHUUHJLRQZKHUHSURIHVVLRQDOOLQJXLVWLF UHVHDUFKLVUHODWLYHO\UHFHQWLVVDOXWDU\ RYHURI$XVWUDOLDQODQJXDJHVKDYHDJUDPPDU DYDLODEOH FRPSDUH WR KDOI WKDW QXPEHU IRU$XVWURQHVLDQ 3DSXDQ 6HFRQG in relative terms, Papua + Austronesian has the lowest proportion of languages with the KLJKHVWUDQNRIGHVFULSWLRQ LHDJUDPPDURISDJHVRUPRUH WKHKLJKHVWSURSRUWLRQ ZLWKRQO\DZRUGOLVWRUOHVVDQGWKHORZHVWDYHUDJHOHYHORIGRFXPHQWDWLRQ7KLUGZKHQ Austronesian and non-Austronesian languages are compared within the above categories, the non-Austronesian languages have lower levels of documentation, making their overall documentation status even lower than that for Papua + Austronesian as a whole (see their 7DEOH In the years to come it is to be hoped that LangDoc will be extended to give more DFFXUDWH PHWULFV LQ D QXPEHU RI ZD\V ± VRPHWKLQJ ZKLFK ZLOO EH DLGHG WR WKH H[WHQW WKDWPRUHOLQJXLVWVKHHGWKHDXWKRUV¶FDOOWRSXWWKHLUUHVXOWVLQWKHSXEOLFGRPDLQ%XW WKHLUSDSHUDOUHDG\SURYLGHVDYHU\FOHDUTXDQWLWDWLYHEDVLVIRURXUFODLPDERYHWKDWWKH 0HODQHVLDQUHJLRQ±DQGSDUWLFXODUO\WKH3DSXDQODQJXDJHVZLWKLQLW±LVIDUDQGDZD\WKH most linguistically diverse part of the planet, and that conversely it suffers from the lowest OHYHORIODQJXDJHGRFXPHQWDWLRQIRXQGLQDQ\TXDUWHURIWKHHDUWK7KHFRPELQDWLRQRI MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Introduction WKHVHIDFWRUVLVZKDWPDNHVWKHVWXG\RI0HODQHVLDQODQJXDJHVDQHQRUPRXVFKDOOHQJH %HIRUHOHDYLQJWKLVSDSHUZHQRWHWZRLPSRUWDQWIXWXUHGHYHORSPHQWV)LUVWO\LWZLOO be crucial to link some form of comprehensive database like LangDoc to actual documents so that it is possible to inspect the actual materials listed there and gradually improve WKHTXDOLWDWLYHUDWLQJVWKURXJKWKHFROOHFWLYHHIIRUWVRIZRUOGVFKRODUVKLS6HFRQGO\LWLV desirable that the structure of the LangDoc database allows inspection of data at a number RIGLIIHUHQWJHRJUDSKLFOHYHOV:KLOHWKHLUSUHVHQWDUWLFOHODUJHO\WUHDWHG0HODQHVLDDVDQ undifferentiated whole, their discussion of one geographical variable (distance from coast) VKRZVKRZPRUH¿QHO\DUWLFXODWHGJHRJUDSKLFDOFKDUDFWHULVDWLRQVFDQEHPDGH±VRWKDW one can compile comparable reports for geographical regions like the Sepik, Bougainville, HWF The staggering linguistic complexity of Melanesia creates special problems for attempts to classify languages into families and subgroups, especially for efforts that try to reduce the large number of independent maximal clades (over forty on any estimate) by JURXSLQJVRPHRIWKHPWRJHWKHU Under these circumstances, the languages of Melanesia have provided a particularly important testing-ground in recent years for new methods which aim to ‘break the time barrier’ of the classical comparative method, by drawing inferences from the signal in assemblages of typological traits rather than simply in the sound-meaning pairings of WKHOH[LFRQDQGJUDPPDWLFDOPRUSKRORJ\7KRXJKFRQWURYHUVLDODQGVWLOOVXEMHFWWR¿HUFH FULWLTXH VHH UHIHUHQFHV LQ 5HHVLQN DQG 'XQQ SDSHU LW LV OLNHO\ WKDW VXFK PHWKRGV DV applied to Melanesia are here to stay, at the very least as a supplement to the comparative PHWKRG,QGHHGWKHVLWXDWLRQLQ0HODQHVLDIRUFHVKLVWRULFDOOLQJXLVWVWRPDNHDYLUWXHRI QHFHVVLW\E\GULYLQJWKHPWRGHYHORSQHZPHWKRGV The article by Reesink and Dunn gives an overview of these methods, focussing on WKHODQJXDJHVRI(DVWHUQ,QGRQHVLDVSRNHQDURXQGWKH%LUG¶V+HDGDUHD$VLQWKHLURWKHU studies, a grave problem with the method is that resemblant signals can signal either shared SK\ORJHQ\RUDUHDOFRQYHUJHQFH,QWKHFHQWUDOSDUWRIWKHLUSDSHUWKH\FRQVLGHUWKHFDVHRI two Papuan languages of the Bird’s Head, Hatam and Meyah, which consistently cluster with the Austronesian language Biak no matter how many ‘founding lineages’ (K values) DUHDVVXPHGRQUXQVRIWKHµ6WUXFWXUH¶DOJRULWKP,QWKLVFDVHWKHQµLWWKXVDSSHDUVWKDW GLIIXVLRQRYHUULGHVSK\ORJHQ\¶DVWKH\SXWLW %XWWKH\WKHQWDNHDIXUWKHUVWHSWHDVLQJRXWWKH¿IWHHQW\SRORJLFDOIHDWXUHV RXWRI 160 altogether) which align with phylogeny rather than areality, opposing the Papuan ODQJXDJHV+DWDPDQG0H\DKDJDLQVWWKH$XVWURQHVLDQODQJXDJH%LDN±VHHWKHLUWDEOH 'RHVHYLGHQFHOLNHWKLVKROGWKHNH\WRUH¿QLQJW\SRORJLFDOVXLWHEDVHGPRGHOVVRWKDWWKH\ FDQ¿OWHURXWDUHDOQRLVHWR¿QGWKHSK\ORJHQHWLFVLJQDO"2EYLRXVO\LIWKHDUJXPHQWDWLRQ proceeds just from a single case, as here, it risks being merely post hoc, but on the other KDQGLWZRXOGEHSRVVLEOHWRLWHUDWHWKLVSURFHGXUHRYHUDQXPEHURIDUHDVDQGVPDOOJURXSV :LOO LWHUDWLRQV RI WKLV W\SH E\ ¿OWHULQJ RXW WKH PRUH IURP WKH OHVV GLIIXVDEOH RYHU LQGHSHQGHQWFDVHVIURPDURXQGWKHZRUOGDOORZXVWR¿QHWXQHDQDOJRULWKPOLNH6WUXFWXUH by weighting the evidentiary value of different typological characters as regards to SK\ORJHQ\YVDUHDOLW\"7KLVZLOOEHDFUXFLDOTXHVWLRQRYHUWKHQH[WGHFDGHVRIVFKRODUVKLS as more extensive documentation of Melanesia’s languages provides us with more LQIRUPDWLRQIRUIHHGLQJLQWRFRPSDUDWLYHHQWHUSULVHVOLNH5HHVLQNDQG'XQQ¶V$WWKHVDPH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Introduction 5 WLPHWKHLUZRUNUHPLQGVXVWKDWWRGUDZPD[LPXPEHQH¿WIURPUHVHDUFKOLNHWKHLUVRXU ODQJXDJHGRFXPHQWDWLRQVQHHGWRHQVXUHWKDWPDWFKHGW\SRORJLFDOGDWDLVREWDLQHG±WKLV QHHGQRWHQWDLOµTXHVWLRQQDLUHVW\OH¶JUDPPDUVEXWIHDWXUHOLVWVOLNHWKRVHLQWKH$SSHQGL[ to their article do lay down a basic checklist of typological points which all descriptions VKRXOGPDNHVXUHWRFRYHU The next two papers each consider regions of Melanesia in which there have been FRPSOH[LQWHUSOD\VEHWZHHQODQJXDJHVEHORQJLQJWRTXLWHGLIIHUHQWIDPLOLHVLQDVRFLDO environment where different types of contact appear to have played a role at different SRLQWVLQWKHSDVW Marian Klamer’s Papuan-Austronesian language contact: Alorese from an areal perspective focuses on Alorese, an Austronesian language abutting the westernmost group RIH[WDQW3DSXDQODQJXDJHVRQWKHLVODQGRI3DQWDU6KHGHGXFHVDFRPSOH[FRQWDFWKLVWRU\ FRPSULVLQJDWOHDVWWZRVWDJHVSOD\HGRXWLQGLIIHUHQWORFDWLRQV 7KH¿UVWSKDVHRQKHUPRGHOZRXOGKDYHWDNHQSODFHRQWKHLVODQGRI)ORUHVRUQHDUE\ DWDWLPHZKHQ3DSXDQODQJXDJHVZHUHVWLOOVSRNHQWKHUH,WLVDWWKLVVWDJHWKDWWKHODQJXDJH DQFHVWUDOWRPRGHUQ/DPRKRORWDQG$ORUZRXOGKDYHDFTXLUHGDVXLWHRIW\SRORJLFDOIHDWXUHV WKDWDUHVHHQDVW\SLFDOO\µ3DSXDQ¶±RUPRUHSUHFLVHO\DVW\SLFDORIWKH3DSXDQODQJXDJHV RIWKH$ORU3DQWDUUHJLRQ±LQFOXGLQJSRVWSUHGLFDWHQHJDWLRQWKHPDUNLQJRISRVVHVVRUV noun-locational order in locative constructions, the presence of a focus particle and the DEVHQFHRIDSDVVLYHYHUEIRUP7KLVµ3DSXDQLVDWLRQ¶RISURWR/DPRKRORWZRXOGKDYHWDNHQ SODFHXQGHUFRQGLWLRQVRIORQJWHUPVWDEOHFRQWDFWLQYROYLQJSUHDGROHVFHQWVDFTXLULQJWKH complexities of both Papuan and Austronesian languages and melding them into a new V\VWHP In a second phase, following the migration of Alorese speakers to Pantar and the separation this entailed from their Lamoholot cousins, a series of further changes would KDYHWDNHQSODFH$ORUHVHFRQWUDVWVGUDVWLFDOO\ZLWK/DPRKRORWLQWHUPVRIPRUSKRORJLFDO FRPSOH[LW\:KHUH /DPRKRORW KDV WZR VHWV RI VXEMHFW DI¿[HV WR WKH YHUE SUH¿[HV IRU WUDQVLWLYHVVXI¿[HVIRULQWUDQVLWLYHV $ORUHVHUHOLHVRQIUHHSURQRXQVZLWKDOOEXWDIHZ IUHTXHQWYHUEVZKLFKUHWDLQIRVVLOLVHGDJHQWSUH¿[HV$QGZKHUH/DPRKRORWKDVDQXPEHU RIGHULYDWLRQDODI¿[HV VRPHSURGXFWLYHVRPHOH[LFDOLVHG $ORUHVHKDVQRGHULYDWLRQDO PRUSKRORJ\DWDOO±UHGXSOLFDWLRQLVLWVRQO\SURGXFWLYHZRUGIRUPDWLRQSURFHVV7KHVH GLIIHUHQFHVVXJJHVWDUDGLFDOSURFHVVRIPRUSKRORJLFDOVLPSOL¿FDWLRQLQWKHSDVVDJHIURP /DPRKRORWWR$ORUHVH.ODPHUK\SRWKHVLVHVWKDWLQWKHLQLWLDOVWDJHVRI$ORUHVHVHWWOHPHQWV of Pantar and Alor, Alorese-speaking men would have taken as their wives women speaking D QXPEHU RI GLIIHUHQW3DSXDQ ODQJXDJHVRI WKHLQODQG(QWHULQJWKHVSHHFKFRPPXQLW\ DVDGXOWVWKH\ZRXOGKDYHOHDUQHGDVLPSOL¿HGIRUPRI$ORUHVHMHWWLVRQLQJDOPRVWDOORI LWVPRUSKRORJ\7KHFRQWDFWEHWZHHQ$ORUHVHDQGORFDO3DSXDQODQJXDJHVKRZHYHUZDV QHLWKHUSURORQJHGQRUFRQVLVWHQWDWWKLVVWDJH7KHQXPEHURIORDQZRUGVIURPORFDO3DSXDQ ODQJXDJHVLVUHODWLYHO\ORZ RQO\$ORUHVHWHUPVRXWRIDZRUGOLVWKDYHDNQRZQ Papuan source) and is moreover distributed evenly across the different Papuan languages RIWKHORFDOLW\7KLVVXJJHVWVDQXPEHURIUHODWLYHO\ZHDNFRQWDFWVDQGQRVWDEOHSDWWHUQRI ELOLQJXDOFRQWDFW This case study illustrates a type of multi-phase contact scenario likely to have been played out between Austronesian and Papuan speakers in a number of parts of Eastern ,QGRQHVLD DW GLIIHUHQW SKDVHV RYHU WKH ODVW WZR WR WKUHH PLOOHQQLD 7KH YHU\ GLIIHUHQW MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Introduction 6 outcomes of the two phases posited in Klamer’s model are a salutary reminder of the social and linguistic complexity that must have been involved between two groups who would KDYHEHHQGHPRJUDSKLFDOO\HTXDOO\SRLVHGDQGLQWHUGHSHQGHQWLQPDQ\ZD\V$WWKHVDPH WLPHDV.ODPHUSRLQWVRXWLWLVRQO\DUHFRQVWUXFWLRQDQGZHZRXOGEHRQPXFK¿UPHU ground if we were able to draw on contemporary sociolinguistic studies of the types of LQWHUDFWLRQ±VRFLDODQGOLQJXLVWLF±WKDWDUHRFFXUULQJEHWZHHQJURXSVDORQJWKH3DSXDQ $XVWURQHVLDQLQWHUDFWLRQ]RQH$VZLWKVRPDQ\RIWKHTXHVWLRQVUDLVHGLQWKLVLVVXHWKH time for this sort of study is running out fast, as the presence of an alternative lingua franca HJ,QGRQHVLDQ UDGLFDOO\DOWHUVWKHW\SHRIOLQJXLVWLFLQWHUDFWLRQEHWZHHQVXFKJURXSV From Nusa Tenggara we then move east to the Southern New Guinea region, the focus of Nicholas Evans’ Even more diverse than we thought: The multiplicity of Trans- Fly languages,QFRQWUDVWWRWKH$XVWURQHVLDQ3DSXDQLQWHUDFWLRQVLQWKHSUHFHGLQJWZR DUWLFOHVKHUHWKHLQWHUDFWLRQVDUHEHWZHHQYDULRXVXQUHODWHG3DSXDQJURXSV6RXWKHUQ1HZ Guinea is an intriguing zone, of great diversity, about which our level of knowledge dips HYHQORZHUWKDQWKHQRUPVIRUHOVHZKHUHLQ0HODQHVLD The Southern New Guinea region is essentially a nucleus of several small language IDPLOLHVVXUURXQGHGE\7UDQV1HZ*XLQHDODQJXDJHVZKLFKVLJQL¿FDQWO\RXWQXPEHUWKHP GHPRJUDSKLFDOO\DQGZKLFKDWWKHWLPHRI¿UVWFRORQLDOGRFXPHQWDWLRQWHQGHGWREHIDU PRUHH[SDQVLYHDQGPLOLWDULVHGWKDQWKHLUQRQ71*FRXQWHUSDUWV,WRIIHUVDQH[FHOOHQW opportunity for historical linguistics to study the mechanisms by which Trans-New Guinea languages have expanded into areas previously characterised by greater levels of deep SK\ORJHQHWLFGLYHUVLW\ Nonetheless, it is clear that all languages of the region share a number of typological FKDUDFWHULVWLFV±WRWKHH[WHQWWKDWVRPHODQJXDJHVZKLFKKDYHEHHQFODVVL¿HGDV71*OLNH Marind, pattern typologically with other Southern New Guinea languages (as well as some ODQJXDJHVIXUWKHUD¿HOGLQFOXGLQJ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Introduction 7 DGGLWLRQDOWULDORUSDXFDO %XWWKHH[DFWURXWHE\ZKLFKVXFKV\VWHPVDUHGHULYHGYDULHV VLJQL¿FDQWO\IURPODQJXDJHWRODQJXDJH7KLVVXJJHVWVWKDWZKDWHYHUVHULHVRISDWKZD\V leads to shared areal features of this type is a long and tortuous one, probably based on the slow patchwork emergence of grammatical solutions to particular semantic targets shared DFURVVODQJXDJHVRIWKHUHJLRQ Staying in Southern New Guinea and sticking to the topic of grammatical number, Wayan Arka’s paper Projecting morphology and agreement in Marori, an isolate of southern New Guinea examines issues of how to represent systems of ‘constructed’ grammatical number featurally, focussing on the TNG-level isolate Marori and other languages with FRPSDUDEOHSKHQRPHQD Marori, in line with the South New Guinea pattern described in Evans’ article, constructs a three-valued number system by combining two binary values in a system of distributed H[SRQHQFH+RZHYHUWKHEDVHIHDWXUHVXVHGWRGHULYHWKLVUHVXOWDUHGLIIHUHQW ZKHUH1HQ crosses a singular vs non-singular with a dual vs non-dual distinction, Marori crosses a VLQJXODUYVQRQVLQJXODUZLWKDSOXUDOYVQRQSOXUDOGLVWLQFWLRQ$UND¶VDUWLFOHVKRZVKRZ WKHXQL¿FDWLRQRIQXPEHUYDOXHVLQ0DURULPRUSKRORJ\FDQEHGHULYHGZLWKLQDPRGHOLQ ZKLFKWKHIHDWXUHVDUHKLHUDUFKLFDOO\VWUXFWXUHGLQGLIIHUHQWZD\VLQGLIIHUHQWODQJXDJHV 7KXVZKHUH0DURULWUHDWVGXDODVWKHQXPEHUYDOXHWKDWLVQHLWKHUVLQJXODUQRUSOXUDO±DQG KHQFHUHOHJDWHVWKHGXDOWRDGHULYHGFDWHJRU\±WKH1HQIHDWXUHVWUXFWXUHEXLOGVLQGXDODVD SULPDU\VSHFL¿HGIHDWXUHEXWWUHDWVSOXUDODVDGHULYHGFDWHJRU\WKDWLVQHLWKHUVLQJXODUQRU GXDO7KLVPRGHOLVDQHOHJDQWLOOXVWUDWLRQRIKRZVRPHFURVVOLQJXLVWLFYDULDELOLW\LQIHDWXUH VWUXFWXUHFDQEHEXLOWLQWRDUREXVWRYHUDOODUFKLWHFWXUH±WKHSUHVHQFHRIDQRYHUDOOIHDWXUH structure, and of a primary singular vs non-singular cut, remain constant, but the internal PDNHXSRIWKHQRQVLQJXODUVXEVSDFHGLIIHUVDVEHWZHHQ0DURULDQG1HQ7KHDYDLODELOLW\ of differing feature architectures then makes it possible to model the differences between languages with similar sets of contrasts, but derived in different ways, within a formalism OLNH/)*±ZHUHIHUWKHUHDGHUWRWKDWFKDSWHUIRUWKHIRUPDOGHWDLOV 7KH GLI¿FXOWLHV LQYROYHG LQ OLQLQJ XS ODQJXDJHVSHFL¿F GHVFULSWLYH FDWHJRULHV ZLWK comparative concepts are nicely illustrated, from a different theoretical perspective, in Mats Exter’s article ‘Realis’ and ‘Irrealis’ in Wogeo: A valid category? Recall that, in 5HHVLQNDQG'XQQ¶VDUWLFOHRQHRIWKHLUTXHVWLRQV DVOLVWHGLQWKHLUDSSHQGL[ LVµ,V a distinction between realis/irrealis mood available as a morphological choice (1: present, 0: absent)’? But how do we decide what ‘realis/irrealis’ actually means? Wogeo offers LQWHUHVWLQJGLI¿FXOWLHVLQDQVZHULQJWKLVTXHVWLRQ Wogeo is a ‘mood-dominated’ Austronesian language, spoken off the north coast of 31*ZLWKDFRPSOH[YHUEDOPRUSKRORJ\LQFOXGLQJVL[SUH¿[DODQGHLJKWVXI¿[DOVORWV $EDVLFRSSRVLWLRQLVEHWZHHQWKHµUHDOLV¶DQGµLUUHDOLV¶IRUPVRIWKHSURQRPLQDOSUH¿[ illustrated by a pair like o-lako ‘I go, I went’ vs go-lako ‘I must go, I want to go, I will JR QRZ ¶,IWKLVZHUHDOOWKHUHZDVWRWKHRSSRVLWLRQWKHFKDUDFWHULVDWLRQZRXOGEHIDLUO\ straightforward, but once we consider more semantically precise combinations problems DULVH:RJHRKDV DGGLWLRQDO SUH¿[DO FRPELQDWLRQV H[SUHVVLQJ VXFK PHDQLQJV DV IXWXUH tentative (‘try doing X’), counterfactual (‘would have done X’), proximal imperfective (‘am/was doing X, nearby’)’ and distal imperfective (‘am/was doing X (further away)’), ZKLFKDUHIROORZHGE\HLWKHUWKHUHDOLVRULUUHDOLVSUH¿[HJm-o-lako>)87VJUHDOLVJR@ µ,ZLOOJR,FDQJR,PD\JR¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Introduction 8 )RUVRPHRIWKHVHWKHFKRLFHRISUH¿[PDNHVVHQVHLQWHUPVRIQRUPDOO\FKDUDFWHULVHG SURSHUWLHVRIWKLVRSSRVLWLRQHJWKHUHDOLVLVXVHGZLWKWKHWZRLPSHUIHFWLYHVHULHV%XW for others, notably the counterfactual, it is the realis series that is chosen rather than the H[SHFWHGLUUHDOLV([WHUWKHQJRHV RQ WR FRQVLGHU ZKDW VXFK FDVHV PHDQIRU WKHRYHUDOO HQWHUSULVH RI WU\LQJ WR GH¿QH WHUPV OLNH UHDOLV DQG LUUHDOLV LQ FURVVOLQJXLVWLF WHUPV +H ends up arguing against the usefulness of a term with as broad a range as the realis-irrealis FRQWUDVW±ZKLFKLIDFFHSWHGUDLVHVWKHSRVVLELOLW\WKDWW\SRORJLFDOFRPSDULVRQVPD\EH PRUH VXFFHVVIXO LI WKH\ ZRUN DW PXFK PRUH VHPDQWLFDOO\VSHFL¿HG OHYHOV ZKHUH FURVV OLQJXLVWLFFRPSDULVRQFDQEHPRUHSUHFLVH The next two papers examine the embedding of language in its sociocultural and SV\FKRORJLFDOFRQWH[WV Darja Hoenigman’s paper From mountain talk to hidden Talk: Continuity and change in Awiakay registers examines the diachronic sociolinguistics of special registers in Awiakay, a language of East Sepik province, and in the process throws a fascinating light on how ideologies of the need for linguistic difference intersect with high levels of PHWDOLQJXLVWLFDZDUHQHVVWRGULYHDG\QDPLFRIOH[LFDOLQQRYDWLRQ3DUWLFXODUO\QRWHZRUWK\ LVWKHFRQWLQXLW\±LQWHUPVRIXWLOLVLQJVSHFLDOUHJLVWHUV±WKDWKROGVLQWKHIDFHRIVLJQL¿FDQW FKDQJH±LQWKHIRUPRI&KULVWLDQVWULFWXUHVDJDLQVWWKHRQJRLQJXVHRIVRPHWUDGLWLRQDO UHJLVWHUV Traditionally, Awiakay people used a special register, known as ‘mountain talk’, to protect themselves from mountain spirits when travelling up into mountain regions; WKLVLQYROYHGWKHVXEVWLWXWLRQRUDYRLGDQFHRIDQXPEHURIOH[LFDOLWHPV7KHDUULYDORI Christianity has arrested the use of ‘mountain talk’, with the recognition it gives to the SRZHURISDJDQVSLULWVDQGNQRZOHGJHDQGXVHRIWKLVWUDGLWLRQDOUHJLVWHULVLQGHFOLQH%XW at the same time, another special register has come into use, kay menda RUµKLGGHQWDON¶ Travelling outside the village to regional centres such as Wewak, especially when it is for commercial purposes which leaves the travellers vulnerable to theft and predation, is regarded as a risky business and speaking a language impenetrable to outsiders provides JRRGVHFXULW\ Though Awiakay is traditionally spoken in just one village, and would therefore normally have been incomprehensible to outsiders, the recent arrival of Tok Pisin ORDQZRUGV FUHDWHV FKLQN LQ WKH DUPRU RI OLQJXLVWLF LPSHQHWUDELOLW\ ,W LV SUHFLVHO\ WKHVH loanwords which get replaced in kay menda, through ingenious native coinages some of which have already won full acceptance in the community and others of which still include ULYDOFRLQDJHV Hoenigman’s paper includes subtitled video footage of a journey from the village to WKHUHJLRQDOFHQWUHGXULQJZKLFKZHFDQZLWQHVVWKHFDPRXÀDJLQJSURFHVVHVRIµKLGGHQ talk’ at work, as well as watching the rehearsal and induction of less experienced members RIWKHSDUW\ZKLOHWUDYHOOLQJWRZDUGVWKHGHVWLQDWLRQ7KLVLVRILQWHUHVWQRWMXVWIRUWKHWRSLF of special registers, but more generally for our understanding of how at least some of the SURFHVVHVRIOLQJXLVWLFGLYHUVL¿FDWLRQLQ0HODQHVLDDUHGULYHQDORQJE\YHU\FRQVFLRXVDQG negotiated processes of change aimed at differentiating one’s language from that of other JURXSV Her paper concludes by surveying the parallels and differences between the new register RIKLGGHQWDONDQGWKHIDGLQJROGUHJLVWHURIµPRXQWDLQWDON¶%RWKDUHXVHGLQXQIDPLOLDU MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Introduction 9 perilous territory where one goes to obtain valued items, encountering dangerous entities (mountain spirits before, rascals now) and dangers (sickness before, robbery and theft now), preventing these dangers through judicious out-of-the-ordinary language use, and SUHGRPLQDQWO\LQYROYLQJPHQZKRDUHWKHRQHVWUDYHOOLQJWRWKHGDQJHURXVGHVWLQDWLRQV The most interesting difference, on her comparison, has to do with who is held to have FUHDWHGWKHVSHFLDOUHJLVWHU,QWKHFDVHRIPRXQWDLQWDONWKLVLVDWWULEXWHGWRµPRXQWDLQ spirits’ deep in the past, whereas in the case of ‘hidden talk’ the process of creation is still taking place, involves contemporary Awiakay individuals, and is therefore a process that LV DPHQDEOH WR GLUHFW UHVHDUFK RQ VXFK TXHVWLRQV DV KRZ ULYDO LQQRYDWLRQV DUH VHOHFWHG EHWZHHQZKLFKLWHPVDUHFKRVHQIRUFDPRXÀDJLQJDQGKRZFKDQJHVDUHSURSDJDWHGIURP LQQRYDWLYHLQGLYLGXDOVWRWKHFRPPXQLW\ Michael Frank’s paper, Cross-cultural differences in representations and routines for exact number, leads us from the known diversity of Melanesian languages to the presumed EXWXQWHVWHGFRJQLWLYHGLYHUVLW\WKLVVXEWHQGV%HOOHU %HQGHU ZKRPKHTXRWHV LQKLVDUWLFOHREVHUYHWKDWµWKHUHPD\EHQRRWKHUGRPDLQLQWKH¿HOGRIFRJQLWLYHVFLHQFHV ZKHUHLWLVVRREYLRXVWKDWODQJXDJH LHWKHYHUEDOQXPHUDWLRQV\VWHP DIIHFWVFRJQLWLRQ LHPHQWDODULWKPHWLF ¶&RPELQLQJWKLVZLWKWKHOLNHOLKRRGWKDW0HODQHVLDQGLYHUVLW\LQ QXPHUDOV\VWHPV /HDQ LVSHUKDSVHYHQJUHDWHULQUHODWLYHDQGDEVROXWLYHWHUPV than in other aspects of the language systems, we have here a fascinating domain for the LQYHVWLJDWLRQRIKRZOLQJXLVWLFGLYHUVLW\VKDSHVFRJQLWLYHGLYHUVLW\±DVZHOODVKRZFXOWXUDO practices like different counting routines themselves select for the emergence of different W\SHVRIQXPHUDOV\VWHP Frank’s paper does not in itself begin the exciting project of investigating how 0HODQHVLDQGLYHUVLW\LQQXPHUDOV\VWHPVSURGXFHV RUGRHVQ¶W VLJQL¿FDQWGLIIHUHQFHVLQ FRJQLWLRQ5DWKHULWVJRDOLVWRFODULI\WKHUHODWLRQVEHWZHHQRQWKHRQHKDQGKRZH[DFW number is represented linguistically or through other types of representational tool such as WKHWKH0HQWDO$EDFXVDQGQXPHULFDOFRJQLWLRQRQWKHRWKHU)UDQNDGGXFHVH[SHULPHQWDO evidence that linguistic systems in the form of numerals, but also non-linguistic systems in the form of the Mental Abacus, both provide a ‘cognitive technology’ enabling the online HQFRGLQJDQGPDQLSXODWLRQRITXDQWLW\LQIRUPDWLRQ)UDQNVKRZVWKDWFXOWXUDOH[SRVXUH alone does not scaffold the manipulation of exact number, that the lack of exact numeral terms in a language impacts negatively on arithmetical manipulations, and that it is not enough to possess a language with appropriate terms but that one must also be able to DFFHVVLWRQOLQHLQRUGHUWRVXFFHVVIXOO\FDUU\RXWDULWKPHWLFDOFDOFXODWLRQV 'LIIHUHQWQXPHUDOV\VWHPVVXFKDVWKHXVHRIGLIIHUHQWEDVHV FDQ be expected to furnish different cognitive strategies in this sense; and so would different QXPHUDOVHWVIRUGLIIHUHQWW\SHVRIFRXQWHGREMHFWV7KLVVXJJHVWVWKDWFROODERUDWLYHUHVHDUFK RQWKHLPSDFWRIQXPHUDOV\VWHPVRQQXPHULFDOFRJQLWLRQZLOO\LHOGULFKUHVXOWV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Introduction 10 LQWHUHVWLQJQXPHULFDOUHSUHVHQWDWLRQV7KLVJHQHUDOL]DWLRQVXJJHVWVWKHEHQH¿WVRI SV\FKROLQJXLVWLF¿HOGZRUNWKDWFRPELQHVH[SHULPHQWDOGHVLJQZLWKFURVVFXOWXUDO RUFURVVOLQJXLVWLFSRSXODWLRQV )UDQNWKLVYROXPH ,QIDFWWKHUHOHYDQWQXPHULFDOV\VWHPVDUHUDWKHUIUDJLOHLQVRPHFDVHVVLJQL¿FDQWO\ more so than other parts of the language system: ‘there are many cases where a language is not endangered, or not particularly endangered, but whose numeral systems are endangered’ &RPULH 2NVDSPLQLVDVDOXWDU\0HODQHVLDQFDVHLQYHVWLJDWHGE\6D[H DQG WKHQ6D[H (VPRQGH 7KRXJKWKHODQJXDJHLVVWLOOKHDOWK\ /RXJKQDQH LWV GLVWLQFWLYHEDVHERG\FRXQWV\VWHPLVJLYLQJZD\WRDQ(QJOLVKVW\OHGHFLPDOV\VWHPLQ DPRGHUQVHWWLQJZKHUHFRXQWLQJLVPRVWFRPPRQO\DSSOLHGWRPRQH\7KLVJUHDWIUDJLOLW\ of numeral systems means that there is an exceedingly narrow time window for carrying out the sort of collaborative work on the impact of numeral systems on numerical cognition ZKLFKLVRXWOLQHGLQ)UDQN¶VDUWLFOH 7KHPDQ\IDVFLQDWLQJTXHVWLRQVDQGUHVHDUFKWKURZQXSE\WKHSUHFHGLQJDUWLFOHV± DQGHYHQPRUHVRWKHIXWXUHUHVHDUFKZKLFKZHKRSHWKH\ZLOOVWLPXODWH±JHQHUDWHDQ HQRUPRXVDPRXQWRISULPDU\GDWDDV¿HOGZRUNHUVRIDUDQJHRILQWHUHVWVDQGQDWLRQDOLWLHV UHFRUGPDWHULDOVRQQXPHURXVODQJXDJHYDULHWLHVLQLQFUHDVLQJO\GDWDULFKIRUPDWV%XW endangered data is a problem we need to take almost as seriously as endangered languages WKHPVHOYHV±¿HOGQRWHVDQGUHFRUGLQJVPD\HQGXSORVWXQFDWDORJXHGXQORFDWDEOHRU GHJHQHUDWHRQROGWDSHVRURWKHUPDWHULDOV(TXDOO\SUREOHPDWLFDUHTXHVWLRQVDERXWZKHUH GDWDVKRXOGEHKRXVHGDQGZKRVKRXOGJHWDFFHVVWRLW0RGHUQGLJLWDODUFKLYHVDUHJLYLQJ XVWKHSRZHUWRDGGUHVVWKHVHLVVXHVLQDQHI¿FLHQWZD\7KH\KDYHWKHSRWHQWLDOWRSUHVHUYH huge amounts of data far into the future while allowing them to be accessible to researchers DQGFRPPXQLW\PHPEHUVIURPDOOORFDWLRQV,WLVZLWKWKHGHVLJQDQGUXQQLQJRIRQHVXFK archive, Paradisec, that the last article, by Nicholas Thieberger and Linda Barwick, is concerned: 7KH3DFL¿FDQG5HJLRQDO$UFKLYHIRU'LJLWDO6RXUFHVLQ(QGDQJHUHG&XOWXUHV (PARADISEC): A resource for Melanesian linguistics. 3$5$',6(& ZDV HVWDEOLVKHG LQ $XVWUDOLD LQ E\ D WHDP RI UHVHDUFKHUV OHG E\7KLHEHUJHU DQG %DUZLFN ,W ZDV ERUQ DV D UHVSRQVH WR WKH FKDOOHQJHV VHW RXW LQ WKH SUHFHGLQJSDUDJUDSKIURPDQDZDUHQHVVWKDWDYDVWERG\RIKDUGZRQ¿HOGGDWDZDVDWULVN RIYDQLVKLQJDOWRJHWKHU±SDUWO\DVDUHVXOWRISRRUIDFLOLWLHVLQORFDODUFKLYHV HJODFNRI air-conditioning to maintain tapes in good condition), partly through technological changes HJWKHGLVDSSHDUDQFHRIPDFKLQHVDEOHWRUHDGROGUHFRUGLQJVRQZD[F\OLQGHUVZLUH UHFRUGHUVHWF SDUWO\WKURXJKDODFNRIHPSKDVLVLQWKH¿HOGRIOLQJXLVWLFVRQWKHSULPDF\ of documentation as opposed to theoretical debate or grammar-writing, and partly through the reluctance of individuals to make their materials available to others until they had DQDO\VHGWKHPWKHPVHOYHV±DPRPHQWZKLFKVRPHWLPHVJHWVRYHUWDNHQE\$O]KHLPHU¶VRU GHDWK Part of their article is devoted to showing how PARADISEC works, in terms of HTXLSPHQW VHWXS EDFNXS DFFHVV ZRUNÀRZ DQG GDWD LQJHVWLRQ %XW 7KLHEHUJHU DQG Barwick also discuss a number of other important concerns raised in discussions of where DUFKLYHVOLNH3$5$',6(&¿WLQDUHJLRQFKDUDFWHULVHGE\VXFKYDVWGLVFUHSDQFLHVEHWZHHQ countries in terms of living standards, technology, access to digital data, and the potential YDOXHRIORFDOLQIRUPDWLRQ$FHQWUDOLVVXHLVWKHPRUDOWHQVLRQFHQWUHGRQWKHIRUHVHHDEOH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Introduction 11 DQGXQIRUHVHHDEOHXVHVRIDUFKLYHGLQIRUPDWLRQ HJLQHVWDEOLVKLQJFODQRZQHUVKLSRI ODQGVULJKWVWRUR\DOWLHVHWF ZKLFKUHTXLUHJUDGXDWHGOHYHOVRIDFFHVVRQWKHSKLORVRSK\ that there should be general commitment to permanent archiving, for future safety’s sake, but that communities and researchers should be supplied with technologies for regulating DFFHVVZKHUHWKLVLVZDUUDQWHG Looking in the other direction, the potential for harnessing the collective knowledge of various kinds of expert through cumulative annotation of archived material by different archive users at different locations is a goal that has great potential to galvanise a more collective and interdisciplinary approach to adding commentary and interpretation to SULPDU\PDWHULDOWKURXJKWLPH The nine contributions we have outlined can do no more than give a tantalising glimpse RIWKHFKDOOHQJHVUDLVHGE\WKHODQJXDJHVRI0HODQHVLD±IRUOLQJXLVWVDQGVFKRODUVLQDOOLHG ¿HOGV EXW DOVR IRU HGXFDWRUV FRPPXQLFDWLRQ WHFKQRORJLVWV DQG GHYHORSPHQW DJHQFLHV ZDQWLQJWRIRFXVRQORFDONQRZOHGJHDQGH[SHUWLVH0RVWLPSRUWDQWO\WKLVLVDFKDOOHQJH of utmost interest to community members wanting to maintain the intellectual wealth held LQ WKHLU OLQJXLVWLF KHULWDJH )RU WKHP FROODERUDWLYH ZRUN ZLWK OLQJXLVWV DQG RWKHUV FDQ offer new ways of integrating that heritage with other sorts of language products such as RUWKRJUDSKLHVGLFWLRQDULHVJUDPPDUVWH[WFROOHFWLRQVDQGGLJLWDOHWKQRHQF\FORSDHGLDV )RUHYHQDIUDFWLRQRIWKHVHFKDOOHQJHVWREHPHWPDQ\WKLQJVPXVWKDSSHQ:HQHHG to attract a new generation of adventurous and capable young scholars to work in this IDVFLQDWLQJGLYHUVHDQGKRVSLWDEOHSDUWRIWKHZRUOG:HQHHGWREXLOGFDSDFLW\DPRQJ local linguists and language workers in the countries where these languages are spoken, so as to reverse the drastic current imbalance between where Melanesian languages are VSRNHQDQGZKHUHIXWXUHUHVHDUFKHUVFDQUHFHLYHDGYDQFHGWUDLQLQJLQKRZWRVWXG\WKHP CELD, the Centre for Endangered Language Documentation in Manokwari, which hosted the conference where most of the papers here were presented, is a promising step in this GLUHFWLRQ There need to be many other developments like this, and international funding DJHQFLHVQHHGWREHFRQYLQFHGWKDWODQJXDJHGLYHUVLW\LVDUHVRXUFHQRWDKDQGLFDS7KLV is particularly relevant at a juncture when key sources of international research funding RYHUWKHODVWWZRGHFDGHV WKH9RONVZDJHQVWLIWXQJ¶V'R%H6SURJUDPWKH+DQV5DXVLQJ Endangered Languages Program, the NWO Bedreigde Talen program and the ESF (XUR%$%(/ SURJUDP DUH GUDZLQJ WR D FORVH RU KDYH DOUHDG\7KHUH LV YDVW SRWHQWLDO in such new approaches as BOLD or Basic Oral Language Documentation (KWWS ZZZ EROGSQJLQIRLZOS) and mobile-phone based crowd-sourcing to assist the data-gathering SURFHVV%XWWKHQHHGIRUORQJWHUPWUDGLWLRQDO¿HOGZRUNGUDZLQJRQWKHNQRZOHGJHRI OLQJXLVWVZKROHDUQWKHODQJXDJHVDQGFXOWXUHVRQVLWHZLOOUHPDLQIXQGDPHQWDO)LQDOO\ while there will always be some divergence of interest between missionary organisations and academically-motivated researchers, the vast extent of missionary enterprises through Melanesia means that the potential for fruitful collaborative work is vast, given JRRGZLOORQERWKVLGHV$QLPSRUWDQWUHFHQWLQLWLDWLYHLVWKHUHHVWDEOLVKPHQWRIWKHMRXUQDO Language and Linguistics in Melanesia, now as an open-access on-line journal (http:// ZZZODQJO[PHODQHVLDFRP), as a forum for publishing peer-reviewed research and book UHYLHZVRQWKHODQJXDJHVRI0HODQHVLD MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Introduction As can be seen from these considerations, and the fact that almost every paper in this collection is an early step in a new research path, the study of Melanesia’s languages offers abundant opportunities to make new discoveries We hope that in the collection of SDSHUVJDWKHUHGKHUH\RXZLOO¿QGPDWHULDOWKDWLQYLWHV\RXLQWRDQHQJDJHGDQGGLYHUVH international community of scholars dedicated to advancing our understanding of a OLQJXLVWLFWHUULWRU\WKDWLVDUJXDEO\WKHOHDVWFKDUWHGRQHDUWK REFERENCES %HOOHU 6 $ %HQGHU 7KH OLPLWV RI FRXQWLQJ 1XPHULFDO FRJQLWLRQ EHWZHHQ HYROXWLRQDQGFXOWXUHScience &RPULH %HUQDUG (QGDQJHUHG QXPHUDO V\VWHPV ,Q -DQ :RKOJHPXWK 7\NR 'LUNVPH\HU HGV Bedrohte Vielfalt: Aspekte des Sprach(en)tods [Endangered Diversity: Aspects of Language Death@%HUOLQ :HLVVHQVHH9HUODJ )UDQN0LFKDHO&7KLVYROXPH&URVVFXOWXUDOGLIIHUHQFHVLQUHSUHVHQWDWLRQVDQGURXWLQHV IRUH[DFWQXPEHU +DVSHOPDWK 0DUWLQ &RPSDUDWLYH FRQFHSWV DQG GHVFULSWLYH FDWHJRULHV LQ FURVV OLQJXLVWLFVWXGLHVLanguage /HDQ* &RXQWLQJV\VWHPVRI3DSXD1HZ*XLQHDDQG2FHDQLD/DH 3DSXD1HZ Guinea University of Technology doFWRUDOGLVVHUWDWLRQ /RXJKQDQH5RE\Q$JUDPPDURI2NVDSPLQ0HOERXUQH 8QLYHUVLW\RI0HOERXUQH 3K'WKHVLV 6D[H * % 'HYHORSLQJ IRUPV RI DULWKPHWLFDO WKRXJKW DPRQJ WKH 2NVDSPLQ RI 3DSXD1HZ*XLQHDDevelopmental Psychology 6D[H*% (VPRQGH,6WXG\LQJFRJQLWLRQLQÀX[ $KLVWRULFDOWUHDWPHQWRIfu in WKHVKLIWLQJVWUXFWXUHRI2NVDSPLQPDWKHPDWLFVMind, Culture, and Activity Nicholas Evans QLFKRODVHYDQV#DQXHGXDX Marian Klamer [email protected] MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY /DQJXDJH'RFXPHQWDWLRQ &RQVHUYDWLRQ6SHFLDO3XEOLFDWLRQ1R 'HFHPEHU Melanesian Languages on the Edge of Asia: Challenges for the 21st Century, HGE\1LFKRODV(YDQVDQG0DULDQ.ODPHUSS± KWWS QÀUFKDZDLLHGXOGFVS 2 KWWS KGOKDQGOHQHW The languages of Melanesia: Quantifying the level of coverage Harald Hammarström Max Plank Institute for Evolutionary Anthopology Sebastian Nordhoff Max Plank Institute for Evolutionary Anthopology The present paper assesses the state of grammatical description of the languages of the Melanesian region based on database of semi- DXWRPDWLFDOO\ DQQRWDWHG DJJUHJDWHG ELEOLRJUDSKLFDO UHIHUHQFHV years of language description in Melanesia has produced at least some grammatical information for almost half of the languages of Melanesia, almost evenly spread among coastal/non-coastal, Austronesian/non- $XVWURQHVLDQ DQG LVRODWHVODUJH IDPLOLHV 1HYHUWKHOHVV RQO\ RI WKHVH ODQJXDJHV KDYH D JUDPPDU DQG DQRWKHU KDYH D JUDPPDU VNHWFK &RPSDUHG WR (XUDVLD $IULFD DQG WKH $PHULFDV WKH 3DSXD Austronesian region is the region with the largest number of poorly documented languages and the largest proportion of poorly documented ODQJXDJHV :H FRQFOXGH ZLWK VRPH GLFXVVLRQ DQG UHPDUNV RQ WKH GRFXPHQWDWLRQDOFKDOOHQJHDQGLWVIXWXUHSURVSHFWV 1. INTRODUCTION. We will take Melanesia to be the sub-region of Oceania extending IURPWKH$UDIXUD6HDDQG:HVWHUQ3DFL¿FLQWKHZHVWWR)LMLLQWKHHDVW±VHHWKHPDSLQ ¿JXUH17KLVUHJLRQLVKRPHWRQRIHZHUWKDQ OLYLQJUHFHQWO\H[WLQFW DWWHVWHGLQGLJHQRXVODQJXDJHVDVSHUWKHODQJXDJHGLDOHFWGLYLVLRQVRI/HZLV ZLWK VPDOODGMXVWPHQWVDQGDGGLQJDWWHVWHGH[WLQFWODQJXDJHVJLYHQLQWDEOH 1 7KHDXWKRUVZLVKWRWKDQNWZRDQRQ\PRXVUHYLHZHUVIRUKHOSIXOFRPPHQWV cc Creative Commons Attribution Non-Commercial No Derivatives Licence Languages of Melanesia ~ quantifying coverage Action Language Location Living/Extinct Brief Rationale Added Bai of PNG, Madang Presumed Extinct Not the same as Dumun Miklucho- =¶JUDJJHQ Maclay Added Nori PNG, Sandaun Extinct Not the same as Warapu &RUULV'RQRKXH &URZWKHU:LONHV Added Kaniet of PNG, Manus Presumed extinct Not the same as Kaniet of Dempwolff Thilenius (Blust 1996) Added O’oku PNG, Northern Presumed Extinct 6HHPLQJO\D MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage 15 Action Language Location Living/Extinct Brief Rationale Removed *XOLJXOL>JOL@ Solomon Extinct Not different from Islands Kazukuru (Dunn & Ross Removed Makolkol PNG, New Possibly Extinct Unattested (Stebbins >]PK@ Britain Removed :DUHV>ZDL@ Indonesia, - Unattested or same Papua DV0DZHV>PJN@ (Wambaliau forthcoming) Removed TABLE 1. Adjustments concerning the languages of Melanesia to the language catalogue of /HZLV :HKDYHQRWDGGHGWRWDOO\XQDWWHVWHGYHU\SRRUO\DWWHVWHGODQJXDJHV HJ Ambermo, attested in two numerals, Fabritius 1855), or once attested languages whose DWWHVWDWLRQKDVGLVDSSHDUHG HJ5XWDQRQO\ZRUGVQRZUHPDLQLQJ&URZOH\E FIGURE 1. 0DS RI 0HODQHVLD DGDSWHG IURP KWWS HQZLNLSHGLDRUJZLNL0HODQHVLD DFFHVVHG -XO\ 7KH FRXQWULHV SUHVHQW LQ 0HODQHVLD DUH 3DSXD 1HZ *XLQHD ,QGRQHVLD)LML)UDQFH 1HZ&DOHGRQLD 6RORPRQ,VODQGVDQG9DQXDWX MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage 16 The present paper seeks to describe the current state of description of the languages of Melanesia in detail (in the online appendix at KWWS VFKRODUVSDFHPDQRDKDZDLLHGX ELWVWUHDPKDQGOH PHODQHVLDBDSSHQGL[SGI) and in general (in the body of WKHSDSHU EDVHGRQDGDWDEDVHRIDQQRWDWHGELEOLRJUDSKLFDOUHIHUHQFHV7KLVGDWDEDVHRI UHIHUHQFHVFDOOHG/DQJ'RF +DPPDUVWU|P 1RUGKRII VSDQVWKHHQWLUHZRUOGEXW ZHUHVWULFWLWWRWKH0HODQHVLDQVXEVHWLQWKHSUHVHQWVXUYH\ 2. ASSESSING STATUS OF DESCRIPTION. 7RDVVHVVVWDWXVRIGHVFULSWLRQZH¿UVWD FROOHFW all relevant bibliographical references, b) annotate them as to (target-)language and type (grammar, wordlist etc), and c) for each language, mark its status of description according WRWKHPRVWH[WHQVLYHRUVXPGHVFULSWLRQLWKDV 2.1. COLLECTING REFERENCES. Language documentation and description is, and has been, a decentralized activity carried out by missionaries, anthropologists, travellers, QDWXUDOLVWVDPDWHXUVFRORQLDORI¿FLDOVDQGQRWOHDVWOLQJXLVWV,QRUGHUWRFRPSUHKHQVLYHO\ collect all relevant such items, we have, in essence, gone through all handbooks and overviews concerning the Melanesian region, in the hope that specialists on families and VXE UHJLRQVKDYHWKHEHVWNQRZOHGJHRQZKDWGHVFULSWLYHPDWHULDOVDFWXDOO\H[LVW7KLV is supplemented by a) intensive searching as to (sub-)regions for which there is no recent expert-written handbook/overview paper and b) whole-sale inclusion of relevant existing bibliographical resources such as the WALS, the SIL Bibliography, SIL Papua Guinea %LEOLRJUDSKLHVWKHOLEUDU\FDWDORJXHRI03,(9$LQ/HLS]LJDQGVRRQ±VHH+DPPDUVWU|P DQG1RUGKRII IRUDOLWWOHPRUHGHWDLOUHJDUGLQJWKLVSURFHGXUHDQGDOWHUQDWLYHV Everything published by a locatable publisher has been included as well as MAs and 3K'VVLQFHWKH\VKRXOGLQSULQFLSOHEH¿QGDEOHYLDWKHQDWLRQDOOLEUDU\RUWKHGHJUHHJLYLQJ LQVWLWXWLRQ+RZHYHU¿HOGQRWHVPDQXVFULSWVVHOISXEOLVKHGLWHPVDQGLWHPVSXEOLVKHGE\ DORFDOELEOHVRFLHW\KDYHQRWEHHQLQFOXGHGVLQFHWKH\FDQQRWEHORFDWHGV\VWHPDWLFDOO\ In our experience, locating manuscripts too often turns out to be a wild goose chase and including them in the current survey would do more harm than good, in particular, it would JLYHDIDOVHSLFWXUHRIWKHVWDWHRI DFFHVVLEOH GHVFULSWLRQ+RZHYHUZHKDYHLQFOXGHGD VPDOOQXPEHURIPDQXVFULSWVDQGRU¿HOGQRWHVZKHUHWKHLWHPLQTXHVWLRQKDVEHHQSRVWHG RQWKHLQWHUQHWDQGRULVYHUL¿HGWREHORFDWHGLQDSXEOLFO\DFFHVVLEOHDUFKLYH HJWKH .,7/9LQ/HLGHQ DQGWKXVPHHWVWKHDFFHVVLELOLW\FULWHULRQ It should be stressed, however, that the amount of original and valuable data sitting in XQSXEOLVKHGIRUPLVKLJKO\VLJQL¿FDQW7RJLYHMXVWDIHZH[DPSOHV&DSHOO FLWHVD large number of missionary manuscripts from the islands east of the Papuan mainland, the DUFKLYHVRIWKH6,/LQ-D\DSXUDDQG8NDUXPSD FI6LO]HU +HLNNLQHQ&ORXVH KROG a huge number of unpublished survey wordlists and/or grammar sketches spanning (in our LPSUHVVLRQ DWOHDVWRIWKHODQJXDJHVRI0HODQHVLDDQGOLQJXLVWV0DUN'RQRKXHDQG :LOOLDP)ROH\KDYHXQSXEOLVKHG¿HOGGDWDIURP,QGRQHVLDQ3DSXDDQGWKH6HSLN5DPX region respectively which is enough for several full grammars and dozens of grammar VNHWFKHV SF 0DUN 'RQRKXH DQG :LOOLDP )ROH\ ,I XQSXEOLVKHG PDWHULDO LVLQFOXGHGWKHGHVFULSWLYHSLFWXUHRIWKHODQJXDJHVRI0HODQHVLDFKDQJHVVLJQL¿FDQWO\ especially on the breadth side, with far more data on the lesser-known languages FI&DUULQJWRQ ,QWRWDOWKHELEOLRJUDSKLFDOGDWDEDVHFRQWDLQV UHIHUHQFHVSHUWDLQLQJWR0HODQHVLD MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage 17 2.2. ANNOTATION. %LEOLRJUDSKLFDO UHIHUHQFHV DUH DQQRWDWHG DV WR LGHQWLW\ LH WKH LVR FRGHRIWKHODQJXDJH V WUHDWHGDQGW\SHRIGHVFULSWLRQLHJUDPPDUZRUGOLVW HWF$VWRW\SHWKHIROORZLQJKLHUDUFK\KDVEHHQXVHG JUDPPDU DQH[WHQVLYHGHVFULSWLRQRIPRVWHOHPHQWVRIWKHJUDPPDU SDJHVDQG beyond JUDPPDUVNHWFK DOHVVH[WHQVLYHGHVFULSWLRQRIPDQ\HOHPHQWVRIWKHJUDPPDU± 150 pages (typically 50 pages) GLFWLRQDU\ SDJHVDQGEH\RQG VSHFL¿FIHDWXUH GHVFULSWLRQRIVRPHHOHPHQWRIJUDPPDU LHQRXQFODVVV\VWHP verb morphology etc) SKRQRORJ\ SKRQRORJLFDOGHVFULSWLRQZLWKPLQLPDOSDLUV WH[W WH[W FROOHFWLRQ ZRUGOLVW DFRXSOHRIKXQGUHGZRUGV PLQLPDO DVPDOOQXPEHURIFLWHGPRUSKHPHVRUUHPDUNVRQJUDPPDU VRFLROLQJXLVWLF GRFXPHQWZLWKGHWDLOHGVRFLROLQJXLVWLFLQIRUPDWLRQ FRPSDUDWLYH LQFOXVLRQLQDFRPSDUDWLYHVWXG\ZLWKRUZLWKRXWFLWHGPRUSKHPHVHJ lexicostatistical survey KDQGERRNRYHUYLHZ GRFXPHQWZLWKPHWDLQIRUPDWLRQDERXWWKHODQJXDJH LHZKHUH VSRNHQQRQLQWHOOLJLELOLW\WRRWKHUODQJXDJHVHWF HWKQRJUDSKLF HWKQRJUDSKLFLQIRUPDWLRQRQWKHJURXSVSHDNLQJDODQJXDJH The hierarchy is an ad-hoc amalgam of existing annotation, automatizability properties and bias towards typologist usage (with grammar at the top, trumping text and dictionary, DQGIRUPIXQFWLRQSDLUVUDWHGKLJKHUWKDQVRFLROLQJXLVWLFLQIRUPDWLRQ ,WLVLQPDQ\ZD\V LPSHUIHFWEXWLWLVPRUHLQIRUPDWLYHWKDQQRWKLQJ2WKHUH[LVWLQJVFKHPDVFRXOGQRWEH IHOLFLWRXVO\DGRSWHGHJ0RRUH LVVLPLODUWRWKHSUHVHQWVFKHPHEXWFUHGLWVWKH H[LVWHQFHRIYDULRXVW\SHV VFLHQWL¿FDUWLFOHVGLVVHUWDWLRQVHWF UDWKHUWKDQWKHLUDFWXDO FRQWHQWDQG$,$76,6 LVDOVRVLPLODUWRWKHSUHVHQWVFKHPHEXWVRPXFK more detailed (several hundred categories including vocabulary/animals, vocabulary/ ERG\SDUWVHWF WKDWLWFRXOGQRWEHDXWRPDWL]HGRUGRQHE\KDQGZLWKLQWKHVFRSHRIWKH SUHVHQWSURMHFW%LEOLRJUDSKLFDOUHIHUHQFHVLQWKHSUHVHQWSURMHFWKDYHEHHQDQQRWDWHGERWK DXWRPDWLFDOO\DQGE\KDQG6RPHH[DPSOHVDUHVKRZQLQ7DEOH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage 18 Reference Language Type Comment /LQGVWU|P(YD 7RSLFVLQ .XRW>NWR@ grammar although it contains WKH*UDPPDURI.XRW6WRFNKROP some text and a University doctoral dissertation, Swadesh word-list at SS the end, it counts as grammar )UDQNOLQ.DUO- &/9RRUKRHYH )DVX>IDD@ overview; There is a discussion (1973) Languages near the )RH>IRL@ comparative; of comparative matters intersection of the Gulf, Southern )LZDJD>¿Z@ minimal and a number of +LJKODQGVDQG:HVWHUQ'LVWULFWV,Q .HZD>NHZ@ morphemes are given .DUO-)UDQNOLQ HG The linguistic IRUHDFKODQJXDJH situation in the Gulf District and adjacent areas, Papua New Guinea 3DFL¿F/LQJXLVWLFV 6HULHV& &DQEHUUD 5HVHDUFK 6FKRRORI3DFL¿FDQG$VLDQ6WXGLHV Australian National University :LU]3DXO Anthropologische =ZDUW9DOOH\ ethnographic; It contains a grammar und ethnologische Ergebnisse der Dani-Western grammar sketch in addition to Central Neu-Guinea Expedition >GQZ@ sketch HWKQRJUDSKLFGDWD Nova Guinea;9, +XJKHV-RFN 7KH 0DULUL>PTL@ overview; No actual words languages of Kei, Tanimbar and East Tarangan comparative or wordlists are $UX /H[LFRVWDWLVWLFFODVVL¿FDWLRQ >WUH@ included, just results of ,Q6RHQMRQR'DUGMRZLGMRMR HG /RUDQJ>OUQ@ FRPSDULQJZRUGOLVWV Miscellaneous studies of Indonesian /ROD>OFG@ and other languages in Indonesia, .RED>NSG@ part 9 (NUSA: Linguistic Studies .RPSDQH>NYS@ of Indonesian and Other Languages %DWXOH\>ED\@ LQ,QGRQHVLD -DNDUWD %DUDNDL>EDM@ Universitas Katolik Indonesia Atma .DUH\>N\G@ -D\D TABLE 2. ([DPSOHVRIWKHDQQRWDWLRQVFKHPHXVHGLQWKHSUHVHQWVXUYH\ Automatic annotation is possible when the title words contain the language name and/ RU ZRUG V UHYHDOLQJ WKH W\SH RI WKH GRFXPHQW HJ ³$ JUDPPDU RI 7DX\D´ FDQ EH DXWRPDWLFDOO\ UHFRJQL]HG DV >W\D@ DQG JUDPPDU ([DFWO\ KRZ WKLV LV GRQH DQG ZKDW SHUFHQWDJHVRIFRUUHFWQHVVDUHWREHH[SHFWHGLVGHVFULEHGLQ+DPPDUVWU|P )RUPRVWUHIHUHQFHVQXPEHURISDJHVLVUHFRUGHGDQGLVXVHGWRUDQNZLWKLQFDWHJRULHV 2.3. STATUS OF DESCRIPTION PER LANGUAGE )RU HDFK ODQJXDJH WKH UHIHUHQFHV concerning it are aggregrated and its status of description is straightforwardly assessed as MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage 19 SHUWKHDQQRWDWLRQKLHUDUFK\,QDGGLWLRQIRUWKHSXUSRVHVRIWKHFXUUHQWSUHVHQWDWLRQLWKDV EHHQVLPSOL¿HGLQWRDPRUHGLVWLOOHGVFKHPHDVSHU7DEOH numerical type distilled type value grammar Grammar grammar sketch grammar sketch 3 dictionary SKRQRORJ\GLFWLRQDU\VSHFL¿FWH[W text SKRQRORJ\GLFWLRQDU\VSHFL¿FWH[W VSHFL¿FIHDWXUH SKRQRORJ\GLFWLRQDU\VSHFL¿FWH[W wordlist wordlist or less 1 minimal wordlist or less 1 sociolinguistic wordlist or less 1 comparative wordlist or less 1 handbook/overview wordlist or less 1 ethnographic wordlist or less 1 TABLE 3.7KHIXOODQGGLVWLOOHGGHVFULSWLRQOHYHOKLHUDUFK\XVHGLQWKHSUHVHQWVXUYH\ There may be missing extant references and manual as well as automatic annotation has JDSVDQGHUURUV7KHFODLPZHDUHDEOHWRPDNHLVWKDWDWOHDVWWKHVWDWXVRIGHVFULSWLRQIRU HYHU\ODQJXDJHVKRXOGEHFRUUHFW7KDWLVWKHRXWFRPHKDVEHHQVFUHHQHGDWWKHODQJXDJH level by an informed human, and inasmuch as errors of omission and annotation remain, WKH\GRQRWDOWHUWKH FRUUHFW VWDWXVRIGHVFULSWLRQRIDQ\ODQJXDJH7KXVIRUDODQJXDJH which only has a published wordlist to its documentation it may be that there are several ZRUGOLVWVSXEOLVKHGEXWRQO\RQHRIWKHPLVDFFXUDWHO\UHÀHFWHGLQWKHGDWDEDVH DFFXUDWHO\ UHÀHFWLQJWKHRWKHUVZRXOGQRWFKDQJHWKHVWDWXVRIGHVFULSWLRQDZD\IURPZRUGOLVW DQG if a language is given a certain status of description, the claim is that there is, in reality, QRRWKHUGHVFULSWLYHSXEOLFDWLRQWKDWZRXOGJLYHLWDKLJKHUPDUN2IWKHSXEOLFDWLRQVWKDW DUHWKHZLWQHVVWRWKHVWDWXVRIGHVFULSWLRQRIDODQJXDJH WKHPRVWVLJQL¿FDQWLWHPVRI GHVFULSWLRQ KDYHEHHQSHUVRQDOO\LQVSHFWHGE\WKHDXWKRUVEXWVLQFHWKLVZDVGRQH over a long period of time it is no guarantee of consistency and we are not in a position to DVVHVVWKHTXDOLW\RIDGHVFULSWLRQ ,WVKRXOGEHQRWHGDJDLQWKDWWKHDERYHKLHUDUFK\UHÀHFWVGHVFULSWLYHVWDWXVDQGKDV DELDVWRZDUGVW\SRORJLVWXVDJH)RUH[DPSOHDODQJXDJHWKDWKDVDJUDPPDUGLFWLRQDU\ and text collection will be ranked the same (grammar) as a language with only a grammar, HYHQWKRXJKWKHIRUPHULVEHWWHUGRFXPHQWHGRYHUDOO$QLQGH[RIRYHUDOOGRFXPHQWDWLRQ HJZLWKSRLQWVVHSDUDWHO\IRUJUDPPDWLFDOOH[LFDODQGWH[WXDOGRFXPHQWDWLRQ FRXOGEH FRPSXWHGIURPWKHVDPHGDWDEDVH:HGRQRWGRWKLVIRUWKHSUHVHQWVXUYH\VLQFHZHFDQQRW YHQWXUHWKHVDPHFODLPRIFRPSOHWHQHVVDVZLWKWKHJUDPPDURULHQWHGVFKHPHDERYH,Q other words, the database screening is likely to have missed cases of missing texts and GLFWLRQDULHVIRUODQJXDJHVZKLFKDOUHDG\KDYHDJUDPPDU VNHWFK 7KHGDWDEDVHLVUHOHDVHG to the public so that others who are more interested in overall documentation can complete WKHGDWDEDVHDQGFRPSXWH¿JXUHVRIWKHLURZQ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage The fact that “grammar” is the highest weighted category of description should not be WDNHQWRPHDQWKDWDODQJXDJHZLWKDJUDPPDULVFRPSOHWHO\GHVFULEHG±LWPHUHO\PHDQV that it is the highest category of grammatical description that is commonly distinguished E\OLQJXLVWVLHWKHUHDUHDV\HWQRGHVFULSWLRQVWKDWDUHFDOOHG³VXSHUJUDPPDUV´RUWKH OLNH+RZHYHUJUDPPDUVFDQEHPRUHRUOHVVFRPSUHKHQVLYHDQGDFRUUHODWHRIWKLV ZLWK validity only on average) may be the number of pages, which is recorded in the present GDWDEDVH1RULVOHQJWKPRUHWKDQDURXJKSUR[\IRUTXDOLW\DQGFRPSUHKHQVLYHQHVV±LW ZRXOGUDQNDUDPEOLQJDQGREWXVHGRFXPHQWDERYHDFRQFLVHDQGHOHJDQWRQH±EXWLWKDV WKHYLUWXHRIEHLQJRSHUDWLRQDOLVDEOHDQGDSSOLFDEOHWRWKHGDWDZHKDYH 3. STATUS OF DESCRIPTION OF MELANESIAN LANGUAGES. Results of the full survey are given in the online appendix (KWWS VFKRODUVSDFHPDQRDKDZDLLHGXELWVWUHDP KDQGOH PHODQHVLDBDSSHQGL[SGI)VRUWHGE\IDPLO\DXWKRUDQGODQJXDJH:H UHYLHZWKHJHQHUDOLWLHVKHUH Total as Living Extinct Total percentage grammar 0 grammar sketch 7 SKRQRORJ\RUVLP 107 109 wordlist or less 756 779 TABLE 4.5DZQXPEHURIODQJXDJHVLQ0HODQHVLDDQGWKHLUOHYHORIGHVFULSWLRQ FIGURE 2.7KHORFDWLRQDQGGHVFULSWLRQOHYHORI0HODQHVLDQODQJXDJHV7KHFRORXUFRGLQJ LVJUDPPDU JUHHQJUDPPDUVNHWFK RUDQJHRUOLJKWJUD\ LIH[WLQFW SKRQRORJ\RUVLP RUDQJHUHGRUVODWHJUD\ LIH[WLQFW ZRUGOLVWRUOHVV UHGRUEODFN LIH[WLQFW MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage 5DZQXPEHUVRIODQJXDJHVGHVFULEHGWRYDULRXVGHJUHHVDUHVKRZQLQWDEOHDQGDPDS LVVKRZQLQ¿JXUH7KHQXPEHUVVSHDNIRUWKHPVHOYHV\HWWKHPRVWFRQVSLFXRXVIDFWLV that more than half of the languages of Melanesia have only a wordlist or less of published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howning 1976, Dutton 1976, Grace 1976, Haudricourt 1971, Healey 1976, Hooley /D\FRFN /D\FRFNDQG9RRUKRHYH /LQFROQ /LWKJRZ /\QFKDQG&URZOH\6FKW] 7D\ORU 7U\RQDQG+DFNPDQ 9RRUKRHYH E=¶JUDJJHQ :HVXSSOHPHQWWKHVHZLWKVRPHTXDQWLWDWLYHUHVXOWVLQ)LJXUH As can be seen, language description in Melanesia takes off in the second half of the 19th FHQWXU\ZLWKWUDYHOOHUVFRORQLDORI¿FHUVDQGPLVVLRQDULHVSURGXFLQJZRUGOLVWV)URPWKHUH GHVFULSWLRQLQFUHDVHVDWDVWHDG\SDFHGXHPRVWO\WRPLVVLRQDULHVDQG*HUPDQVFKRODUV A sharp rise in the number of items produced every year, and a corresponding (but less sharp) increase in the overall descriptive status, happens after 1950, presumably due to the HVWDEOLVKPHQWRIWKH6,/LQ3DSXD1HZ*XLQD +RROH\ )ROH\ 7KHSDFH has since been kept up mainly by SIL missionaries and academic linguists in Australia and RWKHUZHVWHUQFRXQWULHV9HU\OLWWOHKDVVRIDUEHHQSURGXFHGE\0HODQHVLDQVWKHPVHOYHV QRWDEOHH[FHSWLRQVLQFOXGH)ODVV\ 1HNLWHO 6XPEXN 7KHUHDUHPRUH than a dozen languages whose corresponding ethnic groups have a monograph-length HWKQRJUDSKLFGHVFULSWLRQ\HWWKHODQJXDJHVDUHQRWGHVFULEHGEH\RQGDZRUGOLVWHJ*QDX >JQX@ /HZLV RU%DQDUR>E\]@ -XLOOHUDW MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage FIGURE 3. The upper diagram shows the raw number of publications per year FRQFHUQLQJODQJXDJHVRI0HODQHVLD7KHORZHUGLDJUDPVKRZVWKHDYHUDJH GHVFULSWLRQOHYHODVLWLQFUHDVHVWKURXJKWLPH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage ,QWKHHDUO\WLPHVODQJXDJHVQHDUWKHFRDVWZHUHPXFKEHWWHUNQRZQWKDQLQODQGODQJXDJHV $WWKHSUHVHQWWLPHWKLVFRUUHODWLRQLVPXFKGLOXWHG7DEOHVKRZVWKHPHGLDQDQGDYHUDJH GLVWDQFHV DVWKHFURZÀLHV WRWKHFRDVWIRUWKHYDULRXVOHYHOVRIGHVFULSWLRQZKLFKVKRZV OLWWOHGLIIHUHQFH7KHVOLJKWWHQGHQF\IRUJUDPPDUVWREHZULWWHQRIODQJXDJHVQHDUHUWRWKH FRDVWLVQRWVWDWLVWLFDOO\VLJQL¿FDQWIRUDYHUDJHGLVWDQFHVEXWLWLVVRIRUPHGLDQGLVWDQFHV 7KLVPHDQVWKDWKDOIRIWKHODQJXDJHVZLWKJUDPPDUVDUHZLWKLQ NPVWRWKHFRDVW ZKHUHDV KDOI RI WKH ODQJXDJHV RI RWKHU FDWHJRULHV DUH IXUWKHU DZD\ DQG WKDW languages with grammars that are not near the coast (the exceptions) are so far away that WKH\EOXUWKHWHQGHQF\RQDYHUDJH7KLVRYHUDOOODFNRIDVWURQJHUWUHQGPXVWEHWDNHQWR PHDQWKDWÀLJKWDQGULYHUDFFHVVLQODQGEDODQFHVWKHDPRXQWRIQHJOHFWHGODQJXDJHVRQWKH FRDVWDQGLPPHGLDWHFRDVWDOKLQWHUODQGV Average distance S§ Median distance S§ to coast (kms) to coast (kms) grammar grammar sketch SKRQRORJ\RUVLP wordlist or less overall TABLE 5.$YHUDJH DQG PHGLDQ GLVWDQFH DV WKH FURZ ÀLHV IRU ODQJXDJHV RI YDULRXVOHYHOVRIGHVFULSWLRQ6LJQL¿FDQFHWHVWLQJLVE\VHOHFWLQJUDQGRP VXEVHWVRIWKHFRUUHVSRQGLQJVL]HIURPWKHWRWDOSRRORIODQJXDJHVDQG FKHFNLQJKRZPDQ\RIWKRVHKDYHDQDYHUDJHPHGLDQGLVWDQFHORZHUYL]KLJKHU WKDQWKHGLVWDQFHWREHWHVWHG As is well-known, the languages of Melanesia divide into two classes, the Austronesian ODQJXDJHV ODQJXDJHV DQG WKH QRQ$XVWURQHVLDQ ODQJXDJHV ODQJXDJHV 7KH $XVWURQHVLDQODQJXDJHVDUHPRUHFRDVWDO DYHUDJH NPVDQGPHGLDQ NPVIURP WKHFRDVW WKDQWKH3DSXDQRQHV DYHUDJH NPVDQGPHGLDQNPV EXWVLQFH there is only a weak or no trend that favours the description of coastal languages, we can check fairly easily if there is a bias towards the description of Austronesian or non- $XVWURQHVLDQODQJXDJHV)LJXUHVKRZVWKDWKLVWRULFDOO\WKHUHZDVDORQJWLPHGXULQJ which AN languages were better described on average (presumably due to being coastal) DQGLQUHFHQWWLPHVWKHVOLJKWO\KLJKHUOHYHOKDVEHHQUHJDLQHG7KHFXUUHQWDYHUDJHOHYHO RIGHVFULSWLRQIRU$1ODQJXDJHVLQ0HODQHVLDLVDJDLQVWIRUQRQ$1ODQJXDJHV 7KHGLIIHUHQFHLVVOLJKWEXWKLJKO\VLJQL¿FDQWp§7KHGLIIHUHQFHLVKDUGO\GXHWRWKH tendency for full grammars to be coastal, as the AN languages have higher representation at DOOOHYHOV EH\RQGZRUGOLVW DVSHU7DEOH:HGRQRWNQRZZKDWWKHUHDVRQIRUWKLVELDVLV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage FIGURE 4. The average description level for Austronesian (AN, green) and QRQ$XVWURQHVLDQ QRQ$1UHG ODQJXDJHVWKURXJKWLPH Austronesian non-Austronesian total number number number grammar 93 grammar sketch phonology 55 109 or similar wordlist or less 509 779 TABLE 6. Numbers and proportions of Austronesian (AN) and QRQ$XVWURQHVLDQ Q$1 ODQJXDJHVDWGLIIHUHQWOHYHOVRIGHVFULSWLRQ ,WLVGLI¿FXOWWRVD\ZKLFKLVWKHEHVWGHVFULEHGODQJXDJHRI0HODQHVLDDVWKDWZRXOGUHTXLUH DTXDOLW\MXGJPHQWWKDWZHDUHQRWLQDSRVLWLRQWRPDNH+RZHYHUWKHGHVFULSWLRQZLWK WKH ODUJHVW QXPEHU RI SDJHV LV /LFKWHQEHUN ¶V SDJH JUDPPDU RI 7R¶DED¶LWD MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage DQ2FHDQLF$XVWURQHVLDQODQJXDJHRIWKH6RORPRQ,VODQGV ,QIDFWLWLVDOVRWKHORQJHVW grammar of any lesser-known language in the world, in terms of number of pages devoted WRJUDPPDWLFDOGHVFULSWLRQ7KHVHFRQGORQJHVWJUDPPDURIDODQJXDJHRI0HODQHVLDLV $LNKHQYDOG ¶VSDJHJUDPPDURI0DQDPEX D1GXODQJXDJH $VIDUDVFDQEH told with documents accessible to us, the least described languages whose existence seems FHUWDLQ HQRXJK DUH .HKX >NKK@ DQG .HPEUD >[NZ@ WZR VHHPLQJO\ LVRODWHG ODQJXDJHV LQ ,QGRQHVLDQ 3DSXD .HKX LV NQRZQ IURP IURP WZR XQSXEOLVKHG PLQXVFXOH ZRUGOLVWV 0R[QHVV :KLWHKRXVHQG DWOHDVWRQHRIZKLFKLVIURPDQRQQDWLYHVSHDNHUDQG Kembra is known from a minuscule wordlist taken up from a transient speaker by Doriot DWWULEXWHGWRDYLOODJHQDPHG.HPEUDQHDUWKHFRQÀXHQFHRIWKH6REJHUDQG1DZD .LDPEUDDSSHDUVDWWKHULJKWSODFHRQDFRORQLDOPDS+RRJODQG $UJXDEO\WKHPRVWSUROL¿FDXWKRURIGHVFULSWLYHZRUNRQ0HODQHVLDQODQJXDJHVKDV EHHQ WKH 'XWFK &DWKROLF SULHVW 3HWUXV 'UDEEH 9RRUKRHYH ZKR FDQ FRXQW WR KLV QDPHQROHVVWKDQODQJXDJHVZLWKJUDPPDUVDQRWKHU ZLWKJUDPPDUVNHWFKHVDQG ZRUGOLVWVIRUPRUHVSDQQLQJDUDQJHRIGLIIHUHQWIDPLOLHV/LQJXLVW7HUU\&URZOH\ZURWH 6 grammars and 9 grammar sketches of Austronesian languages before his premature death LQ /LQJXLVWV VXFK DV$UWKXU &DSHOO 6WHSKHQ:XUP 6LGQH\ 5D\ 0DOFROP 5RVV -&$QFHDX[-$=¶*UDJJHQ'DUUHOO7U\RQDQG&/9RRUKRHYHKDYHEHWZHHQWKHP published wordlists (or similar bits of information) of several hundred languages, either FROOHFWHGWKHPVHOYHVRUE\RWKHUV $FXUUHQWGLVFXVVLRQDPRQJOLQJXLVWVDVWRSULRULWLHVIRUGRFXPHQWDWLRQ±WKHFRQWH[W EHLQJWKDWWLPHLVUXQQLQJRXW±LVZKHWKHUWRGHVFULEHDQXQGHVFULEHGLVRODWHGODQJXDJHRU ZKHWKHUWRGHVFULEHDQXQGHVFULEHGODQJXDJHIURPDIDPLO\ZLWKRWKHUGHVFULEHGODQJXDJHV $WSUHVHQWZHFRXQWODQJXDJHLVRODWHVIRUWKH0HODQHVLDQUHJLRQ VHH+DPPDUVWU|P DE DSSHQGL[ IRU D MXVWL¿FDWLRQ RI WKLV ¿JXUH 7KH LVRODWHV KDYH DQ DYHUDJH GHVFULSWLRQOHYHORIDQGWKH QRQLVRODWHVKDYH 7KHGLIIHUHQFHKRZHYHULV QRWVWDWLVWLFDOO\VLJQL¿FDQWDWFRQYHQWLRQDOOHYHOVRIVLJQL¿FDQFH p§ 7KDWLVWKHUH is no overall principle at work that has favoured the description of isolates rather than non- LVRODWHV1HYHUWKHOHVVWKHUHLVDFRQVSLFXRXVO\ODUJHabsolute number of underdescribed LVRODWHVDQGVPDOOIDPLOLHVLQWKH0HODQHVLDQUHJLRQHVSHFLDOO\ORZODQG1HZ*XLQHD±VHH +DPPDUVWU|P E IRUGHWDLOV 4. MELANESIAN LANGUAGES IN RELATION TO THE REST OF THE WORLD. The bibliographical database LangDoc spans the entire world in a fairly uniform way, allowing XV WR FRPSDUH 0HODQHVLD WR RWKHU FRQYHQWLRQDO PDFURDUHDV RI WKH ZRUOG 7KH WRWDO database contains over 160 000 references collected and annotated in much the same way DV WKH 0HODQHVLDQ VXESDUW +DPPDUVWU|P DQG 1RUGKRII $OWKRXJK WKH (XUDVLDQ Australian and Meso-American sections have not been screened as thoroughly as the other DUHDV\HWWKHJHQHUDOWUHQGVRIWKHFRPSDULVRQVZLWK0HODQHVLDVKRXOGVWLOOEHWUXVWZRUWK\ For this section, we will consider all Papuan-Austronesian languages together, not just the 0HODQHVLDQRQHVLQRUGHUWRDSSURSULDWHO\FRYHUDOORIWKHZRUOG¶VODQJXDJHV7KLVHQWDLOV WKDWWKH(XUDVLD¿JXUHVGRQRWLQFOXGHWKH$XVWURQHVLDQODQJXDJHVRI6RXWK(DVW$VLDWKH 3KLOLSSLQHVDQG,QGRQHVLD)LJXUHVDUHVKRZQLQ7DEOH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage North South Africa Australia Eurasia America Papua+AN America grammar >@ >@ >@ >@ >@ >@ grammar sketch >@ >@ >@ >@ >@ >@ phonology or >@ >@ >@ > @ >@ >@ VLP wordlist or less >@ >@ >@ > @ >@ >@ Total >@ >@ >@ >@ >@ > @ $YHUDJHGHVF JUDPPDU OLYLQJXQGRF TABLE 7. The number of languages at various levels of description broken up by PDFURDUHDV7KHQXPEHUVRXWVLGHEUDFNHWVUHIHUWRVWULFWO\OLYLQJODQJXDJHVDQG WKRVHZLWKLQEUDFNHWVUHIHUWRH[WLQFW7KHODVWURZJLYHVWKHSURSRUWLRQRIOLYLQJ ODQJXDJHVZLWKRQO\DZRUGOLVWRIOHVV In absolute terms, Papua+Austronesian has the largest number of languages with only a ZRUGOLVW WR WKHLU GRFXPHQWDWLRQ ,Q UHODWLYH WHUPV 3DSXD$XVWURQHVLDQ KDV WKH ORZHVW proportion of grammars, the highest proportion of languages with only a wordlist or less, DQGWKHORZHVWDYHUDJHOHYHORIGRFXPHQWDWLRQ7KH0HODQHVLDVXESDUWVFRUHVVOLJKWO\ORZHU RQDOOUHODWLYHDFFRXQWV7KHUHIRUH3DSXD$XVWURQHVLDQDQGWKHODQJXDJHVRI0HODQHVLD LQSDUWLFXODUFDQULJKWO\EHFDOOHGWKHOLQJXLVWLFDOO\OHDVWNQRZQDUHDRIWKHZRUOG 5. 21ST CENTURY CHALLENGES IN DOCUMENTATION. $VLVFOHDUIURPWKH¿JXUHVDERYH a formidable challenge for linguistic science is to provide descriptions of the vast number RIXQ GHU GHVFULEHGODQJXDJHVLQWKH0HODQHVLDQUHJLRQEHIRUHLWLVWRRODWH On the optimistic side, a) the trend from the past century predicts a continued large production of grammatical descriptions and, b) it seems, impressionistically, that people from a wider array of countries of the world are taking interest in the Melanesian languages, and c) infrastructure in Melanesia is making it easier to reach and live in otherwise remote DUHDV On the pessimistic side, a) at the same pace as infrastructure is developing the languages become endangered, b) violence, tropical diseases, visa/permit-matters and lack RIIXQGLQJFRQWLQXHWRGHWHU:HVWHUQHUVIURPLQVLWX¿HOGZRUNF KDUQHVVLQJRIORFDOWDOHQW and interest, and the training of linguists from the region, remains extremely undeveloped, DQGG ODUJHDPRXQWVRIGHVFULSWLYHZRUNQHYHUUHDFKWKHVFLHQWL¿FFRPPXQLW\DVLIVXFK PDWHULDOVKDGQRVFLHQWL¿FPHULW $IHZFRPPHQWVDUHLQRUGHU MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage The failure of local interest to develop into active descriptive work is not endemic WR0HODQHVLDSHUVHEXWLVZLGHVSUHDGLQDOORIWKHODQJXDJHULFKFRXQWULHVRIWKHZRUOG However, exceptions such as Brazil and Ethiopia show that it is possible for local XQLYHUVLWLHVDQGFRPPXQLWLHVWRWDNHDSURGXFWLYHLQWHUHVWLQORFDOODQJXDJHV In addition to unpublished materials alluded to above, many valuable descriptive works DUHGLI¿FXOWWRDFFHVVLQSDUWLFXODUDODUJHQXPEHURIXQSXEOLVKHG3K'DQG0$WKHVHV3K' DQG0$WKHVHVDUHLQPDQ\LQVWDQFHVWKHPRVWH[WHQVLYHGHVFULSWLRQWKHUHLVRIDODQJXDJH Many universities (for instance, the Australian National University) that regularly keep 0$WKHVHVGRQRWDOORZLQWHUOLEUDU\ORDQVRIWKHPSUHFLVHO\ZKHQWKHLUVLVWKHRQO\FRS\ Other universities, including the convenors of the 3L Language Documentation school, LH /HLGHQ 8QLYHUVLW\ 8QLYHUVLWp /XPLqUH /\RQ ,, DQG 62$6 HLWKHU GR QRW UHJXODUO\ keep awarded MA theses at all, or do not keep them in a manner that allows systematic DFFHVV VXFKDVWKH'HSDUWPHQWOLEUDU\RUWKHPDLQ8QLYHUVLW\OLEUDU\ 3HUKDSVWKHPRVW EODWDQWH[DPSOHRIDXQLYHUVLW\LQDQWLSDWK\RILWVVFLHQWL¿FSURGXFWLRQDFWXDOO\EHLQJXVHG LV8QLYHUVLWp/LEUHGH%UX[HOOHVDVWKH¿UVWDXWKRUH[SHULHQFHGSHUVRQDOO\DIWHUPDNLQJWKH WULSWR%UX[HOOHVWRUHDGWKHSUHVXPDEO\RQO\OLEUDU\FRS\RI/HY\ ¶V3K'JUDPPDU RI1XELD$ZDUE\IDUWKHPRVWH[WHQVLYHGHVFULSWLRQRIWKDWODQJXDJH$FFRUGLQJWR UHJXODWLRQVQRERG\±EHLWUHJLVWHUHGOLEUDU\FDUGKROGHUVRUYLVLWRUV±LVDOORZHGWRread WKLVWKHVLV OHWDORQHERUURZRUSKRWRFRS\IURP ZLWKRXWWKHZULWWHQFRQVHQWRIWKHDXWKRU 6LPLODUO\ ¿QLVKHG GRFXPHQWV DQG UHSRUWV IURP 6,/ 3DSXD 1HZ *XLQHD DQG 6,/ Indonesia cannot be systematically accessed, although many items have been made DFFHVVLEOHLQSXEOLFDWLRQVHULHVDQGRWKHURXWOHWV'LVVHPLQDWLRQLVDVFLHQWL¿FSULQFLSOH DQGVFKRODUO\LQVWLWXWLRQV±EHWKH\PLVVLRQDU\RUJDQL]DWLRQVRUXQLYHUVLWLHV±WKDWDFWLYHO\ RUSDVVLYHO\UHVWULFWDFFHVVWRRUHIIHFWLYHO\OHWVFLHQWL¿FDOO\YDOXDEOHGRFXPHQWVEHWKURZQ DZD\GRQRWIXOO\PHULWWKHODEHO¶VFLHQWL¿FLQVWLWXWLRQ¶,IGHVFULSWLYHZRUNFRQWLQXHVWREH GLVYDOXHGLQWKHDERYHH[HPSOL¿HGZD\VWKHUHLVOHVVLQFHQWLYHIRUPRUHGHVFULSWLYHZRUN WREHSURGXFHG $SDUW IURP ¿UVWKDQG GHVFULSWLYH ¿HOGZRUN WKHUH DUH OHVV REYLRXV ZD\V LQ ZKLFK RQHFDQFRQWULEXWHWRWKHGHVFULSWLRQRI0HODQHVLDQODQJXDJHV$QRQWULYLDOQXPEHURI ODQJXDJHVRI0HODQHVLDKDYHVFULSWXUHWUDQVODWLRQVLHERGLHVRIWH[WZLWKWUDQVODWLRQ EXWQRSXEOLVKHGJUDPPDWLFDOGHVFULSWLRQV7KHODQJXDJHVIRUZKLFKVFULSWXUHWUDQVODWLRQV DUHVDLGWRH[LVWDUHJLYHQLQ/HZLV 3DUWLDOEXWVXEVWDQWLDODQDO\VHVRIJUDPPDU FDQEHGRQHRQWKHEDVLVRIWH[WGDWDIURPVFULSWXUHWUDQVODWLRQVZLWKRXW¿HOGZRUNLQVLWX Comparative and typological work on languages of Melanesia can help generate interest LQSURGXFLQJPRUHGHWDLOHGGHVFULSWLRQV7KHGLJLWDOHUDDOORZVIRUWRROVRQPDQDJHPHQW annotation and interoperability of language resources which can free up time for strictly KXPDQQHHGHGDQDO\VLVIRUODQJXDJHGHVFULSWLRQ$QGLIQRWKLQJHOVHSXEOLVKLQJRUPDNLQJ DYDLODEOHOHJDF\UHVRXUFHVLVDYDOXDEOHFRQWULEXWLRQ3ULPHH[DPSOHVDUHWKHSXEOLFDWLRQ RI$QFHDX[¶VJLJDQWLFZRUGOLVWFROOHFWLRQIURP,QGRQHVLDQ3DSXDE\6PLWVDQG9RRUKRHYH D E DQGWKHGLJLWL]DWLRQRI$UWKXU&DSHOODQG'RQDOG/D\FRFN¶V ¿HOGQRWHV IURP 3DSXD 1HZ *XLQHD E\ 3$5$',6(& VHH 7KLHEHUJHU %DUZLFN WKLV YROXPH 6. CONCLUSION. 150 years of language description in Melanesia has produced at least some grammatical information for almost half of the languages of Melanesia, almost evenly spread among coastal/non-coastal, Austronesian/non-Austronesian and isolates/ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage ODUJHIDPLOLHV1HYHUWKHOHVVRQO\RIWKHVHODQJXDJHVKDYHDJUDPPDUDQGDQRWKHU KDYHDJUDPPDUVNHWFK&RPSDUHGWR(XUDVLD$IULFDDQGWKH$PHULFDVWKH3DSXD Austronesian region is the region with the largest number of poorly documented languages DQGWKHODUJHVWSURSRUWLRQRISRRUO\GRFXPHQWHGODQJXDJHV REFERENCES $,$76,6 AIATSIS Subject Thesaurus KWWS ZZZDLDWVLVJRYDXWKHVDXUXVGDWD 6XEMHFW7KHVDXUXVSGI -XQH $LNKHQYDOG $OH[DQGUD < The Manambu language of East Sepik, Papua New Guinea2[IRUG 1HZ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage 'UDEEH3HWHU Spraakkunst der Fordaatsche taal 9HUKDQGHOLQJHQYDQKHW.RQLQNOLMN %DWDYLDDVFK *HQRRWVFKDS YDQ .XQVWHQ HQ :HWHQVFKDSSHQ /;9,, (HUVWH 6WXN %DWDYLD $OEUHFKW 'UDEEH 3HWHU %LM]RQGHUKHGHQ XLW GH 7DOHQ YDQ )UHGHULN+HQGULN(LODQG .LPDJKDPD1GRPHQ5LDQWDQDBijdragen tot de Taal-, Land- en Volkenkunde ± 'XQQ 0LFKDHO 0DOFROP 5RVV $UH .D]XNXUX ODQJXDJHV UHDOO\$XVWURQHVLDQ" Oceanic Linguistics ± 'XWWRQ7RP )URPSRWVWRSHRSOH )LQHWXQLQJWKHSUHKLVWRU\RI0DLOX,VODQGDQG QHLJKERXULQJFRDVWVRXWKHDVW3DSXD1HZ*XLQHD,Q5RJHU0%OHQFK 0DWWKHZ 6SULJJV HGV Archaeology and LanguageYRO 2QH:RUOG$UFKDHRORJ\ /RQGRQ 1HZ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage 30 +DPPDUVWU|P +DUDOG 'DYLG .DPKRO] $ QRWH RQ 'XYOH:DQR SLGJLQ 3DSHU SUHVHQWHG DW WKH VHFRQG$3L&6 FRQIHUHQFH 0D[ 3ODQFN ,QVWLWXWH IRU (YROXWLRQDU\ $QWKURSRORJ\/HLS]LJ +DPPDUVWU|P+DUDOG 6HEDVWLDQ1RUGKRII/DQJ'RF %LEOLRJUDSKLFLQIUDVWUXFWXUH IRUOLQJXLVWLFW\SRORJ\Oslo Studies in Language ± +DXGULFRXUW$QGUp* 1HZ&DOHGRQLDDQGWKH/R\DOW\,VODQGV,Q6HEHRNLinguistics in Oceania +HDOH\$ODQ 7KH2NODQJXDJHIDPLO\LQ1HZ*XLQHD&DQEHUUD $183K'WKHVLV +HDOH\$ODQ +LVWRU\RIUHVHDUFKLQ$XVWURQHVLDQODQJXDJHV $GPLUDOW\,VODQGVDUHD In Wurm, Austronesian Languages, +RRJODQG- 0HPRULHYDQ2YHUJDYHYDQGH2QGHUDIGHOLQJ+ROODQGLD1DWLRQDDO Archief, Den Haag, Ministerie van Koloniën: Kantoor Bevolkingszaken Nieuw- *XLQHD WH +ROODQGLD 5DSSRUWHQDUFKLHI QXPPHU WRHJDQJ LQYHQWDULVQXPPHU +RROH\%UXFH$ 6,/UHVHDUFKLQ1HZ*XLQHDKivung ± +RROH\%UXFH$ +LVWRU\RIUHVHDUFKLQ$XVWURQHVLDQODQJXDJHV 0RUREHSURYLQFH In Wurm, Austronesian languages -XLOOHUDW%HUQDUG La révocation des Tambaran: les Banaro et Richard Thurnwald revisités &156(WKQRORJLH 3DULV &156 /DXIHU&DUO 3)XWVFKHUV$XI]HLFKQXQJHQEHUGLH%XWDP6SUDFKH 1HXEULWDQQLHQ Anthropos± /D\FRFN'RQDOG& $KXQGUHG\HDUVRI3DSXDQOLQJXLVWLFUHVHDUFK (DVWHUQ1HZ *XLQHDDUHD,Q:XUPPapuan Languages /D\FRFN'RQDOG& +LVWRU\RIUHVHDUFKLQ$XVWURQHVLDQODQJXDJHV 6HSLNSURYLQFHV In Wurm, Austronesian Languages /D\FRFN'RQDOG& &/9RRUKRHYH +LVWRU\RIUHVHDUFKLQ3DSXDQODQJXDJHV,Q Sebeok, Linguistics in Oceania /D\FRFN'RQDOG & -RKQ$ =¶*UDJJHQ 7KH6HSLN5DPXSK\OXP,Q:XUP Papuan Languages /HHGHQ$OH[DQGHU&RUQHOLVYDQGHU Verslag over taalgebieden in het Sarmische van de Ambtenaar van het Kantoor voor Bevolkingszaken YROXPH +ROODQGLD Gouvernement van Nederlands-Nieuw-Guinea, Dienst van Binnenlandse Zaken, .DQWRRUYRRU%HYRONLQJV]DNHQ /HY\&DWKHULQH$WHQWDWLYHGHVFULSWLRQRI$ZDUSKRQRORJ\DQGPRUSKRORJ\ /RZHU 5DPXIDPLO\3DSXD1HZ*XLQHD%UXVVHOV 8QLYHUVLWp/LEUHGH%UX[HOOHVGRFWRUDO GLVVHUWDWLRQ /HZLV*LOEHUW Knowledge of illness in a sepik society: a study of the Gnau, New Guinea/RQGRQ $WKORQH3UHVV /HZLV3DXO0 HG Ethnologue: Languages of the worldWKHGQ'DOODV 6,/ ,QWHUQDWLRQDO /LFKWHQEHUN)UDQWLVHNA grammar of Toqabaqita 0RXWRQ*UDPPDU/LEUDU\ YROV%HUOLQ 'H*UX\WHU /LQFROQ 3HWHU & +LVWRU\ RI UHVHDUFK LQ $XVWURQHVLDQ ODQJXDJHV %RXJDLQYLOOH 3URYLQFH,Q:XUPAustronesian Languages MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage 31 /LWKJRZ'DYLG +LVWRU\RIUHVHDUFKLQ$XVWURQHVLDQODQJXDJHV 0LOQH%D\3URYLQFH In Wurm, Austronesian Languages /\QFK-RKQ 7HUU\&URZOH\Languages of Vanuatu: A new survey and bibliography 3DFL¿F/LQJXLVWLFV &DQEHUUD 5HVHDUFK6FKRRORI3DFL¿FDQG$VLDQ6WXGLHV $XVWUDOLDQ1DWLRQDO8QLYHUVLW\ 0DUVGHQ:LOOLDP2QWKH3RO\QHVLDQRU(DVW,QVXODUODQJXDJHV,QMiscellaneous Works/RQGRQ 3DUEXU\$OOHQDQG&R 0RRUH 'HQQ\ (QGDQJHUHG ODQJXDJHV RI ORZODQG WURSLFDO 6RXWK $PHULFD ,Q 0DWWKLDV %UHQ]LQJHU HG Language Diversity Endangered (Trends in Linguistics: 6WXGLHVDQG0RQRJUDSKV %HUOLQ 'H*UX\WHU >0D[ 0RV]NRZVNL@ :|UWHUYHU]HLFKQLVVH YRQ 3DSXD6SUDFKHQ DXV KROOlQGLVFK 1HXJXLQHDAnthropos9,,,± 0R[QHVV0LNH $EULHIVHFRQGKDQGUHSRUWRQWKH.HKX .HX" ZRUGVIURPWKH PHPRU\RIDQRQQDWLYHVSHDNHU8QSXEOLVKHGPV 1HNLWHO 2WWR 6RFLROLQJXLVWLF$VSHFWV RI$EX¶ D 3DSXDQ /DQJXDJH RI WKH 6HSLN $UHD3DSXD1HZ*XLQHD&DQEHUUD $183K'WKHVLV 3DZOH\$QGUHZ5REHUW$WWHQERURXJK-DFN*ROVRQ 5RELQ+LGH HGV Papuan pasts: Studies in the cultural, linguistic and biological history of the Papuan-speaking peoples 3DFL¿F/LQJXLVWLFV &DQEHUUD 3DFL¿F/LQJXLVWLFV5HVHDUFK6FKRRORI 3DFL¿FDQG$VLDQ6WXGLHV$XVWUDOLDQ1DWLRQDO8QLYHUVLW\ 5D\6LGQH\+ 7KHODQJXDJHVRIWKH1HZ+HEULGHVJournal and Proceedings of the Royal Society of New South Wales;;9,,± ± 5D\6LGQH\+ 1RWHVRQODQJXDJHVLQWKHHDVWRI1HWKHUODQGV1HZ*XLQHD,Q$) 5:ROODVWRQ HG Pygmies and Papuans: The stone age to-day in Dutch New Guinea, /RQGRQ 6PLWK(OGHU &R 5D\6LGQH\+ 7KHODQJXDJHVRIWKH3DSXDQJXOIGLVWULFW3DSXDZeitschrift für Kolonialsprachen,9± 5D\6LGQH\+ 7KHODQJXDJHVRIQRUWKHUQ3DSXDJournal of the Royal Anthropological Institute of Great Britain and Ireland ± 5D\ 6LGQH\ + 7KH 3RO\QHVLDQ ODQJXDJHV LQ 0HODQHVLD Anthropos ± 5D\6LGQH\+ 7KH/DQJXDJHVRIWKH:HVWHUQ'LYLVLRQRI3DSXDJournal of the Royal Anthropological Institute of Great Britain and Ireland± 5D\6LGQH\+ A comparative study of the Melanesian island languages&DPEULGJH &DPEULGJH8QLYHUVLW\3UHVV 5D\6LGQH\+ 7KHODQJXDJHVRIWKH&HQWUDO'LYLVLRQRI3DSXDJournal of the Royal Anthropological Institute of Great Britain and Ireland ± 5D\6LGQH\+ E7KHODQJXDJHVRIWKH(DVWHUQ/RXLVLDGH$UFKLSHODJRBulletin of the School of Oriental and African Studies ± 5D\6LGQH\+ D7KHODQJXDJHVRIWKH(DVWHUQDQG6RXWK(DVWHUQ'LYLVLRQRI3DSXD JRAI± 5HHVLQN*HU30DQVLPDORVWODQJXDJHRIWKH%LUG¶V+HDG,Q*HU35HHVLQN HG Languages of the eastern Bird’s Head 3DFL¿F/LQJXLVWLFV &DQEHUUD 5HVHDUFK6FKRRORI3DFL¿FDQG$VLDQ6WXGLHV$XVWUDOLDQ1DWLRQDO8QLYHUVLW\ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage 5HPLMVHQ $ & / :RUGSURVRGLF V\VWHPV RI 5DMD $PSDW /DQJXDJHV /HLGHQ 5LMNVXQLYHUVLWHLWWH/HLGHQGRFWRUDOGLVVHUWDWLRQ 5RELGpYDQGHU$D3LHWHU-DQ%DSWLVW&DUHO5HL]HQYDQ')YDQ%UDDP0RUULV naar de noordkust van Nederlandsch Nieuw-Guinea eerste vaart op de Amberno- of 5RFKXVVHQ5LYLHUBijdragen tot de Taal-, Land- en Volkenkunde van Nederlandsch- Indië, 4e volg., Deel X± 6FKW]$OEHUW- The languages of Fiji2[IRUG &ODUHQGRQ3UHVV 6HEHRN7KRPDV$ HG Linguistics in Oceania &XUUHQW7UHQGVLQ/LQJXLVWLFV %HUOLQ 'H*UX\WHU 6LO]HU3HWHU- +HOMl+HLNNLQHQ&ORXVH Index of Irian Jaya languages (Special ,VVXH RI ,ULDQ %XOOHWLQ RI ,ULDQ -D\D QG HGQ -D\DSXUD 3URJUDP .HUMDVDPD 8QLYHUVLWDV&HQGHUDZDVLKDQG6,/ 6PLWV/HR &/9RRUKRHYH DThe J. C. Anceaux collection of wordlists of Irian Jaya languages A: Austronesian languages 3DUW, ,ULDQ-D\D6RXUFH0DWHULDO1R 6HULHV% /HLGHQ-DNDUWD '6$/&8/,5,6 6PLWV/HR &/9RRUKRHYH EThe J. C. Anceaux collection of wordlists of Irian Jaya languages A: Austronesian languages 3DUW,, ,ULDQ-D\D6RXUFH0DWHULDO1R 6HULHV% /HLGHQ-DNDUWD '6$/&8/,5,6 6PLWV/HR &/9RRUKRHYH The J. C. Anceaux collection of wordlists of Irian Jaya languages B: Non-Austronesian (Papuan) languages (Part I) (Irian Jaya Source 0DWHULDO1R 6HULHV% /HLGHQ-DNDUWD '6$/&8/,5,6 6PLWV/HR &/9RRUKRHYH The J. C. Anceaux collection of wordlists of Irian Jaya languages B: Non-Austronesian (Papuan) languages (Part II) (Irian Jaya Source 0DWHULDO1R6HULHV% /HLGHQ-DNDUWD '6$/&8/,5,6 6WHEELQV7RQ\D17KH3DSXDQODQJXDJHVRIWKH(DVWHUQ%LVPDUFNV PLJUDWLRQRULJLQV DQG FRQQHFWLRQV ,Q %HWKZ\Q (YDQV HG Discovering history through language: Papers in honour of Malcolm Ross 3DFL¿F /LQJXLVWLFV &DQEHUUD 5HVHDUFK6FKRRORI3DFL¿FDQG$VLDQ6WXGLHV$XVWUDOLDQ1DWLRQDO8QLYHUVLW\ 6XPEXN.HQQHWK0HPVRQ 0RUSKRV\QWD[RI6DUH:DLNDWR 8QLYHUVLW\RI:DLNDWR 3K'WKHVLV 7D\ORU$- +LVWRU\RIUHVHDUFKLQ$XVWURQHVLDQ/DQJXDJHV :HVWHUQSDUWRIVRXWK HDVWHUQPDLQODQG3DSXD,Q:XUP Austronesian Languages 7U\RQ 'DUUHOO 7 % ' +DFNPDQ Solomon Islands languages: An internal FODVVL¿FDWLRQ 3DFL¿F/LQJXLVWLFV& &DQEHUUD 5HVHDUFK6FKRRORI3DFL¿FDQG $VLDQ6WXGLHV$XVWUDOLDQ1DWLRQDO8QLYHUVLW\ YDQGHU/HHGHQ$OH[DQGHU&RUQHOLV VHH/HHGHQ 9RRUKRHYH%HUW(HQPRHLOLMNEHJLQ =HQGLQJPLVVLHHQWDDOZHWHQVFKDSLQ1LHXZ *XLQHD,Q:LOOHPYDQGHU0ROHQ %HUQDUG$USV HGV Woord en schrift in de Oost: De betekenis van zending en missie voor de studie van taal en literatuur in Zuidoost- Azië 6HPDLDQ /HLGHQ 8QLYHUVLWHLW/HLGHQ 9RRUKRHYH&/ D&HQWUDODQG:HVWHUQ7UDQV1HZ*XLQHD3K\OXP/DQJXDJHV,Q Wurm, Papuan Languages 9RRUKRHYH&/ E$+XQGUHG MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Languages of Melanesia ~ quantifying coverage 33 :DPEDOLDX7KHUHVLD)RUWKFRPLQJ'UDIW/DSRUDQ6XUYHLSDGD%DKDVD0DZHVGL3DSXD ,QGRQHVLDSIL Electronic Survey Reports :KLWHKRXVH 3DXO QG 7\SHXS RI DQRQ\PRXV .HKX ZRUGOLVW IURP 6,/ ,QGRQHVLD WKH ZRUGOLVWSUHVXPDEO\FRPHVIURP5RQ%DLUGLQWKH V 8QSXEOLVKHGPV :LONHV-5$GDPV $SSHQGL[% 9RFDEXODU\RIQDWLYHODQJXDJHVAustralian Report on the Administration of New Guinea ± :XUP6WHIDQ 7RQDOODQJXDJHVLQ1HZ*XLQHDDQGWKHDGMDFHQWLVODQGVAnthropos ± :XUP6WHSKHQ$ HG Papuan languages and the New Guinea linguistic scene 3DFL¿F /LQJXLVWLFV & 9RO RI New Guinea area languages and language study. &DQEHUUD 5HVHDUFK6FKRRORI3DFL¿FDQG$VLDQ6WXGLHV$XVWUDOLDQ1DWLRQDO 8QLYHUVLW\ :XUP6WHSKHQ$ HG Austronesian languages 3DFL¿F/LQJXLVWLFV& 9RO of New Guinea area languages and language study&DQEHUUD 5HVHDUFK6FKRRORI 3DFL¿FDQG$VLDQ6WXGLHV$XVWUDOLDQ1DWLRQDO8QLYHUVLW\ =¶JUDJJHQ -RKQ$ The Languages of the Madang District, Papua New Guinea 3DFL¿F/LQJXLVWLFV% &DQEHUUD 5HVHDUFK6FKRRORI3DFL¿FDQG$VLDQ6WXGLHV $XVWUDOLDQ1DWLRQDO8QLYHUVLW\ =¶JUDJJHQ -RKQ $ +LVWRU\ RI 5HVHDUFK LQ $XVWURQHVLDQ /DQJXDJHV 0DGDQJ 3URYLQFH,Q:XUPAustronesian Languages Harald Hammarström KKDPPDUVWURP#OHWUXQO Sebastian Nordhoff VHEDVWLDQBQRUGKRII#HYDPSJGH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY /DQJXDJH'RFXPHQWDWLRQ &RQVHUYDWLRQ6SHFLDO3XEOLFDWLRQ1R 'HFHPEHU Melanesian Languages on the Edge of Asia: Challenges for the 21st Century, HGE\1LFKRODV(YDQVDQG0DULDQ.ODPHUSS± KWWS QÀUFKDZDLLHGXOGFVS 3 KWWS KGOKDQGOHQHW Systematic typological comparison as a tool for investigating language history Ger Reesink Max Planck Institute, Nijmegen Michael Dunn Max Planck Institute, Nijmegen Similarities between languages can be due to 1) homoplasies because of a OLPLWHGGHVLJQVSDFH FRPPRQDQFHVWU\DQG FRQWDFWLQGXFHGFRQYHU- JHQFH7\SRORJLFDORUVWUXFWXUDOIHDWXUHVFDQQRWSURYHJHQHDORJ\EXWWKH\ can provide historical signals that are due to common ancestry or contact (or ERWK )ROORZLQJDEULHIVXPPDU\RIUHVXOWVREWDLQHGIURPWKHFRPSDULVRQRI VWUXFWXUDOIHDWXUHVIURPODQJXDJHV 5HHVLQN6LQJHU 'XQQ we discuss some issues related to the relative dependencies of such features: logical entailment, chance resemblance, typological dependency, phylogeny DQGFRQWDFW7KLVGLVFXVVLRQIRFXVVHVRQWKHFOXVWHULQJRIODQJXDJHVIRXQG in a small sample of 11 Austronesian and 8 Papuan languages of eastern ,QGRQHVLDDQDUHDNQRZQIRULWVKLJKGHJUHHRIDGPL[WXUH 1. INTRODUCTION. The practice of proposing families on the basis of typological FRPSDULVRQLVRQHRIWKHJXLOW\VHFUHWVRIKLVWRULFDOOLQJXLVWLFV,WLVDEDVLFSULQFLSOHRI the historical linguistic tradition that genealogical relationships between languages can only be established by the comparative method, which detects sets of cognates on the basis of regular sound changes and shared irregularities, and thus allows the positing and reconstruction of a proto-language1,QVSLWHRIWKLVVRPHHDUO\FODVVL¿FDWLRQVRIWKH more than 800 Papuan languages are based on just a handful of lexical correspondences, supplemented by observations of structural and typological similarities (Greenberg 1971; 1 The original research conducted for this study was supported by funds from NWO (Netherlands 2UJDQL]DWLRQIRU6FLHQWL¿F5HVHDUFK IRUWKH3URJUDP³%UHDNLQJWKHWLPHEDUULHU 6WUXFWXUDOWUDFHV RIWKH6DKXOSDVW´RI3URIHVVRU3LHWHU0X\VNHQ 5DGERXG8QLYHUVLW\1LMPHJHQDQG 3URIHVVRU6WHSKHQ&/HYLQVRQ0D[3ODQFN,QVWLWXWHIRU3V\FKROLQJXLVWLFV1LMPHJHQ:HWKDQN two anonymous reviewers for comments on an earlier version and Angela Terrill for editorial DVVLVWDQFH cc Creative Commons Attribution Non-Commercial No Derivatives Licence Systematic typological comparison 35 :XUP 7KHVHSURSRVDOVKDYHEHHQVHYHUHO\FULWLFL]HG VHH3DZOH\ IRUDVXPPDU\ EXWWKHLQÀXHQFHRIW\SRORJLFDOGDWDDWWKHVWDJHRIJHQHDORJLFDOK\SRWKHVLV JHQHUDWLRQUHPDLQV Typological features of languages are subject to the same evolutionary processes ZKLFK FUHDWH JHQHDORJLFDO KLVWRU\ LQ RWKHU DVSHFWV RI VDPSOHV RI UHODWHG ODQJXDJHV There is a tendency for more closely related languages to be more similar on the level of OLQJXLVWLFVWUXFWXUHMXVWDVWKH\DUHPRUHVLPLODULQWHUPVRIHJVKDUHGYRFDEXODU\7KH evolutionary and statistical properties of lexical and sound change have been extensively examined: a great deal is known about what kinds of sound changes are likely, as there is too about what kinds of words tend to be lost, replaced, semantically or phonologically PXWDWHGDQGVRIRUWK/HVVLVNQRZQDERXWWKHHYROXWLRQDU\DQGVWDWLVWLFDOSURSHUWLHVRI W\SRORJLFDOVWUXFWXUDOIHDWXUHV(YHQZKHUHOH[LFDOFRJQDWHVFDQQRWEHLGHQWL¿HGEHFDXVHRI phonological and semantic drift, there remains the possibility that other aspects of language retain traces of the historical relations between languages, whether due to genealogical GHVFHQWRUFRQWDFW$UHDVSHFLDOLVWVPD\EHDEOHWRPDNHJHQHUDOL]DWLRQVDERXWODQJXDJHVRI one or another family on the basis of typological features even where comparative method UHFRQVWUXFWLRQKDVQRWEHHQFDUULHGRXW+\SRWKHVLVJHQHUDWLRQRQWKHEDVLVRIVWUXFWXUDO IHDWXUHVRIODQJXDJHUHOLHVLQWULQVLFDOO\RQVWDWLVWLFDODUJXPHQWV As in biology, there are a number of different historical factors that lead languages to be VLPLODU FRPPRQDQFHVWU\FRQWDFW K\EULGL]DWLRQ DQGFKDQFHFRQYHUJHQFH KRPRSODV\ The smaller the design space the higher the probability that convergence is the result of FKDQFHUDWKHUWKDQJHQHDRORJLFDORUJHRJUDSKLFDOIDFWRUV,QELRORJLFDOHYROXWLRQWKHUHIRUH the more degrees of freedom in a given domain, the more powerful is the mutation and VHOHFWLRQSURFHVVUHVXOWLQJLQJUHDWHUGLVSDULW\DQGGLYHUVLW\RIVSHFLHV7KLVVXJJHVWVIRU linguistic evolution that the greater degree of freedom of lexical elements allows for a PRUHH[DFWPHDVXUHRISK\ORJHQHWLFUHODWLRQVKLSRQWKHEDVLVRIFRJQDF\VHWV6WUXFWXUDO features have a much more limited design space, thus convergent evolution will cause homoplasies that need to be distinguished from historical signals, be they phylogenetic or GXHWRK\EULGL]DWLRQ/DUJHVFDOHFKDQFHFRQYHUJHQFHLVOHVVOLNHO\KRZHYHUZKHQDJUHDW QXPEHURIIHDWXUHVDUHFRPSDUHGSURYLGHGWKHVHKDYHDPHDVXUHRILQGHSHQGHQFH6HHIRU DPRUHH[WHQVLYHDUJXPHQWDWLRQ'XQQHWDO ZKHUHZHDQVZHUWKHVNHSWLFLVP H[SUHVVHGE\+DUULVRQ :HFRPHEDFNWRWKLVSRLQWLQWKHFRQFOXVLRQ In this paper we examine the statistical properties of structural features of languages ZLWKDQH\HWRWKHLUSRWHQWLDOLQLOOXPLQDWLQJKLVWRULFDOUHODWLRQV:HXVHWKHODQJXDJHVRI HDVWHUQ,QGRQHVLDSUHYLRXVO\LGHQWL¿HGDVDQLQWHUHVWLQJDUHDLQFOXGLQJERWKGLIIXVLRQDQG LQKHULWDQFHDVDFDVHVWXG\:HLGHQWLI\YDULRXVWUDLWVRIWKHVHODQJXDJHVDVSUHVHQWHLWKHU WKURXJKGLIIXVLRQRUJHQHWLFLQKHULWDQFH :HDGRSWDV\VWHPDWLFSUREDELOLVWLFDSSURDFKXVLQJFRPSXWDWLRQDOPRGHOV7KHUHDUH D QXPEHU RI UHDVRQV IRU WKLV ERWK SUDFWLFDO DQG WKHRUHWLFDO 3UDFWLFDOO\ FRPSXWDWLRQDO models are able to process a multitude of traits for a great number of languages, while minimizing the apophenic effects of observer preconceptions , where ‘apophenic’ refers to the human tendency to see meaningful patterns or connections in random or meaningless data7KHRUHWLFDOO\FRPSXWDWLRQDOPRGHOVSURYLGHXVZLWKFRQVLVWHQWDQGWHVWDEOHUHVXOWV comparable over different hypotheses, and having useful statistical properties such DV H[SOLFLW OLNHOLKRRG VFRUHV $ IXUWKHU DGYDQWDJH RI FRPSXWDWLRQDO PHWKRGV RYHU WKH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 36 Comparative Method is that the former approach allows hypothesis generation and testing LQDZD\QRWSRVVLEOHZLWKWKH&RPSDUDWLYH0HWKRG:HVKRZWKDWZKLOHWKH&RPSDUDWLYH Method illuminates genealogy, structural features can illuminate a long-term history of FRQWDFW The use of structural data in phylogenetic inference has been applied in a few earlier VWXGLHVZKLFKDUHVXPPDUL]HGLQVHFWLRQ,QVHFWLRQZHGLVFXVVWKHQXPEHUDQGQDWXUH RIVWUXFWXUDOIHDWXUHVWKDWKDYHEHHQXVHGLQWKRVHVWXGLHV,QSDUWLFXODUZHSD\DWWHQWLRQ WRWKHLVVXHRIWUDLWLQGHSHQGHQF\6HFWLRQSUHVHQWVWKHUHVXOWVRIDVPDOOVFDOHVWXG\ illustrating how structural features provide some clusterings in a set of genealogically GLYHUVH$XVWURQHVLDQDQG3DSXDQODQJXDJHVRIHDVWHUQ,QGRQHVLD+HUHZHDWWHPSWWRGLVWLOO ZKLFKVHWRIIHDWXUHVFRQWULEXWHVPRVWVWURQJO\WRWKHFOXVWHULQJV7KHFRQFOXVLRQLQVHFWLRQ 5 summarizes discoveries and remaining issues of a standardized approach to typological FRPSDULVRQ 2. PREVIOUS STUDIES EMPLOYING STRUCTURAL FEATURES. The use of structural data in phylogenetic inference has been applied in an investigation into the relationships between WZHQW\WZR ODQJXDJHV RI WKH 2FHDQLF VXEJURXS RI WKH$XVWURQHVLDQ IDPLO\ DQG ¿IWHHQ 3DSXDQODQJXDJHVRI,VODQG0HODQHVLDUHSRUWHGLQWZRSXEOLFDWLRQV 'XQQHWDO 'XQQHWDO $OWKRXJKWKH3DSXDQODQJXDJHVRIWKLVVDPSOHKDGEHHQFODLPHGWRIRUP a genealogical group (the East-Papuan phylum, see Wurm 1975), this genealogical unity KDGEHHQFKDOOHQJHGE\5RVV DQG'XQQHWDO 'XQQ HW DO XVHG D PD[LPXP SDUVLPRQ\ DQDO\VLV RI WKH GLVWULEXWLRQ RI abstract structural features and found a reasonable congruence between the consensus WUHHDQGWKHWUDGLWLRQDOFODVVL¿FDWLRQRIWKH2FHDQLFODQJXDJHVLQWKHLUVDPSOHZKLOHWKH 3DSXDQWUHHVKRZHGVRPHJHRJUDSKLFFOXVWHULQJSRVVLEO\UHÀHFWLQJDQFLHQWUHODWLRQVKLSV GXHWRLQKHULWDQFHRUGLIIXVLRQWKURXJKFRQWDFW )RUDFULWLFDOGHEDWHRQWKHPHULWVRI WKDWVWXG\VHH'RQRKXHDQG0XVJUDYH DQG'XQQHWDO &URIW UHPDUNV³DOWKRXJKWKHUHVXOWIURP'XQQHWDO LVVXUSULVLQJWRDKLVWRULFDOOLQJXLVW LWPD\EHWKDWDFOXVWHURIW\SRORJLFDOWUDLWVZLOOSURYLGHPRUHSUHFLVLRQLQFODVVL¿FDWLRQ WKDQZLOOLQGLYLGXDOWUDLWV DOVRVRPHW\SRORJLFDOWUDLWVDUHTXLWHVWDEOHDQGWKHUHIRUHPD\ EHXVHIXOLQGLFDWRUVRISK\ORJHQ\´ 'XQQHWDO H[SODLQHGYDULRXVFRPSXWDWLRQDOPHWKRGVLQPRUHGHWDLOVKRZHG KRZ WKH\ FDQ EH H[WHQGHG DQG UH¿QHG DQG H[SORUHG KRZ D SK\ORJHQHWLF VLJQDO FDQ EH GLVWLQJXLVKHGIURPSRVVLEOHFRQWDFW7KDWVWXG\XVHGD%D\HVLDQDOJRULWKPWRFDUU\RXWD SK\ORJHQHWLFDQDO\VLVRQDVHWRIDEVWUDFWSKRQRORJLFDODQGJUDPPDWLFDOIHDWXUHV:KLOH a certain degree of possible admixture of structural features was detectable between some Oceanic and some Papuan languages, the overal clustering of the languages distinguished the Papuan languages from the Oceanic languages, and the Papuan languages could be FOXVWHUHGLQWRWKUHH JHRJUDSKLFDOO\DUFKDHRORJLFDOO\ SODXVLEOHVXEJURXSV7KHFOXVWHULQJ of the Papuan languages into three groups was shown not to be the result of degrees of contact with Oceanic languages, leaving as the most plausible hypothesis that the historical signal found on the basis of structural features is most likely due to a common ancestry, DQFLHQWFRQWDFWEHWZHHQ3DSXDQOLQHDJHVRUERWK 2QH RI WKH TXHVWLRQV UDLVHG E\ WKHVH VWXGLHV 'XQQ HW DO ZDV KRZ WKH eastern Papuan languages of Island Melanesia would cluster if a much greater sample MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 37 RI3DSXDQODQJXDJHVZHUHLQYHVWLJDWHG,QWKHLUFULWLTXHRQ'XQQHWDO 'RQRKXH DQG0XVJUDYH ³SURSRVHGWKDWFRPSDULVRQZLWK$XVWURQHVLDQODQJXDJHVVKRXOG include representative Austronesian languages from beyond Island Melanesia, in order to obtain an idea of the degree of diversity of these features that can be expected in a family RYHUD\HDU LQWKH$XVWURQHVLDQFDVH\HDU WLPHIUDPH´ For a follow-up study designed to apply the structural method to a much larger sample of languages, the set of structural features was critically reviewed, revised and H[SDQGHG6HHEHORZIRUDFRPSDULVRQRIVRPHUHYLVHGTXHVWLRQVDQGWKH$SSHQGL[IRU ERWKTXHVWLRQQDLUHV ,Q WKH VHFRQG VWXG\ 5HHVLQN HW DO ZH FRPSDUHG D ODUJH VDPSOH RI ODQJXDJHVIURPWKH6DKXOUHJLRQ LH1HZ*XLQHDDQG$XVWUDOLD PDGHXSRI3DSXDQ $XVWUDOLDQDQG$XVWURQHVLDQODQJXDJHVDQGRQH$QGDPDQHVHODQJXDJHXVLQJWKH UHYLVHGDQGH[SDQGHGVHWRIVWUXFWXUDOIHDWXUHV6LQFHWKHOLQJXLVWLFVLWXDWLRQRI6DKXO is complex, combining great time depth with long-term and intensive contact situations, we used a Bayesian algorithm originally developed to discover population structure on WKHEDVLVRIUHFRPELQLQJJHQHWLFPDUNHUVLHDPRGHORILQKHULWDQFHDQGDGPL[WXUH7KH 6WUXFWXUHDOJRULWKP 3ULWFKDUGHWDO PRGHOVHYROXWLRQDU\FKDQJHDQGDGPL[WXUHDQG simultaneously determines both the most likely number of ancestral groups and the most likely contribution of each of these ancestral populations to each of the observed individuals LQWKLVFDVHODQJXDJHV 7KHUHVXOWVRI5HHVLQNHWDO VWXG\VXJJHVWDQFHVWUDO OLQJXLVWLFSRSXODWLRQVVRPHRIZKLFKODUJHO\FRUUHVSRQGWRFOHDUO\GH¿QHGRUSURSRVHG SK\ORJHQHWLFJURXSV VHH¿JXUH ZKLOHRWKHUVH[KLELWDKLJKGHJUHHRIK\EULGL]DWLRQ Where there are very different degrees of hierarchical relatedness the inferred populations PD\EHQHVWHGZLWKLQNQRZQJHQHDORJLFDOJURXSLQJV7KHDQFHVWUDOSRSXODWLRQVLQIHUUHG by the structure algorithm can be characterized as follows: The Austronesian family is captured by three groups: dark green The Austronesian languages of Borneo and the Phillipines pale blue 2FHDQLFODQJXDJHVRIPDLQODQG1HZ*XLQHD1HZ%ULWDLQDQG9DQXDWX dark purple All other Oceanic languages of the sample 7KH7VRXODQJXDJHRI7DLZDQLVHTXDOO\UHODWHGWRWKHdark green and dark purple groups Other major families dark blue Trans-New-Guinea (note that this does not include some of the languages hypothesised to belong to the TNG periphery, such as the Alor-Pantar languages) light green Pama-Nyungan languages Areal groupings light orange Non-Pama-Nyungan languages dark orange North coast Papuan light purple South coast Papuan MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 38 pink (DVW3DSXDQ SOXV%XNL\LSDQG FIGURE 17KHJHRJUDSKLFSDWWHUQLQJRI6WUXFWXUHUHVXOWVIRUIRXQGLQJSRSXODWLRQV 5HHVLQNHWDO 7KHSLHFKDUWVLQGLFDWHWKHSURSRUWLRQDOFRQWULEXWLRQRIHDFKRIWKH IRXQGLQJSRSXODWLRQVWRHDFKODQJXDJH/DQJXDJHVDUHLGHQWL¿HGE\QXPEHU Legend ()XOOHUGHWDLOVRIWKHLQWHUSUHWDWLRQRIHDFKSRSXODWLRQDUHJLYHQLQ5HHVLQNHWDO 2QJH>RRQ@ ,PRQGD>LPQ@ 7XQJDJ>OFP@ %HODLW>EHJ@ ,VDND>NVL@ 0DQJVHQJ>PEK@ .LPDUDJDQJ>NTU@ $UDPPED>VWN@ 1DNDQDL>QDN@ 6DPD>VVE@ 1DPLD>QQP@ .LOLYLOD>NLM@ 7VRX>WVX@ 7HOHIRO>WOI@ 0HQJHQ>PHH@ ,ORFDQR>LOR@ .XXN7KD\RUUH>WKG@ 0HUDPHUD>P[P@ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 39 7DJDORJ>WJO@ .DOD/DJDZ $PRQJWKHFRQFOXVLRQVWREHGUDZQIURPWKH5HHVLQNHWDO VWXG\DUH 6WUXFWXUDOIHDWXUHVRIODQJXDJHFDQEHXVHGWRKHOSFODULI\KLVWRULFDOUHODWLRQVKLSV In the study, large known groups of languages are recapitulated: MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison – The Austronesian family with Oceanic as subgroup – 7KHSXWDWLYH7UDQV1HZ*XLQHDIDPLO\DVSURSRVHGE\5RVV DSSHDUHGDV a solid block with the exception of the Alor-Pantar languages Klon and Abui and the Marind family (Marind and Inanwatan), separated from various non-TNG clusters – $XVWUDOLDQODQJXDJHVDUHVHSDUDWHGLQ3DPD1\XQJDQYHUVXVDQRQ31FOXVWHU However, some clusters represent hybridization rather than phylogeny, especially the FOXVWHUFRQWDLQLQJERWK3DSXDQDQG$XVWURQHVLDQODQJXDJHVRIHDVWHUQ,QGRQHVLD 6RPHLPSRUWDQWTXHVWLRQVUHPDLQ ZKLFKIHDWXUHVDUHUHVSRQVLEOHIRUWKHFOXVWHULQJ"7R what extent are structural features independent? Is it possible to distinguish phylogeny from lateral transfer? The issue of relative (in)dependence of structural features will be DGGUHVVHGLQVHFWLRQDQGLQVHFWLRQZHZLOOWDNHDFORVHUORRNDWWKHK\EULGFOXVWHURI HDVWHUQ,QGRQHVLDLGHQWL¿HGDERYHDSSO\LQJWKH6WUXFWXUHDOJRULWKPWRDQHZVDPSOHRI $XVWURQHVLDQDQG3DSXDQODQJXDJHVRIWKDWDUHD 3. RELATIVELY (IN)DEPENDENT TRAITS. After chance resemblance of features (due to the limited design space of language structure at the level of granularity that we have GDWDIRU VHH'XQQHWDO'XQQHWDODQG5HHVLQNHWDO WKHPDLQIDFWRUV OHDGLQJWRUHVHPEODQFHVEHWZHHQODQJXDJHVFDQEHGLYLGHGLQWRWZRJURXSV)LUVWO\WKHUH DUHIDFWRUVLQGLFDWLYHRIKLVWRULFDOVLJQDO7KHVHLQFOXGHVKDUHGLQKHULWDQFHIURPDFRPPRQ ancestral language, and diffusion through contact between speakers of different linguistic FRPPXQLWLHV6HFRQGO\WKHUHDUHIDFWRUVZKLFKZKLOHLQVRPHFDVHVKLVWRULFDOO\GHWHUPLQHG GRQRWDOORZXVWRLQIHULQGLYLGXDOODQJXDJHKLVWRULHV7KHVHLQFOXGHORJLFDOHQWDLOPHQW typological dependency (implicational universals), and functionally motivated similarities GXHWRV\VWHPFRQVWUDLQWV &URIW )RUWKHSXUSRVHVRIPDNLQJKLVWRULFDOLQIHUHQFHV about languages, this second set of factors acts as noise at best (obscuring a signal where present), and is misleading at worst (creating the appearance of a signal where one is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¿QHGDQXPEHURITXHVWLRQVZHUHUHPRYHGDQGRWKHUVZHUHDGGHG,QWDEOHDQGWDEOH ZHJLYHVRPHH[DPSOHVRIRULJLQDOTXHVWLRQVZKLFKFRXOGQRWHDVLO\EHDQVZHUHGIRU PDQ\ODQJXDJHVDQGZKLFKZHUHUHSODFHGE\TXHVWLRQVZKRVHWHUPVZHUHEHWWHUGH¿QHG MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison DQGPRUHHDVLO\LGHQWL¿HGLQDJLYHQGHVFULSWLRQ 7KH TXHVWLRQV ZKHWKHU WKHUH DUH DGMHFWLYHV DQG KRZ WKH\ IXQFWLRQ DWWULEXWLYHO\ DQG SUHGLFDWLYHO\ FDXVHG VRPH GLI¿FXOWLHV LQ WKH ¿UVW YHUVLRQ7KLV ZDV VROYHG E\ WKH QHZ IRUPXODWLRQVZKLFKVSHFL¿FDOO\DUHPHDQWWRFDSWXUHZKHWKHUDGMHFWLYDOQRWLRQVDUHQRXQ\ RUYHUE\LQDSDUWLFXODUODQJXDJH $'-(&7,9(6 $'-(&7,9(6 >/$1*8$*(@ >3/26%,2/2*<@ Is there lexical overlap between a 69 'RFRUHDGMHFWLYHV GH¿QHGVHPDQWLFDOO\ VLJQL¿FDQWSURSRUWLRQRIDGMHFWLYHV as property concepts; value, shape, age, and verbs (including zero- dimension) act like verbs in predicative derivation)? position? Does the same lexical set 70 'RFRUHDGMHFWLYHV GH¿QHGVHPDQWLFDOO\ of adjectives function both as property concepts; value, shape, age, attributively and predicatively? GLPHQVLRQ XVHG DWWULEXWLYHO\ UHTXLUH the same morphological treatment as verbs? TABLE 14XHVWLRQVUHODWLQJWR$GMHFWLYHVLQWZRYHUVLRQV 7KHRULJLQDOTXHVWLRQQDLUHXVHGIRU'XQQHWDODQGFRQWDLQHGDQXPEHURI TXHVWLRQVDWWHPSWLQJWRFROOHFWGDWDRQ7HQVH$VSHFW0RRGFDWHJRULHV7KRVHTXHVWLRQV were phrased in terms of “how many pure tenses are distinguished?” and “how many fused WHQVHPRRGFDWHJRULHVDUHGLVWLQJXLVKHG"´,WZDVVWLSXODWHGWR³LQFOXGHDI¿[HVFOLWLFV and satellite particles associated with verbs forming a constituent with the verb on some OHYHOEXWH[FOXGHRSWLRQDODGYHUELDOV´6LQFHWKHWHUPVµSXUH¶YHUVXVµIXVHG¶DUHQRWHDVLO\ interpreted and because the answers were not binary as they are for all other traits, these TXHVWLRQVZHUHQ¶WHYHQXVHGIRUWKRVHVWXGLHV 7KXVRQO\WKHIHZTXHVWLRQVLQFROXPQLQWDEOHZHUHSDUWRIWKHDQDO\VHVLQWKHWZR studies, which meant that potentially important information regarding Tense marking could QRWEHXVHG7KHUHYLVHGTXHVWLRQVLQWDEOH\LHOGPRUHFOHDUO\LQWHUSUHWDEOHFRGHVDQG WKH\UHVWULFWWKHWUDLWVWRFOHDUO\PRUSKRORJLFDOFDWHJRULHVPDUNHGRQWKHYHUE $VDFNQRZOHGJHGLQ5HHVLQNHWDO IRUFRPPHQWVDQGDGGLWLRQVUHVXOWLQJLQWKHODWHVW version we thank Sjef Barbiers, Milly Crevels, Nick Evans, Rob Goedemans, Eva Lindström, 3LHWHU0X\VNHQ*XQWHU6HQIW/HRQ6WDVVHQDQG+HLQYDQGHU9RRUW :RUNVKRS0D\ 5DGERXG8QLYHUVLW\DQG0D[3ODQFN,QVWLWXWHIRU3V\FKROLQJXLVWLFV1LMPHJHQ1HWKHUODQGV ,Q SDUWLFXODUWKHUHIRUPXODWLRQRIWKHTXHVWLRQVUHJDUGLQJDGMHFWLYHVLVGXHWR/HRQ6WDVVHQ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 7$0>/$1*8$*(@ 7$0 >3/26%,2/2*<@ Do the same morphemes 79 'RYHUEVKDYHSUH¿[HVSURFOLWLFVRWKHU systematically encode both TAM WKDQ WKRVH WKDW 21/< PDUN $ 6 RU and person? O (do include portmanteau: A & S + TAM)? 'RYHUEVKDYHSUH¿[HVSURFOLWLFV" 80 'R YHUEV KDYH VXI¿[HVHQFOLWLFV RWKHU WKDQ WKRVH WKDW 21/< PDUN $ 6 RU O (do include portmanteau: A & S + TAM)? 'RYHUEVKDYHVXI¿[HVHQFOLWLFV" 81 &DQ LQ¿[DWLRQ EH XVHG RQ YHUEV IRU derivational, aspectual, or voice- changing purposes? is there present tense regularly morphologically marked on the verb? 83 is there past tense regularly morphologically marked on the verb? is there future tense regularly morphologically marked on the verb? 85 are there multiple past or future tenses, distinguishing distance from Time of Reference, marked on the verb? is a distinction between punctual/ 86 is a distinction between punctual/ continuous aspect available as a continuous aspect available as a morphological choice? morphological choice? 50 is a distinction between realis/ 87 is a distinction between realis/irrealis irrealis mood available as a mood available as a morphological morphological choice? choice? TABLE 2.)HDWXUHVUHODWLQJWR7HQVH$VSHFW0RRGDI¿[DWLRQLQWZRYHUVLRQV For a full comparison of the differences between the two versions we refer to the $SSHQGL[:HFRQWLQXHZLWKDGLVFXVVLRQRIWKHUHODWLYH LQ GHSHQGHQFHRIWUDLWVLQWKH PRVWUHFHQWTXHVWLRQQDLUH 3.2. LOGICAL ENTAILMENT. In spite of our attempt to minimize logical entailment between features in our database, there are some cases where we judge it innocuous to DOORZ IHDWXUHV ZLWK VRPH GHJUHH RI ORJLFDO GHSHQGHQF\ EHWZHHQ WKHP WR UHPDLQ )RU H[DPSOHFRQVLGHUWKHSRVVLEOHYDOXHVIRUWZRTXHVWLRQVUHODWLQJWRWKHSKRQRWDFWLFVRID language in (1): D $UHWKHUHZRUG¿QDOFRQVRQDQWV" MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison (b) Are there consonant clusters (not counting prenasalized consonants) in syllable coda? 7KHWZRTXHVWLRQVDUHFOHDUO\QRWWRWDOO\LQGHSHQGHQWIURPHDFKRWKHUDVSDUWLFXODUYDOXHV of certain features logically entail particular values of others: LI D WKHQ E RU LI D WKHQ E LI E WKHQ D LI E LVWKHQ D RU However, the bias added by this dependency is small as this entailment is only partial, outweighed by the added statistical power we get from including data with the logically LQGHSHQGHQWYDOXHV*LYHQWKHODUJHQXPEHURIIHDWXUHVLQRXUDQDO\VLVLWLVQRWOLNHO\WKDW WKLVRQHFDVHRISDUWLDOGHSHQGHQF\KDVVHULRXVO\DIIHFWHGWKHUHVXOWVLQRXUHDUOLHUDQDO\VHV 3.3. CHANCE RESEMBLANCE DUE TO LIMITED DESIGN SPACE. Most or all of the structural IHDWXUHV RI ODQJXDJH KDYH D IDU PRUH UHVWULFWHG GHJUHH RI IUHHGRP WKDQ OH[LFDO LWHPV 7KH\DUHDIXQGDPHQWDOO\GLIIHUHQWNLQGRIGDWDZLWKGLIIHUHQWVWDWLVWLFDOSURSHUWLHV)RU H[DPSOHWKHWZRTXHVWLRQVDERXWWKHEHKDYLRURIDGMHFWLYDOHOHPHQWVLQSUHGLFDWLYHDQG attributive position (see table 1) were formulated to capture whether a language has verby <WRERWKTXHVWLRQV RUQRXQ\ 1WRERWK DGMHFWLYHVRULQEHWZHHQ <WRSUHGLFDWLYH 1 WRDWWULEXWLYHYHUEOLNHEHKDYLRU $ODQJXDJHZKLFKZRXOGKDYH1WRYHUEOLNHEHKDYLRU LQSUHGLFDWLYHSRVLWLRQEXW<WRYHUEOLNHEHKDYLRULQDWWULEXWLYHSRVLWLRQZDVFRQVLGHUHG DVXQOLNHO\+RZHYHULQRXUVDPSOHZHGR¿QGWKLVDQRPDORXVVLWXDWLRQLQWKHQRQ71* ODQJXDJH,PRQGD7KXVWKHPD[LPXPQXPEHURIIRXUSRVVLELOLWLHVLVDYDLODEOH7KLVKROGV DOVRIRUWKHWZRTXHVWLRQVZKHWKHUDODQJXDJHKDVSUHSRVLWLRQVRUSRVWSRVLWLRQV7KHUHDUH ODQJXDJHVZLWK<RU1WRERWKTXHVWLRQVLQDGGLWLRQWRWKRVHWKDWKDYHRQO\RQHRUWKHRWKHU With regard to the order of Possessor and Possessum, the design space allows for three SRVVLELOLWLHV WKH3RVVHVVRUPD\ SUHFHGHRU IROORZRU PD\GRERWK$QHJDWLYH YDOXHRIERWKTXHVWLRQVLVRIFRXUVHQRWSRVVLEOH :KLOH VXFK OLPLWHG GHJUHHV RI IUHHGRP PD\ FUHDWH KRPRSODVLHV WKDW GR QRW UHÀHFW shared history, large-scale chance convergence is rendered unlikely through the use of a ODUJHQXPEHURIIHDWXUHV 3.4. TYPOLOGICAL DEPENDENCY – IMPLICATIONAL UNIVERSALS. Typological dependencies have been widely discussed since the sixties when Joseph Greenberg launched KLVODQJXDJHXQLYHUVDOVSURMHFW0RVWJHQHUDOL]DWLRQVGHDOZLWKZRUGRUGHUSURSHUWLHVLQ WKHFODXVHDQGWKHQRPLQDOFRQVWLWXHQW)RUH[DPSOHLWLVZHOONQRZQWKDW29RUGHUDQG SRVWSRVLWLRQVDUHFRPPRQO\IRXQGWRJHWKHUDVDUH92RUGHUDQGSUHSRVLWLRQV'XQQHWDO KDVDUJXHGWKDWWKHUHLVDVWURQJOLQHDJHVSHFL¿FHOHPHQWWRWKHVHDSSDUHQWXQLYHUVDOV 'U\HU SUHVHQWVGDWDVKRZLQJWKDWWKHFRUUHODWLRQLVQRWSHUIHFW2IDWRWDORI ODQJXDJHVKDYH29DQGSRVWSRVLWLRQVDQGKDYH92DQGSUHSRVLWLRQVZKLOH ODQJXDJHVFRPELQH29ZLWKSUHSRVLWLRQVDQGKDYH92WRJHWKHUZLWKSRVWSRVLWLRQV,Q DGGLWLRQODQJXDJHVGRQRWIDOOLQWRRQHRIWKHVHIRXUFDWHJRULHV)RUH[DPSOH'XWFK KDVSUHSRVLWLRQVEXWKDVERWK29DQG92RUGHU2QWKHRWKHUKDQG-DErPKDV692RUGHU ZLWKERWKSUHSRVLWLRQVDQGSRVWSRVLWLRQV7KXVZKLOHWKHUHLVDVWURQJW\SRORJLFDOWHQGHQF\ IRU WKH YDOXHV RI WKHVH IHDWXUHV WR EH FRUUHODWHG E\ UHPRYLQJ VRPH RI WKHVH TXHVWLRQV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison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– SYSTEM CONSTRAINTS. Somewhat related to W\SRORJLFDOGHSHQGHQF\LVFRQYHUJHQFHGXHWRV\VWHPFRQVWUDLQWV6RPHRIRXUIHDWXUHV PD\DW¿UVWEOXVKEHPXWXDOO\H[FOXVLYHRULQFOXVLYH)RULQVWDQFHODQJXDJHVWHQGWRKDYH SUHSRVLWLRQV RU SRVWSRVLWLRQV EXW UHODWLYHO\ LQIUHTXHQWO\ KDYH ERWK RU QHLWKHU7KH UDZ counts for these features in our complete database (ignoring for the moment that these REVHUYDWLRQV DUH SK\ORJHQHWLFDOO\ GHSHQGHQW DUH VKRZQ LQ WDEOH 7KH FRQGLWLRQDO SUREDELOLW\ RI KDYLQJ SUHSRVLWLRQV JLYHQ SRVWSRVLWLRQV LV DQG WKH FRQGLWLRQDO SUREDELOLW\RIKDYLQJSRVWSRVLWLRQVJLYHQSUHSRVLWLRQVLVRQO\$GLDFKURQLFDFFRXQW for the development of adpositions predicts that the order of adposition and noun phrase ZLOOW\SLFDOO\EH¿[HG Postpositions Present Absent Prepositions Present 11 87 Absent 61 13 TABLE 3.3RVWSRVLWLRQVDQGSUHSRVLWLRQV +HLQHDQG.XWHYD GHVFULEHW\SLFDOJUDPPDWLFDOL]DWLRQSDWKZD\VVXFKDVrelational noun>adposition, adverb>adposition, or verb+complement>adposition+noun phrase, which each have as their starting point a construction which most commonly already KDV¿[HGRUGHULQJ(YHQLIWZRRUGHUVRIDGSRVLWLRQVDQGQRXQSKUDVHVDUHSRVVLEOHWKH RUGHUZLOOPRVWOLNHO\EH¿[HGZLWKUHVSHFWWRWKHSDUWLFXODUDGSRVLWLRQVHOHFWHG*LYHQWKH constraint on adposition systems that there will usually be only one kind, there is a negative FRUUHODWLRQEHWZHHQKDYLQJSUHSRVLWLRQVDQGKDYLQJSRVWSRVLWLRQV 6LPLODUNLQGVRIV\VWHPFRQVWUDLQWVH[LVWLQRWKHUSDUWVRIWKHJUDPPDU)RUH[DPSOH WKHUH LV D WHQGHQF\ IRU DJUHHPHQW DI¿[HV IRU WUDQVLWLYH DQG LQWUDQVLWLYH VXEMHFWV WR EH PDUNHGWKHVDPHZD\7KXVWKHUHLVDSRVLWLYHFRUUHODWLRQEHWZHHQKDYLQJSUH¿[HVIRU marking transitive subjects (A) and intransitive subjects (S), and likewise there is a positive FRUUHODWLRQEHWZHHQKDYLQJVXI¿[HVIRUVXEMHFWVRIWUDQVLWLYH $ DQGVXEMHFWVRILQWUDQVLWLYH FODXVHV 6 DVVKRZQLQWDEOH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 6VXI¿[ Present Absent $VXI¿[ Present 11 Absent 7 66 6SUH¿[ Present Absent $SUH¿[ Present 35 3 Absent 3 2VXI¿[ Present Absent 2SUH¿[ Present 1 0 Absent 51 TABLE 4:$6DQG2DVSUH¿[HVDQGVXI¿[HV 6RPH RI WKHVH WHQGHQFLHV DUH QHYHUWKHOHVV QRW VWURQJ :KLOH WKHUH LV D QHJDWLYH FRUUHODWLRQEHWZHHQKDYLQJREMHFWVXI¿[HV 2 DQGKDYLQJREMHFWSUH¿[HV VHHWDEOH the amount of the variance this correlation explains of the (phylogenetically uncorrected) GDWDLVEDUHO\VLJQL¿FDQW7KLVLVGHVSLWHDVWURQJSK\ORJHQHWLFELDVLQWKDWQR$XVWURQHVLDQ ODQJXDJHVKDYHDQREMHFWSUH¿[DQGPRVW7UDQV1HZ*XLQHDODQJXDJHVGR7KLVZRXOG be expected to have the effect of exaggerating the apparent negative correlation between REMHFWSUH¿[HVDQGVXI¿[HV 3.6 SHARED INHERITANCE. Correlations between features in a linguistic data set cannot be interpreted as causal with any validity without taking into account the confound introduced E\ SRVVLEOH JHQHDORJLFDO UHODWLRQVKLSV EHWZHHQ WKH ODQJXDJHV 7KLV LVVXH LV NQRZQ DV Galton’s problem: variables in languages related by common descent or diffusion are not VWDWLVWLFDOO\LQGHSHQGHQW$Q\DSSDUHQWFDXVDOFRUUHODWLRQVEHWZHHQIHDWXUHVRIODQJXDJHV linked by shared history might be no more than ‘duplicate copies of the same original’ *DOWRQLQ7\ORU 7KLVZDVDOOXGHGWRDERYH VHFWLRQ ZLWKWKHH[DPSOHRI REMHFWSUH¿[HV2EMHFWSUH¿[HVDUHDEVHQWLQ$XVWURQHVLDQODQJXDJHVDQGKLJKO\IUHTXHQW LQ7UDQV1HZ*XLQHDODQJXDJHV6RDQ\RWKHUIHDWXUHZKLFKLVUDUHLQ$XVWURQHVLDQDQG FRPPRQLQ71*ZLOOFRUUHODWHZLWKWKHSUHVHQFHRUDEVHQFHRIREMHFWSUH¿[HV,QDGDWD set limited to Austronesian and TNG languages we could expect absurd correlations, such DVSRVLWLYHFRUUHODWLRQVEHWZHHQREMHFWSUH¿[HVDQGDOWLWXGHQHJDWLYHFRUUHODWLRQVEHWZHHQ REMHFWSUH¿[HVDQGQDYLJDWLRQDOWHFKQRORJ\7KHVHFRUUHODWLRQVDUHGULYHQE\DFFLGHQWVRI KLVWRU\UDWKHUWKDQDQ\FDXVDOOLQN ,QWKHIXOOVHWRIIHDWXUHVZH¿QGVXEVWDQWLDOFRUUHODWLRQVHLWKHUSRVLWLYHRUQHJDWLYH ZKLFKDUHFOHDUO\WKHSURGXFWRIVKDUHGKLVWRU\)RULQVWDQFHWKHUHDUHSRVLWLYHFRUUHODWLRQV between Decimal counting systems, Prepositions, and the Inclusive/Exclusive distinction IRUQRQVLQJXODU¿UVWSHUVRQ/LNHZLVHWKHUHLVDQHJDWLYHFRUUHODWLRQEHWZHHQWKHVHIHDWXUHV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison DQGYHUEDOSDVWWHQVH )LJXUH LOOXVWUDWHV WKH DFFLGHQWDO QDWXUH RI WKH QHJDWLYH FRUUHODWLRQ EHWZHHQ GHFLPDOFRXQWLQJV\VWHPVDQGYHUEDOSDVWWHQVHPDUNLQJ$GHFLPDOV\VWHPRIFRXQWLQJ SUHGRPLQDWHV LQ $XVWURQHVLDQ DOWKRXJK WKHUH DUH TXLWH D QXPEHU RI $1 ODQJXDJHV ZKLFKH[KLELWD TXLQDU\V\VWHP$QG WKHUHLV LQWHUQDOHYLGHQFHLQPDQ\RI WKH3DSXDQ languages with decimal systems that this occurred through contact with Austronesian VSHDNLQJFRPPXQLWLHV7KHOHIWSDQHORI¿JXUHVKRZVDSRVVLEOHUHFRQVWUXFWLRQRIWKH KLVWRU\RIGHFLPDOFRXQWLQJV\VWHPVLQDVDPSOHRI$XVWURQHVLDQODQJXDJHV7KHFDVHIRU UHFRQVWUXFWLQJGHFLPDOV\VWHPVIRUSURWR$XVWURQHVLDQVHHPVVWURQJ7KHUHDUHWZRVXE branches of languages lacking decimal counting systems and decimal counting systems RFFXURQHYHU\OHYHORIWKHSK\ORJHQ\7KHULJKWSDQHOVKRZVDVRPHZKDWGLIIHUHQWVWRU\ 9HUEDOSDVWWHQVHPDUNLQJRFFXUVVSRUDGLFDOO\WKURXJKRXWWKHWUHHEXWWKHUHDUHQRFDVHV ZKHUHLWPDNHVVHQVHWRUHFRQVWUXFWSDVWWHQVHPDUNLQJWRDQHDUOLHUQRGHRIWKHWUHH7KH negative correlation between these features is apparently because of the relative stability of WKHWZRIHDWXUHVDQGWKHLUVWDWHVLQWKHDQFHVWUDOODQJXDJHV,QRWKHUZRUGVWKHFOXVWHULQJ RIVXFKVWDWLVWLFDOO\GHSHQGHQWIHDWXUHVLVGXHWRDJHQHDORJLFDOVLJQDO 7KH¿QDOSRVVLEOHFDXVHRIW\SRORJLFDOVLPLODULW\EHWZHHQODQJXDJHVGLIIXVLRQWKURXJK FRQWDFWZLOOEHGLVFXVVHGLQVHFWLRQ FIGURE 2'HFLPDOFRXQWLQJV\VWHPVDQGSDVWWHQVHPDUNLQJRQWKHYHUELQ Austronesian tree MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 4. LINGUISTIC POPULATIONS IN EAST INDONESIA. The study examining linguistic traces RIWKH6DKXO3DVW 5HHVLQNHWDO HPSOR\HGWKHSTRUCTUREDOJRULWKP 3ULWFKDUGHWDO DVVKRZQLQVHFWLRQ7KHPHWKRGDVVXPHVDPRGHOLQZKLFKWKHUHDUHDQXPEHU . RIXQVSHFL¿HGRUXQNQRZQSRSXODWLRQVHDFKRIZKLFKLVFKDUDFWHUL]HGE\DVHWRI DOOHOHIUHTXHQFLHVDWHDFKORFXV,QGLYLGXDOVLQDQ\VDPSOHDUHDVVLJQHG SUREDELOLVWLFDOO\ to populations, or jointly to two or more populations if their genotypes indicate that they DUHDGPL[HG7KHGLIIHUHQWYDOXHVRIWKHOLQJXLVWLFFKDUDFWHUVDUHWKHDQDORJLFDOHTXLYDOHQW RIWKHJHQHWLFDOOHOHVZKLOHDODQJXDJHLVWKHHTXLYDOHQWRIDQLQGLYLGXDOLQWKHELRORJLFDO VWXGLHV,QRWKHUZRUGVMXVWDVDQLQGLYLGXDO¶VDXWRVRPDO'1$LVLQKHULWHGIURPDQXPEHU of different ancestors belonging to one or more biological populations, so a language may KDYHLQKHULWHGVWUXFWXUDOIHDWXUHVIURPRQHRUPRUHGLIIHUHQWSRSXODWLRQV7KHVWUXFWXUH algorithm computes the most likely contribution of a given number (K) of ancestral SRSXODWLRQVWRHDFKRIWKHLQGLYLGXDOV $V VWDWHG DW WKH HQG RI VHFWLRQ 5HHVLQN HW DO GLG ¿QG VRPH VWULNLQJ FRUUHVSRQGHQFHEHWZHHQHDUOLHUGH¿QHGOLQJXLVWLFIDPLOLHV+RZHYHUDVDOUHDG\PHQWLRQHG structural features cannot be used to claim or refute genealogical relationships between ODQJXDJHVVHHDOVR&URIW 7KLVLVLOOXVWUDWHGLQWKHIDFWWKDWWKHVWULNLQJFRUUHVSRQGHQFH GRHVQRWDPRXQWWRIXOODJUHHPHQWDPRQJWKHJURXSLQJVIRXQGE\WKHGLIIHUHQWPHWKRGV $UDWKHUUREXVWOLQJXLVWLFSRSXODWLRQLGHQWL¿HGE\WKH6WUXFWXUHDOJRULWKP 5HHVLQNHWDO DVWKHµUHG¶RUµ:HVW3DSXDQ¶FOXVWHU VHH¿JXUH FRQWDLQVDOOWKH3DSXDQODQJXDJHV of eastern Indonesia and the Bird’s Head in the sample: Klon and Abui from the Alor- Pantar family, Tobelo and Tidore from North Halmahera, and Meyah and Hatam from WKH%LUG¶V+HDGDVZHOODVWKHWZR$1ODQJXDJHV7DEDDQG%LDN7KLVFOXVWHUKDVDOVR contributions to Papuan and Austronesian languages along the north coast of New Guinea DQGLQWKH%LVPDUFNDUFKLSHODJR:HFRQFOXGHGLQWKDWVWXG\ ³7KLV¿QGLQJVXJJHVWVDQ area of millennia of contact between AN and Papuan non-TNG speaking groups” (Reesink HWDO Given that earlier studies had shown a great degree of heterogeneity among the Papuan JURXSVLQHDVW,QGRQHVLD VHHIRUH[DPSOH5HHVLQN LWZDVUDWKHUVXUSULVLQJWRVHH WKHPFOXVWHUHGWRJHWKHUZLWKDIHZ$1ODQJXDJHVWKURZQLQ7KXVQHZUHVHDUFKTXHVWLRQV DUHUDLVHG ZKLFKIHDWXUHVDUHUHVSRQVLEOHIRUDFHUWDLQFOXVWHULQJ DQG LVLWSRVVLEOHWR differentiate phylogeny and diffusion? ,Q RUGHU WR DQVZHU WKHVH TXHVWLRQV D QHZ VWXG\ ZDV FRQGXFWHG ZLWK WZR PRUH$1 languages from the same region, Tetun spoken in East Timor and Buru of the Moluccas, ERWKFODVVL¿HGDVPHPEHUVRIWKH&HQWUDO0DOD\R3RO\QHVLDQVXEJURXS7KHYDOLGLW\RI this subgroup, proposed by Blust as a linkage (1993), has been challenged by Donohue DQG*ULPHV DQGUHDI¿UPHGDVPRVWOLNHO\GHVFHQGLQJIURPDGLDOHFWFKDLQE\%OXVW :HQRZUHSRUWWKHUHVXOWVRIWKLVQHZVWXG\ The Structure algorithm was applied this time to just a small sample of Austronesian ODQJXDJHV RI HDVWHUQ ,QGRQHVLD DQG 3DSXDQ ODQJXDJHV RI WKH VDPH DUHD 6LQFH WKH algorithm simultaneously determines both the most likely number of ancestral groups and the most likely contribution of each of these populations to each of the observed individuals, we wanted to focus on the similarities and differences between just these languages, avoiding clustering that might ignore intragroup differences when compared to 3DSXDQDQG$XVWUDOLDQODQJXDJHVZLWKGLIIHUHQWSUR¿OHVDVZDVGRQHLQWKHPDMRUVWXGLHV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison ,Q¿JXUHWKHFOXVWHULQJRIWKHVHODQJXDJHVLVVKRZQIRUWZRWR¿YHDQFHVWUDOSRSXODWLRQV . )RUHDFKVSHFL¿HGQXPEHURIFOXVWHUV . WKHDOJRULWKPDVVLJQVDFHUWDLQZHLJKWWR HDFKDOOHOH LQRXUFDVHWKHVWDWHRIDSDUWLFXODUIHDWXUH 7KLVFOXVWHULQJLVLQGHSHQGHQWIRU each K value, so that individuals may be assigned to different clusters (arbitrarily given a particular colour) on the basis of the amalgamated weights of the feature-states within each . FIGURE 3. Clustering of AN and Papuan languages of eastern Indonesia 7KH.YDOXHVDQGKDUGO\GLIIHULQWKHLUOLNHOLKRRGVFRUH$W.DQG.WKHOLJKW EOXHFOXVWHUFRQWDLQV$1DQG3DSXDQODQJXDJHVZKLOHDW.ZH¿QGVRPHGLIIHUHQWLDWLRQ A new cluster (pink) is detected contributing mainly to Klon and Abui of the Alor-Pantar group, Tobelo of North Halmahera and Mairasi, spoken in the ‘neck’ connecting the Bird’s +HDGWRWKHUHVWRI1HZ*XLQHD7KXVDW.ZH¿QGDVHSDUDWLRQRIDQXPEHURIWKH3DSXDQ ODQJXDJHVEXWVWLOOQRWDOO Clustering by the Structure algorithm is based on differential weighting to each of WKHIHDWXUHVSHUFOXVWHU7KHVDPHIHDWXUHPD\KDYHDKLJKHURUORZHUZHLJKWLQJIRU GLIIHUHQW.YDOXHV6SDFHGRHVQRWDOORZXVWRJLYHDIXOOOLVWRIGLIIHUHQWZHLJKWVRIHDFK IHDWXUHIRUHDFKYDOXHRI.EXWLQWDEOHDVDPSOHSHUWDLQLQJWRZRUGRUGHULVJLYHQ7KHVH YDOXHVVKRZWKDWSUHVHQFHRI9¿QDO3RVWSRVLWLRQVDQG2EMHFW3UH¿[KDYHDORZHUZHLJKW IRUWKHOLJKWEOXHFOXVWHUDW.DQGWKXVWRJHWKHUZLWKWKHYDOXHVRIRWKHUIHDWXUHVFDQQRW GLIIHUHQWLDWH.ORQ$EXL7REHORDQG0DLUDVLIURPWKHRWKHU3DSXDQDQG$1ODQJXDJHV$W .WKHVHIHDWXUHVKDYHDVWURQJHUZHLJKWLQJDQGWKXVDQHZFOXVWHULVLGHQWL¿HG MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison Light blue K4 K5 Pink K5 9HUE¿QDO Postposition 2EMHFW3UH¿[ 9HUEPHGLDO Preposition TABLE 5$OOHOHZHLJKWVIRUIHDWXUHVLQFRQWULEXWLQJSRSXODWLRQV Some values of the features in table 5 may look like a system constraint, or a typological FRUUHODWLRQEXWWKHRYHUDOOFRUUHODWLRQEHWZHHQ2EMHFWSUH¿[DQG9HUE¿QDOZRUGRUGHULQ WKHVDPSOHRIODQJXDJHVLVUDWKHUZHDN U 7KHFRQWULEXWLRQRIWKHVHIHDWXUHVLV WKHUHIRUHUHODWLYHO\LQGHSHQGHQW ,Q¿JXUHLWLVFOHDUWKDWLQDOOLQGHSHQGHQWUXQVDWDOO.YDOXHVWKHWZRXQUHODWHG Papuan languages of the Bird’s Head, Hatam and Meyah, consistently cluster with the AN ODQJXDJH%LDN,WWKXVDSSHDUVWKDWLQWKLVFDVHGLIIXVLRQRYHUULGHVSK\ORJHQ\ Is it possible to differentiate the two historical processes by extant structural features? We know from the comparison of their lexicons that Biak belongs to the South Halmahera- West New Guinea subgroup of the AN family and that Hatam and Meyah belong to two different Papuan families, albeit with perhaps a very remote common ancestor (Reesink $UHWKHUHDQ\WUDFHVLQWKHLUVWUXFWXUDOIHDWXUHVWKDWVWLOOEHWUD\WKHLUJHQHDORJLFDO DI¿OLDWLRQ" ,Q RWKHU ZRUGV WR ZKDW H[WHQW DUH WKHVH ODQJXDJHV GLIIHUHQW LQ WKH VHW RI structural features employed? ,QRUGHUWR¿QGVXFKWUDFHVZHKDYHWRJRLQWRWKHQLWW\JULWW\RIWKHGDWD7DEOHOLVWV DOO¿IWHHQIHDWXUHV RXWRI RQZKLFKWKHWZRXQUHODWHG3DSXDQODQJXDJHV+DWDPDQG Meyah both agree with each other, presumably due to shared diffusion of Papuan traits, and DUHGLIIHUHQWLQYDOXHIURP$XVWURQHVLDQ%LDN Hatam Meyah Biak Weight sensitive stress - - + Syllable position stress - - + 'H¿QLWHVSHFL¿FDUWLFOHV - - + ,QGH¿QLWHDUWLFOHUHTXLUHG - - + Difference comitative vs coordination - - + Gender in third person - - SODQLPDWH 1XPHUDOFODVVL¿HUV + + - 3RVVHVVLRQE\VXI¿[ - - + Quinary counting system + + - MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 50 Hatam Meyah Biak $WWULEXWLYHDGMHFWLYHVUHTXLUHVDPH - - + morphology as verbs Copula for predicative N(P) - - + Aspectual auxiliaries - - + &DXVDWLYHE\6HULDO9HUE&RQVWUXFWLRQ + + - Nouns can be reduplicated - - + 2WKHUHOHPHQWVWKDQ1RU9FDQEH + + - reduplicated TABLE 6+DWDPDQG0H\DKYDOXHVDJUHHDQGGLIIHUIURP%LDN These facts show very faint traces of structural features that may betray phylogenetic DI¿OLDWLRQ)RUH[DPSOHSRVVHVVLRQE\VXI¿[VHHPVWLJKWO\OLQNHGWRWKH$1IDPLO\,Q many AN languages to the north-west of this geographic region the Possessor normally follows the Possessum, and when that is expressed by a pronoun it can easily become HQFOLWLFL]HGRUVXI¿[HG7KLVRUGHULVVWLOOSUHVHQWLQ%LDN,WVKRXOGEHQRWHGWKDWWKHIHDWXUH 3RVVHVVLRQE\SUH¿[ DVHSDUDWHTXHVWLRQLQRXUGDWDEDVH LVQRWSDUWRIWKHOLVWVHSDUDWLQJ Hatam and Meyah from Biak, because for this trait all three languages have a positive YDOXH7KLVRUGHULVW\SLFDORIWKH3DSXDQODQJXDJHVRIWKH%LUG¶V+HDG DQGRWKHUUHJLRQV of east Indonesia), and has diffused to a few AN languages in the Cenderawasih Bay area, in Biak and Ambai for plural possessors, in Waropen for both singular and plural (see .ODPHUHWDO :KLOHDOORWKHU3DSXDQODQJXDJHVRI1RUWK+DOPDKHUDDQGWKH Bird’s Head have a gender distinction for third person singular, the two east BH families WKDW+DWDPDQG0H\DKEHORQJWRGRQRW 5. CONCLUSION. The results of large-scale comparison of structural features in a great QXPEHURIODQJXDJHVIURPGLIIHUHQWOLQHDJHVFDQEHVXPPDUL]HGDVIROORZV ,Q SRSXODWLRQ JHQHWLFV WKH GLVWULEXWLRQ DQG IUHTXHQF\ RI PXWDWLRQV LQ XQUHODWHG LQGLYLGXDOVDUHXVHGWRWUDFHDQFHVWUDOSRSXODWLRQV,QWKHVWXGLHVUHYLHZHGLQVHFWLRQ we practice population linguisticsWKDWLVZHDWWHPSWWR¿QGFOXVWHUVEHWZHHQLQGLYLGXDO ODQJXDJHV WKDW DUH 127 LPPHGLDWH IDPLO\ :KHUH FRJQDWHEDVHG PHWKRGV FDQQRW EH DSSOLHGSUR¿OHVRIDEVWUDFWVWUXFWXUDOIHDWXUHVFDQGLVFRYHUSODXVLEOHJURXSLQJVLQKLWKHUWR XQUHODWHGFOXVWHUV RI ODQJXDJHV7KHVHJURXSLQJV PD\EHWKHUHVXOWRI UHPRWHFRPPRQ DQFHVWU\GLIIXVLRQRUERWK,QWKHFDVHRIDSXWDWLYHIDPLO\OLNHWKH3DSXDQ71*IDPLO\ the result obtained by structural features may strengthen the tentative conclusions based on MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 51 SURQRPLQDOIRUPV:HGRQRWFODLPWKDWZHQRZKDYHFRQFOXVLYHHYLGHQFHIRU71*DVD ERQD¿GHIDPLO\EXWVLPSO\WKDWWKHSURSRVHGXQLW\KDVVRPH¿UPHUIRRWLQJ Chance resemblances due to the limited degrees of freedom structural features have +DUULVRQ VHFWLRQ DERYH FDQ WR VRPH H[WHQW EH RYHUFRPH E\ FRQVLGHULQJ D ODUJHQXPEHURIIHDWXUHV7\SRORJLFDOGHSHQGHQFLHVVXFKDVLPSOLFDWLRQDOXQLYHUVDOVDQG functionally motivated convergences are an empirical matter: how strong are they? They DSSDUHQWO\GLIIHULQGLIIHUHQWOLQHDJHV 'XQQHWDO 7KHUHVXOWVUHSRUWHGLQVHFWLRQVKRZWKDWDODUJHVHWRIVWUXFWXUDOIHDWXUHVGRHVUHYHDO DSK\ORJHQHWLFVLJQDOLQWKDWKLJKHUOHYHOOLQJXLVWLFJURXSLQJVDUHLGHQWL¿HG'XHWRWKHLU OLPLWHGGHVLJQVSDFHDQGUHODWLYHHDVHRIGLIIXVLRQWKH\FDQQRWXQHTXLYRFDOO\LGHQWLI\ORZHU OHYHOODQJXDJHIDPLOLHV$VVKRZQLQVHFWLRQWKH6WUXFWXUHDOJRULWKPFDQQRWVHSDUDWH GLIIHUHQWOLQHDJHVLQHDVWHUQ,QGRQHVLDDWOHDVWQRWZLWKDVWURQJOLNHOLKRRG$PDWWHUIRU IXUWKHUUHVHDUFKZRXOGEHWRLQYHVWLJDWHZKHWKHUDGLIIHUHQWVHWRIIHDWXUHVFRXOGGREHWWHU It may be that a small set of diagnostic traits is masked by a much larger number of features that are shared by languages of different families by a Bayesian inference algorithm such DV6WUXFWXUHDVLOOXVWUDWHGIRU+DWDP0H\DKDQG%LDNLQVHFWLRQ :KLOH VWUXFWXUDO IHDWXUHV FDQ EH GLIIXVHG FRPSOHWH VXEVWLWXWLRQ LV TXLWH UDUH 7KH EDVLFPRUSKRV\QWDFWLFSUR¿OHOLQNHGWRWKHVHPDQWLFSUDJPDWLFZD\RIUHSUHVHQWLQJWKH QDWXUDODQGVRFLDOZRUOGRIDQ\SDUWLFXODUVSHHFKFRPPXQLW\LVTXLWHUREXVWWKURXJKPDQ\ GHVFHQGLQJJHQHUDWLRQV7KHUHIRUHWKHOLQJXLVWLFFOXVWHUVIRXQGRQWKHEDVLVRIIXOOSUR¿OHV SURYLGH LQIRUPDWLRQ DERXW WKHLU KLVWRULFDO SURYHQDQFH ,I LW LV SRVVLEOH WR UHFRQVWUXFW determine the ancestral state of a particular feature in a (putative) family, as for example shown by the presence of a decimal counting system and absence of past tense marking in the Austronesian family, then aberrant values in daughter languages can be accounted for E\K\EULGL]DWLRQ REFERENCES %OXVW5REHUW &HQWUDODQG&HQWUDO(DVWHUQ0DOD\R3RO\QHVLDQOceanic Linguistics %OXVW5REHUW 7KHSRVLWLRQRIWKHODQJXDJHVRIHDVWHUQ,QGRQHVLD $UHSO\WR'RQRKXH DQG*ULPHVOceanic Linguistics &URIW :LOOLDP 7\SRORJLFDO WUDLWV DQG JHQHWLF OLQJXLVWLFV KWWS ZZZXQP HGXaZFURIW3DSHUV7\S*HQSGI 1RYHPEHU &URIW:LOOLDP(YROXWLRQDU\OLQJXLVWLFVAnnual Review of Anthropology ± 'DKOgVWHQ 9HOXSLOODL9LYHND7KH3DVW7HQVH,Q'U\HU +DVSHOPDWKhttp:// ZDOVLQIRFKDSWHU 6HSWHPEHU 'DKOgVWHQ 9HOXSLOODL9LYHND7KH)XWXUH7HQVH,Q'U\HU +DVSHOPDWKhttp:// ZDOVLQIRFKDSWHU 6HSWHPEHU 'RQRKXH0DUN 6LPRQ0XVJUDYH7\SRORJ\DQGWKHOLQJXLVWLFPDFURKLVWRU\RI ,VODQG0HODQHVLDOceanic Linguistics ± 'RQRKXH0DUN &KDUOHV(*ULPHV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 'U\HU0DWWKHZ65HODWLRQVKLSEHWZHHQWKHRUGHURIREMHFWDQGYHUEDQGWKHRUGHU RIDGSRVLWLRQDQGQRXQSKUDVH,Q'U\HU +DVSHOPDWKKWWS ZDOVLQIRFKDSWHU 6HSWHPEHU 'U\HU 0DWWKHZ 6 0DUWLQ +DVSHOPDWK HGV The world atlas of language structures onlineKWWS ZDOVLQIR 'XQQ 0LFKDHO *HU 5HHVLQN $QJHOD 7HUULOO 7KH (DVW 3DSXDQ ODQJXDJHV $ SUHOLPLQDU\W\SRORJLFDODSSUDLVDOOceanic Linguistics ± 'XQQ 0LFKDHO$QJHOD7HUULOO *HU 5HHVLQN 5REHUW$ )ROH\ 6WHSKHQ & /HYLQVRQ 6WUXFWXUDO SK\ORJHQHWLFV DQG WKH UHFRQVWUXFWLRQ RI DQFLHQW ODQJXDJH KLVWRU\ Science ± 'XQQ0LFKDHO5REHUW$)ROH\6WHSKHQ&/HYLQVRQ*HU5HHVLQN $QJHOD7HUULOO 6WDWLVWLFDO UHDVRQLQJ LQ WKH HYDOXDWLRQ RI W\SRORJLFDO GLYHUVLW\ LQ ,VODQG 0HODQHVLD Oceanic Linguistics 'XQQ0LFKDHO6WHSKHQ&/HYLQVRQ(YD/LQGVWU|P*HU5HHVLQN $QJHOD7HUULOO Structural phylogeny in historical linguistics: Methodological explorations applied in ,VODQG0HODQHVLDLanguage 'XQQ 0LFKDHO 6LPRQ - *UHHQKLOO 6WHSKHQ & /HYLQVRQ 5XVVHOO ' *UD\ (YROYHGVWUXFWXUHRIODQJXDJHVKRZVOLQHDJHVSHFL¿FWUHQGVLQZRUGRUGHUXQLYHUVDOV Nature ± )ROH\:LOOLDP$ 7RZDUGXQGHUVWDQGLQJ3DSXDQODQJXDJHV,Q0LHGHPDHWDO± )ROH\:LOOLDP$7KHODQJXDJHVRI1HZ*XLQHDAnnual Review of Anthropology *UHHQEHUJ -RVHSK + 7KH ,QGR3DFL¿F K\SRWKHVLV ,Q 7KRPDV $ 6HEHRN HG Linguistics in Oceania &XUUHQW7UHQGVLQ/LQJXLVWLFV ±7KH+DJXH 0RXWRQ +DUULVRQ632QWKHOLPLWVRIWKHFRPSDUDWLYHPHWKRG,Q%ULDQ'-RVHSK 5LFKDUG '-DQGD HGV The handbook of historical linguistics±/RQGRQ %ODFNZHOO +HLQH%HUQG 7DQLD.XWHYDThe genesis of grammar2[IRUG 2[IRUG8QLYHUVLW\ 3UHVV .ODPHU0DULDQ*HU5HHVLQN 0LULDPYDQ6WDGHQ(DVWHUQ,QGRQHVLDDVDOLQJXLVWLF DUHD ,Q 3LHWHU 0X\VNHQ HG From linguistic areas to areal linguistics $PVWHUGDP %HQMDPLQV /\QFK-RKQ0DOFROP5RVV 7HUU\&URZOH\The Oceanic Languages/RQGRQ &XU]RQ 0LHGHPD-HOOH&HFLOLD2Gp 5LHQ$&'DP HGV Perspectives on the Bird’s Head of Irian Jaya.$PVWHUGDP 5RGRSL 3DZOH\$QGUHZ. 7KH7UDQV1HZ*XLQHDSK\OXPK\SRWKHVLV $UHDVVHVVPHQW,Q 0LHGHPDHWDO 3DZOH\$QGUHZ7KHFKHTXHUHGFDUHHURIWKHWUDQV1HZ*XLQHDK\SRWKHVLV 5HFHQW UHVHDUFKDQGLWVLPSOLFDWLRQV,Q3DZOH\HWDO 3DZOH\$QGUHZ5REHUW$WWHQERURXJK-DFN*ROVRQ 5RELQ+LGH HGV Papuan Pasts, Studies in the cultural, linguistic and biological history of the Papuan-speaking peoples&DQEHUUD 3DFL¿F/LQJXLVWLFV 3ULWFKDUG-RQDWKDQ0DWWKHZ6WHSKHQV 3HWHU'RQQHOO\,QIHUHQFHRISRSXODWLRQ VWUXFWXUHXVLQJPXOWLORFXVJHQRW\SHGDWDGenetics ± MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 53 5HHVLQN*HULanguages of the eastern Bird’s Head&DQEHUUD 3DFL¿F/LQJXLVWLFV 5HHVLQN*HU:HVW3DSXDQODQJXDJHV 5RRWVDQGGHYHORSPHQW,Q3DZOH\HWDO 5HHVLQN*HU5XWK6LQJHU 0LFKDHO'XQQ ([SODLQLQJWKHOLQJXLVWLFGLYHUVLW\RI 6DKXOXVLQJSRSXODWLRQPHWKRGVPloS Biology H 5RVV 0DOFROP ,V WKHUH DQ (DVW 3DSXDQ SK\OXP" (YLGHQFH IURP SURQRXQV ,Q $QGUHZ3DZOH\0DOFROP5RVVDQG'DUUHOO7U\RQ HGV The boy from Bundaberg: Studies in Melanesian linguistics in honour of Tom Dutton,&DQEHUUD 3DFL¿F /LQJXLVWLFV 5RVV0DOFROP3URQRXQVDVDSUHOLPLQDU\GLDJQRVWLFIRUJURXSLQJ3DSXDQODQJXDJHV ,Q3DZOH\HWDO :XUP6WHSKHQ$ 7KH(DVW3DSXDQSK\OXPLQJHQHUDO,Q6WHSKHQ$:XUP HG Papuan languages and the New Guinea linguistic scene&DQEHUUD 3DFL¿F /LQJXLVWLFV :XUP6WHSKHQ$ Papuan languages of Oceania $UV/LQJXLVWLFD 7ELQJHQ *XQWHU1DUU Ger Reesink JHUUHHVLQN#KFFQHWQO Michael Dunn PLFKDHOGXQQ#PSLQO MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison APPENDIX ,QWKLVWDEOHWKHOLVWVRIFKDUDFWHUVXVHGIRU'XQQHWDO LQLanguage and for Reesink HWDO LQPloS BiologyDUHFRPSDUHG characters ‘Language 2008’ characters ‘PloS 2009’ 1 Are there as many points of articulation for nasals as there are for stops? Only consider points of articulation where a nasal is phonetically possible (1: present, 0: absent) Is there contrast between heterorganic DQG KRPRUJDQLF VHTXHQFH RI QDVDO and velar stop? For example, does the language permit a phonetic contrast EHWZHHQ QN DQG ƾN FOXVWHUV (1: present, 0: absent) 1 Are there fricative phonemes? 3 Are there fricative phonemes? (1: present, 0: absent) Are there phonemic prenasalised Are there phonemic prenasalised stops? stops? (1: present, 0: absent) 3 Is there a phonemic distinction 5 Is there a phonemic distinction between l/r? between l/r? (1: present, 0: absent) Is there a phonemic velar fricative or 6 Is there a phonemic velar fricative or glide? glide? (1: present, 0: absent) 5 Is there a voicing contrast between 7 Is there a voicing contrast between RUDO LHQRQSUHQDVDO VWRSV" RUDO LH QRQSUHQDVDO VWRSV" present, 0: absent) 8 Is there a laminal/apical contrast? (1: present, 0: absent) 9 $UHWKHUHUHWURÀH[HGFRQVRQDQWV" present, 0: absent) 6 Is there phonemic consonant length? 10 Is there phonemic consonant length? (1: present, 0: absent) 7 Is there phonemic vowel length? 11 Is there phonemic vowel length? (1: present, 0: absent) 8 Are there contrastive phonation W\SHVIRUYRZHOV" HJQDVDOFUHDN\ etc) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 55 characters ‘Language 2008’ characters ‘PloS 2009’ Are there two or more contrastive central vowels Do not include length contrasts (1: present, 0: absent) 9 Is there lexically determined 13 Is there lexically determined suprasegmental prominence? suprasegmental prominence? suprasegmental prominence can be suprasegmental prominence can be loudness, duration, pitch, i.e. stress loudness, duration, pitch, i.e. stress or tone phenomena (don’t include or tone phenomena (don’t include phonemic vowel length) phonemic vowel length) (1: present, 0: absent) Is there weight-sensitive suprasegmental prominence suprasegmental prominence can be loudness, duration, pitch, i.e. stress or tone phenomena (1: present, 0: absent) 15 Is there syllable position sensitive suprasegmental prominence? suprasegmental prominence can be loudness, duration, pitch, i.e. stress or tone phenomena (1: present, 0: absent) 16 Is there a tonal system? I.e. two or more contrastive tones (1: present, 0: absent) 10 $UHWKHUHZRUG¿QDOFRQVRQDQWV" 17 $UHWKHUHZRUG¿QDOFRQVRQDQWV" present, 0: absent) 11 Are there consonant clusters? 18 Are there consonant clusters (not counting prenasalized consonants) in syllable onset? (1: present, 0: absent) 19 Are there consonant clusters (not counting prenasalized consonants) in syllable coda? (1: present, 0: absent) $UHWKHUHGH¿QLWHRUVSHFL¿FDUWLFOHV" $UHWKHUHGH¿QLWHRUVSHFL¿FDUWLFOHV" (1: present, 0: absent) 13 $UH WKHUH LQGH¿QLWH RU QRQVSHFL¿F ,V DQ LQGH¿QLWH 13 REOLJDWRULO\ articles? DFFRPSDQLHGE\DQLQGH¿QLWH RUQRQ VSHFL¿F DUWLFOH"Disregard if only on personal names (1: present, 0: absent) Is the order of NP elements Art N? Are there prenominal articles? (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 56 characters ‘Language 2008’ characters ‘PloS 2009’ 15 Are NPs N-initial (except for Are there postnominal articles? (1: articles)? present, 0: absent) What is the relative position of numeral and noun in the NP? (multistate 1XP1 11XP ERWK What is the relative position of demonstrative and noun in the NP? (multistate 'HP1 1'HP ERWK Are there ‘discontinuous noun phrases’? Can an argument be expressed by multiple N/NP throughout the clause > i.e. the Australian type. (1: present, 0: absent) Is there a difference between the marking of NP coordination (‘John and Mary went to market’) and the marking of comitative phrases (‘John went to market with Mary’)? (1: present, 0: absent) 16 Is there an inclusive/exclusive Is there an inclusive/exclusive distinction? distinction? (1: present, 0: absent) Is there a minimal-augmented system? i.e. four basic pronominal forms for 1sg, 2sg, 3sg and 1+2, which each FDQEHDI¿[HGIRUSOXUDO RUGXDOHWF (1: present, 0: absent) 30 Is there a gender distinction in 3rd person pronouns (or demonstratives, if no 3rd person pronouns)? either two- or threefold (1: present, 0: absent) 31 Is there a dual (or unit augmented) in addition to a plural (or augmented) number category in pronouns? (1: present, 0: absent) 17 $UHVWDQGQGSHUVRQVFRQÀDWHGLQ $UHVWDQGQGSHUVRQVFRQÀDWHGLQ any context? any context? (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 57 characters ‘Language 2008’ characters ‘PloS 2009’ 18 $UH QG DQG UG SHUVRQV FRQÀDWHG 33 $UH QG DQG UG SHUVRQV FRQÀDWHG in non-singular numbers? in non-singular numbers? (Morphologically in any paradigm. morphologically in any paradigm. Disregard pragmatics/politeness) Disregard pragmatics/politeness (1: present, 0: absent) Are person categories neutralized under some conditions? e.g. in non- singular, under NEG, in certain TAM (1: present, 0: absent) 19 $UH PRUH WKDQ GHJUHHV RI 35 Is there an opposition between distance morphologically marked in three or more distance terms in the demonstratives? demonstrative system? (1: present, 0: absent) Are any of the spatial demonstratives not speaker-based? Speaker- based spatial demonstratives are demonstratives that take as their decitic centre the speaker. By contrast, some demonstratives take not the speaker but the addressee as the deictic centre, for example a demonstrative might mean ‘close to the speaker’; and some take both speaker and addressee as the deictic centre e.g. ‘far from speaker and addressee’. Is elevation morphologically marked 36 Is elevation morphologically marked in demonstratives? in demonstratives? (1: present, 0: absent) 37 Is the opposition visible-non-visible marked on demonstratives? (1: present, 0: absent) $UHGHPRQVWUDWLYHVFODVVL¿HG" 38 $UH GHPRQVWUDWLYHV FODVVL¿HG" present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 58 characters ‘Language 2008’ characters ‘PloS 2009’ Are there declensions (partly) 39 Are there declensions (partly) determined by number of the noun? determined by number of the noun? By noun declensions is meant e.g By noun declensions is meant e.g nouns divided into groups which nouns divided into groups which have formally different sets of have formally different sets of morphological marking. Do not morphological marking. Do not include place names which can act include place names which can act as as bare adjuncts bare adjuncts (1: present, 0: absent) Are there declensions (partly) Are there declensions (partly) determined by gender of the noun? determined by gender of the noun? By noun declensions is meant e.g By noun declensions is meant e.g nouns divided into groups which nouns divided into groups which have formally different sets of have formally different sets of morphological marking. Do not morphological marking. Do not include place names which can act include place names which can act as as bare adjuncts bare adjuncts (1: present, 0: absent) Are there nouns which are suppletive Are there nouns which are suppletive for number? (Only yes if present for for number? Only answer yes if more than 2 (basic) kin terms) present for more than 2 (basic) kin terms Can singular number be marked on the noun itself? Number marking on noun does not count phrase level clitic or reduplication; absence of plural marking does not count as singular marking; exclude derivational forms (e.g. deverbal, deadjectival) (1: present, 0: absent) Can dual number be marked on the Can dual number be marked on the noun itself? Number-marking on N noun itself? number-marking on N does not count phrase-level clitic or does not count phrase-level clitic or reduplication reduplication (1: present, 0: absent) Can plural number be marked on the noun itself? number-marking on N does not count phrase-level clitic or reduplication (1: present, 0: absent) Is number marking prohibited on Is number marking prohibited on certain (types of) nouns? (do not certain (types of) nouns? (do not include proper nouns, e.g. place include proper nouns, e.g. place names or personal names) names or personal names) (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 59 characters ‘Language 2008’ characters ‘PloS 2009’ Are there associative plurals? e.g. Mary-PL = Mary and her family (1: present, 0: absent) Is there a productive morphologically marked Action/state nominalization (arrive-arrival)? if a language is precategorial, include the morphological mechanisms to produce such ‘nominalizations’ (1: present, 0: absent) Is there a productive morphologically marked Agentive nominalization (sing-er)? (1: present, 0: absent) Is there a productive morphologically marked Object nominalization (sing; song)? (1: present, 0: absent) Are there noun classes/genders? 50 Are there noun classes/genders? By noun classes/genders is meant By noun classes/genders is meant a system of dividing all or almost a system of dividing all or almost all of the nouns of a language all of the nouns of a language into morphological classes which into morphological classes which determine agreement phenomena determine agreement phenomena beyond the noun itself. beyond the noun itself. (1: present, 0: absent) 51 Is sex a relevant category in the noun class/gender system? (1: present, 0: absent) Is shape a relevant category in the noun class/gender system? (1: present, 0: absent) 53 Is animacy (without reference to sex) a relevant category in the noun class/ gender system? (1: present, 0: absent) Is plant status a relevant category in the noun class/gender system? (1: present, 0: absent) 55 Does the language only have a gender distinction in 3rd person pronouns? (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 60 characters ‘Language 2008’ characters ‘PloS 2009’ 56 ,V WKHUH FRQFRUG ZLWKLQ WKH 13 LH agreement of elements within the NP with the noun class of a noun? related to class/gender (1: present, 0: absent) $UH WKHUH QXPHUDO FODVVL¿HUV" i.e. 57 $UH WKHUH QXPHUDO FODVVL¿HUV" i.e. free or bound morphemes which are free or bound morphemes which are non-agreeing, noun categorization non-agreeing, noun categorisation devices, the choice of which is devices, the choice of which is determined by lexical selection determined by lexical selection (1: present, 0: absent) 30 $UHWKHUHSRVVHVVLYHFODVVL¿HUV"i.e. 58 $UH WKHUH SRVVHVVLYH FODVVL¿HUV" i.e. free or bound morphemes which are free or bound morphemes which are non-agreeing, noun categorisation non-agreeing, noun categorisation devices, the choice of which is devices, the choice of which is determined by lexical selection determined by lexical selection (1: present, 0: absent) 31 Are there possessive classes? i.e. different nouns treated differently in possession according to semantically-based groupings. Include alienable/inalienable Is alienable/inalienable a relevant 59 Is alienable/inalienable a relevant distinction? distinction? (1: present, 0: absent) 33 Are there different possessive 60 Are there different possessive constructions? constructions? (1: present, 0: absent) 61 &DQSRVVHVVLRQEHPDUNHGE\DSUH¿[" even if only on a restricted numer of kin terms. Emphasis is on *can* (1: present, 0: absent) &DQSRVVHVVLRQEHPDUNHGE\DVXI¿[" even if only on a restricted numer of kin terms. Emphasis is on *can* (1: present, 0: absent) Can possession be marked on the 63 Can possession be marked on the nominal possessor? nominal possessor? (1: present, 0: absent) 35 Can possession be marked on the Can possession be marked on the nominal possessee? nominal possessee? (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 61 characters ‘Language 2008’ characters ‘PloS 2009’ 36 If the order of elements in a 65 What is the relative position of SRVVHVVLYHFRQVWUXFWLRQLV¿[HGLVLW possessor and possessed in the possessor-possessed? attributive possessive construction? (multistate 1:Possessor-Possessed; 3RVVHVVHG3RVVHVVRU ERWK 66 Are there different orders of elements in a possessive phrase for different classes of possession? emphasis on *for different types of possession* (1: present, 0: absent) 37 Is there a decimal counting system? 67 What is the counting system? (i.e. elements of decimal; even (multistate 'HFLPDO 4XLQDU\ lexical 10, 10+5 qualify %RG\SDUW WDOO\LQJ PLQLPDO [Other systems, like senary, are not VFRUHG@ 38 Is there evidence for any element RIDTXLQDU\FRXQWLQJV\VWHP" e.g. expressions for 5+1, 10+5+1 39 Are there words for particular amounts of a thing? (e.g. ten possums) Is there lexical overlap between a 68 'R FRUH DGMHFWLYHV GH¿QHG VLJQL¿FDQW SURSRUWLRQ RI DGMHFWLYHV semantically as property concepts; and verbs (including zero- value, shape, age, dimension) act derivation)? like verbs in predicative position? (1: present, 0: absent) Does the same lexical set of 69 'R FRUH DGMHFWLYHV GH¿QHG adjectives function both attributively semantically as property concepts; and predicatively? value, shape, age, dimension) XVHG DWWULEXWLYHO\ UHTXLUH WKH VDPH morphological treatment as verbs? (1: present, 0: absent) Is there case marking for core 70 Is there case marking for core nominal QRPLQDO 13V LH 6 $ RU 2 13V LH6$RU2IXQFWLRQ "for case function)? For case marking, PDUNLQJLQFOXGHDQ\DI¿[DOPDUNLQJ LQFOXGH DQ\ DI¿[DO PDUNLQJ ZKLFK which appears in the NP and shows appears in the NP and shows the the function of the NP in the clause; function of the NP in the clause; do not count adpositions (1: present, adpositions are not counted. 0: absent) 71 Is there case marking for core pronouns? (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison characters ‘Language 2008’ characters ‘PloS 2009’ ,V WKHUH FDVH PDUNLQJ IRU REOLTXH ,V WKHUH FDVH PDUNLQJ IRU REOLTXH nominal NPs ? e.g. locationals, nominal NPs ? e.g. locationals, instrumentals, etc.; adpositions are instrumentals, etc. do not count not counted. adpositions (1: present, 0: absent) 73 ,V WKHUH FDVH PDUNLQJ IRU REOLTXH pronouns? (1: present, 0: absent) Are there prepositions? Are there prepositions? (1: present, 0: absent) Are there postpositions? 75 Are there postpositions? (1: present, 0: absent) 76 Are there adpositions to mark core NPs? (1: present, 0: absent) 77 $UHWKHUHDGSRVLWLRQVWRPDUNREOLTXH NPs? (1: present, 0: absent) 78 Is there a distinction between locational and directional adpositions? (1: present, 0: absent) Do the same morphemes 79 'R YHUEV KDYH SUH¿[HVSURFOLWLFV systematically encode both TAM RWKHUWKDQWKRVHWKDW21/<PDUN$6 and person? or O (do include portmanteau: A & S + TAM)? $6DQG2DI¿[HVDUHGHDOW with in 3.3 (1: present, 0: absent) 'RYHUEVKDYHSUH¿[HVSURFOLWLFV" 'RYHUEVKDYHVXI¿[HVHQFOLWLFV" 80 'RYHUEVKDYHVXI¿[HVHQFOLWLFVRWKHU WKDQ WKRVH WKDW 21/< PDUN$ 6 RU O (do include portmanteau: A & S + TAM)? (1: present, 0: absent) 81 &DQ LQ¿[DWLRQ EH XVHG RQ YHUEV IRU derivational, aspectual, or voice- changing purposes? (1: present, 0: absent) Is there present tense regularly morphologically marked on the verb? (1: present, 0: absent) 83 Is there past tense regularly morphologically marked on the verb? (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 63 characters ‘Language 2008’ characters ‘PloS 2009’ Is there future tense regularly morphologically marked on the verb? (1: present, 0: absent) 85 Are there multiple past or future tenses, distinguishing distance from Time of Reference, marked on the verb? (1: present, 0: absent) Is a distinction between punctual/ 86 Is a distinction between punctual/ continuous aspect available as a continuous aspect available as a morphological choice? morphological choice? (1: present, 0: absent) 50 Is a distinction between realis/irrealis 87 Is a distinction between realis/irrealis mood available as a morphological mood available as a morphological choice? choice? (1: present, 0: absent) 88 Is there an apprehensive modal category marked on the verb also known as ‘evitative’, ‘lest’, etc (1: present, 0: absent) 51 Is the S participant (at least 89 Is the S participant (at least sometimes) VRPHWLPHV PDUNHG E\ D VXI¿[ PDUNHGE\DVXI¿[HQFOLWLF"pertains enclitic? pertains to verb morphology to verb morphology (1: present, 0: absent) Is the S participant (at least 90 Is the S participant (at least sometimes) VRPHWLPHV PDUNHG E\ D SUH¿[ PDUNHGE\DSUH¿[SURFOLWLF"pertains proclitic? pertains to verb to verb morphology (1: present, 0: morphology absent) 53 Is the A participant (at least 91 Is the A participant (at least sometimes) VRPHWLPHV PDUNHG E\ D VXI¿[ PDUNHGE\DVXI¿[HQFOLWLF"pertains enclitic? pertains to verb morphology to verb morphology (1: present, 0: absent) Is the A participant (at least Is the A participant (at least sometimes) VRPHWLPHV PDUNHG E\ D SUH¿[ PDUNHGE\DSUH¿[SURFOLWLF"pertains proclitic? pertains to verb to verb morphology (1: present, 0: morphology absent) 55 Is the O participant (at least 93 Is the O participant (at least VRPHWLPHV PDUNHG E\ D VXI¿[ VRPHWLPHV PDUNHG E\ D VXI¿[ enclitic? pertains to verb morphology enclitic? pertains to verb morphology (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison characters ‘Language 2008’ characters ‘PloS 2009’ 56 Is the O participant (at least Is the O participant (at least sometimes) VRPHWLPHV PDUNHG E\ D SUH¿[ PDUNHGE\DSUH¿[SURFOLWLF"pertains proclitic? pertains to verb to verb morphology (1: present, 0: morphology absent) 57 Are variations in marking strategies 95 Are variations in marking strategies of core participants based on TAM of core participants based on TAM distinctions? distinctions? this question refers to variations (if they occur) in 89-94 (1: present, 0: absent) 58 Are variations in marking strategies 96 Are variations in marking strategies based on verb classes? based on verb classes? this question refers to variations (if they occur) in 89-94 (1: present, 0: absent) 59 Are variations in marking strategies 97 Are variations in marking strategies EDVHG RQ FODXVH W\SH HJ PDLQ YV EDVHG RQ FODXVH W\SH HJ PDLQ YV subordinate? subordinate? this question refers to variations (if they occur) in 89-94 (1: present, 0: absent) 60 Are variations in marking strategies 98 Are variations in marking strategies based on person distinctions? based on person distinctions? this question refers to variations (if they occur) in 89-94 (1: present, 0: absent) 61 Do verb stems alter according to the number of a core participant? Do verb stems alter according to the 99 Do verb stems alter according to person of a core participant? the person of a core participant? (1: present, 0: absent) 63 Is number ever marked separately 100 Is number ever marked separately from person on the verb? from person on the verb? (1: present, 0: absent) Are person, number and any TAM 101 Are person, number and any TAM FDWHJRU\ LH 3 or more categories FDWHJRU\ LH RU PRUH FDWHJRULHV in all) marked by portmanteau in all) marked by portmanteau morphemes on verbs? morphemes on verbs? (1: present, 0: absent) 65 Are categories such as person, Are categories such as person, number, gender related to a single number, gender related to a single participant discontinuously marked participant discontinuously marked on a verb? on a verb? (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 65 characters ‘Language 2008’ characters ‘PloS 2009’ 66 Is a non-core participant marked on 103 Are benefactive nominals marked on the verb? ,QFOXGHDI¿[HVFOLWLFVDQG the verb? (1: present, 0: absent) satellite particles associated with verbs forming a constituent with the verb on some level, but exclude optional adverbials. Can instruments be marked on the verb? (1: present, 0: absent) 67 Can recipients be treated as a 105 Can recipients be treated as a WUDQVLWLYH REMHFW LH DV 'LUHFW WUDQVLWLYHREMHFWLHDV'LUHFW2EMHFW" object? (1: present, 0: absent) 68 Are there syntactically ditransitive 106 Are there syntactically ditransitive verbs? verbs? (1: present, 0: absent) 69 Is negation marked morphologically 107 Is negation marked morphologically on the verbs? LH DI¿[DWLRQ VWHP on the verbs? LH DI¿[DWLRQ VWHP alternation, neutralization of some alternation, neutralization of some LQÀHFWLRQ LQÀHFWLRQ (1: present, 0: absent) 70 Is direction marked on verbs 108 Can locative or direction be ,QFOXGHVDI¿[HVFOLWLFVDQGVDWHOOLWH morphologically marked on the verb? particles associated with verbs Locative as Direct Object (‘she sleeps forming a constituent with the verb mat’) does not qualify (1: present, 0: on some level, but excludes optional absent) adverbials. 71 Are there suppletive verbs for 109 Are there suppletive verbs for number number of participants of participants? (list them all if feasible, otherwise give an estimate of the number and/or proportion of nouns) (1: present, 0: absent) 110 Are there suppletive verbs for tense or aspect? (1: present, 0: absent) Are there conjugation classes? 111 Are there conjugation classes? (1: present, 0: absent) 73 Are there (several) verbs which Are there (several) verbs which can be used either transitively or can be used either transitively or intransitively with no morphological intransitively with no morphological marking? say no if it’s only one or marking? say no if it’s only one or two two stems; Intended here is the stems; Intended here is the ‘break’ ‘break’ and ‘open’ type; not John and ‘open’ type; not John eats/ eats eats/ eats the bread the bread (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 66 characters ‘Language 2008’ characters ‘PloS 2009’ Is there transitivising morphology 113 Is there transitivising morphology (include clitics)? (include clitics)? (1: present, 0: absent) 75 Is there morphology (include clitics) Is there morphology (include clitics) WRPDUNDUHÀH[LYHDFWLRQ"free word/ WRPDUNDUHÀH[LYHDFWLRQ"free word/ particle does not count; neither a particle does not count; neither a default P/N co-reference default P/N co-reference (1: present, 0: absent) 76 Is there morphology (include 115 Is there morphology (include clitics) clitics) to mark a reciprocal action? to mark a reciprocal action? free free word/particle does not count; word/particle does not count; neither neither a default P/N co-reference a default P/N co-reference (1: present, 0: absent) 77 Do verbs classify the shape, size, 116 Do verbs classify the shape, size, consistency or position of absolutive consistency or position of absolutive arguments by means of incorporated arguments by means of incorporated QRXQV YHUEDO DI¿[HV RU VXSSOHWLYH QRXQV YHUEDO DI¿[HV RU VXSSOHWLYH verb stems? not included here are verb stems? not included here are positional verbs that classify a positional verbs that classify a referent in such terms referent in such terms - covered by 127 (1: present, 0: absent) 78 Is there a copula for predicate nouns? 117 Is there a copula for predicate nouns? e.g. John is a teacher e.g. John is a teacher (1: present, 0: absent) 79 Are there serial verb constructions? 118 Are there serial verb constructions? (i.e. two or more verbs in (i.e. two or more verbs in juxtaposition, juxtaposition, functioning as a single functioning as a single predicate, predicate, with no morphology with no morphology to mark their to mark their relationship with relationship with each other. Each of each other. Each of the verbs is a the verbs is a separate phonological separate phonological word but word but the construction as a whole the construction as a whole is is expressed in one intonational unit. expressed in one intonational unit. Morphology is shared to a greater or Morphology is shared to a greater lesser extent.) (1: present, 0: absent) or lesser extent.) 80 Is there one or more auxiliary? 119 Are there modal auxiliaries? (1: present, 0: absent) Are there aspectual auxiliaries? (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 67 characters ‘Language 2008’ characters ‘PloS 2009’ Are there tense auxiliaries? (1: present, 0: absent) 81 Is verb compounding a regular Is verb compounding a regular process? (i.e. two or more verb process? (i.e. two or more verb stems stems acting as one phonological acting as one phonological and and grammatical word) grammatical word) (1: present, 0: absent) Are there verb-adjunct (aka Are there verb-adjunct (aka light-verb) light-verb) constructions? (i.e. constructions? (i.e. constructions constructions involving a non- involving a non-predicating element predicating element expressing the expressing the lexical meaning of the lexical meaning of the construction, construction, in conjunction with a in conjunction with a semantically semantically fairly empty verb, which fairly empty verb, which enables the enables the element to function as a element to function as a predicate by predicate by providing the necessary providing the necessary morphology, morphology, e.g. eye do for ‘see’; or e.g. eye do for ‘see’; or sneeze hit for sneeze hit for ‘sneeze’) (1: present, 0: ‘sneeze’) absent) 83 Is there incorporation of any element into verbs? Is there incorporation of nouns into verbs a productive intransitivizing process? (1: present, 0: absent) Is there productive incorporation of other elements (adjectives, locatives, HWF LQWRYHUEV" SUHVHQW DEVHQW Is there one or more existential verb? Is there one or more existential verb? exclude e.g. positional verbs exclude e.g. positional verbs (3.8.02) (1: present, 0: absent) $UH WKHUH SRVLWLRQDO FODVVL¿FDWRU\ verbs? (i.e. in answer to a question ‘Where is the X’, does the verb used in the answer depend on the type of referent (e.g. do you have to say ‘The X sits/stands/lies/etc on the table’). List them all. (1: present, 0: absent) 85 Is the verb ‘give’ morphologically Is the verb ‘give’ morphologically peculiar (different from most peculiar (different from most other other verbs)? e.g. stem suppletion, verbs)? e.g. stem suppletion, different GLIIHUHQWDI¿[DWLRQ DI¿[DWLRQ (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 68 characters ‘Language 2008’ characters ‘PloS 2009’ 86 ,VWKHUHDQRWDEO\VPDOOQXPEHULH ,V WKHUH D QRWDEO\ VPDOO QXPEHU LH about 100 or less, of verbs in the about 100 or less, of verbs in the language? language? (1: present, 0: absent) 87 Is a pragmatically unmarked 130 What is the pragmatically unmarked FRQVWLWXHQWRUGHU69IRULQWUDQVLWLYH RUGHU RI 6 DQG 9 LQ LQVWUDQVLWLYH clauses? clauses? (multistate 69 96 both) 88 Is a pragmatically unmarked FRQVWLWXHQWRUGHU96IRULQWUDQVLWLYH clauses? 89 Is a pragmatically unmarked 131 Is a pragmatically unmarked constituent order verb-initial for constituent order verb-initial for transitive clauses? transitive clauses? (1: present, 0: absent) 90 Is a pragmatically unmarked Is a pragmatically unmarked constituent order verb-medial for constituent order verb-medial for transitive clauses? transitive clauses? (1: present, 0: absent) 91 Is a pragmatically unmarked 133 Is a pragmatically unmarked FRQVWLWXHQW RUGHU YHUE¿QDO IRU FRQVWLWXHQW RUGHU YHUE¿QDO IRU transitive clauses? transitive clauses? (1: present, 0: absent) Is the order of constituents the same in main and subordinate clauses? (1: present, 0: absent) 135 Do clausal objects occur in the same position as nominal objects? (1: present, 0: absent) ,V FRQVWLWXHQW RUGHU ¿[HG" Do not 136 ,V FRQVWLWXHQW RUGHU ¿[HG" Do not consider ‘left or right-dislocation’, consider ‘left or right-dislocation’, accompanied by intonational signals accompanied by intonational signals (1: present, 0: absent) 93 Can negation be marked clause- 137 Can negation be marked clause- ¿QDOO\" 7KLV LQFOXGHV VXI¿[HV RQ ¿QDOO\"7KLVLQFOXGHVVXI¿[HVRQYHUE YHUE¿QDO FODXVHV SUH¿[HV RQ ¿QDOFODXVHV SUH¿[HVRQFODXVH¿QDO FODXVH¿QDO YHUEV GR QRW FRXQW verbs do not count; Don’t include Don’t include elliptical ‘Pete didn’t’ elliptical ‘Pete didn’t’ (1: present, 0: absent) Can negation be marked clause- 138 Can negation be marked clause- initially? Don’t include elliptical initially? Don’t include elliptical ‘Not ‘Not Mary’ Mary’ (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 69 characters ‘Language 2008’ characters ‘PloS 2009’ 95 Is there a difference between 139 Is there a difference between imperative and declarative negation? imperative and declarative negation? (1: present, 0: absent) 96 Are verbal and non-verbal predicates Are verbal and non-verbal predicates marked by the same negator? marked by the same negator? (1: present, 0: absent) 97 $UH 6 DQG 2 FRQÀDWHG $UH6DQG2FRQÀDWHGPRUSKRORJLFDOO\ morphologically in at least some in at least some basic constructions, EDVLFFRQVWUXFWLRQVLHVLPSOHPDLQ LHVLPSOHPDLQFODXVHV" SUHVHQW clauses? 0: absent) 98 $UH 6 DQG $ FRQÀDWHG $UH6DQG$FRQÀDWHGPRUSKRORJLFDOO\ morphologically in at least some in at least some basic constructions, EDVLFFRQVWUXFWLRQVLHVLPSOHPDLQ LHVLPSOHPDLQFODXVHV" SUHVHQW clauses? 0: absent) 99 $UH 6 DQG 2 FRQÀDWHG $UH6DQG2FRQÀDWHGPRUSKRORJLFDOO\ morphologically across clause DFURVVFODXVHERXQGDULHVLHDFWLQJDV ERXQGDULHV LH DFWLQJ DV V\QWDFWLF syntactic pivot? (1: present, 0: absent) pivot? 100 $UH 6 DQG $ FRQÀDWHG $UH6DQG$FRQÀDWHGPRUSKRORJLFDOO\ morphologically across clause DFURVVFODXVHERXQGDULHVLHDFWLQJDV ERXQGDULHV LH DFWLQJ DV V\QWDFWLF syntactic pivot? (1: present, 0: absent) pivot? 101 Do S and O operate in the same Do S and O operate in the same way, way, and differently from A, and differently from A, for the purpose for the purpose of any syntactic of any syntactic construction? (1: construction? present, 0: absent) Is there a morpho-syntactic Is there a morpho-syntactic distinction distinction between predicates between predicates expressing expressing controlled versus controlled versus uncontrolled events uncontrolled events or states? or states? (1: present, 0: absent) Is there a morphologically marked passive construction? morphological marking includes some verbal DI¿[DWLRQ RU VRPH SHULSKUDVWLF element in the VP or clause (1: present, 0: absent) Is there a morphologically marked antipassive? morphological marking LQFOXGHV VRPH YHUEDO DI¿[DWLRQ RU some periphrastic element in the VP or clause (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 70 characters ‘Language 2008’ characters ‘PloS 2009’ Is there a morphologically marked inverse? i.e. different marking by YHUEDODI¿[DWLRQRUSURQRPLQDOFOLWLFV referring to A and O, depending on SHUVRQ DQLPDF\ RU GH¿QLWHQHVV (1: present, 0: absent) 103 Is there clause chaining? i.e. chains 150 Is there clause chaining? i.e. chains of morphologically stripped-down of morphologically stripped-down medial clauses which are dependent medial clauses which are dependent on a single clause (usually, but not on a single clause (usually, but not QHFHVVDULO\¿QDO IRUWKHLU7$0RU QHFHVVDULO\ ¿QDO IRU WKHLU 7$0 RU SDUWLFLSDQWPDUNLQJVSHFL¿FDWLRQ SDUWLFLSDQW PDUNLQJ VSHFL¿FDWLRQ (1: present, 0: absent) 151 Is there a morphologically-marked switch reference system? (1: present, 0: absent) Is there a morphologically marked Is there a morphologically marked distinction between simultaneous distinction between simultaneous and DQGVHTXHQWLDOFODXVHV" VHTXHQWLDO FODXVHV" SUHVHQW absent) 105 ,V WKH YHUE µVD\¶ RU D TXRWDWLYH 153 ,V WKH YHUE µVD\¶ RU D TXRWDWLYH construction used in desiderative construction used in desiderative constructions? (e.g. ‘I said for him constructions? (e.g. ‘I said for him to go’ for ‘I wanted him to go’) to go’ for ‘I wanted him to go’) (1: present, 0: absent) 106 $UH WKHUH SXUSRVLYH QRQ¿QLWH subordinate clauses? 107 $UH WKHUH WHPSRUDO QRQ¿QLWH subordinate clauses? 108 Are there complement clauses? 109 Are causatives formed by serial verb Are causatives formed by serial verb constructions? constructions? (1: present, 0: absent) 110 Are causatives formed by bound 155 Are causatives formed by bound DI¿[HVFOLWLFV" DI¿[HVFOLWLFV" SUHVHQW DEVHQW 111 Are causatives formed by 156 Are causatives formed by constructions involving ‘say’? constructions involving ‘say’? (1: present, 0: absent) Is topic or focus marked morphologically? LH E\ DI¿[HV RU clitics MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Systematic typological comparison 71 characters ‘Language 2008’ characters ‘PloS 2009’ 113 Is there tail-head linkage? (i.e. a 157 Is there tail-head linkage? (i.e. a GLVFRXUVHVWUDWHJ\LQZKLFKWKH¿QDO GLVFRXUVH VWUDWHJ\ LQ ZKLFKWKH ¿QDO verb of one sentence is repeated as verb of one sentence is repeated as WKH¿UVWYHUERIWKHQH[WVHQWHQFH WKH¿UVWYHUERIWKHQH[WVHQWHQFH (1: present, 0: absent) Are verbs reduplicated? 158 Are verbs reduplicated? (1: present, 0: absent) 115 Are nouns reduplicated? 159 Are nouns reduplicated? (1: present, 0: absent) 160 Are elements apart from verbs or nouns reduplicated? (1: present, 0: absent) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY /DQJXDJH'RFXPHQWDWLRQ &RQVHUYDWLRQ6SHFLDO3XEOLFDWLRQ1R 'HFHPEHU Melanesian Languages on the Edge of Asia: Challenges for the 21st Century, HGE\1LFKRODV(YDQVDQG0DULDQ.ODPHUSS KWWS QÀUFKDZDLLHGXOGFVS 4 KWWS KGOKDQGOHQHW Papuan-Austronesian language contact: Alorese from an areal perspective Marian Klamer Leiden University This paper compares the grammar and lexicon of Alorese, an Austronesian language spoken in eastern Indonesia, with its closest genealogical relative, Lamaholot, spoken on east Flores, as well as ZLWK LWV JHRJUDSKLFDO QHLJKERXUV WKH 3DSXDQ ODQJXDJHV RI 3DQWDU ,W IRFXVVHVRQWKHTXHVWLRQKRZ$ORUHVHFDPHWRKDYHWKHJUDPPDUDQG OH[LFRQLWKDVWRGD\,WLVVKRZQWKDW$ORUHVHDQG/DPDKRORWVKDUHD QXPEHU RI V\QWDFWLF IHDWXUHV ZKLFK VLJQDO 3DSXDQ LQÀXHQFHV WKDW must have been part of Proto-Lamaholot, suggesting (prehistoric) Papuan presence in the Lamaholot homeland in east Flores/Solor/ $GRQDUD/HPEDWD7KHGDWDLQGLFDWHWKDW3URWR/DPDKRORWKDGDULFK morphology, which was completely shed by Alorese after it split from /DPDKRORW$WWKHVDPHWLPHOH[LFDOFRQJUXHQFHEHWZHHQ$ORUHVHDQG its current Papuan neighbours is limited, and syntactic congruence YLUWXDOO\DEVHQW&RPELQLQJWKHFRPSDUDWLYHOLQJXLVWLFGDWDZLWKZKDW little is known about the history of the Alorese, I propose a scenario ZKHUHE\/DPDKRORWZDVDFTXLUHGDVQRQQDWLYHODQJXDJHE\VSRXVHV from different Papuan clans who were brought into the Lamaholot FRPPXQLWLHVWKDWVHWWOHGRQWKHFRDVWRI3DQWDUDWOHDVW\HDUVDJR 7KHLU PRUSKRORJLFDOO\ VLPSOL¿HG ODQJXDJH ZDV WUDQVIHUUHG WR WKHLU FKLOGUHQ7KHKLVWRU\RI$ORUHVHDVUHFRQVWUXFWHGKHUHVXJJHVWVWKDWDW different time depths, different language contact situations had different outcomes: prehistoric contact between Papuan and Proto-Lamaholot LQWKH)ORUHVDUHDUHVXOWHGLQDFRPSOH[L¿FDWLRQRI3URWR/DPDKRORW ZKLOHSRVWPLJUDWLRQFRQWDFWUHVXOWHGLQVLPSOL¿FDWLRQ,QERWKFDVHV the contact was intense, but the prehistoric contact with Papuan in the Flores area must have been long-term and involve pre-adolescents, while the post-migration contact was probably of shorter duration and LQYROYHGSRVWDGROHVFHQWOHDUQHUV cc Creative Commons Attribution Non-Commercial No Derivatives Licence Papuan-Austronesian language contact 73 1. INTRODUCTION. This article is about Alorese (Alor), an Austronesian language in HDVWHUQ,QGRQHVLD1,WIRFXVVHVRQWKHTXHVWLRQKRZ$ORUHVHFDPHWRKDYHWKHJUDPPDUDQG OH[LFRQLWKDVWRGD\%\FRPSDULQJ$ORUHVHZLWKLWVFORVHVWUHODWLYH/DPDKRORWDVZHOODV with its non-Austronesian neighbouring languages, we reconstruct some of its history and VWUXFWXUDOIHDWXUHV Alorese (also refered to as Bahasa Alor, Alor, Coastal Alorese,%DUQHV LV VSRNHQE\VSHDNHUVLQSRFNHWVDORQJWKHFRDVWVRIZHVWHUQ3DQWDUDQGWKH%LUG¶V Head of Alor island, as well as on the islands Ternate and Buaya (Stokhof 1975:8-9, Grimes HWDO /HZLV VHH¿JXUH.ODPHU LVDVNHWFKJUDPPDURIWKHODQJXDJH $ORUHVHLVWKHRQO\LQGLJHQRXV$XVWURQHVLDQODQJXDJHVSRNHQLQWKH$ORU3DQWDUDUFKLSHODJR ,WVKRZVVLJQL¿FDQWGLDOHFWDOYDULDWLRQ IRUH[DPSOHOH[LFDOGLIIHUHQFHVH[LVWVEHWZHHQWKH GLDOHFWRI%DUDQXVD 3DQWDULVODQG DQGWKHGLDOHFWVSRNHQRQ$ORUThe data discussed in WKLVSDSHULVPDLQO\IURPWKH%DUDQXVDGLDOHFW$OOGDWDDUHSULPDU\GDWDFROOHFWHGGXULQJ ¿HOGZRUNLQ FIGURE 1. Alorese as spoken Alor, Pantar, Buaya and Ternate (dark grey areas); Lamaholot YDULHWLHVDVVSRNHQRQ)ORUHV6RORU/HPEDWD In earlier sources, it has been suggested that Alorese is a dialect of Lamaholot (Stokhof .HUDI 6WHLQKDXHU DQGOLNHZLVHWKHPDSLQ%OXVW D 1 I would like to thank the two anonymous reviewers of this volume and Nick Evans as co-editor IRULQVLJKWIXOFRPPHQWVDQGGHWDLOHGVXJJHVWLRQVIRULPSURYHPHQW0DQ\WKDQNVDOVRWR6DQGHU Adelaar, Antoinette Schapper, Ger Reesink, and Hein Steinhauer who commented on earlier YHUVLRQVRIWKLVSDSHU MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 74 LQGLFDWHVWKDW/DPDKRORWLVVSRNHQRQ$ORUDQG3DQWDU$UHFHQWKLVWRULFDOFRPSDULVRQE\ 'R\OH VXJJHVWVWKDWJHQHDORJLFDOO\$ORUHVHLVLQGHHGFORVHO\UHODWHGWR/DPDKRORW Lamaholot (abbreviated as LMH) is spoken on the eastern part of Flores, and on the LVODQGVRI6RORUDQG/HPEDWD/DPDKRORWKDVVSHDNHUV$OWKRXJKLWLV XVXDOO\UHIHUUHGWRDVDVLQJOHODQJXDJHLWLVEHWWHUWKRXJKWRIDVDGLDOHFWFKDLQ.QRZQ YDULHWLHVLQFOXGHWKHIROORZLQJ VHH¿JXUH (i) LMH-Lewotobi, spoken in Wulunggitang and Ile Bura, in the western-most part of WKH/DPDKRORWVSHDNLQJUHJLRQRQ)ORUHV 1DJD\D DE (ii) LMH-Lewolema, spoken in the village Belogili-Balukhering, north of the town /DUDQWXNDRQHDVW)ORUHV3DPSXV DUHZRUGOLVWVRIWKLVYDULHW\ (iii) LMH-Lewoingu (Lewolaga), spoken in the village Leworook, south of Larantuka DQGGHVFULEHGLQ1LVKL\DPDDQG.HOHQ (iv) LMH-Solor, spoken on Solor island and described by Arndt (1937) and Bouman OH[LFDOVXUYH\GDWDFROOHFWHGE\.ODPHU (v) /0+/DPDOHUD VSRNHQ RQ VRXWK /HPEDWD .HUDI LV D GHVFULSWLRQ RI WKH PRUSKRORJ\RIWKLVYDULHW\ To the west, the Lamaholot speaking area is bordered by the language Sika (Lewis & *ULPHV $QHLJKERULQJODQJXDJHLQWKHHDVWLV.HGDQJ 6DPHO\ VSRNHQRQ QRUWK/HPEDWD:KLOH.HGDQJLVJHRJUDSKLFDOO\FORVHWRERWK/DPDOHUD VRXWK/HPEDWD and Alorese (north-west Pantar), it is genealogically only remotely related to either variety 'R\OH $ FRPSDULVRQ RI EDVLF ZRUG OLVWV RI /0+/HZRLQJX /0+6RORU DQG /0+ Lamalera with Alorese renders lexical similarity percentages of Alorese versus these three RWKHUYDULHWLHVWKDWUDQJHEHWZHHQDQG .ODPHU 7KLVVXJJHVWV WKDW$ORUHVHLVOH[LFDOO\GLVWLQFWHQRXJKWREHTXDOL¿HGDVDODQJXDJHRILWVRZQ,QDGGLWLRQ VLJQL¿FDQWPRUSKRORJLFDOGLIIHUHQFHVH[LVWEHWZHHQ/DPDKRORWDQG$ORUHVH VHHVHFWLRQ EHORZ DQG .ODPHU )RU WKHVH UHDVRQV WKH FXUUHQW SDSHU FRQVLGHUV$ORUHVH D ODQJXDJHRQLWVRZQDQGGLIIHUHQWIURP/DPDKRORWLQDQ\RILWVYDULHWLHVOLVWHGDERYH ,QWKLVSDSHU,¿UVWFRPSDUHWKHV\QWD[PRUSKRORJ\DQGEDVLFYRFDEXODU\RI$ORUHVH ZLWK/DPDKRORWLQVHFWLRQVDQGIROORZHGE\DFRPSDULVRQZLWKLWVQRQ$XVWURQHVLDQ QHLJKERXUVLQVHFWLRQ)RUWKHV\QWDFWLFFRPSDULVRQ$ORUHVHZLOOEHFRQWUDVWHGPRVWO\ with the LMH-Lewoingu variety, as 1LVKL\DPD .HOHQ KHQFHIRUWK1 . LVWRGDWHWKHRQO\SXEOLVKHGVRXUFHRQD/DPDKRORWYDULHW\WKDWFRQWDLQVV\QWDFWLFGHWDLOV (When possessive structures are compared I also refer to LMH- Lamalera, as Keraf 1978 $EEUHYLDWLRQV $/ DOLHQDEOH ',67 GLVWDO ),1 ¿QDO )2& IRFXV ,1$/ LQDOLHQDEOH ,1' ,QGRQHVLDQ /2& ORFDWLRQ 1(* QHJDWLRQ 2%/ REOLTXH 3/ SOXUDO 3266 SRVVHVVRU 35) SHUIHFWLYH5'3 UHGXSOLFDWLRQ5($/ UHDOLV6(4 VHTXHQWLDO6* VLQJXODU MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 75 FRQWDLQVLQIRUPDWLRQRQWKLVWRSLF )RUWKHPRUSKRORJLFDOFRPSDULVRQ$ORUHVHZLOOEH FRQWUDVWHGZLWKERWK/0+/HZRLQJX 1 . DQG/0+/DPDOHUD .HUDI )RU the comparison of the Alorese lexicon and syntax with its non-Austronesian Alor-Pantar neighbours, I refer to the Alor Pantar Lexical Database (listed as such in the references) and IRUWKHJUDPPDWLFDOFRQVWUXFWLRQV,SUHVHQWSXEOLVKHGDQGXQSXEOLVKHG¿HOGGDWDFROOHFWHG E\FROOHDJXHVDQGP\VHOIDVLQGLFDWHGLQWKHWH[W 7KHVWUXFWXUHRIWKHSDSHULVDVIROORZV,QVHFWLRQ,identify a number of ‘Papuan’ features found in both Alorese and Lamaholot, and investigate what these suggest about WKHVKDUHGKLVWRU\RIWKHWZRODQJXDJHV,QVHFWLRQ,LQYHVWLJDWHWRZKDWH[WHQW$ORUHVH and Lamaholot are syntactically or morphologically different, and what these differences VXJJHVWDERXWWKHKLVWRU\RI$ORUHVHDIWHULWVSOLWIURP/DPDKRORW,QVHFWLRQ,LQYHVWLJDWH some lexical and syntactic changes that occurred after its speakers settled on Pantar and Alor, by comparing the Alorese lexicon and grammar with the lexicon and grammar of LWV QRQ$XVWURQHVLDQ QHLJKERXUV ,Q VHFWLRQ , SUHVHQW VRPH QRWHV RQ WKH KLVWRU\ DQG ethnography of the Alorese speakers and in section 6, I summarize the reconstruction of the history of the Alorese language and its speakers, and suggest a scenario how it developed LQWRWKHODQJXDJHLWLVWRGD\ 2. PAPUAN FEATURES IN ALORESE AND LAMAHOLOT 2.1. INTRODUCTION. The term ‘Papuan’ is often used to refer to the perhaps 800 languages spoken in New Guinea and its vicinity that do not belong to the Austronesian language IDPLO\3 In this paper I use ‘Papuan’ to refer to languages that are not Austronesian and DUHVSRNHQLQHDVWHUQ,QGRQHVLD7KH3DSXDQODQJXDJHVVSRNHQLQWKH$ORUDUFKLSHODJR just north of Timor are geographically closest to the Lamaholot speaking region, and will WKHUHIRUHEHIRFXVVHGRQLQWKHGLVFXVVLRQRIµ3DSXDQ¶IHDWXUHVLQWKLVVHFWLRQ7KH$ORU 3DQWDUODQJXDJHVIRUPDFORVHNQLWIDPLO\ +ROWRQHWDO DQGDUHLQWXUQUHODWHGWRWKH non-Austronesian languages of Timor and Kisar, with whom they form the Timor-Alor- 3DQWDUIDPLO\ 6FKDSSHUHWDO $KLJKHURUGHUDI¿OLDWLRQRIWKH7LPRU$ORU3DQWDU IDPLO\WRDQRWKHU3DSXDQJURXSFDQQRWEHHVWDEOLVKHG +ROWRQHWDO5RELQVRQ +ROWRQ DOWKRXJKDORQJVWDQGLQJDVVXPSWLRQEHJLQQLQJZLWK:XUPHWDO KDVLWWKDWWKH7LPRU$ORU3DQWDUODQJXDJHVEHORQJWRWKH7UDQV1HZ*XLQHDIDPLO\ The non-Austronesian populations in eastern Indonesia must have predated the arrival RIWKH$XVWURQHVLDQVSHDNHUV FI3DZOH\ 5RVV EXWWKHUHLVQRUHDVRQ to assume that Papuan languages spoken in eastern Indonesia today descend from a single SUHKLVWRULFJURXS,WLVIDUPRUHSODXVLEOHWKDWWKH\GHULYHIURPDFRPSOH[PL[RISUHKLVWRULF SRSXODWLRQVDQGYDULRXVZDYHVRILPPLJUDQWV Over the past decade a body of literature has appeared which argued for the relevance of certain particular structural features in the typological characterization of the languages of HDVWHUQ,QGRQHVLD VHHDOVR5HHVLQN 'XQQWKLVYROXPH ,QWKH$XVWURQHVLDQODQJXDJHV RI WKLV DUHD FHUWDLQ IHDWXUHV DUH FRQVLGHUHG WR UHSUHVHQW D µ3DSXDQ¶ LQÀXHQFH HJ WKH H[LVWHQFHRIDSRVWSUHGLFDWHQHJDWRU5HHVLQN VHH)ORUH\IRUDPRGL¿FDWLRQ 3 As in Tryon (1995:3): “The term ‘Papuan’ is a convenient term for the non-Austronesian languages RI3DSXD1HZ*XLQHDDQGHDVWHUQ,QGRQHVLDQRWDOORIZKLFKDUHGHPRQVWUDEO\UHODWHG´ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 76 ZKLOHRWKHUIHDWXUHVIRXQGLQ3DSXDQODQJXDJHVDUHVXJJHVWLYHRI$XVWURQHVLDQLQÀXHQFH HJYHUEREMHFWRUGHUFRUUHODWLQJZLWKWKHW\SLFDOKHDGLQLWLDOSKUDVHVWUXFWXUHIRXQGLQ $XVWURQHVLDQODQJXDJHV &ODUN 7U\RQ :RUNVGLVFXVVLQJ$XVWURQHVLDQ3DSXDQ contact in eastern Indonesia proposing features that diffused as the result of this contact LQFOXGH*ULPHV 5HHVLQN .ODPHU 'RQRKXH +LPPHOPDQQ .ODPHU5HHVLQN 9DQ6WDGHQ DQG.ODPHU (ZLQJ 7KHFXUUHQWVHFWLRQLGHQWL¿HVDQXPEHURIIHDWXUHVWKDWDUHSDUWRIWKH$XVWURQHVLDQ languages Alorese and Lamaholot, but at the same time are generally recognized as features that are typical for a ‘Papuan’ language, not an Austronesian one (in the general sense of µ3DSXDQ¶LQWKHVRXUFHVMXVWPHQWLRQHG ,LQYHVWLJDWHZKDWWKHSUHVHQFHRIWKHVHIHDWXUHV VXJJHVW DERXW WKH KLVWRU\ RI WKHVH ODQJXDJHV +LJKOLJKWHG IHDWXUHV DUH SRVWSUHGLFDWH QHJDWLRQ VHFWLRQ WKHPDUNLQJRISRVVHVVRUV VHFWLRQ WKHQRXQORFDWLRQDORUGHULQ ORFDWLYHFRQVWUXFWLRQV VHFWLRQ WKHSUHVHQFHRIDIRFXVSDUWLFOH VHFWLRQ DQGWKH DEVHQFHRIDSDVVLYHYHUEIRUPDQGFRQVWUXFWLRQ VHFWLRQ 7KHUHVXOWVDUHVXPPDULVHG DQGGLVFXVVHGLQVHFWLRQ7KH3DSXDQODQJXDJHVFORVHVWWR/DPDKRORWDUHWKH$ORU 3DQWDUODQJXDJHVVSRNHQRQZHVW3DQWDUVHH¿JXUH7KH3DSXDQIHDWXUHVGLVFXVVHGLQWKH following sections will therefore be illustrated with examples from languages spoken on 3DQWDU 7HLZD%ODJDU$GDQJ6DUDQG.DHUD,WLVKRZHYHULPSRUWDQWWREHDULQPLQGWKDW LQWKH/DPDKRORWVSHDNLQJUHJLRQLWVHOIQR3DSXDQODQJXDJHLVFXUUHQWO\VSRNHQ 2.2. POST-PREDICATE NEGATION. The canonical Austronesian position for negations is to precede the predicate, but in the Papuan languages in the Alor Pantar region it follows the predicate, as illustrated for Teiwa in ( (1) Na iman ga-pak-an iman suk-an maan. TEI 1SG they 3SG -call- REAL 3PL H[LWFRPHGRZQ REAL NEG µ,FDOOHGWKHP>EXW@WKH\GLGQ¶WFRPHRXW¶ .ODPHU %RWK$ORUHVH $ORU DQG/DPDKRORW /0+ DOVRKDYHD¿QDOµSRVWSUHGLFDWH¶QHJDWLRQ DVVKRZQLQ DQG Akhirnya, kujo ha no nele n-ei tobo kaha lang Alor ¿QDOO\ IND) crab this 3SG crawl 3SG-go sit FRFRQXWVKHOO under ‘Finally, this crawled to sit underneath a coconut shell mu no pana ha n-ei tahi lahe. SEQ 3SG walk this 3SG-go sea NEG WKHQKHGLGQRWJRWRWKHVHD>DJDLQ@¶ (3) Go EΩULQ na hala’. LMH I hit him NEG µ,GRQ¶WKLWKLP¶ 1 . The Alorese negator lahe is a metathetised form of the Lamaholot negator hala’ found in MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 77 /0+/HZRLQJX/0+/HZROHPDDQG/0+6RORU,5 2.3. POSSESSIVE MARKING. In the nominal domain, three Papuan features relating to SRVVHVVLYHVWUXFWXUHVDUHUHOHYDQW L WKHUHSODFHPHQWRISRVVHVVLYHVXI¿[HVE\SRVVHVVRU pronouns that precede the possessed noun, (ii) the marking of distinct classes of alienable DQGLQDOLHQDEOHQRXQVDQG LLL WKHUHODWLYHRUGHURISRVVHVVRUDQGSRVVHVVHH 5HSODFLQJ SRVVHVVLYH VXI¿[HV E\ SUHQRPLQDO SRVVHVVRU SURQRXQV In Papuan languages, possessors typically precede the possessed, and the person and number features RIDSRVVHVVRUDUHHQFRGHGDVDSUH¿[RQWKHQRXQDVLOOXVWUDWHGIRU7HLZDLQ ): ) Rai ga-yaf TEI king 3SG-house ‘The king’s house’ The possessor pronouns of Alorese and LMH-Lewoingu and LMH-Lamalera are given in ( ,Q/0+/HZRLQJXDQG/0+/DPDOHUDSRVVHVVRUVDUHHQFRGHGDVVXI¿[HVRUDVIUHH pronouns following the possessee, as illustrated in ( $ORUHVHKDVQRSRVVHVVLYHVXI¿[HV and uses a free possessor pronoun, which precedes the possessee, as illustrated in ( 6HH DOVRVHFWLRQ 3URQRXQVDQGDI¿[HVWRHQFRGHSRVVHVVRUV6,7 Alorese LMH-Lewoingu LMH-Lamalera 1 . (Keraf 1978:85-95) 1SG go go’en NԥQ goe -k, -ka 2SG mo mo’en -ko moe -m, ma 3SG.AL ni6 na’en QԥQ nae non-segmental7 3SG.INAL no na’en QԥQ nae QRVXI¿[ &¿QDOVWHP QRQVHJPHQWDO 9¿QDO stem) 1PL.EXCL kame kame’en NԥQ kame -kem 1PL.INCL ite tete’en -te tite -te 2PL mi mion -ke mio -kre, re 3PL fe / fereng ra’en -ka rae -ri Identical metathesis patterns occur in other words; compare Alorese mareng ‘night’ with LMH- Lewolema remã, LMH-Lewoingu UϷPDQ; and Alorese kamore ‘rat’ with LMH-Lewolema NΩURPH LMH-Lewoingu NΩURPH 5 The LMH-Lamalera negation listed in Keraf (1978) is take. This word functions in LMH- /HZRLQJXDVQHJDWLYHDQVZHUµQR¶ 6 Alternative pronunciation ne 7 UGSHUVRQSRVVHVVRUVXI¿[HVGLIIHUIRUVWHPVHQGLQJLQDFRQVRQDQWRULQDYRZHO,QDOLHQDEOH QRXQV HQGLQJ LQ D FRQVRQDQW KDYH QR VXI¿[ )RU DOO WKH RWKHU VWHPV UG SHUVRQ VLQJXODU possessor features are expressed as lengthening of the stem vowel and/or consonant, and/or vowel MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 78 (6) D /DQJRNΩQ LMH house-1SG µ0\KRXVH¶ 1 . E Lango go’en house 1SG µ0\KRXVH¶ 1 . (7) Mato kete ni ning anang labi. Alor frog that 3SG POSS child many ‘That frog has many children’ or ‘That frog’s children are many’ 2.3.2. Marked distinction between alienable and inalienable nouns%RWK/DPDKRORWDQG $ORUHVHKDYHDPDUNHGGLVWLQFWLRQEHWZHHQDOLHQDEOHDQGLQDOLHQDEOHQRXQV7KLVGLVWLQFWLRQ is not a typical feature of the Austronesian family as a whole, although it is found in some $XVWURQHVLDQODQJXDJHVRIHDVWHUQ,QGRQHVLD VHH.ODPHUIRUH[DPSOHV 7KH3DSXDQ ODQJXDJHVRI$ORU3DQWDUDOOPDUNWKHGLVWLQFWLRQ,Q%ODJDUIRULQVWDQFHLQDOLHQDEOHVKDYH DQ REOLJDWRU\ SRVVHVVRUSUH¿[(a), while alienables have a free possessor pronoun (b): (8) D N-amal E Ne quu Blagar 1SG.INAL-voice 1SG.AL tuber ‘My voice’ ‘My tuber’ (Steinhauer 1993:150-151) ,Q/0+/DPDOHUDWKHGLVWLQFWLRQLVDOVRPDUNHGWKLVWLPHE\WKHREOLJDWRU\YVRSWLRQDO XVHRIDSRVVHVVRUPRUSKHPH LQDOLHQDEOHQRXQVPXVWDOZD\VKDYHDSRVVHVVRUVXI¿[ZKLOH DOLHQDEOHQRXQVFDQRFFXUZLWKRXWDSRVVHVVRU%RWKLQDOLHQDEOHDQGDOLHQDEOHSRVVHVVLRQ are expressed by the same morphemes, except for the 3rd person singular possessor, as shown in the rightmost column of ( DERYH In Alorese and in LMH-Lewoingu, inalienable possession is expressed by a dedicated VXI¿[ WKDW DWWDFKHV WR ERG\ SDUW QRXQV ,Q$ORUHVH WKH IRVVLOL]HG VXI¿[ LV D URRW¿QDO FRQVRQDQW±ng >ƾ@,Q/0+/HZRLQJXLWLV–(’V)n >ݦ9Q@8 Examples are given in (9); most RIWKHIRUPVLQ$ORUDQG/DPDKRORWDUHFRJQDWHV5HFRQVWUXFWHG3URWR&HQWUDO0DOD\R 3RO\QHVLDQ 3&03 DQG 3URWR0DOD\R 3RO\QHVLDQ IRUPV DUH LQFOXGHG IRU FRPSDULVRQ9 QDVDOL]DWLRQDQGRUVWUHVVVKLIW VHH.HUDI IRUGHWDLOV 8 7KH9VWDQGVIRUDQ\YRZHO GHSHQGLQJRQWKHRSHQFORVHQHVVRIWKH¿QDOURRWV\OODEOHWKH¿QDO YRZHORIWKHURRWLVFRSLHGDVVXI¿[YRZHO 9 Central Malayo Polynesian (CMP) and Eastern Malayo Polynesian (EMP) languages together form the Central Eastern Malayo Polynesian (CEMP) subgroup, a daughter node of Proto Malayo- 3RO\QHVLDQ 303 ZKLFKLQWXUQLVDGDXJKWHUQRGHRI3URWR$XVWURQHVLDQ 3$1 303LQFOXGHV DOOWKHODQJXDJHVRI,QGRQHVLD7KH&03QRGH RUµOLQNDJH¶ ZDVSURSRVHGE\%OXVW DQG /DPDKRORWLVDVVXPHGWREHDI¿OLDWHGWRLW7KHH[LVWHQFHRIWKH&03QRGHLVWKHWRSLFRIDGHEDWH 'RQRKXH *ULPHV%OXVW E ZKLFK,ZLOOQRWJRLQWRKHUH7KHSURWRIRUPVFLWHG KHUHDUHWDNHQIURPWKHRQOLQH$XVWURQHVLDQ%DVLF9RFDEXODU\'DWDEDVH *UHHQKLOO%OXVW *UD\ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 79 0DQ\PRGHUQ$ORUDQG/DPDKRORWZRUGVGRQRWUHÀHFWWKHVHSURWRIRUPVEXWWKRVHWKDWGR VXFKDVµKDQGDUP¶µPRXWK¶DQGµH\H¶ FRQWDLQDQRQHW\PRORJLFDO¿QDOQDVDO+RZHYHU the body part nouns in ( G GRFRQWDLQDQHW\PRORJLFDO¿QDOQDVDO,Q/0+WKHVXI¿[LV optional (9a), obligatory (9b), or absent ( F 7KLVVXJJHVWVWKDWLQ/0+WKHVXI¿[GLGQRW OH[LFDOL]HUHJXODUO\,Q$ORUHVHWKHVXI¿[KDVEHHQFRPSOHWHO\OH[LFDOLVHG (9) %RG\SDUWQRXQVZLWK IRVVLOL]HG SRVVHVVLYHVXI¿[HVLQ$ORUDQG/0+ Lewoingu Alor LMH-Lewoingu PCMP PMP Meaning 1 . D limang lima(n) *lima >TD@OLPD ‘hand/arm’ fofang ZԥZD n) EDEDT EDTEDT ‘mouth’ ratang rata(n) EXT TXOX *buhek ‘hair’ fuling wuli(n) (no data) OLTH5 ‘neck’ E kotung NRWԥn TXOX TXOX ‘head’ aleng kola’an *mudi *likud ‘back’ leing lein *wai TDTD\ ‘foot, leg’ F matang mata *mata *mata ‘eye’ fefeleng wewel O ԥD PD GLODT ‘tongue’ G tilung tilun *talinga ‘ear’ nirung irun *(i/u)jung ‘nose’ ulong ipԥ(’ԥn) QLSԥn *(n)ipen ‘tooth’ The modern LMH-/HZRLQJXSRVVHVVRUVXI¿[ OLVWHGLQ(5) above) is in complementary GLVWULEXWLRQZLWKWKHIRVVLOL]HGVXI¿[ LQDOLHQDEOH VXI¿[±n in ( D 7KLVLVVKRZQE\WKH pair (DE DGDSWHGIURP1 . ,WLVQRWSRVVLEOHWRFRPELQHERWKVXI¿[HV (F 1RWHDOVRWKDWWKHIRVVLOL]HGQDVDOVXI¿[KDVQRWEHHQLQWHJUDWHGFRPSOHWHO\LQWRWKH nominal root form: it can attach to the adjective and have scope over the nominal phrase, compare (GH ZKLFKOLVWVWKHVRXUFHDXWKRUDV%OXVW MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 80 (10) D mata-n ‘eye’ LMH eye-POSS E PDWDNΩQ ‘my eye’ eye-1SGPOSS F *mata-n-kΩn eye-POSS-1SGPOSS G PDWDEHOΩ ‘big eye’ eye big H PDWDEHOΩQ ‘big eye’ eye big-POSS In sum, Alorese and LMH-Lewoingu both distinguish inalienable body part nouns from DOLHQDEOHQRXQVE\WKHSUHVHQFHRID¿QDOYHODUQDVDOVXI¿[,Q/0+/HZRLQJXWKHQDWXUH RIWKLVHOHPHQWYDULHVEHWZHHQDVXI¿[DQGDFOLWLFDQGLWPD\EHUHSODFHGE\DPRGHUQ SRVVHVVRUVXI¿[,Q$ORUHVHKRZHYHULWLVDFRPSOHWHO\DQGUHJXODUO\IRVVLOL]HG¿QDOURRW FRQVRQDQW10,Q/0+/DPDOHUDLQDOLHQDEOHQRXQVODFNDSRVVHVVRUVXI¿[HQWLUHO\RUKDYH DQRQVHJPHQWDOSRVVHVVRU Unlike any of the LMH varieties, an additional strategy has been innovated in Alorese to mark the alienable-inalienable distinction by choice of free pronoun: alienable nouns take ni as 3SG possessor, while inalienable nouns take no7KLVLVLOOXVWUDWHGLQ( D ni uma E no amang Alor 3SG. AL house 3SGINAL father ‘his house’ ‘his father’ I consider ni as cognate with LMH 3rd singular possessor pronouns na’en / nae, while no is an innovation (possibly harmonizing the vowel with the vowels in 1st singular go and nd singular mo) as a dedicated form to mark a 3SGLQDOLHQDEOHSRVVHVVRU 2.3.3. Possessor-possessed order. The third non-Austronesian feature in the nominal GRPDLQLVWKHUHODWLYHRUGHURISRVVHVVRUDQGSRVVHVVHGLQ$ORUHVHDQG/DPDKRORW7KH 3DSXDQ RUGHU >SRVVHVVRUSRVVHVVHG@ VHH ) above) is the reverse of the [possessed- SRVVHVVRU@RUGHUW\SLFDOO\IRXQGLQ$XVWURQHVLDQODQJXDJHVIRULQVWDQFH,QGRQHVLDQrumah- ku ‘house-1SG¶µP\KRXVH¶ In LMH-Lamalera, a possessor may be expressed as a free pronoun and replace the SRVVHVVRUVXI¿[ .HUDI $IUHHSRVVHVVRUSURQRXQIROORZVWKHSRVVHVVHGUHQGHULQJ WKHRUGHU>SRVVHVVHGSRVVHVVRU@DVLQlango goe ‘house 1SG’µP\KRXVH¶ .HUDI ,QRWKHUZRUGV/0+/DPDOHUDFRQVLVWHQWO\GLVSOD\VWKH$XVWURQHVLDQRUGHU 10 This analysis also implies that not all inalienables end in a velar nasal, as only those inalienable nouns whose historical root ends in a vowel could take the ng DVVXI¿[ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 81 %\FRQWUDVW$ORUHVHRQO\DOORZVWKHUHYHUVHG>SRVVHVVRUSRVVHVVHG@RUGHUDVLOOXVWUDWHG in ( ,IWKHSRVVHVVRULVH[SUHVVHGDVDSURSHUQDPHDVLQ(13), it must be accompanied E\DSURQRXQDQGERWKQDPHDQGSRVVHVVRUSURQRXQSUHFHGHWKHSRVVHVVHG7KH$ORUHVH RUGHUWKXVPLUURUVWKHSRVVHVVRUSRVVHVVHGRUGHURI3DSXDQODQJXDJHVDVH[HPSOL¿HGE\ Teiwa in ( D Ni uma Alor 3SG.AL house ‘his house’ E *uma ni; *uma-ni; *uma=ni house 3SG.POSS (13) [Bapa John ni uma@ being. Alor Mr John 3SG. AL house big µ0U-RKQ¶VKRXVH>LV@ELJ¶ [Kri John ga-yaf@ uwaad. TEI Mr John 3SG-house big µ0U-RKQ¶VKRXVH>LV@ELJ¶ .ODPHUQG The position between LMH-Lamalera and Alorese is taken up by LMH-Lewoingu, which allows either order of possessor and possessee, and employs free possessor pronouns DVZHOODVSRVVHVVRUDI¿[HV7KH$XVWURQHVLDQ>SRVVHVVHGSRVVHVVRU@RUGHULVWKHXQPDUNHG RUGHU LQ /0+/HZRLQJX FI 1 . DQG LV LOOXVWUDWHG LQ ( 9DULRXVNLQGV of possessors may follow the possessed noun: free possessor pronouns (15a), possessor VXI¿[HV(15b), or lexical possessors (F $VXI¿[DQGIUHHSRVVHVVRUFDQQRWFRRFFXUDV shown in (G ZKLFKVXJJHVWVWKDWWKH\KDYHWKHVDPHUHIHUHQWLDOIXQFWLRQ2QWKHRWKHU hand, a nominal and a pronominal possessor can co-occur, as shown in (H (15) D Lango go’en LMH house 1SG µ0\KRXVH¶ 1 . E /DQJRNΩQ house-1SG.POSS µ0\KRXVH¶ 1 . F Lango guru house teacher µ$WHDFKHU¶VKRXVH¶ 1 . G /DQJRNΩQ go’en house-1SG.POSS 1SG.POSS MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 82 H Lango guru na’en house teacher 3SG µ7KHWHDFKHU¶VKRXVH¶ 1 . ,Q DGGLWLRQ WR WKH >SRVVHVVHGSRVVHVVRU@ RUGHU /0+/HZRLQJX DOVR H[KLELWV WKH µUHYHUVHG¶>SRVVHVVRUSRVVHVVHG@RUGHU7KLVRUGHULVXVHGZKHQWKHSRVVHVVRULVHQFRGHG DVDVXI¿[DQGWKH13FRQWDLQVDFRUHIHUHQWLDOQRXQ,QWKDWFDVHWKHQRXQLVSUHSRVHGDV illustrated in (16): (16) D Guru lango-nԥn LMH teacher house-3SG.POSS µ$WHDFKHU¶VKRXVH¶ 1 . E Guru lango-ka teacher house-3PL.POSS ‘The teachers’ (PL KRXVH V IDFXOW\UHVLGHQFH¶ 1 . Of the two possessor marking strategies, the free possessor pronoun is more regular DQGSURGXFWLYHLQ/0+/HZRLQJXWKDQWKHSRVVHVVRUVXI¿[)RUH[DPSOH1 . QRWHWKDWVRPH/DPDKRORWVSHDNHUVFDQQRWXVHSRVVHVVRUVXI¿[HVZLWKZRUGVOLNHoto ‘car’ and bapa µIDWKHU¶/RDQZRUGV OLNHoto DQGIUHTXHQWO\XVHGZRUGV OLNHbapa) thus appear WRSUHIHUIUHHSRVVHVVRUVWRERXQGRQHV11 This suggest a development where the possessor VXI¿[LQJVWUDWHJ\LVORVLQJJURXQGWRWKHIUHHSURQRXQVWUDWHJ\LQ/0+/HZRLQJX In conclusion, the Lamaholot varieties and Alorese share some Papuan structural features LQ WKH SRVVHVVLYH GRPDLQ )LUVW LQ /0+/HZRLQJX D prenominal possessor pronoun VWUDWHJ\LVUHSODFLQJSRVVHVVLYHVXI¿[LQJ7KLVFKDQJHKDVEHHQ¿QDOLVHGLQ$ORUHVHZKLFK KDVRQO\IUHHSRVVHVVRUSURQRXQVOHIW%RWKODQJXDJHVPDUNLQDOLHQDEOHERG\SDUWQRXQVDV DGLVWLQFWFODVVE\PHDQVRIDIRVVLOL]HGQDVDOVXI¿[ DQG$ORUHVHLQQRYDWHGDQDGGLWLRQDO dedicated 3rd person singular inalienable possessor pronoun no %RWK$ORUHVHDQG/0+ /HZRLQJX EXWQRW/0+/DPDOHUD VKRZWKH>SRVVHVVRUSRVVHVVHG@RUGHUWKDWLVW\SLFDO IRU3DSXDQODQJXDJHV,Q/0+/HZRLQJXWKLVLVDPDUNHGRUGHUZKLOHLQ$ORUHVHLWLVWKH RQO\RUGHUDOORZHG7KH3DSXDQIHDWXUHVZKLFKDUHSUHVHQWLQWKH/DPDKRORWYDULHWLHVDQG LQ$ORUHVHKDYHWKXVGHYHORSHGWRDIXUWKHUVWDJHLQ$ORUHVH 2.4. [NOUN-LOCATIONAL] ORDER IN LOCATIVE EXPRESSIONS. In Alorese and LMH- Lewoingu, locative expressions are constructed of a noun, followed by a locational lexeme RIQRPLQDORULJLQ ZKLFKPD\IXQFWLRQDVSRVWSRVLWLRQLQFHUWDLQFRQWH[WV $QH[DPSOHDUH the locational nouns unung ‘inside’ (Alor), illustrated in (17), and ono’on ‘inside’ (LMH), 11 7KLVH[SODQDWLRQGLIIHUVIURPWKHRQHVXJJHVWHGE\1 . ZKRUHIHUWRoto as a “less familiar” word, and bapa DVD³UHVSHFWIXONLQVKLSWHUP´7RFKDUDFWHULVHWKHVHZRUGVDVVXFKGRHV not seem to be true to fact: oto is a loan from Indonesian (which borrowed it from Dutch auto < French 1897 auto µFDU¶ DQGLVNQRZQWRHYHU\RQHBapa ‘father; Mr’ is not only a kinship term EXWDOVRXVHGIUHTXHQWO\DVWKHSROLWHWHUPRIDGGUHVVIRUPDOHDGXOWV FI,QGRQHVLDQBapak µ0U¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 83 illustrated in ( (17) Pa ru oro uma unung? Alor what FOC LOC house inside ‘What is in(side) the house?’ (18) Busan to’u pe dos ono’on. LMH cat one at box inside µ7KHUHLVDFDWLQWKHER[¶ 1 . Both unung and ono’on are cognate to ’oné LQ.HRVSRNHQLQ&HQWUDO)ORUHV %DLUG ,Q.HRWKLVOH[HPHLVV\QFKURQLFDOO\DSUHSRVLWLRQ,Q,QGRQHVLDQWRRORFDWLRQDO QRXQVRFFXULQSUHQRPLQDOSRVLWLRQ FI,QGRQHVLDQdi dalam rumah ‘LOC inside house’ versus *di rumah dalamµ/2&KRXVHLQVLGH¶ ,QOLQHZLWKWKHVHREVHUYDWLRQV,DVVXPH that the position of the lexeme unung/ono’on in the Austronesian languages of Flores was originally prenominal, and that it moved to postnominal position in Lamaholot and Alorese EHFDXVHRI3DSXDQLQÀXHQFH$7HLZDH[DPSOHRID3DSXDQQRXQORFDWLRQDOQRXQRUGHULV given in (19), where the locational noun gom ‘its inside’ follows the noun yaf: (19) Na [yaf g-om@ ma gi. TEI 1SG house 3SG-inside LOC go µ,JRLQVLGHWKHKRXVH¶/LWµ,JRLQWR>WKHKRXVH¶VLQVLGH@¶ .ODPHUQG 2.5. FOCUS PARTICLE. Alorese and LMH-Lewoingu both have an information structure particle, ru and ke respectively. 7KLV SDUWLFOH IXQFWLRQV WR PDUN FRQWUDVWLYH IRFXV 7KH contrast between an unfocused constituent and a focused one in Alorese is illustrated in (DE DQRWKHULOOXVWUDWLRQLV( D No lelang batang. Alor 3SG make break µ+HEURNH>WKHP@¶ E No ru lelang batang. 3SG FOC make break µ+(EURNH>WKHP@ QRWPH ¶ No maring aleng keleng maring mo ru hela. Alor 3SG say back slender say 2SG FOC climb µ+HVDLGWR6OHQGHU%DFN ³<28FOLPELW´>QRW,@¶ The particle ke marks contrastive focus in Lamaholot, as illustrated in ( MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 84 D Go-ke hope buku pi’in. LMH 1SG-FOC buy book this µ,W¶VPHZKRERXJKWWKLVERRN¶ 1 . E Go hope-ke buku pi’in. 1SG buy-FOC book this µ,%28*+7WKLVERRN¶ 1 . F Go hope buku pi’in-ke. I buy book this-FOC µ,ERXJKW7+,6%22.¶ 1 . Many Papuan languages have particles marking contrastive focus; an illustration from a Pantar language is Teiwa la ‘FOC’, illustrated in ): (3) D Rai na-soi ga-kamadal ga-buxun tas. TEI king 1SG-order 3SG-belt 3SG-guard stand µ7KHNLQJRUGHUHGPHWRJXDUGKLVEHOW¶ .ODPHU E Rai la na-soi ga-kamadal ga-buxun tas. king FOC 1SG-order 3SG-belt 3SG-guard stand µ7KH.,1*RUGHUHGPHWRJXDUGKLVEHOW¶ F Rai [na la@ soi ga-kamadal ga-buxun tas. king 1SG FOC order 3SG-belt 3SG-guard stand µ,ZDVRUGHUHGE\WKHNLQJWRJXDUGKLVEHOW¶ Focus particles encode new information, and are typically followed by propositions WKDWDUHSUDJPDWLFDOO\SUHVXSSRVHG,QPDQ\ODQJXDJHVUHODWLYHFODXVHVDUHLQVWUXPHQWDOLQ FRGLQJSUHVXSSRVHGSURSRVLWLRQV7KHIRFXVPDUNHUWKXVIXQFWLRQVLQDZD\WKDWLVVLPLODU WRDUHODWLYHFODXVHPDUNHU,WLVSODXVLEOHWKDWEHFDXVHWKH\KDYHDIRFXVPDUNHU$ORUHVH DQG/DPDKRORWODFNDGHGLFDWHGLQGLJHQRXVUHODWLYHFODXVHFRQVWUXFWLRQ8QGHULQÀXHQFH of Indonesian, however, both languages have borrowed a relative clause construction that is marked with Indonesian yang µUHODWLYHPDUNHU¶%RUURZHGyang is used optionally, in DGGLWLRQWRWKHIRFXVPDUNHU VHH1 . 2.6. ABSENCE OF A PASSIVE VOICE VERB AND CONSTRUCTION. A passive construction is GH¿QHGKHUHDVDFODXVHZKHUHWKHYHUEFDUULHVVSHFLDOPRUSKRORJ\WRPDUNWKHSURPRWLRQRI the verb’s underlying patient argument to become the grammatical subject, while demoting WKHRULJLQDODJHQWVXEMHFWLQWRDQREOLTXHSKUDVH While the languages of Taiwan and the Philippines have fully developed systems with more than two voices, the western Malayo-Polynesian languages of Indonesia usually MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 85 KDYHWZR 5RVV ,QHDVWHUQ,QGRQHVLDWKLVYRLFHV\VWHPLVUHGXFHG and many ODQJXDJHVODFNERWKSDVVLYHPRUSKRORJ\DQGDGHGLFDWHGSDVVLYHFRQVWUXFWLRQ([DPSOHV LQFOXGH7DED$OXQH/HWL5RWL7HWXQ)HKDQ%LPD.DPEHUDDQG.HR FI.ODPHU In the Papuan languages of Alor and Pantar a passive is also generally lacking; examples LQFOXGH.ORQ %DLUG $EXL .UDWRFKYtO DQG7HLZD .ODPHUD ,Q7HLZD WKHIXQFWLRQDOHTXLYDOHQWRIDSDVVLYHLVDFODXVHZLWKDIURQWHG3IROORZHGE\DJHQHULF noun hala ‘someone, unknown person’ expressing the (backgrounded) Agent; compare ( ZLWKEDVLF$39FRQVWLWXHQWRUGHUZLWK( ZLWK3$9RUGHUDQG$JHQWhala: ( P A 9 TEI Uy ga’an yivar ga-far. person that dog 3SG-kill µ7KDWSHUVRQNLOOHGDGRJGRJV¶ (5) P A 9 TEI Uy ga’an hala ga-far. person that someone 3SG-kill µ7KDWSHUVRQZDVNLOOHG¶ OLWµ7KDWSHUVRQVRPHRQHNLOOHG¶ $ORUHVH DQG /DPDKRORW WRR ODFN D SDVVLYH 1 . 1DJD\D %RWK ODQJXDJHVKDYHEDVLF$JHQW9HUE3DWLHQW $93 FRQVWLWXHQWRUGHUDVLQ ) and $ IXQFWLRQDOHTXLYDOHQWWRDGHGLFDWHGSDVVLYHLVWKHIURQWLQJRI3DVLQ( DQG( 6) A 9 P Alor Ama kali g-ang fata. father that 3SG-eat rice µ7KDWPDQHDWVULFH¶ P A 9 Alor Ume ape g-ang mungga. house ¿UH 3SG-eat while µ7KHKRXVHLVRQ¿UH¶ (8) A 9 P LMH Na KΩER ana’ pe’en. 3SG bathe child the µ6KHEDWKHVWKHFKLOG¶ 1 . 1RWHWKDWWKHZHVWDQGFHQWUHRI,QGRQHVLDDUHPRUHYDULHJDWHG LQSDUWLFXODU%RUQHRDQG6XODZHVL MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 86 (9) P A 9 LMH Nolo pe’en tahan tite gΩta hala’ past that rice we harvest NEG µ,QWKHSDVWULFHZDVQ¶WDFURS¶ OLWµZHGLGQ¶WKDUYHVWULFH¶ 1 . In neither of the languages does the fronting of P involve a change in the verbal PRUSKRORJ\ QRUGRHVWKHRULJLQDO$EHFRPHSDUWRIDQREOLTXHFRQVWLWXHQWDQGDOOWKH QRPLQDOFRQVWLWXHQWVUHWDLQWKHLURULJLQDOVKDSH,QVXP/DPDKRORWDQG$ORUHVHODFNWKH passive constructions and voice morphology found in most of the western Austronesian ODQJXDJHVZKLFKDUHVLPLODUO\ODFNLQJLQWKH3DSXDQODQJXDJHVRI$ORUDQG3DQWDU 2.7. SUMMARY AND DISCUSSION. Lamaholot and Alorese share a number of features that are atypical for Austronesian languages in general, but do exist in Papuan languages of the region: they lack a passive, place the negation in post-predicate position, have [possessor- SRVVHVVHG@RUGHUDIRUPDOGLVWLQFWLRQEHWZHHQDOLHQDEOHDQGLQDOLHQDEOH ERG\SDUW QRXQV D>QRXQORFDWLRQDOQRXQ@RUGHULQORFDWLYHH[SUHVVLRQVDQGDIRFXVSDUWLFOH The hypothesis I submit is that these features arose in Lamaholot and Alorese as a result RI LQWHQVLYH FRQWDFW ZLWK RQH RU PRUH 3DSXDQ ODQJXDJHV$V VLPLODU VWUXFWXUDO IHDWXUHV arose in both Lamaholot and Alorese, I assume that they did not arise independently, but ZHUH SDUW RI WKHLU VKDUHG DQFHVWRU ODQJXDJH 3URWR/DPDKRORW 7KLV LPSOLHV WKDW PRVW of the Papuan features found in today’s Alorese are not due to contact with its current Papuan neighbors on Pantar and Alor, but rather entered the language before it split from /DPDKRORW No written or oral records exist of a history of contact between Lamaholot speakers DQG VSHDNHUV RI D 3DSXDQ ODQJXDJH V 1HLWKHU GR ZULWWHQ RU RUDO UHFRUGV H[LVW RI 3DSXDQODQJXDJHVVSRNHQLQHDVW)ORUHVZKHUH/DPDKRORWLVVSRNHQWRGD\13 However, there is general consensus among linguists that Papuan (non-Austronesian) populations predated the Austronesians, who arrived in the eastern Indonesian region some 3,500 years DJR 3DZOH\5RVV'RQRKXH *ULPHV(ZLQJ .ODPHU 7KH Papuan structural features I have reconstructed here for Proto-Lamaholot constitute further evidence that Austronesian and Papuan speakers were once in contact in the Lamaholot KRPHODQG7KLVKRPHODQGPD\KDYHEHHQDQ\ORFDWLRQZHVWRI3DQWDU LWFRXOGKDYHEHHQ 6RORU/HPEDWDDQGRUHDVW)ORUHVEXWDOVRDQRWKHUORFDWLRQ VHHVHFWLRQ 13 $OWKRXJK 'RQRKXH DUJXHV WKDW H[WLQFW 7DPERUD ZDV D 3DSXDQ ODQJXDJH VSRNHQ RQ 6XPEDZDZHVWRI)ORUHVLVODQG While the Lamaholot speakers currently live in east Flores, Solor and Lembata, the homeland RI3URWR/DPDKRORWFRXOGDOVRKDYHEHHQVRPHZKHUHHOVH$VRQHUHYLHZHUUHPDUNHGWKHRUDO traditions of most communities in East Flores record that they originally came from elsewhere, DOWKRXJKLWUHPDLQVXQFOHDUIURPZKHUHH[DFWO\ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 87 3. CONTRASTING LAMAHOLOT AND ALORESE 3.1. INTRODUCTION. This section investigates to what extent Alorese and Lamaholot DUHGLIIHUHQWV\QWDFWLFDOO\ VHFWLRQ RUPRUSKRORJLFDOO\ VHFWLRQ DQGZKDWWKHVH GLIIHUHQFHVVXJJHVWDERXWWKHKLVWRU\RIWKH$ORUHVH VHFWLRQ 3.2. SYNTACTIC DIFFERENCES. The syntactic differences between Alorese and Lamaholot DUHPLQLPDO)LUVWO\WKHRUGHURI>SRVVHVVRUSRVVHVVHG@LVDPDUNHGRUGHULQ/DPDKRORW ZKLOHLWLVWKH¿[HGRUGHULQ$ORUHVH WKLVZDVGLVFXVVHGLQVHFWLRQ6HFRQGO\/DPDKRORW KDV RQO\ FODXVHLQLWLDO FRQMXQFWLRQV HJ NΩGLQ in (30), while Alorese has at least one FRQMXQFWLRQOLNHHOHPHQWWKDWLVFODXVH¿QDOWKHVHTXHQWLDOPDUNHUmu in ( (30) Na VΩED laran nΩnΩn ga’e nΩOΩ bisa ai topi pe’en. LMH 3SG search way how so can get hat the µ6KHZRQGHUHGKRZWRJHWWKDWKDW¶ NΩGLQ Mince mari hi topi pe’en mΩNR pe. then Mince say ah hat that ugly that µ7KHQ0LQFHVDLG³$KWKDWKDWLVXJO\´¶ 1 . (31) Tiba-tiba aho ning kotung maso toples unung mu, Alor suddenly (IND) dog POSS head enter jar inside SEQ ‘Suddenly the dog’s head got into the jar then no goka oro tana lulung. 3SG fall LOC earth on µKHIHOORQWKHJURXQG¶ 7KLUGO\WLPHH[SUHVVLRQVIROORZWKHSUHGLFDWHLQ/DPDKRORWDQGSUHFHGHLWLQ$ORUHVH This is illustrated with the cognate forms ZLD¿DQJ ‘yesterday’ in ( ( 7KH,QGRQHVLDQ H[DPSOHLQ LOOXVWUDWHVWKHW\SLFDOKHDGLQLWLDORUGHUWKDWLVW\SLFDOIRUDQ$XVWURQHVLDQ ODQJXDJH7KLVLVWKHRUGHUIRXQGLQ/DPDKRORW( Ra sΩga wia. LMH they come yesterday µ7KH\FDPH\HVWHUGD\¶ 1 . (33) Ama kali ¿DQJ ho. Alor father that yesterday come µ7KDWPDQFDPH\HVWHUGD\¶ Mereka datang kemarin. IND they come yesterday µ7KH\FDPH\HVWHUGD\WRPRUURZ¶ In sum, I have not found evidence that the syntactic differences between Lamaholot and MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 88 $ORUHVHUHODWHWRPRUHWKDQMXVWDIHZVPDOOGLIIHUHQFHVLQZRUGRUGHU 3.3. MORPHOLOGICAL DIFFERENCES 3.3.1. Introduction. 0RVW$XVWURQHVLDQODQJXDJHVRIHDVWHUQ,QGRQHVLDDQGWKH3DFL¿F have morphological systems that are less elaborate than the Austronesian languages spoken LQ7DLZDQWKH3KLOOLSLQHVRUZHVWHUQ,QGRQHVLD FI%OXVW D 6RPHH[WUHPH morphological empoverishment is found in languages spoken in central and eastern Flores, LQFOXGLQJ 0DQJJDUDL 1JDGD /LR DQG .HR %DLUG +RZHYHU QRW DOO WKH )ORUHV ODQJXDJHVXQGHUZHQWVXFKPDVVLYHPRUSKRORJLFDOORVV/DPDKRORWEHLQJDFDVHLQSRLQW 7KHPRUSKRORJLFDOV\VWHPRI/DPDKRORWKDVSURGXFWLYHUHÀH[HVRIDVLJQL¿FDQWQXPEHU RI3URWR$XVWURQHVLDQ3URWR0DOD\R3RO\QHVLDQPRUSKHPHV,QWKLVVHFWLRQ,SUHVHQW DVXPPDU\RI/DPDKRORWLQÀHFWLRQDOPRUSKRORJ\FRPSDUHGWR$ORUHVH VHFWLRQ /DPDKRORWGHULYDWLRQDOPRUSKRORJ\FRPSDUHGWR$ORUHVH VHFWLRQ IROORZHGE\D VXPPDU\DQGGLVFXVVLRQ VHFWLRQ ,QÀHFWLRQDOPRUSKRORJ\/DPDKRORWKDVTXLWHDORWRIDJUHHPHQWPRUSKRORJ\ subject (A and S) agreement is marked on verbs, adverbs as well as on the conjunctive element o’on ‘and, with’,15 while adjectives agree in person and number with the (pro)noun WKH\PRGLI\ 1 . 7KHUH DUH WZR GLIIHUHQW VXEMHFW SDUDGLJPV RQH LV D VHW RI FRQVRQDQWDO SUH¿[HV WKH RWKHUDVHWRIVXI¿[HVDVJLYHQLQ( /0+/HZRLQJXDQG/0+/DPDOHUDXVHWKHVDPH $SUH¿[HVEXWGLIIHUHQW6VXI¿[HV%HORZ,GLVFXVVVXEMHFWPDUNLQJLQ/0+/HZRLQJX similar (though not identical) observations can be made for LMH-Lamalera, which is not GLVFXVVHGKHUHIRUUHDVRQVRIVSDFH VHH.HUDI 6XEMHFWDI¿[HVLQ/DPDKRORW $SUH¿[ 66XI¿[ LMH- Lewoingu LMH-Lamalera 1 . (Keraf 1978:73,76) 1SG k- NԥQ -ka 2SG m- -ko, -no16 -ko, -o 3SG n- -na,QԥQ -fa/ra, -a 1PL.EXCL m- NԥQ -kem 1PL.INCL t- -te -te 2PL m- -ke/-ne -kre, -re 3PL r- -ka -ri, i ,Q/0+/HZRLQJXWKH$SUH¿[REOLJDWRULO\PDUNVWKHDJHQW $ RIYRZHOLQLWLDOWUDQVLWLYH YHUEV 1 . ([DPSOHVLQFOXGHa’an ‘make’, LWΩ ‘sleep with’, -olin ‘improve’ 1 . +RZHYHU WKHUH DUH DOVR YRZHOLQLWLDO YHUEV ZKLFK FDQQRW WDNH DQ 15 7KLVVXJJHVWVWKDWWKLVHOHPHQWPD\EHDQDO\]HGDVDYHUEUDWKHUWKDQDFRQMXQFWLRQ 16 1 .OLVWERWKIRUPVRQSEXWRQO\-koRQS MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 89 DJUHHPHQWSUH¿[ HJRSΩQµOLHWRVRPHRQH¶1 . VRWKDWWKHXVHRIWKH$SUH¿[ LVQRWSXUHO\SKRQRORJLFDOO\FRQGLWLRQHGEXWDOVROH[LFDOO\VWLSXODWHG Many Lamahalot verbs can be used both transitively and intransitively with no GLIIHUHQFHLQYHUEIRUP 1 . :KHQWKH\DUHXVHGLQDWUDQVLWLYHFRQVWUXFWLRQ$ and P are expressed as free noun phrases; when they occur in an intransitive construction, 6LVHQFRGHGDVDYHUEDOVXI¿[ 1 . 17$Q6VXI¿[LVDOVRIRXQGRQ adjectives in predicative or adverbial function, in which case the adjective gets an excessive LQWHUSUHWDWLRQ 1 . ,Q VXP /DPDKRORW 6 DQG$ DUH RIWHQ H[SUHVVHG DV SURQRPLQDO DI¿[HV RQ YHUEV ,Q contrast to this, verbal arguments in Alorese are almost universally expressed as free SURQRXQV([FHSWLRQVDUHDIHZIUHTXHQWYHUEVZLWKDIRVVLOLVHG$SUH¿[WKDWDUHXVHGLQ FRPELQDWLRQZLWKDQ REOLJDWRU\ IUHHVXEMHFWSURQRXQ([DPSOHVSRLQWHGRXWWRPHE\ speakers are -oing ‘to know’ and -enung ‘to drink’, as in go g-oing ‘I 1SG-know’ and mo m-enung µ\RXSG-GULQN¶ 3.3.3. Derivational morphology. /0+/HZRLQJXKDVVHYHQGHULYDWLRQDODI¿[IRUPV as listed in ( /0+/DPDODUDKDVVL[GHULYDWLRQDODI¿[HVDVOLVWHGLQ( 7KHOLVWV VXPPDUL]HWKHGHULYDWLRQVDQGWKHLUVHPDQWLFVSUHVHQWHGLQ1 . 6RPH RI WKH /0+ GHULYDWLRQDO DI¿[HV DUH UHJXODU DQG SURGXFWLYH ZKLOH RWKHUV DUH OH[LFDOLVHGWRDVPDOORUODUJHH[WHQW2IWHQDVLQJOHSUH¿[KDVGHYHORSHGPRUHWKDQRQH PHDQLQJ ,Q DOO FDVHV WKH VHPDQWLF UHODWLRQ EHWZHHQ WKH EDVH DQG WKH GHULYHG IRUP LV WUDQVSDUHQWHQRXJKWRHVWDEOLVKDWOHDVWDJHQHULFFRUHPHDQLQJRIWKHGHULYDWLRQDOPRUSKHPH 1RWHWKDWWKHPDQ\QRPLQDOL]LQJSUH¿[HVGHULYHGLIIHUHQWVHPDQWLFW\SHVRIQRPLQDOVDQG ,UHIHUWRWKHRULJLQDOVRXUFHVIRUDGGLWLRQDOGHVFULSWLYHGHWDLOVDERXWLQGLYLGXDOGHULYDWLRQV Anticipating a reconstruction of Proto-Lamaholot morphology, I provide the possible PAN 303DI¿[HVDORQJVLGHWKHLUPRGHUQ/DPDKRORWUHÀH[HVDVDK\SRWKHVLVDERXWWKHOLNHO\ HW\PRORJLFDOUHODWLRQEHWZHHQWKHP (36) Derivational morphology in LMH-Lewoingu 3UH¿[be(C)- ‘nominalizer’,18 e.g. linonµUHÀHFW¶!be-linonµPLUURU¶ 1 . 303 SDƾµLQVWUXPHQWDOQRXQ¶RU SD5µGHYHUEDOQRXQ¶ %OXVW D 3UH¿[pΩ- µYHUEDOL]HU¶HJtua ‘palm wine’ > SΩ-tuak ‘taste like palm wine’ (N&K 303 SDNDµWUHDWOLNH;¶ %OXVW D 3UH¿[ Sԥ- µQRPLQDOL]HU¶ HJ tutu’ ‘speak’, SΩ-nutu’ ‘speaker, speaking’ (N&K ) < PMP SD5µGHYHUEDOQRXQ¶ %OXVW D 3UH¿[NΩHJpasa ‘swear’ > NΩ-pasa ‘oath’ ‘nominalizer’ (1 . 303 NDµIRUPDWLYHIRUDEVWUDFWQRXQV¶ %OXVW D ,Q¿[ΩQµQRPLQDOL]HU¶HJtali ‘add’ > WΩQDOLµDGGHGWKLQJ¶ 1 . 3$1 XPµ$FWRUYRLFH¶ %OXVW D 17 There are also intransitive verbs that cannot be used as transitives, and they express S as a free QRXQSKUDVH 1 . 18 1 . UHIHUWRWKLVSUH¿[DVbeN- which is realised as b-, be’-, ben- or ber- MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 90 3UH¿[PΩQ‘nominalizer’,19HJba’at ‘heavy’ > PΩQD¶DW ‘something heavy’ (N&K 3$1 PDµVWDWLYH¶ %OXVW D 3UH¿[ JΩ1 ‘nominalizer’, HJ balik ‘to return’ > JΩZDOLN ‘return (N)’ (N&K 303 NDµDEVWUDFWQRXQIRUPDWLYH¶ %OXVW D &RQVRQDQWUHSODFHPHQWHJpet ‘bind’ > metµEHOW¶µUHVXOWQRPLQDOL]HU¶ 1 . 3$1 PDµVWDWLYH¶ %OXVW D (37) Derivational morphology in LMH-Lamalera 3UH¿[ b-/be- µGHYHUEDO QRPLQDOL]HU¶ HJ udur ‘push’ > b-udur ‘pusher’ (Keraf 1978:188), doru ‘rub’ > be-doru ‘someone rubbing’ (Keraf 1978:193); fai ‘water’ > be- faiµKDYHZDWHU¶ .HUDI < 303 SDƾµLQVWUXPHQWDOQRXQ¶RU SD5µGHYHUEDO QRXQ¶ %OXVW D 3UH¿[ n- µGHYHUEDO QRPLQDOL]HU¶ HJ hau ‘sew’ > nau ‘something sewn’ (Keraf XQFOHDUHW\PRORJ\ &LUFXP¿[pΩ-k, HJtana ‘earth’ > pe-tana-k ‘taste like earth’ .HUDI ) < PMP SDNDµWUHDWOLNH;¶ %OXVW D ,Q¿[-en- µLQVWUXPHQWDOQRPLQDOL]HU¶HJtika ‘divide’ > t-en-ika ‘instrument to divide’ .HUDI 3$1 XPµ$FWRUYRLFH¶ %OXVW D 3UH¿[me-µQRPLQDOL]HU¶HJnange ‘swim’ > me-nange ‘swimmer’ (Keraf 1978:197) < PAN *ma ‘stative’ %OXVW D &RQVRQDQWUHSODFHPHQWHJpota ‘add’ > mota ‘addition’ ‘result nominalizer’ (Keraf 3$1 PDµVWDWLYH¶ LELG ,Q FRQWUDVW WR /DPDKRORW$ORUHVH KDV QR GHULYDWLRQDO PRUSKRORJ\ DW DOO 7KH RQO\ productive word formation process in Alorese is reduplication: verbs and adverbs undergo full reduplication to indicate iterative or intensive activity, as in (38); while nominal UHGXSOLFDWLRQVGHQRWHSOXUDOGLYHUVLW\µDOOVRUWVRI1¶6LPLODUUHGXSOLFDWLRQWDNHVSODFHLQ /DPDKRORW (38) No geki-geki sampai no neing aleng bola. Alor 3SG RDP-laugh until (IND) 3SG POSS back break µ+HODXJKHGDQGODXJKHGWLOOKLVEDFNEURNH¶ The loss of derivational morphological categories in Alorese can be seen as a kind of IRUPDOVLPSOL¿FDWLRQRUUHJXODUL]DWLRQ DI¿[HVWKDWGRQRWVKRZDUHJXODUDQGWUDQVSDUHQW IRUPPHDQLQJUHODWLRQVKLSDUHORVW 3.4. CONCLUSIONS. :KLOH PRGHUQ UHÀH[HV RI 3$1 303 PRUSKRORJ\ DSSHDU LQ abundance in the Lamaholot varieties, and the Lamaholot varieties do not show a gradual decline of morphology that is related to a geographical West-East spread, Alorese has ORVWDOORILWVPRUSKRORJ\$VPRUSKHPHVDUHPRUHHDVLO\ORVWWKDQJDLQHG,DVVXPHWKDW 19 :LWKQRQKRPRUJDQLFQDVDOL]DWLRQRILQLWLDOURRWFRQVRQDQW WKHSURFHVVPD\LQYROYHH[WUD¿QDO QDVDORUV\OODEOH VHH1 . 7KHQDVDOLQWKHSUH¿[FKDQJHVSE!PE!ZK!QDQGLVXQUHDOL]HGEHIRUHUO MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 91 Proto-Lamaholot, the shared ancestor of Alorese and Lamaholot, had at least the amount RIPRUSKRORJ\RIWRGD\¶V/DPDKRORWYDULHWLHV7KLVLPSOLHVWKDW3URWR/DPDKRORW L KDG VXEMHFW DQG SRVVHVVRU DI¿[HV LL GLVWLQJXLVKHG DJUHHPHQW RI$ SUH¿[ DQG 6 VXI¿[ DQG LLL KDG DW OHDVW VHYHQ GLIIHUHQW GHULYDWLRQDO SUH¿[HV$IWHU WKH /DPDKRORW$ORUHVH VSOLW$ORUHVHORVWDOORIWKLVPRUSKRORJ\6XFKPDVVLYHUHGXFWLRQRIPRUSKRORJ\LVRIWHQ taken to suggest that a language has gone through a stage of imperfect or second language OHDUQLQJ 4. ALORESE AND ITS PAPUAN NEIGHBORS 4.1. INTRODUCTION. This section investigates the lexical and syntactic change that occurred in Alorese after it split from Lamaholot, by comparing the Alorese lexicon and JUDPPDU ZLWK WKH OH[LFRQ DQG JUDPPDU RI WKH 3DSXDQ ODQJXDJHV LQ WKH QHLJKERUKRRG /H[LFDO ERUURZLQJ LV LQYHVWLJDWHG LQ VHFWLRQ IROORZHG E\ D V\QWDFWLF FRPSDULVRQ IRFXVVLQJRQWKHH[SUHVVLRQRIWKUHHW\SHVRISUHGLFDWHDUJXPHQWUHODWLRQVLQVHFWLRQ 4.2. LEXICAL COMPARISON. In order to estimate the amount of lexical borrowing in $ORUHVH , FRPSDUHG D LWHP EDVLF ZRUG OLVWV RI$ORUHVH ZLWK SXEOLVKHG OH[LFDO GDWD IURP /0+/DPDOHUD /0+/HZRLQJX /0+6RORU /0+/HZROHPD I focussed on WKH$ORUHVHZRUGVWKDWDUHIRUPDOO\GLVVLPLODUWRWKHLUVHPDQWLFHTXLYDOHQWLQDOOIRXURI WKH/DPDKRORWYDULHWLHV)LIW\¿YHVXFKZRUGVZHUHIRXQG7KUHHRIWKHVHDUHUHÀH[HVRI an Austronesian or Proto Malayo-Polynesian word (which has not been retained in the /DPDKRORWYDULHWLHV 7KHUHPDLQLQJZRUGVLQZKLFK$ORUHVHGLIIHUVIURPDQ\/DPDKRORW YDULHW\FRXOGEHOH[LFDOLQQRYDWLRQVRUORDQZRUGV2IWKHVHZRUGVZHUHLGHQWL¿HGDV loans from an Alor Pantar language (see ( DUH0DOD\,QGRQHVLDQORDQV VHH DQGKDYHDQXQNQRZQHW\PRORJ\RUVRXUFH7KHGRQRUODQJXDJHRIWKHLGHQWL¿DEOH ORDQVZDVIRXQGWKURXJKWKH$ORU3DQWDU/H[LFDO'DWDEDVHZKLFKFRQWDLQV EDVLF OH[LFDOGDWDIURP3DSXDQYDULHWLHVRIWKH$ORU3DQWDUIDPLO\ +ROWRQHWDO )RU comparison, words of the source language(s), LMH-Lamalera, LMH-Lewoingu and PMP DUHLQFOXGHGLQ As the focus of this article is on the changes that took place in Alorese, I do not investigate lexical ERUURZLQJLQWKH/DPDKRORWYDULHWLHV'R\OH SUHVHQWVDQLQLWLDOFRPSLODWLRQDQGDQDO\VLV RIFRPSDUDWLYHOH[LFDOGDWDRIWKH/DPDKRORWYDULHWLHV 7KHLWHPVIRU3URWR0DOD\R3RO\QHVLDQ 303 DUHIURPWKHRQOLQH$XVWURQHVLDQ%DVLF9RFDEXODU\ 'DWDEDVH *UHHQKLOO6-5%OXVW 5'*UD\ ZKLFKOLVWVWKHVRXUFHDXWKRUDV%OXVW MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 92 $ORUHVHZRUGVZLWKLGHQWL¿HG$ORU3DQWDUGRQRUODQJXDJH V LMH- LMH- Alorese Meaning ———— Source ———— Lamalera Lewoingu PMP tor ‘road’ tor W Pantar larã laran *zalan baling ‘axe’ bali W Pantar, Sar hepe soru no data duri ‘knife’ duir Adang hepe hepe no data NRQGݤR ‘clothing’ kondo Blagar alelolo no data no data bire kari ‘children’ biar kariman Teiwa ana DQDݦ *anak KDݦm ‘this’ ƫDݦD Teiwa pi SLSLݦQ *i ni kar-to ‘ten’ Proto AP 5HÀH[HVDFURVV pulo pulo, no data kar-ua ‘twenty’ TDU AP pulu rua PDEDVHT ele ‘wet’ TDORݦ Sar sԥ’nԥbԥ GԥPDQ kalok Teiwa xolo Kaera kari ‘thin’ kira Blagar, Pԥ¶QLSL Pԥ¶QLSL *ma-nipis Kaera, Teiwa laming ‘to wash’ laming W Pantar ba, pu baha no data kalita ‘dirty’ NOLWDݦ Teiwa milã milan *cemed klitak Blagar tobang ‘to push’ tobung Kaera uruk gehan no data doho ‘to rub’ dahok Blagar doru GRVXݦ no data $ORUHVHZRUGVERUURZHGIURP,QGRQHVLDQ0DOD\ Alorese Meaning ————— Source ————— UHNLƾ ‘to count’ reken Malay < Dutch kali ‘river’ kali Malay/Indonesian danau ‘lake’ danau Indonesian EXƾD µÀRZHU¶ EXƾD Indonesian hati ‘liver’ hati Indonesian 7KHGDWDLQ VXJJHVWWKUHHWKLQJV)LUVW$ORUHVHERUURZHGZRUGVIURP$ORU3DQWDU languages from right across the island of Pantar: Teiwa and Sar are spoken in the north- west, Western Pantar is spoken in the west and south, and Blagar and Kaera in the east, see ¿JXUH7KDWDOOWKHVHGRQRUODQJXDJHVDUHVSRNHQRQ3DQWDULVQRVXUSULVHJLYHQWKDWWKH $ORUHVHZRUGOLVWLQYHVWLJDWHGKHUHLVIURPWKH%DUDQXVDGLDOHFWVSRNHQLQZHVW3DQWDU 6HH6FKDSSHU .ODPHU PV Compare Kupang Malay rekenµWRFRXQW¶ -DFRE *ULPHV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 93 FIGURE 2/DQJXDJHVIURPWKH$ORU3DQWDUIDPLO\WKDWDUHGLVFXVVHGLQWKHWH[W Second, among the Alor-Pantar donor languages, there is not one that is particularly GRPLQDQW 7KLV VXJJHVWV WKDW FRQWDFWV RI D VLPLODU NLQG H[LVWHG ZLWK GLIIHUHQW VSHHFK FRPPXQLWLHVUDWKHUWKDQZLWKRQHLQSDUWLFXODU 7KLUGRIDOOWKHGRQRUODQJXDJHV0DOD\,QGRQHVLDQDSSHDUVWKHPRVWGRPLQDQWRQH 7KLVLVH[SHFWHGRIDQDWLRQDOODQJXDJHXVHGLQHGXFDWLRQDQGLQWHUHWKQLFFRPPXQLFDWLRQ 4.3. SYNTACTIC COMPARISON. Alorese and its Alor-Pantar neighbours have a different constituent order: in Alorese the verb precedes the object, as in (), while the AP languages DUHDOOYHUE¿QDODVLOOXVWUDWHGIRU7HLZDLQ (1) Aho gaki be kae kali. Alor dog bite child small that µ$GRJELWWKDWFKLOG¶ () Yivar bif waal ga-sii. TEI dog child WKDWPHQWLRQHG 3SG-bite µ$GRJELWWKDWFKLOG¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 94 The expression of predicate-argument relations is an area where Papuan and Austronesian ODQJXDJHVRIWHQVKRZJUDPPDWLFDOFRQWUDVWV,QWKH$3ODQJXDJHVVHULDOYHUEFRQVWUXFWLRQV are pervasive, and do much of the work that is done in (western) Austronesian languages HLWKHUE\FDXVDWLYHDSSOLFDWLYHRUUHFLSLHQWDI¿[HVRQYHUEV FI+LPPHOPDQQ RUE\DGSRVLWLRQDOSKUDVHV In this section, I compare a small part of the syntax of Alorese with its AP neighbors, to see if there is evidence of syntactic convergence with local Papuan languages after Alorese VSOLWIURP/DPDKRORW$VDSUHOLPLQDU\LQYHVWLJDWLRQ,FRQVLGHUWKUHHW\SHVRISUHGLFDWH argument relations: ‘give’ events with an agent (A), recipient (R), and a displaced theme 7 VHFWLRQ LQVWUXPHQWDOFRQVWUXFWLRQVZLWKDQ$SDWLHQW 3 DQGLQVWUXPHQW , VHFWLRQ DQGFDXVDWLYHFRQVWUXFWLRQVZKHUHDQRULJLQDO6EHFRPHVWKHFDXVHH 3 DQG DQHZFDXVHU $ LVLQWURGXFHG VHFWLRQ Alorese is compared with languages spoken in its vicinity: Teiwa (west Pantar), Kaera (east Pantar), Sar (central-west Pantar), Blagar (east Pantar, Pura, Tereweng), and Adang (spoken north of Kalabahi on the Alor peninsula),VHH¿JXUH&RQWDFWEHWZHHQ$ORUHVH and the Papuan languages spoken in south, central or east Alor is much less plausible, so WKHVHODQJXDJHVZLOOQRWEHFRQVLGHUHGKHUH 4.3.1. ‘Give’ constructions. In the Papuan languages investigated here, the verb ‘give’ is a mono-transitive verb which has the Recipient (R) as its single object, while T is introduced ZLWKLWVRZQSUHGLFDWH DYHUERUD GHYHUEDO REOLTXHSDUWLFOH7KHFRQVWLWXHQWRUGHULVµ$7 5JLYH¶DQGWKHSURQRPLQDOSUH¿[RQµJLYH¶HQFRGHVSHUVRQDQGQXPEHURI5([DPSOHV DUH ( GDWDDUHP\RZQ¿HOGQRWHVXQOHVVLQGLFDWHGRWKHUZLVH ) Uy ga’an u sen ma n-oma’ g-an. TEI person 3SG DIST money OBL 1SG-father 3SG-give µ7KDWSHUVRQJDYHPRQH\WRP\IDWKHU¶ () Ui fo seng ma na-manak g-an. Sar person that money OBL 1SG-father 3SG-give µ7KDWSHUVRQJDYHPRQH\WRP\IDWKHU¶ %DLUGVXUYH\GDWD () Ui gu gang doi mi na-mam g-eng. Kaera person that 3SG money OBL 1SG-father 3SG-give µ7KDWSHUVRQJDYHPRQH\WRP\IDWKHU¶ 6DU GDWD DUH IURP D VXUYH\ FDUULHG RXW E\ /RXLVH %DLUG LQ %ODJDU GDWD DUH IURP +HLQ 6WHLQKDXHUSF7HLZDDQG.DHUDGDWDDUHP\RZQ¿HOGQRWHV $GDQJGDWDDUH IURP+DDQ XQOHVVLQGLFDWHGRWKHUZLVH For further data and discussion of the typology and history of the ‘give’ construction in Timor Alor 3DQWDUVHH.ODPHU 6FKDSSHU MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 95 (6) Na vet nu metma n-oҌal ҌQDQJ Blagar 1SG coconut one OBL 1SG-child 3SG-give µ,JLYHDFRFRQXWWRP\FKLOG¶ 6WHLQKDXHUSF ) Ella seng med ҌRPDQJ ҌHQ Adang Ella money take 3-father 3-give µ(OODJDYHPRQH\WRKHUIDWKHU¶ +DDQ In contrast to these, an Alorese ‘give’ construction employs a ditransitive verb with two EDUHREMHFW13VZLWKFRQVWLWXHQWRUGHUµ$JLYH57¶DVLOOXVWUDWHGLQ Ama kali ning go bapa seng. Alor man that JLYH WR 1SG father money µ7KDWPDQJDYHP\IDWKHUPRQH\¶ In most Austronesian languages, ‘give’ events involve just a single verb which may be D PRUSKRORJLFDOO\ VLPSOH RU GHULYHG IRUP DQG ERWK REMHFWV IROORZ WKH YHUE ,I RQH RI WKHREMHFWVLVSDUWRIDQREOLTXHFRQVWLWXHQWLWLV57KLVLVDOVRWKHSDWWHUQDWWHVWHGLQ /DPDKRORW ZKHUH D EDUH GRXEOH REMHFW FRQVWUXFWLRQ OLNH LV SRVVLEOH DV ZHOO DV D FRQVWUXFWLRQZKHUH5LVDQREOLTXHFRQVWLWXHQW pe inawae to’u µWRDJLUO¶1 . Go nein inawae to’u bunga to’u. LMH 1SG give girl one ÀRZHU one µ,JDYHDJLUODÀRZHU¶ 1 . 4.3.2. Instrumental constructions. Instrumental expresssions involve an agent (A), SDWLHQW 3 DQGLQVWUXPHQW , ,QWKH3DSXDQODQJXDJHVFRPSDUHGKHUHLQVWUXPHQWVDUH either introduced in a serial verb construction with the verb ‘take’ or hold’, or with a GHYHUEDOREOLTXHSDUWLFOH,QVWUXPHQWVDOZD\VSUHFHGHWKHPDLQYHUE (50) Na ped mat ma man taxar. TEI 1SG machete take OBL grass cut µ,FXWWKHJUDVVZLWKDPDFKHWH¶ (51) Ui nuk peed ma tai gor. Sar person one machete OBL tree cut µ6RPHRQHFXWZRRGZLWKDPDFKHWH¶ %DLUGVXUYH\GDWD () Ui gu gang ped mi tei patak-o Kaera person that he machete OBL wood/tree cut- FIN µ7KDWSHUVRQFXWZRRGZLWKDPDFKHWH¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 96 (53) Na hemering medi-t sal ҌXWXNDQJ Blagar 1SG knife take-t rope 3SG-CAU-short µ,VKRUWHQWKHURSHZLWKDNQLIH¶ (5 Name nu sapad puin tiboҌ tatoҌ Adang person one machete hold wood cut µ6RPHRQHFXWZRRGZLWKDPDFKHWH¶ %DLUGVXUYH\GDWD Alorese, in contrast, marks instruments with the preposition nong ‘and, with’, in a prepositional phrase following the main verb, as in (55 (55) Ama to tari kaju nong peda. Alor father one FXWGRZQ wood with machete µ$PDQFXWWKHZRRGZLWKDPDFKHWH¶ Proto-Austronesian and Proto-Malayo Polynesian derived instrumental verbs with an LQVWUXPHQWDO LQ¿[ 7KHUH DUH DOVR PDQ\$XVWURQHVLDQ ODQJXDJHV ZKHUH LQVWUXPHQWV DUH HQFRGHG E\ DQ LQVWUXPHQWDO SUHSRVLWLRQDO SKUDVH RU DV FRPSOHPHQW RI WKH YHUE µXVH¶ /DPDKRORWHPSOR\VWKHODWWHUVWUDWHJ\ (56) Go EΩULQ Bala pake mΩnΩngo mi’in. LMH- 1SG hit Bala use stick this Lewoingu µ,KLW%DODXVLQJWKLVVWLFN¶ 1 . 4.3.3. Causative constructions. Highlighted here are causative constructions based on a non-active intransitive verb, whose original subject (S) becomes the causee (P) of the FDXVDWLYHFRQVWUXFWLRQZKLOHDQHZFDXVHUDJHQW $ LVLQWURGXFHG Two languages of Pantar (Teiwa, Sar) employ lexical causatives, as illustrated in (57) DQG (57) D Wat nuk ba’-an suk. TEI coconut one fall-REAL FRPHGRZQ µ$FRFRQXWIHOOGRZQ¶ E A wat u pua-n moxod-an gula’. 3SG coconut DIST snap-REAL drop-REAL ¿QLVK µ+HSLFNHGDQGGURSSHGWKDWFRFRQXW¶ LHKHKDGFOLPEHGWKH coconut tree) Teiwa also analytical causative constructions where the verb er ‘make’ takes P as its argument, as in (58 7KHUHIHUHQWRI3LVLGHQWLFDOWRWKHUHIHUHQWRIWKH6RIWKHIROORZLQJ 7KH%ODJDUFDXVDWLYHSUH¿[LVHLWKHUDFRS\RIWKH¿UVWVWHPYRZHORULWLVWKHYRZHOD)RU H[DPSOH WKHFDXVDWLYHRIWLDµVOHHS¶LVLWLDLQQRUWK%ODJDUDQGDWLDLQVRXWK%ODJDU MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 97 YHUE (58) Na motor er-an *(a) sig. TEI 1SG motorbike (IND) make- REAL 3SG live µ,VZLWFKRQWKHPRWRUELNH¶ OLWµ,PDNHWKHPRWRUELNHOLYH¶ 7KHOH[LFDOFDXVDWLYHRI6DULVLOOXVWUDWHGLQ 7KH3LVLQWURGXFHGDVDUJXPHQWRI the verb ma µFRPH¶ZKLFKLVSDUWRIDVHULDOYHUEFRQVWUXFWLRQ:KHWKHU6DUDOVRKDVDQ DQDO\WLFDOFDXVDWLYHOLNH7HLZDUHPDLQVWREHLQYHVWLJDWHG (59) D Wat fo baak. Sar coconut that fall µ7KDWFRFRQXWIHOO¶ %DLUGVXUYH\GDWD E A wat ma hod. 3SG coconut come drop ‘He drops coconuts’ ,Q.DHUDDFDXVDWLYHYHUELVGHULYHGE\VXI¿[LQJWKHLQWUDQVLWLYHYHUEZLWKDFDXVDWLYH VXI¿[ng7KHFDXVHHLVSUH¿[HGWRWKHGHULYHGYHUEDVLQ E (60) D Wat nuk ba sero. Kaera coconut one fall descend µ$FRFRQXWIHOOGRZQ¶ E Gang e-wat ga-ba-ng. 3SG 2SG-coconut 3SG-fall-CAU µ+HGURSV\RXUFRFRQXW¶ %ODJDUDQG$GDQJDOVRHPSOR\DFDXVDWLYHVXI¿[ ng in Blagar (61b), -ing in Adang ZKLOHWKH\DOVRKDYHDFDXVDWLYHSUH¿[7KHFDXVDWLYHSUH¿[FRQVLVWVRIDVLQJOH vowel (a- 7KHFDXVDWLYHYHUEPD\EHSUHFHGHGE\DQREMHFWSUH¿[HQFRGLQJWKHFDXVHH as illustrated for Blagar in (61b), and for Adang in ( :KLOHDOO$GDQJFDXVDWLYHVKDYHD SUH¿[QRWDOOKDYHVXI¿[HVDVLOOXVWUDWHGLQ(E IRUPRUHGLVFXVVLRQVHH+DDQ (61) D Vet ҌDQJX ba-t hera. Blagar coconut that fall-t down µ$FRFRQXWIHOOGRZQ¶ +HLQ6WHLQKDXHUSF E ҌDQD vet ҌDEDQJ 3SG coconut 3SG-CAU-fall-CAU µ+HGURSV D FRFRQXW¶ +HLQ6WHLQKDXHUSF MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 98 (6 D John ҌRO Adang John IDOORYHU µ-RKQIDOOVRYHU¶ E John na-ri DҌRO John 1SG-ACC CAUIDOORYHU µ-RKQPDGHPHIDOORYHU¶ +DDQ (63) Ella Ani ҌDPLKLQJDP Adang Ella Ani 3-CAU-sit-CAU- PRF µ(OODKDVPDGH$QLVLWGRZQ¶ +DDQ In contrast to the lexical and morphological causatives found in the AP languages discussed above, Alorese employs analytical causatives: one with the verb n(e)ing ‘give’ ( DQGRQHZLWKWKHYHUElelang ‘make’ ( (6 D Tapo goka. Alor coconut fall µ$FRFRQXWIHOO¶ E No neing goka mo tapo. 3SG give fall 2SG coconut µ+HGURSV\RXUFRFRQXW¶ (65) Mo lelang bola meja ni leing. Alor 2SG make break table POSS leg µ In Proto-Austronesian, a causative of dynamic verbs was marked with pa- and DFDXVDWLYHRIVWDWLYHYHUEVZLWKWKHSUH¿[HVpa-ka- %OXVW D 0DQ\PRGHUQ $XVWURQHVLDQODQJXDJHVVWLOOXVHUHÀH[HVRIpa-(ka-) VRPHWLPHVDORQJZLWKRWKHUDI¿[HV WRGHULYHFDXVDWLYHYHUEV+RZHYHUPDQ\PRGHUQODQJXDJHVDOVRXVHOH[LFDOFDXVDWLYHVRU periphrastic constructions with ‘give’, for example in spoken Indonesian and many Malay YDULHWLHV ,Q /DPDKRORW FDXVDWLYHV DUH H[SUHVVHG E\ DQDO\WLFDO FRQVWUXFWLRQV ZLWK nein µJLYH¶ 1 . RU–a’anµPDNH¶ 1 . LQSDWWHUQVLGHQWLFDOWRWKRVH IRXQGLQ$ORUHVH 4.3.4. Summary of syntactic comparison. The structural contrasts discussed above are represented in ( $ DJHQW7 GLVSODFHGWKHPH5 UHFLSLHQW, LQVWUXPHQW3 SDWLHQW (66) D Give construction ‘A gives T to R’ Teiwa A T OBL R give Sar A T OBL R give MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 99 Kaera A T OBL R give Blagar A T OBL R give Adang A T TAKE R give Alorese A give R T Lamaholot A give R T E Instrument ‘A cuts/shortens P with I’ Teiwa A I OBL P cut Sar A I OBL P cut Kaera A I OBL P cut Blagar A I take P shorten Adang A I hold P cut Alorese A cut P with I Lamaholot A cut P use I F &DXVDWLYHµ$FDXVHV3WR9¶ Lexical Morphological Analytical Teiwa yes >$3PDNH@>69@ Sar yes Kaera yes? 6XI¿[ Blagar yes? 3UH¿[ VXI¿[ Adang yes? 3UH¿[ VXI¿[ SUH¿[ Alorese >$JLYH93@ >$PDNH93@ Lamaholot >$JLYH93@ >$PDNH93@ While the AP languages compared here all express ‘give’ and instrument constructions LQDVLPLODUZD\$ORUHVHHPSOR\VGLIIHUHQWFRQVWUXFWLRQV,QWKHH[SUHVVLRQRIFDXVDWLYHV the AP languages show much internal variation, but the pattern used in Alorese does not DSSHDU WR EH UHODWHG WR DQ\ RI WKH FRQVWUXFWLRQV IRXQG LQ WKH$3 ODQJXDJHV 1RWH WKDW both Teiwa and Alorese have an analytic causative with ‘make’, but the word orders are GLIIHUHQW ,QJHQHUDO$ORUHVHGRHVQRWGLYHUJHIURPWKHSDWWHUQVIRXQGLQ/DPDKRORW In sum, the data presented here provide no evidence that Alorese borrowed grammatical constructions from its Papuan neighbors (and neither did the neighbors borrow from $ORUHVH $WWKHVDPHWLPHZH¿QGWKDWWKH$ORUHVHFRQVWUXFWLRQVDUHYLUWXDOO\LGHQWLFDO ZLWK/DPDKRORWVXJJHVWLQJWKDW$ORUHVHUHWDLQHGWKHV\QWD[RI/DPDKRORW 4.4. CONCLUSIONS. A comparison of the Alorese lexicon and grammar with the lexicon and grammar of neighbouring Papuan languages suggests: (i) Alorese borrowed a small set of words from the basic vocabulary of different AP Papuan languages across Pantar island, no language being more dominant the others; (ii) Alorese did not borrow any of the grammatical constructions to express ‘give’ events, instrumentals or causatives (and neither did the QHLJKERXUVERUURZIURP$ORUHVH ,QVWHDG$ORUHVHKDVUHWDLQHGWKHV\QWD[RI/DPDKRORW MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 100 5. HISTORICAL AND ETHNOGRAPHIC NOTES. In this section I summarize the historical and ethnographic evidence from which we may infer (i) that the speakers of Alorese moved away from the area where Lamaholot is spoken today (and not the other way around), and LL WKHGDWHEHIRUHZKLFKWKHVSOLWPXVWKDYHRFFXUUHG IQ$QRQ\PRXV a distinction is made between the mountain populations RI$ORUDQG3DQWDUDQGWKHSRSXODWLRQVRQWKHFRDVW7KHFRDVWDOSHRSOHDUHFRQVLGHUHGµQLHW LQKHHPVFK¶ µQRQLQGLJHQRXV¶S 7KHSDSHUDOVRUHSRUWVWKHORFDOOHJHQGWKDW3DQGDL LQQRUWKZHVW3DQWDUZDVWKH¿UVWFRDVWWREHSRSXODWHGE\WKHVHQRQLQGLJHQRXVFRDVWDO SHRSOH FIGURE 33DQWDULVODQGZLWKWKHORFDWLRQRI%DUDQXVD3DQGDL0XQDVHOLDQG$ORU Island with the town Alor Besar Today, Alorese speaking communities are only found in coastal areas of Alor and 7KHSURSRVHGGDWHLVQRWDQDEVROXWHGDWHEXWDµWHUPLQXVDQWHTXHP¶ WKHVSOLWPD\KDYHRFFXUUHG DQ\WLPHEHIRUHWKLVGDWH Major sources of this article were (i) the ‘Militaire Memories’ (reports on military expeditions that WRRNSODFHRQWKHLVODQGVLQ DQG DQG LL DUHSRUWRIDJHRORJLFDOH[SHGLWLRQE\5'0 9HUEHHNLQ SXEOLVKHG DVµ0ROXNNHQ9HUVODJ¶LQ-aarboek van het Mijnwezen in Ned. Oost-Indie. MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 101 3DQWDU7KH\DUHVHDRULHQWHGDQGIRUVXEVLVWHQFHWKH\WUDGLWLRQDOO\UHO\RQ¿VKLQJ WKH PHQ DQGZHDYLQJ WKHZRPHQ 7KH\FXUUHQWO\DGKHUHWRWKH,VODPUHOLJLRQ In contrast, speakers of the Papuan languages on Alor and Pantar are inland-oriented, have their traditional villages up in the mountains rather than the coast,30 adhere to the WUDGLWLRQDODQLPLVWUHOLJLRQRUDUH&KULVWLDQV7KH\DUHIDUPHUVDQGGRQRWUHO\RQ¿VKLQJ RUZHDYLQJIRUVXEVLVWHQFH 7UDGLWLRQDOO\WKHFRDVWDO$ORUHVHFODQVH[FKDQJHG¿VKDQGZRYHQFORWKIRUIRRGFURSV ZLWKWKH3DSXDQLQODQGSRSXODWLRQV FI$QRQ\PRXV 7KH$ORUHVHFODQV ZHUH DW OHDVW LQLWLDOO\ TXLWH VPDOO$V DQ H[DPSOH$QRQ\PRXV PHQWLRQV FODQVRISHRSOH$VQHZFRPLQJFODQVLQKDELWLQJFRDVWDOORFDWLRQVJHRJUDSKLFDOO\ remote from each other, many Alorese clans must have been outnumbered by their Papuan QHLJKERUVDQGLWLVSODXVLEOHWKDWWKH\DFTXLUHGWKHLUVSRXVHVIURPWKHH[DJDPRXV3DSXDQ FODQVLQWKHLULPPHGLDWHYLFLQLW\UDWKHUWKDQIURPWKH$ORUHVHFODQVWKDWZHUHPRUHUHPRWH31 $FFRUGLQJ WR D OHJHQG UHSRUWHG LQ$QRQ\PRXV D ³FRORQ\ RI -DYDQHVH´ settled on Pandai, in north west Pantar, some “500 to 600 years ago” [as the article appeared LQ WKLVZRXOGQRZEHWR\HDUVDJRLHWKHFRORQ\VHWWOHGRQ3DQGDLDURXQG $'@+RZHYHUWKHVDPHVRXUFHLQFOXGHVDIRRWQRWH S ZKLFKH[SODLQVWKDW the notion orang djawa OLWµ-DYDQHVHSHRSOH¶ DSSOLHVWRHYHU\RQHZKRFRPHVIURPRWKHU SDUWVRIWKHDUFKLSHODJR In other words, the “Javanese” coastal settlers mentioned in the OHJHQG ZHUH SHRSOH IURP ³RYHUVHDV´ EXW QRW QHFHVVDULO\ IURP -DYD ,QVWHDG , SURSRVH that the close linguistic and cultural ties between today’s Alorese and Lamaholot speakers suggest that the colony of orang djawa that settled on Pandai according to the legend were LQIDFW/DPDKRORWVSHDNHUVIURP)ORUHV6RORUDQGRUVRXWK/HPEDWD 7KHOHJHQGRIWKHIRXQGLQJRI3DQGDLLQQRUWK3DQWDUUHIHUUHGWRLQ$QRQ\PRXV is also reported in Lemoine (1969) and cited in later sources such as Barnes (1973:86, DQG 5RGHPHLHU ,W UHFRXQWV WKDW WZR -DYDQHVH EURWKHUV $NL $L DQG his younger brother Mojopahit, sailed to Pantar, where Aki Ai treacherously abandoned 0RMRSDKLW 0RMRSDKLW¶V GHVFHQGDQWV HYHQWXDOO\ FRORQL]HG 3DQGDL %DUDQXVD DQG $ORU %HVDU$VHFRQGOHJHQGLQ/HPRLQH UHFRXQWVRIDQRWKHUNLQJGRPRQ3DQWDUWKH NLQJGRPRI0XQDVHOLORFDWHGPRUHHDVWZDUGVRQWKHQRUWKHUQFRDVW,QWKHOHJHQG-DYDQHVH 30 $OWKRXJKPDQ\KDYHQRZPRYHGWRYLOODJHVRQWKHFRDVWIRUSUDFWLFDOSXUSRVHV 31 &ODQVH[FKDQJHGZLYHVEXWSHRSOHZHUHDOVRVROGRUJLYHQDZD\DVVODYHV)RUH[DPSOH7HLZD (north-west Pantar) has a word yu’al ZKLFKLVWUDQVODWHGDVµWRJLYHDZD\ SHRSOH ¶ FI7HLZD¶an ‘to sell’), and it refers to an “old custom” of “sending or giving away people that are useless to WKHFODQ´6SHDNHUVQRWHGWKDWIRUPHUO\\X¶DOZDVDOVRXVHGWRUHIHUWRVHOOLQJSHRSOH LQFOXGLQJ ZRPHQ WRWKH%DUDQXVDSHRSOH .ODPHUD IQ %DUDQXVDLVDQ$ORUHVHVSHDNLQJDUHD Compare Kambera (Sumba) tau Jawa µVWUDQJHU¶ OLW µ-DYDQHVH SHUVRQ¶ DQG tau Jawa bara µZHVWHUQHU¶ OLWµZKLWH-DYDQHVH¶ ZKHUHJawaDOVRGHQRWHVµVWUDQJHU¶ 2QYOHH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 102 immigrants who are allied to the kingdom of Pandai, kill the king of Munaseli and destroy KLVNLQJGRPVRPHWLPHEHWZHHQ$'7KHGHIHDWHG0XQDVHOLSRSXODWLRQÀHGWR $ORU%HVDURQWKH$ORUSHQLQVXOD VHH¿JXUH 2WKHUVRXUFHVFRQ¿UPWKDWDURXQG$'WKHLQÀXHQFHRIWKH+LQGX-DYDQHVH kingdom Majapahit indeed extended to Pantar: the Javanese Nagarakertagama chronicles (1365) contain a list of places in the east that were in the Majapahit realm, including µ*DOL\DKR¶ +lJHUGDO 33 The name Galiyahu or Galiyao occurs in a number of 16th and 17th century maps and descriptions by Europeans, and general consensus exists that *DOL\DKX*DOL\DRUHIHUVWR3DQWDU VHH/H5RX[ %DUQHV 'LHWULFK 5RGHPHLHU %DUQHV 5RGHPHLHU+lJHUGDO 5HFHQWOLQJXLVWLF research by Gary Holton and myself on Pantar island revealed that Galiyao is used in various local languages as the indigenous name to refer to the island of Pantar; the name originates from Western Pantar language Gale AwaOLWHUDOO\µOLYLQJERG\¶ +ROWRQ 7RGD\3DQGDLDQG0XQDVHOLDUH$ORUHVHVSHDNLQJDUHDVLQQRUWKHUQ3DQWDUTanjung Muna (‘Cape Muna’) in North Pantar is still considered the location of the mythical NLQJGRP0XQDVHOL7KHODQJXDJHVSRNHQWKHUHLVUHIHUUHGWRLQ,QGRQHVLDQDVBahasa Muna ‘the Muna language’, an abbreviation of Munaseli6SHDNHUVUHIHUWRWKHLURZQODQJXDJHDV Kadire Senaingµ6SHHFKZH8QGHUVWDQG¶ 5RGHPHLHU DQGWKHBahasa Muna or Kadire SenaingUHSRUWHGLQ5RGHPHLHULV DGLDOHFWRI $ORUHVH Alorese is currently also spoken along the coast of the Alor Bird’s Head peninsula, and WKHDQFHVWRUVRIWKHVHVSHDNHUVDUHSUREDEO\UHODWHGWRWKH0XQD VHOL SRSXODWLRQWKDWÀHG WR$ORUDIWHUWKHLUGHIHDWLQ3DQWDUE\HDUO\ In sum, from historical, ethnographic and linguistic observations we can infer that *DOL\DKXZDV3DQWDUWKDW3DQWDUZDVXQGHUWKHLQÀXHQFHRIWKH0DMDSDKLWNLQJGRPLQ $'ZKLFKLVHYLGHQFHWKDWWKHLVODQGZDVNQRZQIDUEH\RQGLWVLPPHGLDWHQHLJKERULQJ WHUULWRULHV%RWKWKH3DQGDLDQG0XQDVHOLNLQJGRPLQ3DQWDUZHUHLQSODFHDURXQG $'LQ1RUWK1RUWKHDVWHUQ3DQWDUKDYLQJEHHQHVWDEOLVKHGE\LPPLJUDQWVVSHDNLQJ DQ$XVWURQHVLDQODQJXDJH,QWKHHDUO\WKFHQWXU\DWOHDVWRQHJURXSÀHGIURP3DQWDUWR $ORUWRVHWWOHLQ$ORU%HVDURQWKH$ORUSHQLQVXOD7RGD\WKHVHWWOHPHQWV3DQGDL0XQDVHOL Alor Besar and Baranusa still exist, and all of them coincide with locations where Alorese is spoken, so we can safely assume that today’s Alorese populations are descendants from FODQVWKDWVHWWOHGRQ3DQWDU 337KHLQÀXHQFHRI0DMDSDKLWLQWKH/HVVHU6XQGD,VODQGVGLGQRWLPSO\DFWXDOSROLWLFDORUFXOWXUDO LQYROYHPHQWDVQR0DMDSDKLWDUFKHRORJLFDOUHPDLQVKDYHEHHQIRXQGLQWKHDUHD “The appropriateness of this name is evidenced by the presence of an active volcano which GRPLQDWHVVRXWKHUQ3DQWDU7KLVYROFDQRUHJXODUO\HUXSWVRIWHQUDLQLQJDVKDQGS\URFODVWLFÀRZV RQWRYLOODJHVRIWKHUHJLRQ(YHQZKHQLWLVQRWHUXSWLQJWKHYROFDQRRPLQRXVO\YHQWVVXOIXUJDV DQGVPRNHIURPLWVFUDWHU,QDYHU\UHDOVHQVHWKHYROFDQRLVDOLYLQJERG\´ +ROWRQ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 103 Given the close linguistic and cultural ties between Alorese and Lamaholot, I conclude that the ancestors of the Alorese were Lamaholot speakers from Solor, Lembata, Adonara DQGRUHDVW)ORUHV7KH\DUULYHGDWWKHFRDVWVRI3DQWDUEHIRUHRUDURXQG$' 6. SUMMARY AND DISCUSSION. A number of shared syntactic features which signal 3DSXDQLQÀXHQFHVDUHIRXQGLQERWK/DPDKRORWDQG$ORUHVHDQGPXVWKDYHEHHQSDUWRI 3URWR/DPDKRORW7KLVVXJJHVWV SUHKLVWRULF 3DSXDQSUHVHQFHLQWKH/DPDKRORWKRPHODQG ZKLFKPD\KDYHEHHQORFDWHGLQHDVW)ORUHVDQGRUWKHLVODQGV6RORU/HPEDWDDQG$GRQDUD 7KH3DSXDQLQÀXHQFHRQ3URWR/DPDKRORWZDVVWURQJHQRXJKWRLQFUHDVHWKHFRPSOH[LW\ of Proto-Lamaholot: an increase in word order patterns, the introduction of an inalienable noun distinction and variable possessor marking structures, as well as a new functional LWHPWKHIRFXVPDUNHU:KHUHODQJXDJHFRQWDFWOHDGVWRDQLQFUHDVHGOLQJXLVWLFFRPSOH[LW\ with additive features, the language is likely to have been spoken in a community with high GHJUHHVRIRXWVLGHFRQWDFWV 7UXGJLOO 7KHFRQWDFWPXVWKDYHEHHQORQJWHUP DQGKDYHLQYROYHGODQJXDJHDFTXLVLWLRQRISUHDGROHVFHQWV µSUHFULWLFDOWKUHVKROGFRQWDFW VLWXDWLRQV¶7UXGJLOO 3URWR/DPDKRORWKDGDIDLUO\ULFKPRUSKRORJ\LQFOXGLQJSRVVHVVRUVXI¿[HVGLVWLQFW SURQRPLQDODI¿[HVIRU$DQG6DQGDWOHDVWVHYHQGHULYDWLRQDOSUH¿[HV$IWHULWVSOLWIURP /DPDKRORW$ORUHVHXQGHUZHQWDSURFHVVRIVLPSOL¿FDWLRQ LWORVWDOORIWKH3URWR/DPDKRORW GHULYDWLRQDODQGLQÀHFWLRQDOPRUSKRORJ\LQFOXGLQJWKHPDUNHGGLVWLQFWLRQEHWZHHQ$DQG 6 WKHYDULDEOHSRVVHVVRUPDUNLQJVWUXFWXUHVZHUHUHJXODUL]HGDQGWKH¿QDOQDVDOPRUSKHPH RQLQDOLHQDEOHQRXQVZDVUHLQWHUSUHWHGDVDURRW¿QDOFRQVRQDQWVHJPHQW $IWHUWKH\DUULYHGRQ3DQWDULVODQGHLWKHUEHIRUHRUGXULQJWKHth century, the Alorese GLGQRWERUURZPXFKYRFDEXODU\IURPWKHLU$ORU3DQWDUQHLJKERXUV7KHOLPLWHGQXPEHURI LGHQWL¿HGORDQVFRPHIURPGLIIHUHQW$3ODQJXDJHVDFURVV3DQWDUQRQHRIZKLFKDSSHDUVWR KDYHEHHQGRPLQDQW$ORUHVHUHWDLQHGWKHV\QWD[RI/DPDKRORWVLPSOLI\LQJDQGUHJXODUL]LQJ VRPHRILWVLUUHJXODULWLHVDQGWKHLQÀXHQFHRIORFDO$3V\QWD[RQ$ORUHVHDSSHDUVWRKDYH been minimal: Alorese moved its time adverb to postverbal position, and adopted a clause ¿QDOFRQMXQFWLRQOLNHHOHPHQW 7KHOLPLWHGOH[LFDOFRQJUXHQFHDQGYLUWXDODEVHQFHRIV\QWDFWLFLQÀXHQFHVVXJJHVWVD contact scenario that neither involved prolonged stable bilingualism, nor Papuan speaking FRPPXQLWLHVVKLIWLQJWR$ORUHVH+RZHYHUWKHPRUSKRORJLFDODQGV\QWDFWLFVLPSOL¿FDWLRQ RI$ORUHVHVXJJHVWVWKDWWKHODQJXDJHZHQWWKURXJKDVWDJHRIVHFRQGODQJXDJHOHDUQLQJ 7KLVFRPELQDWLRQRIIDFWVLVLQGHHGSX]]OLQJ There is evidence that Alorese was spoken as non-native language: it was used as a UHJLRQDOWUDGHODQJXDJH $QRQ\PRXV 6WRNKRI DQGLQWHQVLYHWUDGHUHODWLRQV existed between the coastal Alorese and the Papuan populations living in the Pantar PRXQWDLQVH[FKDQJLQJHJZRYHQFORWKIRUIRRG FI$QRQ\PRXV $VWKH$ORUHVHVHWWOHPHQWVRQWKHFRDVWVRI3DQWDUDQG$ORUZHUHLQLWLDOO\TXLWHVPDOODQG JHRJUDSKLFDOO\UHPRWHIURPHDFKRWKHULWLVOLNHO\WKDWLQLWLDOO\WKH$ORUHVHPHQDFTXLUHG their spouses from one of the various exogamous communities in their vicinity where an $3ODQJXDJHZDVVSRNHQ$VDUHVXOWZRPHQVSHDNLQJ$3ODQJXDJHVZHUHEURXJKWLQWRD FRPPXQLW\WKDWVSRNHDODQJXDJHVLPLODUWR3URWR/DPDKRORW7U\LQJWROHDUQWKLVODQJXDJH DVDGXOWVWKHZRPHQVLPSOL¿HGLWVPRUSKRORJ\DQGWKHLUOHDUQHU¶VRPLVVLRQVEHFDPHSDUW MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 104 RIDPRUSKRORJLFDOO\VLPSOL¿HGYDULHW\WKDWGHYHORSHGLQWRWKHPRUSKRORJLFDOO\LVRODWLQJ $ORUHVHODQJXDJHDVDFTXLUHGE\WKHLUFKLOGUHQ,QÀHFWLRQDOPRUSKRORJ\LVNQRZQWREH seriously problematic for post-adolescent second language learners who have passed the µFULWLFDOWKUHVKROG¶ /HQQHEHUJ IRUODQJXDJHDFTXLVLWLRQ .XVWHUV FLWLQJ &ODKVHQ DQG 0X\VNHQ 0HLVHO $QG GHULYDWLRQDO PRUSKRORJ\ EHLQJ SDUWO\ OH[LFDOLVHGLUUHJXODUDQGVHPDQWLFDOO\RSDTXHUHSUHVHQWVDUELWUDU\JUDPPDWLFDOSDWWHUQV ZKLFKPXVWEHOHDUQHGZLWKRXWDQ\JHQHUDOL]DWLRQSRVVLEOHZKLFKLVHTXDOO\GLI¿FXOWIRU SRVWWKUHVKROGODQJXDJHOHDUQHUV 7KH ORVV RI LQÀHFWLRQDO DQG GHULYDWLRQDO PRUSKRORJLFDO FDWHJRULHV LQ $ORUHVH FDQ WKXVEHVHHQDVDQLQVWDQFHRIVLPSOL¿FDWLRQWKDWRFFXUUHGDVDUHVXOWRIQRQQDWLYHDGXOW ODQJXDJH OHDUQLQJ 7UXGJLOO ,Q JHQHUDO VLPSOL¿FDWLRQ LV PRVW OLNHO\ WR occur in intense contact situations that are short-term and post-critical threshold (Trudgill 7KHTXHVWLRQVWKDWDUHQRWDQVZHUHGE\WKLVVFHQDULRLQFOXGHWKHIROORZLQJ'LGWKH Papuan mothers introduce more of their Papuan words and syntax into the Alorese they VSRNHDVVHFRQGODQJXDJH",IWKH\GLGZK\GLGWKHLUFKLOGUHQQRWDTXLUHWKLVDORQJZLWK WKHLU PRUSKRORJLFDOO\ VLPSOL¿HG $ORUHVH" 2U ZDV WKHUH FRPPXQLW\ SUHVVXUH WR VSHDN Alorese in its lexically and syntactically ‘pure’ form, while omitting its morphology was allowed? Additional sociolinguistic research on the social position and language attitude as well as studies of actual speech of newcomers into Alorese communities may help to shed VRPHOLJKWRQWKLV In the history of Alorese reconstructed here, we see that at different time depths, GLIIHUHQWODQJXDJHFRQWDFWVLWXDWLRQVKDGGLIIHUHQWFRQVHTXHQFHVIRUWKHVWUXFWXUHRIWKH ODQJXDJH3UHKLVWRULFGHHSWLPHFRQWDFWEHWZHHQD3DSXDQVXEVWUDWHDQG3URWR/DPDKRORW UHVXOWHG LQ D FRPSOH[L¿FDWLRQ RI 3URWR/DPDKRORW ZKLOH ODWHU SRVWPLJUDWLRQ FRQWDFW UHVXOWHG LQ D VLPSOL¿FDWLRQ:KLOH ERWK RXWFRPHV VXJJHVW WKDW WKH FRQWDFW ZDV LQWHQVH the sociolinguistic situations were presumably different: prehistoric contact with Papuan languages in the Flores area was long-term and involved pre-adolescents, while the post- migration contact that took place after settlement on Pantar was short-term, and involved SRVWDGROHVFHQWOHDUQHUV7KHUHLVQRHYLGHQFHWKDWVLQFHWKDWSHULRGOLQJXLVWLFFRQWDFWV between Alorese and the speakers of AP languages around them have been any more than VXSHU¿FLDO REFERENCES $QRQ\PRXV 'HHLODQGHQ$ORUHQ3DQWDU5HVLGHQWLH7LPRUHQ2QGHUKRRULJKHGHQ Tijdschrift van het Koninklijk Nederlandsch Aardrijkskundig Genootschap $ORU 3DQWDU /H[LFDO 'DWDEDVH A compilation of LWHP VXUYH\ ZRUGV OLVWV RI ODQJXDJHVRI$ORUDQG3DQWDUFROOHFWHGEHWZHHQE\/RXLVH%DLUG*DU\ +ROWRQ0DULDQ.ODPHU)UDQWLãHN.UDWRFKYLO/DXUD5RELQVRQ $QWRLQHWWH6FKDSSHU /HLGHQ8QLYHUVLW\DQG8QLYHUVLW\RI$ODVND)DLUEDQNV $UQGW3DXO3 Grammatik der Solor-Sprache(QGH)ORUHV $UQROGXV'UXNNHULM %DLUG/RXLVHA grammar of Keo: An Austronesian language of East Nusantara. Canberra: $183K'WKHVLV %DLUG/RXLVHA grammar of Klon. &DQEHUUD 3DFL¿F/LQJXLVWLFV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 105 %DUQHV5REHUW+ 7ZRWHUPLQRORJLHVRIV\PPHWULFSUHVFULSWLYHDOOLDQFHIURP3DQWDU DQG $ORU LQ (DVWHUQ ,QGRQHVLD Sociologus; Zeitschrift für Völkerpsychologie und Soziologie %DUQHV 5REHUW + 7KH 0DMDSDKLW 'HSHQGHQF\ *DOL\DR Bijdragen tot de Taal-, Land-, en Volkenkunde %DUQHV 5REHUW + Sea hunters of Indonesia: Fishers and weavers of Lamalera 2[IRUG 6WXGLHV LQ 6RFLDO DQG &XOWXUDO$QWKURSRORJ\ 2[IRUG 2[IRUG 8QLYHUVLW\ 3UHVV %DUQHV5REHUW+$OOLDQFHDQGZDUIDUHLQDQ(DVWHUQ,QGRQHVLDQSULQFLSDOLW\.HGDQJ LQWKHODVWKDOIRIWKHQLQHWHHQWKFHQWXU\Bijdragen tot de Taal-, Land-, en Volkenkunde %OXVW5REHUW &HQWUDODQGFHQWUDOHDVWHUQ0DOD\R3RO\QHVLDQOceanic Linguistics ± %OXVW5REHUW DThe Austronesian languages. &DQEHUUD 3DFL¿F/LQJXLVWLFV %OXVW 5REHUW E 7KH SRVLWLRQ RI WKH ODQJXDJHV RI (DVWHUQ ,QGRQHVLD $ UHSO\ WR 'RQRKXHDQG*ULPHVOceanic Linguistics &ODKVHQ+DUDOG 3LHWHU0X\VNHQ +RZDGXOWVHFRQGODQJXDJHOHDUQLQJGLIIHUVIURP FKLOG¿UVWODQJXDJHGHYHORSPHQWBehavioural and Brain Sciences &ODUN5 7KH$XVWURQHVLDQODQJXDJHV,Q%HUQDUG&RPULH HG The major languages of East and South East Asia/RQGRQ 5RXWOHGJH 'LHWULFK6WHIDQ $QRWHRQ*DOL\DRDQGWKHHDUO\KLVWRU\RIWKH6RORU$ORU,VODQGV Bijdragen tot de Taal-, Land-, en Volkenkunde 'RQRKXH0DUN7\SRORJ\DQGOLQJXLVWLFDUHDVOceanic Linguistics 'RQRKXH0DUN7KH3DSXDQODQJXDJHRI7DPERUDOceanic Linguistics 'RQRKXH0DUN &KDUOHV(*ULPHV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 106 +lJHUGDO+DQV&DQQLEDOVDQGSHGODUV(FRQRPLFRSSRUWXQLWLHVDQGSROLWLFDODOOLDQFH LQ$ORUIndonesia and the Malay World +LPPHOPDQQ1LNRODXV37KH$XVWURQHVLDQODQJXDJHVRI$VLDDQG0DGDJDVFDU 7\SRORJLFDO FKDUDFWHULVWLFV ,Q .$ $GHODDU DQG 13 +LPPHOPDQQ HGV The Austronesian languages of Asia and Madagascar./RQGRQ 5RXWOHGJH +ROWRQ*DU\An etymology for Galiyao$ODVND 8QLYHUVLW\RI)DLUEDQNV0$WKHVLV +ROWRQ *DU\ 0DULDQ .ODPHU )UDQWLVHN .UDWRFKYtO /DXUD 5RELQVRQ $QWRLQHWWH 6FKDSSHU7KHKLVWRULFDOUHODWLRQRIWKH3DSXDQODQJXDJHVRI$ORUDQG3DQWDU Oceanic Linguistics -DFRE-XQH &KDUOHV(*ULPHVKamus Pengantar Bahasa Kupang. Kupang: Artha :DFDQD3UHVV -RQHV5XVVHOO HG Loan words in Indonesian and Malay/HLGHQ .,7/93UHVV .HUDI*UHJRULXV Morfologi dialek Lamalera. Jakarta: Universitas Indonesia PhD GLVVHUWDWLRQ .ODPHU 0DULDQ .DPEHUD KDV QR SDVVLYH ,Q 0DULDQ .ODPHU HG Voice in Austronesian (NUSA Linguistic Studies of Indonesian and Other Languages in ,QGRQHVLD -DNDUWD 8QLYHUVLWDV$WPD-D\D .ODPHU 0DULDQ 7HQ \HDUV RI V\QFKURQLF $XVWURQHVLDQ OLQJXLVWLFV Lingua .ODPHU0DULDQDA grammar of Teiwa%HUOLQ 'H*UX\WHU .ODPHU0DULDQE'LWUDQVLWLYHVLQ7HLZD,Q$QGUHM0DOFKXNRY0DUWLQ+DVSHOPDWK %HUQDUG&RPULH HGV Studies in ditransitive constructions%HUOLQ 'H*UX\WHU .ODPHU 0DULDQ A short grammar of Alorese (Austronesian) 0XQLFK /LQFRP (XURSD .ODPHU0DULDQQGTeiwa corpus/HLGHQ /HLGHQ8QLYHUVLW\ .ODPHU0DULDQ)RUWKFRPLQJ)URP/DPDKRORWWR$ORUHVH 6LPSOL¿FDWLRQZLWKRXWVKLIW,Q 'DYLG*LO -RKQ0F:KRUWHU HGV Austronesian undressed: How and why languages become isolating. .ODPHU0DULDQ 0LFKDHO(ZLQJ7KHODQJXDJHVRI(DVW1XVDQWDUD $QLQWURGXFWLRQ In Ewing & Klamer, East Nusantara .ODPHU0DULDQ )UDQWLVHN.UDWRFKYtO$EXL7ULSDUWLWH9HUEV ([SORULQJWKHOLPLWV RIFRPSRVLWLRQDOLW\,Q-DQ:RKOJHPXWK 0LFKDHO&\VRXZ HGV Rara & Rarissima: Documenting the fringes of linguistic diversity (Empirical Approaches to Language 7\SRORJ\ %HUOLQ 'H*UX\WHU .ODPHU0DULDQ*HU5HHVLQN 0LULDPYDQ6WDGHQ(DVW1XVDQWDUDDVDOLQJXLVWLF DUHD ,Q 3LHWHU 0X\VNHQ HG From linguistic areas to areal linguistics $PVWHUGDP %HQMDPLQV .ODPHU0DULDQ $QWRLQHWWH6FKDSSHU7KHGHYHORSPHQWRIµJLYH¶FRQVWUXFWLRQVLQ WKH3DSXDQODQJXDJHVRI7LPRU$ORU3DQWDULinguistic Discovery .UDWRFKYtO)UDQWLVHNA grammar of Abui>/HLGHQ8QLYHUVLW\3K'WKHVLV@8WUHFKW /27 .XVWHUV :RXWHU Linguistic complexity. >/HLGHQ 8QLYHUVLW\ 3K' WKHVLV@ 8WUHFKW /273XEOLFDWLRQV /HQQHEHUJ(ULF Biological foundations of language1HZ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 107 /H 5RX[ &&)0 De Elcano’s tocht door den Timorarchipel met Magelhaes’ VFKLS 9LNWRULD Feestbundel Koninklijk Bataviaasch Genootschap van Kunsten en Wetenschappen 1778-1928, II%DWDYLD .ROII /HZLV03DXO HG Ethnologue: Languages of the world, 16th editionKWWS ZZZ HWKQRORJXHFRP $XJXVW) /HZLV('RXJODV &KDUOHV*ULPHV 6LND ,QWURGXFWLRQDQGZRUGOLVW,Q7U\RQ Comparative Austronesian dictionary /\QFK-RKQ0DOFROP5RVV 7HUU\&URZOH\The Oceanic languages. Richmond, 6XUUH\ &XU]RQ 0HLVHO-XUJHQ 7KHDFTXLVLWLRQRIWKHV\QWD[RIQHJDWLRQLQ)UHQFKDQG*HUPDQ FRQWUDVWLQJ¿UVWDQGVHFRQGODQJXDJHGHYHORSPHQWSecond Language Research 1DJD\D 1DRQRUL D 6SDFH DQG PRWLRQ LQ /DPDKRORW 3DSHU SUHVHQWHG DW WKH WK ,QWHUQDWLRQDO&RQIHUHQFHRQ$XVWURQHVLDQ/LQJXLVWLFV$XVVRLV)UDQFH 1DJD\D 1DRQRUL E 6XEMHFW DQG WRSLF LQ /DPDKRORW 3DSHU SUHVHQWHG DW WKH WK ,QWHUQDWLRQDO&RQIHUHQFHRQ$XVWURQHVLDQ/LQJXLVWLFV$XVVRLV)UDQFH 1LVKL\DPD.XQLRDQG+HUPDQ.HOHQA grammar of Lamaholot, Eastern Indonesia: The morphology and syntax of the Lewoingu dialect0XQLFK /LQFRP(XURSD 2QYOHH/RXLV Kamberaas-Nederlands woordenboek/HLGHQ .,7/9 3DPSXV .DUO+HLQ] Koda Kiwa: Dreisprachiges Worterbuch des Lamaholot (Dialekt von Lewolema) $EKDQGOXQJHQ IXU GLH .XQGH GHV 0RUJHQODQGHV 6WXWWJDUW 6WHLQHU 3DPSXV.DUO+HLQ] ZLWKKHOSRI MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Papuan-Austronesian language contact 108 6DPHO\ 8UVXOD Kedang, (Eastern Indonesia): Some aspects of its grammar. +DPEXUJ %XVNH Schapper, $QWRLQHWWH -XOLHWWH +XEHU $RQH YDQ (QJHOHQKRYHQ 7KH KLVWRULFDO UHODWLRQVRIWKH3DSXDQODQJXDJHVRI7LPRUDQG.LVDULanguage and Linguistics in Melanesia 6SHFLDOLVVXH3DUW 6FKDSSHU$QWRLQHWWH 0DULDQ.ODPHUPV&DUGLQDOQXPHUDOVLQWKH3DSXDQODQJXDJHVRI $ORU3DQWDU +LVWRU\DQGW\SRORJ\/HLGHQ8QLYHUVLW\8QSXEOLVKHGPV 6WHLQKDXHU +HLQ %DKDVD %ODJDU 6HOD\DQJ 3DQGDQJ 3HQ\HOLG %DKDVD GDQ 3HUNHPEDQJDQ:DZDVDQQ\D, -DNDUWD 0DV\DUDNDW/LQJXLVWLN,QGRQHVLD 6WRNKRI :$/ Preliminary notes on the Alor and Pantar languages (east Indonesia) 3DFL¿F/LQJXLVWLFV% 3DFL¿F/LQJXLVWLFV5HVHDUFK6FKRRORI3DFL¿F 6WXGLHV7KH$XVWUDOLDQ1DWLRQDO8QLYHUVLW\ 7UXGJLOO3HWHU&RQWDFWDQGVRFLROLQJXLVWLFW\SRORJ\,Q5D\PRQG+LFNH\ HG The handbook of language contact, 0DOGHQ0$ :LOH\%ODFNZHOO 7U\RQ'DUUHOO7 7KH$XVWURQHVLDQODQJXDJHV,Q7U\RQComparative Austronesian dictionary 7U\RQ'DUUHOO7 HG Comparative Austronesian dictionary: An introduction to Austronesian studies 7UHQGV LQ /LQJXLVWLFV 'RFXPHQWDWLRQ SDUWV %HUOLQ 1HZ Marian Klamer [email protected] MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY /DQJXDJH'RFXPHQWDWLRQ &RQVHUYDWLRQ6SHFLDO3XEOLFDWLRQ1R 'HFHPEHU Melanesian Languages on the Edge of Asia: Challenges for the 21st Century, HGE\1LFKRODV(YDQVDQG0DULDQ.ODPHUSS ± KWWS QÀUFKDZDLLHGXOGFVS 5 KWWS KGOKDQGOHQHW Even more diverse than we had thought: The multiplicity of Trans-Fly languages Nicholas Evans Australian National University Linguistically, the Trans Fly region of Southern New Guinea is one RIWKHOHDVWNQRZQSDUWVRI1HZ*XLQHD 1. INTRODUCTION.1 The distribution of linguistic diversity is highly informative, about 1 My thanks to two anonymous referees and to Marian Klamer for their usefully critical comments RQ DQ HDUOLHU YHUVLRQ RI WKLV SDSHU , JUDWHIXOO\ DFNQRZOHGJH WKH VXSSRUW RI WKH $XVWUDOLDQ National University (Professorial Setup Grant) and the Australian Research Council (Discovery 3URMHFWµ/DQJXDJHVRI6RXWKHUQ1HZ*XLQHD¶ IRUVXSSRUWLQJP\¿HOGZRUNLQ6RXWKHUQ31* the Linguistics Society of America for support to teach a Field Methods course on Idi at the %RXOGHU/LQJXLVWLFV,QVWLWXWHLQDVZHOODVWKH$15&IRUIXQGLQJHQDEOLQJPHWRDWWHQG the Manokwari conference, to members of the audience there for their helpful discussion, and Jeff Siegel, Christian Döhler, Grahame Martin and Garrick Hitchcock for access to unpublished PDWHULDOVGUDZQRQKHUH0RVWLPSRUWDQWO\,WKDQNP\1HQDQG,GLWHDFKHUVSDUWLFXODUO\0LFKDHO Binzawa, †Aramang Wlila, Jimmy Nébni and Wasang Baiio, for their insightful and dedicated HIIRUWV WR WHDFK PH WKHLU ODQJXDJHV 0DWHULDO RQ ,GL FRPHV SUHGRPLQDQWO\ IURP WZR VRXUFHV recordings made with Wasang Baiio during a Field Methods course at the LSA Institute in %RXOGHU &RORUDGR LQ -XO\$XJ DQG PDWHULDO UHFRUGHG IURP 0U *XV ,DPDWWD cc Creative Commons Attribution Non-Commercial No Derivatives Licence Multiplicity of Trans-Fly languages 110 KLVWRU\VRFLDOFRQ¿JXUDWLRQVDQGLGHRORJLHVRIODQJXDJHXVH2YHUWKHODVWIRXUGHFDGHV of scholarship, New Guinea’s position as the most linguistically diverse region of the planet has not changed, but received views of where the most deep-level diversity lies ZLWKLQ1HZ*XLQHDKDYHPRYHGVXEVWDQWLDOO\9DULRXVYHUVLRQVRIWKH7UDQV1HZ*XLQHD hypothesis have led to hundreds of languages centred on the cordillera being joined into a single family of (sometimes only distantly) related languages, whereas the progression of UHVHDUFKRQWKH6HSLNKDVIRXQGDPRVDLFRIVPDOOIDPLOLHVDQGLVRODWHV±DSDWWHUQWDNHQWR be more representative of New Guinea as a whole before the spread of Trans-New Guinea ODQJXDJHV 6RXWKHUQ1HZ*XLQHD±DQGPRUHSDUWLFXODUO\WKDWSDUWRILWNQRZQDVWKH7UDQV)O\ ¿J ±KDVQRW\HW¿JXUHGSURPLQHQWO\LQDVVHVVPHQWVRIZKHUHWKHPRVWGLYHUVLW\OLHV 7KRXJKLWKDVVRPHWLPHVEHHQPHQWLRQHG HJ3DZOH\ DVµDVPDOOHUUHJLRQRIKLJK diversity’, existing assessments tend to lump together several families on little evidence: ERWK3DZOH\ DQG5RVV HVVHQWLDOO\UHSURGXFH:XUP¶VHDUOLHUOXPSHU FODVVL¿FDWLRQRIZKDW,ZLOODUJXHDUHVHYHUDOGLVWLQFWIDPLOLHVLQWKH7UDQV)O\UHJLRQ 0HWKRGV FRXUVH HOLFLWHG VRPH RI WKH PDWHULDO FLWHG KHUH LQ VPDOOJURXS VHVVLRQV 0DWHULDO RQ :DUWD7KXQGDLFRPHVIURPD¿HOGPHWKRGVFRXUVHWDXJKWLQ)HEUXDU\ZLWK6HPEDUD'LEDUD ZKRP,WKDQNIRUKLVHQWKXVLDVWLFSDUWLFLSDWLRQ 7KHUHLVQRXQLYHUVDOO\DFFHSWHGGH¿QLWLRQRIWKHH[WHQWRIµ7UDQV)O\¶ LWWHQGVWREHZHOOGH¿QHG at its eastern and northern extremities (by the Fly River) and to the south by the Torres Strait but, as one moves west, geographical boundaries give way to political ones, as in Williams’ (1936) µWKHVRXWKZHVWFRUQHURI3DSXD¶ ZKHUH3DSXDPHDQWWKH>WKHQ@$XVWUDOLDQWHUULWRU\RI3DSXD From the ecological point of view, however, it makes sense to consider the Trans-Fly Region as H[WHQGLQJVRPHZKDWIXUWKHUZHVWWDNLQJLQ.RORSRP,VODQGDVLVGRQHLQ)LJDQGWKLVLVWKH WHUPWKDWKDVEHHQDGRSWHGE\FRQVHUYDWLRQLVWJURXSVOLNHWKH:RUOG:LOGOLIH)XQG)RUSUHVHQW purposes I will take it to extend across the (modern) national boundary to the Merauke River, in WUDGLWLRQDO0DULQGWHUULWRU\6RXWKHUQ1HZ*XLQHDLVRIFRXUVHELJJHUWKDQWKLVZLWKPDQ\RWKHU OLQJXLVWLFJURXSLQJVZKLFK,GRQRWGLVFXVVKHUHVXFKDV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 111 FIGURE 1 7KH7UDQV)O\5HJLRQDVGH¿QHGE\WKH::) ::)7UDQVÀ\ 7HDP In this paper I will argue that Southern New Guinea in fact contains more deep diversity WKDQKDVKLWKHUWREHHQUHDOLVHGZLWKVRPHZKHUHEHWZHHQ¿YHDQGHLJKWXQUHODWDEOHIDPLOLHV WDNLQJLQIRUW\RUVRODQJXDJHVLQDQDUHDDERXWWKHVL]HRIWKH1HWKHUODQGV2QWRSRI that, there are major typological differences between the languages of these families, and PDQ\RIWKHP VXFKDVWKH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages XQGHUVWRRGWKHQLQJRRQWRJLYHEULHISRUWUDLWVRI1HQDQG,GLWZRODQJXDJHVZKLFK although close geographical neighbours bound by relations of marriage exchange and PXOWLOLQJXDOLVPGLYHUJHVLJQL¿FDQWO\RQDZLGHUDQJHRIPHDVXUHV±QRWRQO\DUHWKH\ LQGLIIHUHQWSK\ORJHQHWLFJURXSVEXWWKHLUW\SRORJLFDOSUR¿OHVDOVRGLIIHUPDUNHGO\%XW GLYHUJHQFHRIWKLVW\SHGRHVQRWPHDQWKHUHDUHQRVLJQL¿FDQWDUHDOIHDWXUHVDFURVV6RXWKHUQ 1HZ*XLQHDDQGLQ,LOOXVWUDWHWKLVSRLQWZLWKRQHVXFKIHDWXUH±WKUHHYDOXHGQXPEHU V\VWHPV±ZKLOHHPSKDVLVLQJWKDWWKHPHDQVRIFRPSRVLQJWKHGXDOYDOXHYDU\VLJQL¿FDQWO\ IURPRQHODQJXDJHJURXSWRDQRWKHU,FORVHWKHDUWLFOHLQE\VXPPDULVLQJWKHNH\ VFLHQWL¿FFKDOOHQJHVIDFLQJOLQJXLVWVDVZHFRQIURQWD]RQHWKDWLVVLPXOWDQHRXVO\RQHRI WKHPRVWGLYHUVHDQGRQHRIWKHOHDVWNQRZQUHJLRQVRIWKHORJRVSKHUH 2. SOUTHERN NEW GUINEA AS A GEOGRAPHICAL AND CULTURAL REGION. In its biota, such as its vegetation of eucalypts, melaleuca, acacia and banksias combined with wallabies, bandicoots, goannas, taipans and termite mounds, Southern New Guinea is more OLNHQRUWKHUQ$XVWUDOLDWKDQOLNHWKHUHVWRI1HZ*XLQHD Geographically, much of it is new, low land, a kind of tropical Netherlands built up RYHUWKHODVWIHZPLOOHQQLDDVWKHJLDQW)O\5LYHUWRWKHHDVW ¿J DQGWKH'LJXODQGRWKHU ULYHUVWRWKHZHVWKDYHFDUULHGGRZQDQGGHSRVLWHGVHGLPHQWIURPWKHFHQWUDOFRUGLOOHUD Compared to most of present-day New Guinea and Australia (except for the Sepik), it has KDGDWXUEXOHQWJHRPRUSKRORJLFDOSDVWRYHUWKHODVW\HDUV7KHDQFLHQWODQGEULGJH WR$XVWUDOLDZDVVHYHUHGE\WKHULVLQJVHDVDURXQG ESDQGIRUDZKLOHKLJKHUVHD levels than today meant that some of what is now land was then submerged, before being UHEXLOWE\SURJUDGDWLRQ The northern parts are characterised by vast tracts of rainforest, with only the occasional FOHDULQJIRUDYLOODJHVZLGGHQJDUGHQRUVDJR ¿J 0RYLQJVRXWKWKLVJLYHVZD\WR HXFDO\WXVDQGPHODOHXFDVDYDQQDKUHPLQLVFHQWRIQRUWKHUQ$XVWUDOLD ¿J DQG±DURXQG ULYHUVOLNHWKH%HQVEDFK±H[WHQVLYHÀRRGSODLQVVXSSRUWLQJPDVVLYHSRSXODWLRQVRIELUGV ZDOODELHVDQG QRZ GHHU7KHUHLVDPDUNHGPRQVRRQDOF\FOHZLWKDORQJGU\VHDVRQ -XO\1RYHPEHU DOWHUQDWLQJZLWKDQLQWHQVHZHWVHDVRQ 'HFHPEHU±-XQH 7KHOHQJWKRI WKHZHWVHDVRQLQFUHDVHVDVRQHKHDGVQRUWK 6WDSOHIRRGVYDU\VRPHZKDWDFURVVWKHDUHD,QWKH0RUHKHDGGLVWULFW\DPVDQGRWKHU root crops predominate, based on swidden (slash-and-burn) agriculture which yields one year of fertile soils, followed by one or two years for less demanding crops like cassava and pineapples, then gradual reversion of the cleared area to forest over around 17 years, ZLWKPDWXUHFRFRQXWWUHHVWKHQWKHRQO\VLJQRISULRUFXOWLYDWLRQ/DQJXDJHVRIWKHUHJLRQ FRQWDLQQXPHURXVWHUPVIRUGLIIHUHQWSKDVHVRIFXOWLYDWLRQ±LQDGGLWLRQWRWKHJHQHULFZRUG D\DJ ‘new garden’, kkp get kr ‘old garden’ and duۆ kkp for ‘garden’, Nen distinguishes µDEDQGRQHGRYHUJURZQJDUGHQ¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 113 FIGURE 2$HULDOYLHZRIWKH)O\5LYHUWDNHQIURPWKHVRXWKZHVWZLWKWKH FHQWUDOFRUGLOOHUDYLVLEOHIDUWRWKHQRUWK 3KRWR 1(YDQV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages FIGURE 3.6DJRFOHDULQJLQUDLQIRUHVWEHWZHHQ.LULZRDQG)O\5LYHU 3KRWR 1(YDQV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 115 FIGURE 4. Jimmy Nébni in open savannah country with mixed melaleuca and HXFDO\SWXVYHJHWDWLRQVRXWKHUQSDUWRI1HQVSHDNLQJDUHD 3KRWR 1(YDQV In this region great social value is placed on the accumulation of yams through expert gardening, with large traditional yam-feasts (Williams 1936) and counting-ceremonies based on powers of six, along with social stipulations also reckoned in powers of six, such DVWKDWDKRXVHKROGQHHGVWRKDYH ) stored in its yamhouse to feed it from one \HDUWRWKHQH[W6HQDU\SRZHUWHUPVUHSUHVHQWLQJSRZHUVRIVL[XSWR5 or 66 are found WKURXJKRXWWKH 3 5HVWULFWHGWRVRPHYHU\OLPLWHGXVHWHUPVLQ$J|EDQG,GLZKLFKDSSHDUWREHERUURZLQJV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 116 9DOXH Power Nen Keraakie Arammba Kanum Agöb (base ) exponential (Buzi village) term 6 61 pus (eembru) for nimbo put 36 6 prta ferta (eembru) feté ptae purta [peta] 63 taromba taromba tarumba tarwmpao tarumba [tarumba] 6 damno daameno ndamno ntamnao damuno [dameno] 7,776 65 wärämaka werameka wermeke wrmaekr waramakai 66 [] wi wi 67 meemee wemb TABLE 1%DVHVL[SRZHULQVRPHODQJXDJHVRIWKH In the swampier, more low-lying areas around the Bensbach and Torassi Rivers, there is evidence for the earlier use of mound-and-ditch agriculture to cultivate taro (Hitchcock $QGDVFRQGLWLRQVJHWZHWWHUWRWKHQRUWKDQGQRUWKZHVWPDNLQJEXUQLQJRIIPRUH GLI¿FXOW \DP JDUGHQV JLYH ZD\ WR VDJR DV WKH PDLQ VWDSOH +XQWLQJ LV DOVR LPSRUWDQW throughout the region, with cassowary, wallabies, bandicoots, wild pigs and (in modern WLPHV GHHUDOOSUHVHQWLQODUJHQXPEHUV LQWKHVDYDQQDKDUHDV¿UHGULYHVZHUHXVHGWRKXQW ZDOODELHVLQPXFKWKHVDPHZD\DVLQQRUWKHUQ$XVWUDOLD$FFRUGLQJWRORFDOWUDGLWLRQVRPH peoples, such as some Pahoturi River groups, were until recently hunter-gatherers rather WKDQJDUGHQHUV In addition to the great cordillera-fed rivers, there are numerous shorter rivers running VRXWK LQWR WKH 7RUUHV 6WUDLW IURP WKH ORZO\LQJ 7UDQV)O\ SODWHDX +LVWRULFDOO\ WKHVH were important as supplementary waterways permitting war-canoes to penetrate far into the interior, thus playing a key role in depredations effected on speakers of the smaller language groups by huge war-parties of Marind from the west (as well as Kiwais to the east DQG7RUUHV6WUDLWSHRSOHVWRWKHVRXWK 2.1. PRECONTACT. Colonial contact began late in the region, and it was only early in the twentieth century that the respective colonial powers (at that time the Netherlands in the west and the British in the east) began to assert some control over large and ferocious armed groups such as the Marind (aka Tugere) to the west, the Kiwai to the east, and the 6XNLWRWKHQRUWK,QGHHGLWZDV%ULWLVKGHPDQGVWRWKH'XWFKWKDWWKH\WDNHUHVSRQVLELOLW\ for pacifying raids carried out by peoples within the latter’s territory that led to the joint $QJOR'XWFKH[SHGLWLRQLQ ZKLFK¿[HGWKHERUGHUWKDWKDVGLYLGHGWKHLVODQGRI1HZ *XLQHDHYHUVLQFH There were clear discrepancies in the size of social units in the region which opposed relatively large and complex polities (numbering up to 10,000 or more) employing MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 117 H[SDQVLRQLVWLF PLOLWDU\ SROLFLHV WR VPDOO XQLWV QXPEHULQJ LQ WKH KXQGUHGV DW PRVW7KH 0DULQG±GHVFULEHGLQGHWDLOLQ9DQ%DDO¶V FODVVLFHWKQRJUDSK\RI'HPD±ZHUHWKH most successful of the former groups, in demographic and military terms, able to muster SDUWLHVRIVFRUHVRIZDUFDQRHVHDFKFRQWDLQLQJRUPRUHZDUULRUV7KHLUSROLFLHVLQFOXGHG the forming of alliances with immediate neighbours to allow them safe passage to raid groups beyond, the assimilation of non-Marind neighbours (such as the Marori and the Kanum) into an expansive system of allied clans aligned with Marind cultural norms, and WKHIXOOVRFLDODVVLPLODWLRQRIFKLOGUHQFDSWXUHGLQKHDGKXQWLQJUDLGVWR0DULQGHWKQLFLW\ It is not hard to see how the power imbalances in this situation would have driven GHPRJUDSKLFDQGOLQJXLVWLFH[SDQVLRQRI0DULQGDWWKHH[SHQVHRIWKHLUVPDOOHUQHLJKERXUV The greater retention of Marind with respect to highly endangered smaller languages like 0DURULDQG.DQXPLQWKHPRGHUQHUDLVVLPSO\DFRQWLQXDWLRQRIDPXFKROGHUG\QDPLF Without us yet being able to put details to this scenario, it suggests a situation where rapid expansion of some larger groups at the expense of smaller ones was interrupted by WKHLQWHUYHQWLRQRI(XURSHDQFRORQLDOSRZHUV±DQGZHPD\QRWEHH[DJJHUDWLQJWRVD\ that without the arrival of colonial governments (and missionary endeavours eliminating headhunting and overt warfare) many of the small languages of the Trans-Fly may not have VXUYLYHGLQWKHZD\WKH\KDYH $IXUWKHUIDVFLQDWLQJHOHPHQWLQWKLVG\QDPLFFRPHVIURPWKHOLQJXRJHQHWLFDI¿OLDWLRQV RI WKH JURXSV LQYROYHG$OO RI WKH ODUJH H[SDQVLYH JURXSV KDYH EHHQ FODVVL¿HG WR EH PHPEHUV RI WKH 7UDQV1HZ *XLQHD JURXSLQJ 7KHVH LQFOXGH 0DULQG VSHDNHUV .LZDL DQG6XNL WKRXJKERWK0DULQGDQG.LZDLGHYLDWHVLJQL¿FDQWO\IURP W\SLFDO7UDQV1HZ*XLQHDODQJXDJHVW\SRORJLFDOO\ VHHIRRWQRWHIRUDQHODERUDWLRQRI WKLV SRLQW DV LW SHUWDLQV WR 0DULQG OLNHO\ UHÀHFWLQJ SULRU VXEVWUDWH OLQJXLVWLF LQÀXHQFH IURPDXWRFKWKRQRXV6RXWKHUQ1HZ*XLQHDODQJXDJHV$OORIWKHDERYHODQJXDJHVERDVW speaker populations an order of magnitude higher than languages in the Morehead district, ZLWKSRSXODWLRQVOLNH 1DPEX 1HQ RU±DWWKHODUJHUHQG ,GL 7KLVLV QRWWRVD\WKDWWKHUHDUHQRWDOVRVPDOO7UDQV1HZ*XLQHDODQJXDJHV±0DURUL $UNDWKLV YRO LVDFOHDUFDVHZLWKDFXUUHQWSRSXODWLRQRIXQGHUSUREDEO\UHÀHFWLQJDORQJSHULRG RIUHVWULFWHGGHPRJUDSK\%XWDOOWKHELJODQJXDJHVLQWKHUHJLRQDUH7UDQV1HZ*XLQHD, 3DUWLFXODUO\LQWKHFDVHRI0DULQGWKHUHLVHYLGHQFHIRUVLJQL¿FDQWW\SRORJLFDODVVLPLODWLRQWR their Southern New Guinea neighbours, so it is useful to say a little more here about the Marind FDVH Along with Kuni and other languages around the southern end of Lake Murray, with which it forms a clear subgroup, Marind has been considered by most investigators to be a branch of the 7UDQV1HZ*XLQHDIDPLO\ HJ3DZOH\5RVV 7KRXJKWKHVHVRXUFHVEDVHGWKHFODLP primarily on free pronouns, supplemented by a few possible lexical cognates, their argument has UHFHQWO\EHHQVWUHQJWKHQHGE\6XWHU¶V ¿QGLQJVRIFRJQDF\ZLWKLQWKHERXQGSURQRPLQDO REMHFWV\VWHPDVZHOORQDVXEVHWRIWUDQVLWLYHYHUE6XWHURULJLQDOO\EDVHGKLVUHFRQVWUXFWLRQVRI this subsystem on languages of the Huon Peninsula, but has more recently extended it upward WRDSUREDEOHS71*OHYHO+HUHFRQVWUXFWVVJ2na-VJ2ga-, 3sgO wa- and 3pl ya- for proto Huon Peninsula; in Marind the corresponding forms are na-, ha-, wa- and e- as illustrated by the verbs n-esov ‘follow me’, h-esov ‘follow you’, w-esov ‘follow him/her’ and y-esov ‘follow them’ 'UDEEH $VLQWKH+XRQ3HQLQVXODODQJXDJHVLQYHVWLJDWHGE\6XWHUDVZHOODVLQPDQ\ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 118 VXJJHVWLQJWKDWWKLVDUHDZLOOEHDSDUWLFXODUO\IUXLWIXOSODFHWRORRNDWWKHTXHVWLRQRIZK\ and how speakers of Trans-New Guinea languages have expanded across much of New Guinea, carpeting what was presumably once a much more diverse region with relative OLQJXLVWLFKRPRJHQHLW\ FIGURE 5. 7KHVSHFLDODI¿QDOWHUPVWKDWUHVXOWLQ1HQIURPGLUHFWVLVWHUH[FKDQJH Following the consummation of a full exchange, brothers stop calling their sisters ‘sibling’ and instead call them tampreWKHWHUPIRUµVLEOLQJLQODZ¶6SHFLDOWHUPV mitadma and miti DOVR H[LVW IRU WKH VLPXOWDQHRXVO\ DI¿QDO DQG FRQVDQJXLQHDO UHODWLYHVSURGXFHGE\VXFKDQH[FKDQJH±mitadma denotes both parents’ opposite sex siblings, just in case they were a party to a direct sister exchange, and the term miti GHQRWHVMXVWWKRVHFURVVFRXVLQVERUQWRVXFKDQH[FKDQJH RWKHU71*ODQJXDJHVLWLVRQO\DVXEVHWRIWUDQVLWLYHYHUEVWKDWWDNHREMHFWSUH¿[HV 'UDEEH OLVWV 2QWKHRWKHUKDQGWZRLPSRUWDQWSXEOLFDWLRQV 5HHVLQNHWDO DQG'H9ULHV SODFH 0DULQGRXWVLGH71*RQWKHEDVLVRILWVW\SRORJLFDOSUR¿OH'H9ULHV VXJJHVWHGDOLQN ZLWKWKH,QDQZDWDQIDPLO\$QG5HHVLQNHWDO XVLQJWKH%D\HVLDQWUHHEXLOGLQJDOJRULWKP Structure, single out Marind as one of four languages in their sample (along with Inanwatan again, but also Klon and Abui of the Timor-Alor-Pantar group) which had been considered as TNG in H[LVWLQJFODVVL¿FDWLRQVEXWZKLFKGRQRWSDWWHUQZLWK71*LQDSUR¿OHRIW\SRORJLFDOFKDUDF- WHUV 7KHPRVWOLNHO\UHFRQFLOLDWLRQIRUWKHVHFRQÀLFWLQJDI¿OLDWLRQVLVWKDW0DULQGLVLQIDFWD7UDQV 1HZ*XLQHDODQJXDJHSK\ORJHQHWLFDOO\EXWKDVXQGHUJRQHH[WHQVLYHW\SRORJLFDOUHFRQ¿JXUDWLRQ DVLWVDQFHVWUDOVSHDNHUVPRYHGLQWR6RXWKHUQ1HZ*XLQHD7KLVZRXOGPDNHLWDQLQWHUHVWLQJFDVH of a Trans-New Guinea language assimilating structurally to substrate Papuan languages from RWKHUIDPLOLHV,QIDFW:XUP DOUHDG\VXJJHVWHGVRPHWKLQJDORQJWKHVHOLQHV KHFRQ- sidered Marind and its relatives, while members of the ‘Trans-New Guinea phylum’, to ‘display a number of aberrant features which are probably attributable to a strong substratum, with several RIWKHVHDEHUUDQWIHDWXUHVFRPSDUDEOHWRFKDUDFWHULVWLFVRIODQJXDJHVRIWKH7UDQV)O\6WRFN¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 119 FIGURE 6. 6LVWHUH[FKDQJHDQGPXOWLOLQJXDOLVPLQWKH0RUHKHDG5HJLRQThe Morehead region is famed anthropologically for its practice of direct sister-exchange resulting in YLULORFDO UHVLGHQFH VHH :LOOLDPV DQG $\UHV IRU FODVVLF DQWKURSRORJLFDO DFFRXQWV )LJXUHVKRZVKRZVXFKGLUHFWH[FKDQJHVLPSDFWRQDVSHFWVRIWKHNLQVKLS WHUPLQRORJ\ LQ 1HQ 6LQFH H[FKDQJHG ZRPHQ VKRXOG FRPH IURP GLIIHUHQW FODQV DQG there is a strong chance that different clans will speak different languages, this makes it highly likely that a child’s mother will have married into the village from another language background, adopting her husband’s language after marriage (though possibly knowing it IDLUO\ZHOOEHIRUHWKURXJKSULRUH[SRVXUH 6LQFHGLIIHUHQWJHQHUDWLRQVLQDOLQHDJHW\SLFDOO\ exchange women with different clans, this regularly brings a large set of languages into WKHKRXVHKROGDQGLQWRWKHH[SHULHQFHRIWKHJURZLQJFKLOG)RUH[DPSOHD1HQVSHDNLQJ PDQ8PD\KDYHD1HQVSHDNLQJIDWKHU9ZKRPDUULHGDQ,GLVSHDNLQJZRPDQ:DQGLQ WXUQPDUULHVD1DPEXVSHDNLQJZLIH;8ZRXOGEHH[SHFWHGWRKDYHJRRGPDVWHU\RI1HQ (the language of his father’s clan), Idi (the language of his mother’s clan, whom he would visit regularly) and Nambu (the language of his wife, with whose clan he needs to maintain UHJXODUFRQWDFW ,WLVHYLGHQWWKDWE\FRQWLQXDOO\FUHDWLQJPXOWLOLQJXDOKRXVHKROGVLQD stable and recurring way, direct sister-exchange engenders conditions that favour language FRQWDFW DQG PXWXDO LQÀXHQFH VHH ¿JXUH ± ZH ORRN DW VRPH RI WKH FRQVHTXHQFHV LQ VHFWLRQ 2.2. IMPACT OF MODERN POLITICAL UNITS ON LANGUAGE USE. The impact of modern polities on the Southern New Guinea region has had very different effects on the two sides of the border, so that it is now one of the steepest economic and demographic gradients DFURVVDQDWLRQDOERXQGDU\WREHIRXQGDQ\ZKHUHLQWKHZRUOG ¿JV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages FIGURES 7 AND 8.9LHZVORRNLQJZHVWLQWR,QGRQHVLDDQGHDVWLQWR31*IURPWKHERUGHU SRLQWDW6RWDLQ,QGRQHVLD 3KRWRV 1(YDQV 2QWKH31*VLGHWKH7UDQV)O\LVDIRUJRWWHQUHJLRQ±SHUKDSVWKHSRRUHVWDQGPRVW LVRODWHGLQWKHFRXQWU\ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages DFTXLUHWUDGHJRRGVQRWDYDLODEOHLQWKH0RUHKHDGGLVWULFWLWVHOI7KHRYHUDOOSLFWXUHWKHQ is of solid retention of traditional language as part of a subsistence economy, traditional land rights, and a culture of multilingualism in both local languages and those of wider FRPPXQLFDWLRQ FIGURE 95LFHSDGG\LQDUHDRIFOHDUHGPHODOHXFDIRUHVW EHWZHHQ0HUDXNHDQG:DVXU 3KRWR 1(YDQV On the Indonesian side, rapid economic development and environmental change DFFRPSDQ\LQJ WKH LQÀX[ RI WUDQVPLJUDQWV LV SURFHHGLQJ DW D UDSLG SDFH DQG 0HUDXNH LV D ERRPLQJ ORFDO FHQWUH 0XFK ODQG KDV EHHQ FOHDUHG IRU ULFH FXOWLYDWLRQ ¿J E\ transmigrants from Java and other parts of Indonesia; roads have been established and are now lined with tokos (Indonesian-style roadside stores); there are police posts in every village and houses in villages like Wasur or Poo are now mostly built by the government UDWKHUWKDQE\ORFDOVWKHPVHOYHV6SHDNHUVRIWUDGLWLRQDOODQJXDJHVRIWKHDUHDDUHQRZ VLJQL¿FDQWO\ RXWQXPEHUHG E\ WUDQVPLJUDQWV IURP HOVHZKHUH LQ ,QGRQHVLD 2Q WKH RWKHU hand, access to education, electricity, health care and the means of earning money are all far ahead of what is available on the PNG side, so much so that some young Papua New Guineans are undertaking courses, such as in agriculture, on the Indonesian side of the ERUGHU,QWHUPVRIWKHHIIHFWRQODQJXDJH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 3. MAIN LINGUISTIC GROUPINGS IN SOUTHERN NEW GUINEA. The Southern New *XLQHDUHJLRQLVKRPHWRDURXQGODQJXDJHVVSOLWEHWZHHQVRPHQLQHODQJXDJHIDPLOLHV± UHSUHVHQWLQJRQRXUFXUUHQWNQRZOHGJH¿YHRUVL[PD[LPDOFODGHV LHXQUHODWDEOHXQLWV An indication of the relevant families (though not including all members of each family) LVJLYHQLQ¿JXUHDORQJZLWKDOLVWLQJRIVRXUFHVLQWDEOH7KHVSDWLDOGLVWULEXWLRQ of families suggests a sort of historical pincer movement by which Trans-New Guinea languages came down the Fly River to the north and east, and the Digul to the West, trapping the much more diverse languages of the Trans-Fly region between these rivers DQGWKHVRXWKHUQFRDVW7KXV6XNL*RJRGDODDQG7LULRWRWKHQRUWK.LZDLWRWKHHDVWDQG 0DULQG DQG0DURUL WRWKHZHVWDUHDOOSODXVLEOHEUDQFKHVRIWKH7UDQV1HZ*XLQHDIDPLO\ FIGURE 10. The (focal) Trans-Fly region, showing the main language groups DQG VHOHFWHG ODQJXDJHV IURP HDFK .DQXP MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages Family Main members $I¿OLDWLRQVDQGFRXQWHU Main sources on claims DI¿OLDWLRQ Marind 0DULQG TABLE 2. Main linguistic groupings in Southern New Guinea To the south, in the western part of the Torres Strait, is the language known in its GLDOHFWDOYDULDQWVDV.DOD.DZDZ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages The largest of these, with around 15 languages depending on how the language/dialect boundary is negotiated, has traditionally been called the ‘Morehead-Upper Maro family’, EXWLQWKLVDUWLFOH,ZLOOUHIHUWRLWE\WKHPRUHFRPSDFWWHUPµ 4. NEN AND IDI: DIVERGENT NEIGHBOURS. To give a feeling for how languages of the region work, as well as the balance of sameness and difference across neighbouring ODQJXDJHIDPLOLHV,ZLOOEULHÀ\VNHWFKWKHIXQFWLRQLQJRIWZRODQJXDJHV±1HQDQG,GL± which belong to different non-TNG families in the region, yet are spoken in neighbouring YLOODJHV DQG OLQNHG E\ FORVH WLHV RI LQWHUPDUULDJH DQG PXOWLOLQJXDOLVP 1HQ LV WKH HDVWHUQPRVWPHPEHURIWKH 5 ,H MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages Pahoturi River family and has around 1,600 speakers in several villages, such as Dimsisi DQG6LELGLUL There are close ties between speakers of these languages, reinforced by sister-exchange across the language boundary which produces widespread knowledge of each other’s languages and other interesting manifestations including place-names that are said to mix Nen and Idi elements, such as Sugäl (said to be comprised of Nen su ‘belly’ plus Idi gäl ‘canoe’) or Dudumae (Nen Dudu>ROGJDUGHQSODFHQDPH@SOXV,GLmae µKRXVH¶ $VLVWKH case more widely in the Morehead district, these languages are named after their respective word for ‘what’ (nen in Nen, idi in Idi), as if English were called Whattish, German Wasisch, French quoiais, and Russian štoskij$YDULDQWYHUVLRQRIWKHVHQDPHVLVWRXVHWKHIRUPIRU ‘what is it’, such as Nen Ym>ZKDWVJ EH@RUIdi Da>ZKDWVJ EH@VRPHRIWKHODQJXDJH QDPHVUHSRUWHGLQ5D\ DUHUHQGLWLRQVRIQDPHVRIWKLVW\SHVXFKDVµ1HQLXP¶ 5D\ IRUNen Ym7KHXVHRIVXFKVKLEEROHWKQDPLQJLVRQO\RQHPDQLIHVWDWLRQRID sophisticated metalinguistic awareness of structural, phonological and lexical differences IRXQGTXLWHZLGHO\RYHUWKHUHJLRQ 'HVSLWHWKLVWKHODQJXDJHVGLIIHUVLJQL¿FDQWO\RQPDQ\GLPHQVLRQVLQGHHGVRWKDWLI Nen’s relationship to its westerly neighbour Nambu is like Spanish to Portuguese or German WR'XWFKLWVUHODWLRQVKLSWRLWVHDVWHUO\QHLJKERXU,GLLVOLNH6SDQLVKWR%DVTXHRU*HUPDQ WR+XQJDULDQ,ZLOOLOOXVWUDWHWKLV¿UVWZLWKDEULHIVNHWFKRIKRZHDFKODQJXDJHORRNVRQ LWVRZQWKHQFRPSDUHDQXPEHURIUHOHYDQWW\SRORJLFDOIHDWXUHVPRUHV\VWHPDWLFDOO\ 4.1. NEN (ETHNOLOGUE CODE NQN). 6 Nen’s phonological inventory is given in tables DQG,WKDVUHODWLYHO\IHZSODFHVRIDUWLFXODWLRQQRYHODUQDVDODYRLFLQJFRQWUDVWDQG HLJKWYRZHOV LQFOXGLQJDFRXSOHRIVKRUWYRZHOVSOXVWZRPDUJLQDOQDVDOYRZHOV 7KH only somewhat unusual phonemes are the labial-velars, which are coarticulated at labial and velar places of articulation, though phonemes of this type are of course found in many RWKHUSDUWVRI0HODQHVLD HJ+XRQ3HQLQVXOD2QLQ3HQLQVXOD9DQXDWX $VLQDQXPEHU RIRWKHU3DSXDQODQJXDJHVVXFKDV.DODP %OHYLQV 3DZOH\'RQRKXH PDQ\ V\OODEOHVODFNVSHFL¿HGYRZHOQXFOHL WKHVHDUH¿OOHGLQZLWKEULHIHSHQWKHWLFVFKZDVZKLFK DUHQRWVKRZQLQWKHSUDFWLFDORUWKRJUDSK\ 6 'DWDSUHVHQWHGKHUHZHUHJDWKHUHGE\WKHDXWKRURYHU¿HOGWULSVWRWDOOLQJZHHNVEHWZHHQ DQG MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages Bilabial Alveolar Palatal 9HODU Labial-velar Glottal /dental 9RLFHOHVVVWRS p ৰ TABLE 3. The Nen Phoneme inventory: consonants Front Back Non-short Short (Short)7 Non-short High i (i) ܻ p u (u) Mid e (e) ܣ i o (o) Low aܭ l a (a) PDUJLQDOӁLQӁµ\HV¶DQGJӁKӁµRYHUWKHUH¶ TABLE 47KH1HQ3KRQHPHLQYHQWRU\ YRZHOV In terms of its grammatical typology, Nen has the following features: D SUHIHUHQFHIRUYHUE¿QDO (b) no verb-chaining but widespread use of true subordinate constructions using a QRPLQDOLVHGYHUEXVXDOO\LQÀHFWHGIRUFDVHDVLQ 7 This vowel can almost be eliminated as a phoneme, except in a couple of words, má and mái ‘still’ where the presence of á FDQQRWEHPRWLYDWHGE\HSHQWKHVLV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages (1) Ynd yergb-at one-s-t w-ng-m. 1ABS river-AL ¿VKZLWKQHWNLZR-AL 1sgU8:ĮDZD\EH µ,¶PJRLQJWRWKHULYHUWRQHW¿VK¶ F XVHRIVXI¿[HVRQWKH¿QDO13HOHPHQWWRPDUNDQDEVROXWLYHHUJDWLYHFDVHV\VWHPSOXV DQRWKHUGR]HQRUVRFDVHGLVWLQFWLRQV7KHVHVXI¿[HV DQGDOVRIUHHSURQRXQV DOVRHQFRGH DVLQJXODUQRQVLQJXODUGLVWLQFWLRQLQDOOEXWWKHDEVROXWLYHFDVH 1RWHWKDW1'VWDQGV for ‘non-dual’ (more on this below), and by not glossing the number of yndLHZULWLQJLW DVµ$%6¶,LQGLFDWHWKDWLWLVXQVSHFL¿HGIRUQXPEHU togetoge-yäbem ynd w-aka-ta-t / FKLOGUHQ3/(5*$%6 VJ8 ĮVHH1'QVJ$ yn-aka-ta-t QVJ8 ĮVHH1'QVJ$ µ7KHFKLOGUHQVHHPHXV RUPRUH ¶ G FRPSOH[ YHUE PRUSKRORJ\ LQYROYLQJ ERWK SUH¿[HV DQG VXI¿[HV DQG LQFOXGLQJ GRXEOHDJUHHPHQW DFWRUVXI¿[HVDQGXQGHUJRHUSUH¿[HVWKRXJKVRPHWLPHVDSDUWLFXODU FRPELQDWLRQ RI DFWRU DQG XQGHUJRHU ZLOO EH VKRZQ DW MXVW WKH VXI¿[DO RU SUH¿[DO VLWH GLUHFWLRQ WRZDUGV DZD\ QHXWUDO DQG GLDWKHVLV D UDQJH RI YDOHQF\FKDQJLQJ SUH¿[HV WRWKHURRW $FRPSOH[7$0V\VWHPFRPELQHVLQIRUPDWLRQIURPWKHYHUEDOVXI¿[HV GLVWLQFWLRQVHDFKIRUSHUIHFWLYHLPSHUIHFWLYHDQGQHXWUDODVSHFW WKHYHUEDOSUH¿[HV GLVWLQFWLRQVFRGHGE\GLIIHUHQWVHULHVRIXQGHUJRHUSUH¿[HV DQGSUHYHUEDOSDUWLFOHV (e) Monovalent verbs split in their agreement patterns, though not their case, according WRZKHWKHUWKHSUHGLFDWHLVVWDWLFRUG\QDPLF7KHVXEMHFWVRIVWDWLFYHUEVXVHXQGHUJRHU SUH¿[HV D DQGWKHVXEMHFWVRIG\QDPLFYHUEVXVHDFWRUVXI¿[HVDQGDSHUVRQLQYDULDQW µPLGGOH¶SUH¿[ E (3a) Ynd w-aki-ngr (3b) Ynd n-owab-ta-n $%6 VJ8 ĮEHVWDQGLQJ67$7$%6 0 ĮWDON1' ,3)9VJ$ µ,DPVWDQGLQJ¶ µ,DPWDONLQJ¶ 7KHXQGHUJRHUSUH¿[HVKDYHWKUHHVHULHVZKRVHVHPDQWLFVLVWRRFRPSOH[WRFDSWXUHZLWK DJORVVDQGIRUZKLFK,XVHWKH*UHHNOHWWHUVĮȕȖ,IZHMXVWORRNDWWKHLPSHUIHFWLYH VHULHVĮȕDQGȖZRUNEDFNZDUGVIURPWRGD\LQWRWKHIXWXUH WKHĮIRUPnowabtan is ‘imperfective non-past’ (roughly) and refers to me talking any time from this morning’s GDZQ RQZDUGV ZLWK ¿QHU VSHFL¿FDWLRQ E\ SUHYHUEDO SDUWLFOHV WKH ȕIRUP k-owab-ta-n >0 ȕWDON1' ,3)9VJ$@ LV µLPSHUIHFWLYH \HVWHUGD\ SDVW¶ DQG UHIHUV WR PH WDONLQJ 8 $ $FWRU VXEMHFWRIWUDQVLWLYHRURIG\QDPLFLQWUDQVLWLYH 8 XQGHUJRHU REMHFWRIWUDQVLWLYH VXEMHFWRIVWDWLYH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages \HVWHUGD\RUDIHZGD\VDJRZKLOHWKHȖIRUPg-owab-taw-n>0 ȖWDON1' 5(0,3)9 VJ$@LVµLPSHUIHFWLYHUHPRWHSDVW¶DQGUHIHUVWRPHWDONLQJDWDQ\WLPHEHIRUHWKDW ODVW PRQWKODVW\HDUHWF ,IWKDWZDVDOOWKHUHZDVWRWKHWKUHHVHULHVJORVVLQJWKHPZRXOGEHHDV\%XWLIZH look at other functions of each series the picture becomes muddier: in addition to its LPSHUIHFWLYHQRQSDVWXVHWKHĮVHULHVLVXVHGIRUSHUIHFWLYHVLQWKHSDVW LHWKHGLUHFWLRQ RIWLPHUHFNRQLQJÀLSVRYHULQWKHSHUIHFWLYH IRUIXWXUHLPSHUDWLYHV GRLWODWHU DQGWZR RIWKHµQHXWUDODVSHFW¶FDWHJRULHV ZKLFKLQFOXGHDFRXSOHPRUHUHPRWHSDVWV 7KHȕVHULHV in addition to its yesterday past use in the imperfective, is used for present imperatives GRLWQRZ DQGZLWKDIHZYHUEVIRUSHUIHFWLYHVGHQRWLQJXQH[SHFWHGRFFXUUHQFHV7KH ȖVHULHVXVHGIRUWKHUHPRWHSDVWLQWKHLPSHUIHFWLYHLVXVHGLQWKHSHUIHFWLYHIRUIXWXUHV DQRWKHUWLPHÀLS DVZHOODVIRUPHGLDWHGLPSHUDWLYHVWUDQVPLWWHGYLDDPHVVHQJHU ; VKRXOGGRLW FRQYH\P\FRPPDQGWR; DQGIRUWKHLUUHDOLV Given the semantic disparities between these uses, the best treatment is to regard the FKRLFHRISUH¿[DOVHULHVSOXVWKH7$0VXI¿[DVIRUPLQJDVLQJOHFLUFXP¿[DOVLJQ VHH UHPDUNVODWHURQFLUFXP¿[DOSDUDGLJPV DQGRQFHZHDGRSWWKDWWUHDWPHQWWKHJORVVLQJ GLI¿FXOWLHVYDQLVKVLQFHWKHSUH¿[VHULHVDUHQRWUHTXLUHGWRKDYHDQ\PHDQLQJRIWKHLURZQ )RUIXUWKHUUHPDUNVRQWKLVSUREOHPVHH(YDQV IRUWKFRPLQJE (f) the existence of a large set of positional verbs (around 30), which in addition to PHDQLQJVOLNHµEHVWDQGLQJ¶LQ D RIWHQKDYHYHU\VSHFL¿FVHPDQWLFV HJµEHLQDWUHH IRUN¶ µEH LPPHUVHG¶ DQG ZKLFK IRUP WKH OLRQ¶V VKDUH RI WKH VWDWLYH SUHGLFDWHV )URP these, transitives (‘cause to be in position X’) and middles (‘become in position X’) are WKHQGHULYHG$OOSRVLWLRQDOYHUEVDUHSUH¿[LQJYHUEVLQWKHVHQVHRIXVLQJRQO\SUH¿[HVWR VLJQDOSHUVRQDJUHHPHQWLQIRUPDWLRQ (g) an unusual ‘constructive’9 number system within the verb which obtains three values10 by crossing the singular vs non-singular contrast of the agreement morphology with a dual YVQRQGXDOFRQWUDVWRQWKHURRW )LJD RUWKHYHUEWKHPDWLF )LJE 9 7KHRULJLQDOWHUPXVHGIRUWKLVW\SHRIV\VWHPZDVµFRQVWUXFWHG¶ &RUEHWW EXWLQPRUH UHFHQWSXEOLFDWLRQV HJ$UND WKHWHUPµFRQVWUXFWLYH¶KDVEHHQXVHGDQG,IROORZWKDWYDULDQW KHUH 10 There is also an incipient but much more irregular system of distinguishing small vs large, or partial vs exhaustive plurals, which I don’t discuss here, but which underlies my reluctance to use µSOXUDO¶KHUHDVLILWZHUHDQXQSUREOHPDWLFWHUP MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages Root 83UH¿[ patterning ,QÀHFWHGIRUP sg w- m w-m ‘I am’ pl yn- m yn-m ‘we (more than two) are’ du yn- ren yn-ren ‘we two are’ FIGURE 11A. 8QL¿FDWLRQRIDI¿[DOVLQJXODUYVQRQVLQJXODUDJUHHPHQWYDOXHVZLWKGXDOYV non-dual suppletive root of ‘be’ to give a three-valued basic number system Thematic 83UH¿[ patterning ,QÀHFWHGIRUP sg -n nowabta nowabtan ‘I talk’ pl -m nowabta nowabtam ‘we (more than two) talk’ du -m nowab nowabm ‘we two talk’ FIGURE 11B. 8QL¿FDWLRQRIDI¿[DOVLQJXODUYVQRQVLQJXODUDJUHHPHQWYDOXHVZLWKGXDOYV non-dual thematic forms of ¥RZDE ‘talk’ to give a three-valued basic number system; n- LVDSHUVRQQXPEHULQYDULDQWµPLGGOHSUH¿[¶ Though constructive number systems are not all that unusual cross-linguistically (see &RUEHWW $UNDWKLVYROXPH WKHXVHRIDSHUYDVLYHGXDOYVQRQGXDO RSSRVLWLRQLVDVIDUDV,NQRZXQLTXHWR1HQDQGLWVFORVHUHODWLYHV (h) a general tendency to exploit distributed, paradigmatic, and FRQVWUXFWLYHXQL¿FDWLRQDO DUFKLWHFWXUHVWRJLYHFRPSOHWHJUDPPDWLFDOIHDWXUHVSHFL¿FDWLRQV It is distributed because there is a strong tendency to underspecify information at one VLWH HJJLYLQJSHUVRQEXWQRWQXPEHULQWKHDEVROXWLYHSURQRXQIRUPV ZKLFKLVWKHQ ¿OOHGLQE\XQL¿FDWLRQZLWKLQIRUPDWLRQDWDQRWKHUVLWH HJWKHYHUEFRQWULEXWHVQXPEHU LQIRUPDWLRQZKLOHWKHSURQRXQFRQWULEXWHVSHUVRQLQIRUPDWLRQ Complete feature value VHWVDUHQRWSUHVHQWXQWLOPDWHULDOIURPERWKDI¿[SRVLWLRQVDQGIURPIUHHSURQRXQVKDV EHHQ XQL¿HG WDEOH $V FDQ EH VHHQ WKH DEVROXWLYH SURQRXQV RQO\ VKRZ SHUVRQ QRW QXPEHU±yndµVWSHUVRQDEV DQ\QXPEHU ¶bmµQGSHUVRQDEV DQ\QXPEHU ¶11 bä 11 $ SHFXOLDULW\ RI 1HQ LV WKDW WKH VJ DQG VJ IRUPV ynd and bm respectively, neutralise the DEVROXWLYHYVHUJDWLYHFDVHGLVWLQFWLRQIRXQGHYHU\ZKHUHHOVHLQWKHV\VWHP7KLVDSSHDUVWRUHVXOW IURPDUHFHQWVRXQGFKDQJHE\ZKLFKWKHRULJLQDOHUJDWLYHVLQJXODUSURQRPLQDOVXI¿[o was ORVWIURPWKHVHSURQRXQVDVSDUWRIDJHQHUDOORVVRIZRUG¿QDOo±FI1DPDZKLFKFRQWUDVWV absolutive \ΩQG and IΩPto ergative \ΩQGR and IΩPR and Nambu which contrasts absolutive \ΩQG and EΩP with ergative \ΩQG and EΩPR 1RWHLQSDVVLQJWKDWWKHORVVRI¿QDOo is one of the main ;ۆVRXUFHVRIWKHFRDUWLFXODWHGODELDOYHODUSKRQHPHVLQ1HQ±FI1DPEXPΩQJR ‘house’, Nen PQ leave’, Nambu ingo, Nama injo- ‘catch sight of, see’, Nen‘ -ۆNama frango- ‘leave’, Nen EUDQ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 130 µUGSHUVRQDEV DQ\QXPEHU ¶&RQYHUVHO\DI¿[HVUHOLDEO\VKRZQXPEHUEXWQRWSHUVRQ V\QFUHWLVPV PHUJH WKH QG DQG UG SHUVRQV LQ WKH$ DQG 8 DI¿[ SRVLWLRQV ya-~yä- is µ_QVJ8 Į¶ DQG -e LV µ_VJ$¶2QFH IUHH SURQRXQV DQG LQÀHFWHG YHUEV DUH XQL¿HG DOO ambiguities are eliminated: Free WDON _VJ$ WDON _GX WDON _SO pronoun nowabte nowabt nowabtat bm bm nowabte bm nowabt bm nowabtat ‘you (sg) talk’ ‘you two talk’ ‘you (3(+)) talk’ 3 bä bä nowabte bä nowabt bä nowabtat ‘(s)he talks’ ‘they two talk’ ‘they (3(+)) talk’ TABLE 5 8QL¿FDWLRQRIXQGHUVSHFL¿HGSURQRXQDQGDJUHHPHQWLQIRUPDWLRQWRJLYH SUHFLVHSHUVRQQXPEHUVSHFL¿FDWLRQ It is paradigmatic (and sometimes even circumparadigmatic) because the information IURPSUH¿[DQGVXI¿[RIWHQQHHGVWREHWUHDWHGDVSDUWRIDVLQJOHSDUDGLJPZLWKIRUPV KDYLQJYHU\GLIIHUHQWYDOXHVDFFRUGLQJWRWKHLUSODFHLQWKHSDUDGLJP7KXVZLWKµQHXWUDO DVSHFW¶7$0VXI¿[HVWKHVXI¿[DOSDLU-nd vs -t FRQWUDVWVSOYVGXEXWWKHLUYDOXHVDUH VZDSSHG LHGXYVSO ZLWKSHUIHFWLYHDVSHFW7$0VXI¿[HV/LNHZLVHWKHȖVHULHVRI XQGHUJRHUSUH¿[HVLQGLFDWHVUHPRWHSDVWZKHQFRPELQHGZLWKLPSHUIHFWLYHYHUEVXI¿[HV EXWIXWXUHZKHQFRPELQHGZLWKSHUIHFWLYHRQHV $QG±LQWLPDWHO\OLQNHGWRWKHSUHFHGLQJFKDUDFWHULVWLFV±LWLVFRQVWUXFWLYHXQL¿FDWLRQDO because the full range of categories once combinations are taken into account is much JUHDWHUWKDQWKDWIRXQGDWDQ\FRQWUDVWVLWH1RWHWKDWVXFKXQL¿FDWLRQQHHGVWRWDNHSODFH ERWKZLWKLQWKHZRUG HJEHWZHHQWKHSUH¿[LQJDQGVXI¿[LQJVLWHVRIWKHYHUEV DQGEHWZHHQ WKHYHUEDQGIUHHSURQRXQV HJLQZRUNLQJRXWWKHIXOOSHUVRQQXPEHUVSHFL¿FDWLRQVIRU XQGHUJRHUV 1RWHWKDWWKHVHFKDUDFWHULVWLFVFUHDWHGLI¿FXOWLHVIRULQWHUOLQHDUJORVVLQJ DVWKH\GRLQ ,GL DQG,DGRSWWKHIROORZLQJWZRQRQVWDQGDUGFRQYHQWLRQVLQWKHH[DPSOHVWKDWIROORZ )LUVW,XVHWKHSLSH _ WRMRLQGLVMXQFWIHDWXUHYDOXHV ZKLFKDUHWKHQGLVDPELJXDWHGWKURXJK IHDWXUHXQL¿FDWLRQ VXFKDV_VJIRUµVHFRQGRUWKLUGSHUVRQVLQJXODU¶LQ 6HFRQGDV DOUHDG\PHQWLRQHGDERYH,XVH*UHHNOHWWHUV ĮȕȖ IRUFRQWUDVWLQJSUH¿[VHULHVZLWKRXW FOHDUO\VSHFL¿DEOHVHPDQWLFVRIWKHLURZQZKHUHWKLVLVRQO\µFDVKHGLQ¶DIWHUXQLI\LQJWKLV LQIRUPDWLRQZLWKRWKHUSDUWVRIWKHSDUDGLJP VXFKDVWKHVXI¿[HV A further salient feature of Nen, particularly important for historical and comparative WKHSUHQDVDOLVHGSDODWDO µVHHFDWFKVLJKWRI¶+RZHYHUWKLVRQO\RFFXUVZLWK¿QDOYHODUVۆLQ affricate nj in Nama results in this case from palatalisation after the preceding i $IWHURWKHU VHJPHQWVVXFKDVQGRUP¿QDORVLPSO\GLVDSSHDUHGZLWKRXWWUDFH %\ZKLFK,PHDQWKDWSUH¿[HVDQGVXI¿[HVQHHGWREHFRPELQHGLQWRDVLQJOHSDUDGLJPWKDWLVRQO\ SDUWLDOO\IDFWRULVDEOHLQWRVHSDUDWHSUH¿[DODQGVXI¿[DOSDUDGLJPV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 131 purposes, is the almost total disconnect13 between the form of free pronouns and that of agreement morphology for verbs: table 6 compares the absolutive and possessive IUHH SURQRXQV ZLWK WKH WKUHH VHULHV RI XQGHUJRHU SUH¿[HV DQG WKH DFWRU VXI¿[HV EDVLF LPSHUIHFWLYHDQGSDVWSHUIHFWLYHVHWV ,QIDFWZKHQRQHORRNVULJKWDFURVVWKH $EV 3RVV 8SUH¿[ 8SUH¿[ 8SUH¿[ $VXI¿[ $VXI¿[ Į ȕ Ȗ LPSHUI SDVWSHUI n -n- ۆ -1sg ynd tande w- q 1nsg ynd tbende yn- tn- dn- -m -m VJ bm bende n- k- g- -e -ø QVJ bm bbende ya-~yä- ta-~tä- da- ~dä- -t -t/-nd14 3sg bä yande y- t- ~ d- -e -a 3nsg bä ybende ya-~yä- ta- ~ tä- da- ~ dä- -t -t/-nd TABLE 6. Free pronouns and corresponding verbal agreement forms in Nen We will mention a few further typological features below (see also Evans forthcoming a,b), but this is now a good point to give a global overview of the language by tackling the following mini-text, which can be heard on KWWS VFKRODUVSDFHPDQRDKDZDLLHGX ELWVWUHDPKDQGOH1HQ%ODFNVQDNH([FHUSWZDY,WLVDQDFFRXQWRIWKHGDQJHUV of being bitten by a Papuan Black Snake, recorded by the present author from the late $UDPDQJ:OLOD WKHQDJHGLQKLVHDUO\V LQ6HSWHPEHU According to the story, when a Papuan Black Snake bites you, blood starts pouring out of your eyes, and people check how bad you are by asking: snamb bnz aba ya-wakae-w-ng KRZBPDQ\ ¿UH ,PP3VW _QVJ8 ĮVHH,03)'8_VJ$!'8 ,03) snamb är aba ya-wakae-w-ng KRZBPDQ\ SHUVRQ ,PP3VW _QVJ8 ĮVHH,03)'8_VJ$!'8 ,03) µ³+RZPDQ\¿UHVGLG\RXVHH"+RZPDQ\SHRSOHGLG\RXVHH"¶ >SHUKDSVEHWWHUWUDQVODWHGDVµ'LG\RXVHHVRPDQ\¿UHVWZR"¶'LG\RXVHHVRPDQ\ SHRSOHWZR"@ 13 2I WKH IRUPV JLYHQ LQ WDEOH RQO\ WKH QVJ XQGHUJRHU SUH¿[ VKRZV DQ\ SODXVLEOH IRUPDO FRQQHFWLRQWRWKHIUHHSURQRXQV -XVWLQDFRXSOHRIWKHSHUIHFWLYHVHULHVWKHQGDQGUGSHUVRQDFWRUVXI¿[HVGLVWLQJXLVKGXDO KHUH -nd) from plural (here -t MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages a snamb kesär ya-waka-t-e DQG KRZBPDQ\ VXQ _QVJ8 ĮVHH,03) 1'_VJ ,03) ‘”and how many suns can you see?”’ dene geä g-owab-ta-ng-a deneya[més] WKXV &21' 0 ȖWDON,03)1' )XW3IVJ$ OLNHBWKLV ‘If he says like this:’ “sombes-ngama, sombes-ngama y-ng-aka-ta-n” WZR$%/ WZR$%/ VJ8 Į$:$<VHH1' ,03)QVJ$ µ³,VHHWZRRIHDFK´¶ aa gn-anma-ng-a… XP VJ8 Ȗ %(1 FDOO1' )XW3IVJ)XW3I$ ‘Um, the man will call out to you…’ är-t da-w-anma-nga de yna är-äm geym ti bä da-w-anma-ng-a WKLV PDQ(5* )2& "$%6 _QVJ8 Ȗ%(1FDOO1' )XW3IVJ$ µ7KHSHUVRQZLOOFDOOWKHSHRSOH¶ “tä-n-m, wgd zer-s _QVJ8 ȕKLWKHUEH 1' SURSHU ELWH,1) aba y-ze-n-e, ,003VW VJ8 ĮELWH,03)1'_VJ$ ‘“Come, it has bitten him good and proper’ “a kr kaka y-m” DQG GHDWK QHDU VJ8 ĮEH µ´DQGLVDERXWWRGLH´¶ These few lines of text illustrate many of the salient features of Nen morphosyntax, MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 133 UHLQIRUFLQJWKHVLPSOL¿HGH[DPSOHVJLYHQDERYHZLWKUHDOWH[WXDOPDWHULDO (a) the existence of two alignment systems±DQDEVROXWLYHHUJDWLYHV\VWHPIRUFDVHDQGD split-S system for agreement (on the basis of a stative vs dynamic contrast rather than an DJHQF\FRQWUDVW ¿J 7KHFDVHV\VWHPLVHUJDWLYHDEVROXWLYHRSSRVLQJDQHUJDWLYHIRUP for the agent of transitives (är-ämµPDQ(5*¶LQ WRDQ XQPDUNHG DEVROXWLYHIRUP for the patient of transitives (ärLQ DQGWKHVROHDUJXPHQWRILQWUDQVLWLYHVZKHWKHU G\QDPLFRUVWDWLYH WKHUHLVDOVRDGDWLYHIRUUHFLSLHQWVEHQH¿FLDULHV7KHHUJDWLYHLVIXOO\ VSHFLDOLVHGIRUWKLVIXQFWLRQDQGGRHVQRWPDUNDQ\REOLTXHIXQFWLRQ HJLQVWUXPHQWRU VRXUFH 7KHYHUEDOLQGH[LQJV\VWHPHPSOR\VDQµXQGHUJRHU¶SUH¿[IRUSDWLHQWVRIWUDQVLWLYHV HJy- in yzene ‘it bit him¶ LQ DQGWKHVROHDUJXPHQWRIVWDWLYHV HJy- in ym ‘he LV¶LQ DQGDQµDFWRU¶SUH¿[IRUDJHQWVRIWUDQVLWLYH HJ-eµ_VJ¶LQyawakate in DQGLQyzeneLQ DQGRIG\QDPLFPRQRYDOHQWYHUEV HJ-a µVJ)3I$LQ 7KHµXQGHUJRHU¶SUH¿[ REYLRXVO\WKHWHUPLVQRWSHUIHFW LVDOVRXVHGIRUWKHUHFLSLHQWRU EHQH¿FLDU\RIGLWUDQVLWLYHYHUEV Ditransitive Case marking Dative Ergative 9HUEDOLQGH[LQJ U-: IO -A: A Transitive Case marking Absolutive Ergative 9HUEDOLQGH[LQJ U-: O -A: A Intransitive Stative: Dynamic: Case marking Absolutive Absolutive 9HUEDOLQGH[LQJ U-: Sstat -A: Sdyn FIGURE 12 5ROH VSOLWV DQG PHUJHUV FDVHPDUNLQJ DQG YHUEDO LQGH[LQJ 8 DQG $ UHSUHVHQW WKH XQGHUJRHU SUH¿[ DQG DFWRU VXI¿[ UHVSHFWLYHO\ V\QWDFWLF UROHV are represented by A, Sstat, Sdyn,2DQG,2,QDGGLWLRQWRWKHUROHVVKRZQKHUHLQ ditransitives there is an O, marked with the absolutive case, which is not indexed RQWKHYHUE (b) DVSOLWLQPRUSKRORJLFDORUJDQLVDWLRQEHWZHHQSUH¿[LQJYHUEV (monovalent, stative, HJym LQ 15 and DPEL¿[LQJYHUEV ,XVHWKLVWHUPIRUYHUEVZKLFKWDNHERWKSUH¿[HV DQGVXI¿[HV16 7KHODWWHUPD\EHGLYDOHQWOLNHdawanmanga ‘he will call out to them’ in 15 7KHVWDWLYHFKDUDFWHULVDWLRQOHDNVVOLJKWO\,WKROGVRIWKHEDVHIRUPµEH¶SOXVDURXQGWKLUW\µSRVDOV¶ JLYLQJSRVLWLRQ EHLQDWUHHIRUN RUSRVWXUH EHVLWWLQJ %XWWKUHHYHUEVGHI\WKHFKDUDFWHULVDWLRQ RIWKLVFDWHJRU\DVVWDWLYH±µFRPH¶DQGµJR¶ZKLFKDUHWKHµWRZDUGV¶DQGµDZD\¶IRUPVRIµEH¶DQG hence may simply be inheriting the morphology of their source verb, but also utanµZDON¶ 16 $QRWHRQWKLVWHUPLQRORJLFDOFKRLFH WKHUHDVRQ,GRQ¶WXVHµFLUFXP¿[LQJ¶KHUHLVWKDWDPEL¿[LQJ DOORZV IRU WKH SRVVLELOLW\ WKDW FKRLFHV LQ WKH SUH¿[ DQG VXI¿[ DUH LQGHSHQGHQW LH UHSUHVHQW RUWKRJRQDO FDWHJRULHV ZKHUHDV FLUFXP¿[LQJ LPSOLHV WKDW PDWHULDO IURP SUH¿[ DQG VXI¿[ JHWV LQWHJUDWHG LQWR D VLQJOH VHPDQWLF YDOXH 2I FRXUVH WKH PRUSKRORJLFDO IDFW RI D YHUE EHLQJ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages RUPRQRYDOHQWDQGG\QDPLFOLNHgowabtanga µKHZLOOVD\¶LQ DVZHOODVWULYDOHQW OLNHµJLYH¶ QRH[DPSOHVLQWKLVWH[W 1HQKDVDQXQXVXDOO\ODUJHQXPEHURIPLGGOHYHUEV (Evans forthcoming a), assigning virtually all dynamic one-place predicates to this class HJ WDON ZRUN DVFHQG DV ZHOO DV PRUH W\SLFDO PLGGOHV OLNH GHULYHG UHÀH[LYHV DQG UHFLSURFDOV (c) Four sites for encoding TAM: Coding of tense/aspect/mood is split across Time adverbsDOORIZKLFKDUHELGLUHFWLRQDOHJkae ‘yesterday, tomorrow; one day from today’ Preverbal particlesZKLFKDUHXQLGLUHFWLRQDOHJabaµMXVWQRZYHU\UHFHQWO\¶ geä µLIZKHQ¶ 8QGHUJRHUSUH¿[VHULHVZKLFKKDYHWKUHHVHWVHQFRGLQJ7$07KHVHPDQWLFVRIWKHVHLV QRWVWUDLJKWIRUZDUGDQGFDQQRWEHVSHFL¿HGXQWLOWKH\FRPELQHZLWK7$0VXI¿[HVDQG SUHYHUEDOSDUWLFOHV,QRXUVDPSOHWH[WWKH_QVJ8SUH¿[LVH[HPSOL¿HGZLWKDOOWKUHH YDOXHV ĮIRUPya-LQ DQG ȕIRUPtä- (an allomorph of ta- LQDQGȖIRUP da-LQDQG$VWKHVHIRUPVLOOXVWUDWHWKHĮVHULHVDUHJOLGHVRUQDVDOVWKHȕVHULHV DUHWKHFRUUHVSRQGLQJYRLFHOHVVVWRSVDQGWKHȖVHULHVDUHWKHYRLFHGFRUUHVSRQGHQWVRIWKH ȕIRUPV ,QWKHVHH[DPSOHVWKHĮVHULHVLVDVVRFLDWHGZLWKSUHVHQWDQGUHFHQWSDVWWKHȕVHULHV ZLWK WKH LPSHUDWLYH DQG WKH ȖVHULHV ZLWK WKH IXWXUH %XW WKLQJV DUH QRW DOZD\V VR VWUDLJKWIRUZDUG ZLWKLPSHUIHFWLYHLQÀHFWLRQVWKHȖVHULHVVLJQDOVUHPRWHSDVWUDWKHUWKDQ IXWXUHDQGWKHȕVHULHVVLJQDOVWKHSDVWRI\HVWHUGD\RUDFRXSOHRIGD\VDJR 6XI¿[VHULHV, expressing TAM + number + actor person/number (it is usually possible to VSOLWWKHVHIXUWKHULQWRDµWKHPDWLF¶IROORZHGE\DµGHVLQHQFH¶ VHH(YDQVIRUWKFRPLQJE )RUDPEL¿[LQJYHUEVWKHVHIRUPQLQHVHWVGLYLVLEOHLQWRWKUHHDVSHFWVHULHV SHUIHFWLYH LPSHUIHFWLYHDQGQHXWUDO HDFKFRQWDLQLQJWKUHHYDOXHV )RUSUH¿[LQJYHUEVWKHSRVVLELOLWLHV DUHPXFKPRUHOLPLWHG 7KHFXUUHQWWH[WH[HPSOL¿HVVRPHRIWKHVH WKH EDVLF LPSHUIHFWLYH ZKLFKFRYHUVDOOLPSHUIHFWLYHLQGLFDWLYHVH[FHSWWKHUHPRWHDQGWKH IXWXUH (d) HPSOR\PHQWRILQ¿QLWLYHIRUPV1HQGRHVQRWKDYHDQ\IRUPRIYHUEFKDLQLQJRUVZLWFK UHIHUHQFH5DWKHULWPDNHVIUHTXHQWXVHRILQ¿QLWLYHIRUPVIRUDZKROHUDQJHRIIXQFWLRQV such as complement clauses of various types, as well as a sort of emphatic construction, H[HPSOL¿HGLQ LQZKLFKWKHLQ¿QLWLYHIRUPRIWKHYHUE zers ‘to bite’) is combined ZLWKDQLQÀHFWHGIRUP yzene µKHELWLW¶ WRPHDQVRPHWKLQJOLNHµKHUHDOO\ELWKLP¶ OLWµKH DPEL¿[LQJGRHVQRWSUHFOXGHWKDWVRPHRUDOORIWKHSUH¿[VXI¿[FRPELQDWLRQVIXQFWLRQDV FLUFXP¿[HVEXWLWDOVROHDYHVRSHQWKHSRVVLELOLW\WKDWWKH\DUHLQGHSHQGHQW MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 135 ELWKLPWRELWH¶ ,Q¿QLWLYHVDUHIRUPHGE\DGGLQJ-sWRWKHYHUEURRW17 $QLPSRUWDQWXVHRIWKHLQ¿QLWLYHLQ1HQQRWLOOXVWUDWHGLQWKHWH[WIUDJPHQWLVDVWKH FRPSOHPHQWRISKDVDOVVXFKDVµEHJLQWR9¶RUµ¿QLVK9LQJ¶H[SUHVVHGE\FRPELQLQJWKH LQ¿QLWLYH VXI¿[HGZLWKDQDSSURSULDWHFDVH ZLWKDSKDVDODX[LOLDU\7KHDX[LOLDU\FDUULHV DOOLQÀHFWLRQDOPDWHULDOWKDWWKHOH[LFDOYHUEZRXOGKDYHERUQHKDGLWEHHQ¿QLWH±PLGGOH SUH¿[SOXVDFWRUVXI¿[ZLWKµUHWXUQ LWU ¶LQ D XQGHUJRHUSUH¿[SOXVDFWRUVXI¿[ZLWK µVWDQGXS WU ¶LQ E DQGXQGHUJRHUSUH¿[EHQHIDFWLYHSUH¿[DQGDFWRUVXI¿[ZLWKµJLYH¶ LQ F VW QRSDSQGPۆ5a) .LVW y-a-pap-nd-nۆ5b) Ynd bä ZQ) VJ$ $%6 75VWDQGXS1/=5$/ VJ8 Į&$8EHJLQ1' 3)9 367VJ$ µ,DPEHJLQQLQJWRDERXWWRWU\LQJWRVWDQGKLPXS¶ (5c) Ahã Gbae ynd begta tande yép KHUH\RXDUH >QDPH@ VJ$ VJ '$7 VJ3266 EDJ $%6 räm-s-t n-ng-a-wa-pap-nd-n. JLYH1/=5$/ VJ8DZD\%(1&$8EHJLQ1' 3)9VJ$ µ+HUH*EDH,¶PDERXWWRJLYH\RXP\EDJ¶ 7KLVFRQFOXGHVRXUVKRUWVNHWFKRI1HQ)RUFRPSDULVRQZHQRZWUDYHODERXWNPHDVW IURP%LPDGEQWRWKHQHLJKERXULQJYLOODJHRI'LPVLVL6LQFHWKHUHLVQHJOLJLEOHSXEOLVKHG material on languages of the Pahoturi River family, this will also give a chance to give the SXEOLFDWOHDVWDVPDOOJOLPSVHRIKRZODQJXDJHVLQWKDWIDPLO\ZRUN To give an initial idea of the degree of difference between the languages, we can compare their paradigms of free pronouns, which show negligible18 resemblances; for comparison the free pronouns are also given (in blue and red respectively) for Nama and Nambu19 NPDQGNPWRWKHZHVWRI1HQ WDEOH 17 7KHVJ$IRUPLQ yzene, replaces the r with n7KLVLVDUHJXODUSURFHVVZLWKYHUEVZKRVH stems end in -rEHIRUHQRQGXDO%XWWKHr of the imperative can be seen clearly in imperfective QRQGXDOIRUPVHJyzertµWKHWZRRIWKHPELWKLP¶DQGLQSHUIHFWLYHLPSHUDWLYHVHJtzer ‘bite KLP QHZO\LQLWLDWLQJWKHDFWLRQ ¶ 18 One could seize on the presence of b- LQ QG DQG UG SHUVRQ IRUPV DV D YHVWLJH RI SRVVLEOH UHODWHGQHVV ,Q RWKHU FDVHV DSSDUHQW VLPLODULWLHV HJ 1HQ QVJ DEV bm ,GL QVJ DFF bibim) are coincidental in the sense that the m in Nen is part of the root whereas the -m in Idi marks DFFXVDWLYH 19 ,WKDQN-HII6LHJHOIRUVXSSO\LQJPHZLWKWKHVHIRUPV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 136 Nen (with Nama in small blue and Nambu Idi in small red) Abs Erg Poss Nom Acc Poss 1sg ynd ynd tande ƾΩQ bom bo \ԥQG \ԥQGR tane \ԥQG \ԥQGR tande 1nsg ynd yndbem tbende bi ba ba \ԥQG \ԥQGIHP WԥIHQH \ԥQG \ԥQGYHP WԥYHQGH QVJ ynd yndbem tbende ybi yba yba \ԥQG \ԥQGIHP WԥIHQH \ԥQG \ԥQGYHP WԥYHQGH VJ bm bm bende be babom béna fom IԥPR fene EԥP EԥPR bende QVJ bm bmbem bbende be bibim béna fom IԥPRIHP IԥIHQH EԥP EԥPRYHP EԥYHQGH 3sg bä ymam yande bo obom obo fæ \ԥPR yæne bæ \ԥPR yænde 3nsg bä ymabem ybende bo ubim oba fæ \ԥPRIHP \ԥIHQH bæ \ԥPRYHP \ԥYHQGH TABLE 7)UHHSURQRXQVLQ1HQ1DPD VPDOOEOXHIRQW 1DPEX VPDOOUHGIRQW DQG,GL 4.2. IDI (ETHNOLOGUE CODE IDI). Idi is spoken in the three villages of Dimsisi, 6LELGLUL DQG 'LPLUL E\ D SRSXODWLRQ RI DURXQG SHRSOH7RJHWKHU ZLWK WKUHH RWKHU QDPHGYDULHWLHV±(QGH$J|EDQG7DHPH±LWIRUPVWKH3DKRWXUL5LYHU)DPLO\&RPSDUHG WRWKH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 137 ourselves to phonemes used by all speakers then there do not appear to be labial-velars, WKRXJKWKHUHDUHYHODUVZLWKDUDWKHUOD[URXQGHGUHOHDVH7KHFRQVRQDQWLQYHQWRU\LVVKRZQ LQWDEOH MANNER / RETRO- LAMINO- LABIO- CO-ARTICULATED PLACE BILABIAL ALVEOLAR FLEX PALATAL VELAR VELAR LABIAL-VELAR 9RLFHGVWRS b d t TABLE 8. Idi Consonant inventory (with proposed orthographic symbols in angle brackets) ,QWHUPVRIJUDPPDUWKHUHDUHVRPHJURVVW\SRORJLFDOVLPLODULWLHVZLWK1HQ%RWKDUH YHUE¿QDOERWKLQÀHFWWUDQVLWLYHYHUEVZLWKERWKSUH¿[HVDQGVXI¿[HVERWKKDYH7$0 VHQVWLYHIRUPVRIWKHSUH¿[VHULHVDQGERWKKDYHLQ¿QLWLYHSOXVDX[LOLDU\FRQVWUXFWLRQV LQZKLFKWKHDX[LOLDU\LQGH[HVDOODUJXPHQWVRIWKHLQ¿QLWLYHYHUE+RZHYHUWKHUHDUHQR YHUEVZKLFKXVHSUH¿[HVDORQHWRVLJQDOVXEMHFWDJUHHPHQWLQWKHZD\WKDWLVIRXQGZLWK µSUH¿[LQJYHUEV¶OLNHWKHFRSXODRUWKHSRVLWLRQDOYHUEVLQ1HQ DOOLQWUDQVLWLYHYHUEVLQ,GL LQFOXGLQJWKHLQWUDQVLWLYHDX[LOLDU\DQGWKHFRSXODPDNHH[FOXVLYHXVHRIVXI¿[DWLRQIRU DJUHHPHQWSXUSRVHV The complex architectural relationship between free pronouns and agreement morphology also shows typological similarities to Nen: there is a severe disconnect between both the forms and the categories of free pronouns and verbal agreement, with ZLGHVSUHDGEXWQRQFRUUHODWHGV\QFUHWLVPVLQHDFKV\VWHPZKLFKUHTXLUHWKHXQL¿FDWLRQRI LQIRUPDWLRQIURPERWKIUHHSURQRXQVDQGLQÀHFWHGYHUEVEHIRUHWKHSUHFLVHIHDWXUHYDOXHV FDQEHNQRZQDVZHVKDOOVHHIURPH[DPSOHVWREHJLYHQEHORZ Table 9 gives the free pronoun forms plus intransitive auxiliary forms for two tenses (present and far past); note again the lack of any formal connection between the free SURQRXQIRUPVDQGWKHLQÀHFWHGDX[LOLDULHV1RWHDOVRWKHODFNRIDQ\IRUPDOVLPLODULW\ between the person/number forms of the auxiliary in Idi and those given for Nen verbs in WDEOH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 138 NOMINATIVE ACCUSATIVE POSSESSIVE INTRANS. INTRANS. AUXILIARY: PRES AUXILIARY: PAST 1sg ƾΩQ bom bo wala ZDJΩQ 1du bi ba ba ZDƾDPD gwaga walala GX ybi yba yba ZDƾDPD gwagma 1pl bi ba ba ZDƾDPD gwaga SO ybi yba yba ZDƾDPD gwagma VJ be babom béna walale gwege GX be bibim béna walala gwaga SO be bibim béna ZDƾDPD gwagma 3sg bo obom obo wala gwaggen 3du bo ubim oba walalo gwago 3pl bo ubim oba ZDƾDPR gwagmo TABLE 9&RPSDULVRQRI,GLIUHHSURQRXQVDQGLQÀHFWHGIRUPVRIWKHLQWUDQVLWLYH auxiliary (present and past forms) ,Q,GLWKHLQ¿QLWLYHSOXVDX[LOLDU\FRQVWUXFWLRQLVPXFKPRUHZLGHVSUHDGWKDQLQ1HQ ,Q1HQLWLVXVHGIRUSKDVDOFRQVWUXFWLRQVµEHJLQWR9 ¿QLVK9LQJ¶DQGWKLVLVDOVRWKHFDVH LQ,GL H[DPSOHVWREHJLYHQODWHULQ %XWLQ,GLLWVXVHLVH[WHQGHGIXUWKHU±LWLVWKH QRUPDOFRQVWUXFWLRQLQWKHSUHVHQWWHQVHIRUH[DPSOH DE ±DQGLWLVRQO\LQDVXEVHWRI 7$0YDOXHV HJSDVWSHUIHFWLYHVHWWLQJV WKDWWKHPDLQYHUELVGLUHFWO\LQÀHFWHG FIFG Note that valency alternations shown by auxiliary choice in the periphrastic construction DUHVKRZQE\WKHFKRLFHRISUH¿[RQWKH¿QLWHYHUE (6a) SHODҀD SDOGDE ZDOD SODWH&25 EUHDN ,1) ,175$8; _VJ 35 µ7KHSODWHLVEUHDNLQJ¶ (6b) ҀLҀLPH SHODҀD SDOGDE \HUD JLUO&25 SODWH',5 EUHDN ,1) 75$8; _VJ!VJ 35 µ7KHJLUOLVEUHDNLQJWKHSODWH¶ (6c) ҀLҀLPH SHODҀD \DSDOGDEHQ JLUO&25 SODWH&25 367 VJ2EUHDN_VJ!QSO µ7KHJLUOEURNHWKHSODWH¶ (6d) SHODҀD ZDSDOGDEHQ SODWH&25 367 55EUHDN_VJ µ7KHSODWHEURNH¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 139 Examples (7) and (8) compare intransitive clauses, using the intransitive auxiliary wala, with transitive clauses using the transitive auxiliary yera6RPHWLPHV DVLQWKHFDVHRI (7c) vs (8c), or (7d) vs (8d)) this effects the difference between intransitive/causative or UHÀH[LYHWUDQVLWLYHGRXEOHWV7KHVHH[DPSOHVDOVRLOOXVWUDWHDQRWKHULQWHUHVWLQJIHDWXUHRI ,GL$µFRUHFDVH¶PDUNVDOOFRUHQRPLQDODUJXPHQWV±VXEMHFWV WUDQVLWLYHRULQWUDQVLWLYHV DQGREMHFWV±HYHQWKRXJKQRXQVXVHGLQLVRODWLRQ HJLQQRPLQDWLRQ DSSHDUZLWKRXWLW HJged ‘child’ or ҀLҀLPµJLUO¶ LQDQHOLFLWDWLRQFRQWH[W ,WLVRQO\SHUVRQDOSURQRXQVZKLFK GLVWLQJXLVKFRUHDUJXPHQWVYLDDQRPLQDWLYHYVDFFXVDWLYHFDVHGLVWLQFWLRQ E ±FRQWUDVW ƾΩQ ‘1sgNOM’ vs bom ‘1sgACC’; bo ‘3sgNOM’ vs obomµVJ$&&¶ (7a) ged-e PpȜ wala FKLOG&25 VFUHDP ,175$8; _VJ6 35 µ7KHFKLOGLVVFUHDPLQJ¶ (7b) ҀLҀLPH ZDODƾJDZD ELVL ZDOD JLUO&25 IRUHVW$// JR ,175$8; _VJ6 35 µ7KHJLUOLVJRLQJWRWKHIRUHVW¶ (7c) lu-e ]DƾJ wala WUHH',5 EXUQ ,1) ,175$8; _VJ6 35 µ7KHWUHHLVEXUQLQJ¶ (7d) ҀLҀLPH oboobo tetu wala JLUO',5 VJ55 ZDVK ,175$8; _VJ6 35 µ7KHJLUOLVZDVKLQJKHUVHOI¶ (8a) ged-e lu-e NDNȜ yera FKLOG',5 WUHH',5 FOLPE ,1) 75$8; _VJ!VJ 35 µ7KHFKLOGLVFOLPELQJWKHWUHH¶ (8b) ƾΩQ obom яΩQяΩJ yera VJ120 VJ$&& ELWH ,1) 75$8; _VJ!VJ 35 µ,DPELWLQJKLPKHU (8c) lu-e JHGH ]DƾJ yera WUHH',5 FKLOG',5 EXUQ ,1) 75$8; _VJ!VJ 35 µ7KHFKLOGLVEXUQLQJWKHWUHH¶ (8d) ҀLҀLPH RER JHGH tetu yera JLUO',5 VJ3266 FKLOG',5 ZDVK 75$8; _VJ!VJ 35 µ7KHJLUOLVZDVKLQJKHUFKLOG¶ Other case morphology includes locative -me (NΩOΩPPH ‘in the swamp’), allative -awa (NΩOΩPDZD ‘to the swamp’), ablative D Ҁ (ZDODƾJDҀ‘from the forest’), dative -ble (JΩG ble ‘to the boy’) instrumental -enda (sabor-enda ‘with a spade’ (sabor (QJµVKRYHO¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages As with the verbal morphology and the free pronouns, there are no formal resemblances EHWZHHQ WKH IRUPV RI DQ\ RI WKH FDVH VXI¿[HV DQG WKRVH LQ 1HQ RU RWKHU ODQJXDJHV RI WKH (9a) bi NRPEOHEH ERP QDQяΩJOD 120 WZR VJ$&& 72'367VJ2VHHQSO$!VJ2 µ (9b) be komblebe obom QDQяΩJOD QVJ120 WZR VJ$&& 72'367VJ2VHHQSO$!VJ2 µ (9c) \EL ҀD\HELEL RERP QDQяΩJOD 120 PDQ\ VJ$&& 72'367VJ2VHHQVJ$!VJ2 µ:H \RXPHDQGRWKHUV VDZKLPKHU HDUOLHUWRGD\ ¶ (10a) EL NRPEOHEH ELELP xDQяΩJOD QVJ120 WZR QVJ$&& 72'367QVJ2VHHQVJ$!QVJ2 µ:HWZR H[FO VDZ\RX QRQVLQJXODU HDUOLHUWRGD\ ¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages (10b) bi komblebe obim xDQяΩJOD QVJ120 WZR QVJ$&& 72'367QVJ2VHHQVJ$!QVJ2 µ:HWZR H[FO VDZWKHP HDUOLHUWRGD\ ¶ (10c) be NRPEOHEH ELP xDQяΩJOD QVJ120 WZR QVJ$&& 72'367QVJ2VHHQVJ$!2 µ (10d) ybi komblebe obim xDQяΩJOD QVJ120 WZRQVJ$&& 72'367QVJ2VHHQVJ$!QVJ2 µ:HWZR LQFOXVLYH VDZWKHP HDUOLHUWRGD\ ¶ $VLQ1HQGLDWKHWLFFKDQJHVVXFKDVUHÀH[LYHUHFLSURFDODUHVLJQDOOHGE\YHUEDOSUH¿[ The verb boku ‘cut’ (far past stem kon), for example, normally takes various forms of SUH¿[DFFRUGLQJWRREMHFWYDOXHV HJgakon for ‘I cut you (sg)’, bekon for ‘I/her cut him/ KHU¶ %XWWKHUHÀH[LYHUHFLSURFDOHPSOR\VDSHUVRQQXPEHULQYDULDQWSUH¿[IRUPgwa-, DORQJZLWKDSHUVRQVHQVLWLYHUHÀH[LYHSURQRXQIRUPHGE\WKHSRVVHVVLYHSURQRXQSOXV яDJΩPHQGHHJREDяDJΩPHQGHµWKHPVHOYHV¶RUDUHFLSURFDOUHÀH[LYHSURQRXQIRUPHGE\ UHGXSOLFDWLQJWKHSRVVHVVLYHSURQRXQ HJbabaµRXUVHOYHV H[F HDFKRWKHU¶ ([DPSOHV are: (11a) ƾΩQ ERяDJΩPHQGH JZDNRQ WΩWΩP VJ120 VJ32665()/ 55 5HP3VWFXW_VJ$ \HVWHUGD\ µ,FXWP\VHOI\HVWHUGD\¶ (11b) EH EHQHяDJΩPHQGH JZDNR\D WΩWΩP 120 VJ32665()/ 55 5HP3VWFXWVJ$ \HVWHUGD\ µ (11c) bi baba gwa-ko-ma WΩWΩP 1sgNOM 1nsgRR RR:RemPst-cut-1nsgA yesterday µ:H H[FOXVLYH FXWHDFKRWKHU\HVWHUGD\¶ (11d) bo komblebi obaoba gwa-ko-yo WΩWΩP 3NOM two 3nsgRR RR:RemPst-cut-3duA yesterday µ7KH\WZRFXWHDFKRWKHUWKHPVHOYHV\HVWHUGD\¶ 7RFRQFOXGHWKLVEULHIVNHWFKZHLOOXVWUDWHWKHXVHRILQ¿QLWLYHYHUEVLQÀHFWHGIRUFDVH LQSKDVDOFRPSOHPHQWVZKLFKSDUDOOHO1HQLQWKHLUVWUXFWXUH7KHSKDVDODX[LOLDU\DJUHHV ZLWKERWKDUJXPHQWVRIWKHYHUEDQGWKHLQ¿QLWLYHSODFHGEHIRUHLWLVLQÀHFWHGIRUDQ DSSURSULDWHFDVHVXFKDVWKHDOODWLYHLQFRQVWUXFWLRQVPHDQLQJµWREHDERXWWR¶ DE D Bi babom koko-awa deada nalala SO120 VJ$&& FXW ,1) $// EHDERXWWR 7U$8; QVJ!VJ µ:HWZRDUHDERXWWRFXW\RX¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages E ƾΩQ ELELP NRPEODEH SO120 QVJ$&& WZR koko-awa deada ñere FXW ,1) $// EHDERXWWR 7U$8; VJ!GX µ,DPDERXWWRFXW\RXWZR¶ $VVWDWHGHDUOLHUWKRXJKFODLPHGDVUHODWHGWR1HQDQGWKHRWKHU 7KHXVHRISUH¿[HVDQGVXI¿[HVLVDOVRIRXQGHOVHZKHUHLQ1HZ*XLQHD±IRUH[DPSOHLQ*RURND .DLQDQWXODQJXDJHVRIWKH7UDQV1HZ*XLQHDIDPLO\ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages YHUE LQ VLPSOH FRQVWUXFWLRQV DQG DX[LOLDU\ YHUE LQ QRQ¿QLWH FRQVWUXFWLRQV $X[LOLDU\ FRQVWUXFWLRQVDUHPRUHH[WHQVLYHLQ,GLWKDQLQ1HQ7KLVUHÀHFWVWKHIDFWWKDWLQ,GLWKH\ are the basic construction in the present, and the auxiliary indicates ongoing aspect (as well as serving as a light verb for many verb lexemes) whereas in Nen the auxiliary is reserved IRUSKDVDOFRQVWUXFWLRQV EHJLQWR¿QLVK G ERWKODQJXDJHVDUHYHUE¿QDOEXWWKLVLVVRZLGHVSUHDGLQ1HZ*XLQHDWKDWLWKDV OLWWOHRUQRGLVWLQFWLYHYDOXH 3DVVLQJQRZWRGLYHUJHQWIHDWXUHVZKLFKDUHPXFKPRUHQXPHURXVWKHPRVWVLJQL¿FDQW among them are: D WKH GLIIHUHQW RUJDQLVDWLRQ RI FDVH ERWK RQ SURQRXQV DQG RQ QRXQV 1HQ KDV DQ absolutive/ergative system throughout (apart from the neutralisation of absolutive and HUJDWLYH IRU VW DQG QG VLQJXODU SURQRXQV ,GL KDV D QRPLQDWLYHDFFXVDWLYH V\VWHP for pronouns and a highly unusual system opposing a ‘direct’ case (used in A, S and O IXQFWLRQV WRD]HURIRUP XVHGLQQRPLQDWLRQDQGQRPLQDOSUHGLFDWHV E 1HQODFNVDQLQFOXVLYHH[FOXVLYHGLVWLQFWLRQ ,GLKDVRQH F 1HQIRUPVLWVLQ¿QLWLYHVE\VXI¿[DWLRQWRWKHVWHP HJ¥HVU ‘descend’, esrs ‘to GHVFHQG¶ ,GLIRUPVLWVLQ¿QLWLYHVHLWKHUE\UHGXSOLFDWLRQ HJ¥ҀPH‘close’, ҀPHҀPH ‘to close, closing’; ¥NR ‘cut’, kokoµWRFXWFXWWLQJ¶ E\XVLQJWKHEDUHVWHP HJ¥WUHP ‘open WU ¶trem µWRRSHQRSHQLQJ¶ RULQWURGXFLQJVRPHRWKHUPRGL¿FDWLRQWRWKHVWHP HJ ¥QяRJ‘burn’ яRƾJµWREXUQ¶ (d) Nen has an indigenous power-based senary system; in Idi these are extremely marginal and clearly borrowed H 1HQKDVDULFKVHWRISRVWXUDOSRVLWLRQDOYHUEV±DERXWWKLUW\YHUEVZLWKPHDQLQJV OLNHµEHWKHHQGRIVRPHWKLQJ¶µEHXSKLJK¶µEHZHGJHG¶µEHLQDWUHHIRUN¶DQGVRRQ± which have a cluster of distinct morphosyntactic characteristics and are a central part of the JUDPPDWLFDOV\VWHP,GLDSSHDUVWRKDYHQRVXFKSKHQRPHQRQ I LQ WKH XQPDUNHG FDVH ± DEVROXWLYH IRU 1HQ QRPLQDWLYH IRU ,GL ± 1HQ GRHVQ¶W distinguish number for any person, whereas Idi distinguishes number for all persons except second (g) the dominant person syncretism within the Nen verbal agreement system is second SHUVRQZLWKWKLUG QRWXQXVXDOLQ3DSXDQODQJXDJHV ZKHUHDVLQ,GLLWLV¿UVWSHUVRQZLWK WKLUG PXFKPRUHXQXVXDO DVH[HPSOL¿HGLQPDQ\H[DPSOHVLQ DQG K LQWHUPVRISKRQRORJLFDOLQYHQWRULHV1HQKDVQRYHODUQDVDOQRUHWURÀH[HVDVLQJOH ODWHUDODQGDFRDUWLFXODWHGODELDOYHODUVHULHV,GLKDVDVWULNLQJO\µ$XVWUDOLDQ¶SKRQHPH LQYHQWRU\ZLWKLQLWLDOYHODUQDVDOVDUHWURÀH[VHULHVDQGWZRODWHUDOV±VRPHVSHDNHUVKDYH FRDUWLFXODWHGODELDOYHODUVLQVRPHORDQZRUGVEXWRWKHUZLVHWKLVVHULHVLVDEVHQW Short and incomplete as it is, this list should demonstrate how many typological isoglosses separate Nen from Idi, and show that widespread bilingualism and intermarriage between speakers of these two languages has not produced strong convergences of VWUXFWXUH DOWKRXJKWKHUHDUHDIHZDVRXWOLQHG $WWKHSUHVHQWVWDJHRIUHVHDUFKLWLVWRR early to tell whether this bespeaks relatively recent contact, or rather indicates that long- VWDQGLQJFRQWDFWKDVOHIWWKHEDVLFDOO\GLIIHUHQWW\SRORJLFDOSUR¿OHVRIWKHWZRODQJXDJHV DQGODQJXDJHIDPLOLHV XQWRXFKHG 5. AREALITY IN SOUTHERN NEW GUINEA: THE CASE OF THE DUAL. 'HVSLWHWKHVLJQL¿FDQW W\SRORJLFDOYDULHW\RIWKHODQJXDJHVIRXQGLQ6RXWKHUQ1HZ*XLQHD±VRPHWKLQJLOOXVWUDWHG MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages LQDYHU\ORFDOLVHGZD\E\WKHFRPSDULVRQRI1HQDQG,GLLQWKHODVWVHFWLRQ±WKHUHDUH some common typological themes running through the whole region (see also Reesink & 'XQQWKLVLVVXH ,QWKLVVHFWLRQ,IRFXVRQMXVWRQH±WKHSUHVHQFHRIGXDOQXPEHURQWKH verb, which runs through the region from Marori (Arka this issue) in the west to Kiwai LQWKHHDVW 5D\ WKRXJKDSSDUHQWO\QRWLQ0DULQGDVIDUDV,FDQGHWHUPLQHIURP 'UDEEH ZKRRQO\PHQWLRQVVLQJXODUDQGSOXUDO,QIDFWPRVWODQJXDJHVRIWKHUHJLRQ KDYHDQDGGLWLRQDOQXPEHUGLVWLQFWLRQ±DGGLQJDWULDORUSDXFDORUH[WHQGLQJWKHSOXUDOXS WRDODUJHSOXUDO%XWIRUUHDVRQVRIVSDFH,VNLUWWKDWDGGLWLRQDOFRPSOH[LW\KHUHVLQFHP\ JRDOLVWRIRFXVRQWKHUDWKHUGLIIHUHQWZD\VWKDWWKHVDPHUHVXOW±DJUDPPDWLFDOFDWHJRU\ H[SUHVVLQJGXDOQXPEHURQWKHYHUE±FDQEHSXWWRJHWKHULQLQWHUHVWLQJO\GLIIHUHQWZD\VLQ GLIIHUHQWIDPLOLHV One of Greenberg’s well-known universals about morphological categories states that: 1RODQJXDJHKDVDWULDOQXPEHUXQOHVVLWKDVDGXDO1RODQJXDJHKDVDGXDO XQOHVVLWKDVDSOXUDO *UHHQEHUJ $PRUSKRORJLFDOFRQVHTXHQFHRQHPLJKWH[SHFWIURPWKLVZRXOGEHWKDWGXDOVDUHEXLOWXS IURPSOXUDOV7KLVLVLQGHHGWKHFDVHLQPDQ\ODQJXDJHVHJWKHSURQRPLQDOREMHFWSUH¿[ V\VWHPLQ%LQLQM*XQZRN (YDQV DQGLWLVIRXQGLQVRPHODQJXDJHVRIWKH6RXWKHUQ 1HZ*XLQHDUHJLRQ7KH,GLFRSXODSURYLGHVDFOHDUH[DPSOH WKHVLQJXODUIRUPLVda, the plural is built up from this (dag), and the dual in turn is built up from the plural (dago $VHFRQGSRVVLELOLW\LVWR¿UVWGLVWLQJXLVKVLQJXODUIURPQRQVLQJXODUWKHQWRGLVWLQJXLVK GXDOIURPSOXUDOLQDQHTXLSROOHQWZD\LHWKHUHLVQRREYLRXVZD\RIGHULYLQJHLWKHUQRQ VLQJXODUIRUPIURPWKHRWKHU.DOD.DZDZ Singular Non-singular 1 w- Q 3m s- 3f w- 1 n- _¡\ be:NPl -iyene wiyene niyene siyene wiyene niyene iyene 1sg:be VJ EH VJPEH VJIEH GXEH _GXEH be:Pl -ero QΩUHL yero SOEH _SOEH TABLE 10. Composing the dual of ‘be’ in Warta Thundai by crossing singular vs non- VLQJXODUDQGQRQSOXUDOYVSOXUDOGLVWLQFWLRQV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages $IRXUWKSRVVLELOLW\DOUHDG\LOOXVWUDWHGIRU1HQLQLVWRGHULYHDWKUHHZD\QXPEHU V\VWHP E\ FURVVLQJ D VLQJXODU YV QRQVLQJXODU ZLWK D GXDO YV QRQGXDO V\VWHP7KLV LV highly unusual typologically, but found in several languages of the Nambu branch of the 1sgA: -n _VJ$ -ø 1nsgA: -m _QVJ$ -nd nd -ø- yinjoyn yinjoy yinjoym yinjoynd du -ea- yinjoeaym yinjoyeaynd TABLE 11. Partial verb paradigm for the past perfective of y-injo-_-y_>VJ8FDWFK VLJKWRIB3VW3HUIB@µFDXJKWVLJKWRILW¶ 6LHJHO 7KXVyinjoyn is ‘I caught sight of it’, yinjoyeaymLVµZHWZRFDXJKWVLJKWRILW¶HWF ,QWULJXLQJO\WKLVSDWWHUQLVQRWFRQ¿QHGWRWKH1DPEXODQJXDJHV:LWKLQWKH(DVWHUQ Trans-Fly branch, Meryam Mir (Piper 1989) exhibits a very similar pattern, though the distribution of information is different: the singular vs non-singular contrast is found in WKHIUHHSURQRXQVZKLOHWKHGXDOYVQRQGXDOFRQWUDVWLVIRXQGLQWKHSURQRPLQDOSUH¿[HV There are two further interesting twists: the dual is also used for paucals and many verb VWHPVVXSSOHWHRQDVLQJXODU_GXDOYVSDXFDO_SOXUDOSDWWHUQ7ZRYLHZVRIWKHZRUNLQJVRI WKLVV\VWHPDUHLOOXVWUDWHGLQWDEOHVDQG GDWDIURP3LSHU QRWHWKDW(r)edi LVWKHSUHVHQWWHQVHVXI¿[WRWKHYHUEe and wi are the third singular and third non-singular free pronouns, and (i)mi and (e)mrDUHWKHVLQJXODU_GXDODQGSDXFDO_SOXUDOVWHPVRIµVLW¶ VJ_SOø- GX_SDXFna- VJ_GX(i)mi imiredi ‘he is sitting’ na-miredi ‘they (two) are sitting’ SDXF_SO(e)mr emredi ‘they (pl) are sitting’ na-mredi ‘they (pauc) are sitting’ TABLE 12µ6LW¶DQGQXPEHULQ0HU\DP0LU,QÀHFWHGYHUERQO\ DOOIRXUQXPEHUV VKRZLQJVXSSOHWLYHVWHP Though these terms are not used in descriptions, it would make sense to talk of an ‘outer’ vs µLQQHU¶FRQWUDVWLQQXPEHUZKHUHRXWHULVVLQJXODURUSOXUDODQGLQQHULVGXDORUSDXFDO MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages VJ_SOø-, GX_SDXFna- 3sg e e ø-imiredi ‘he is sitting’ 3nsg wi wi ø-emredi ‘they (pl) are sitting’ wi na-mredi ‘they (pauc) are sitting’ TABLE 13. ‘Sit’ and number in Meryam Mir, showing interaction with free pronouns EXWRPLWWLQJSDXFDO From the examples considered in this section, it is clear that having a dual category on YHUEVLVDFOHDUW\SRORJLFDOIHDWXUHRIWKH6RXWKHUQ1HZ*XLQHDUHJLRQ+RZHYHUWKHPHDQV by which languages build this up span a radically varied range of methods (including some extremely rare ones typologically), suggesting a large number of individual convergence SDWKZD\VEURNHUHGE\DFRPPRQVHPDQWLFWDUJHW)XUWKHUFRQVLGHUDWLRQRIWKLVTXHVWLRQ± taking into account more languages, more patterns within each (for the sake of exposition I have picked particular illustrative patterns which are by no means the only ones in a JLYHQODQJXDJH DQGWKHIXUWKHUFRPSOLFDWLRQVEURXJKWLQE\DIRXUWKQXPEHU±LVOLNHO\ WRUHYHDODQHYHQPRUHLQWULFDWHVHWRIGHYHORSPHQWV,WPD\DOVRVXJJHVWHDUOLHUFRQWDFW VFHQDULRV±LVLWSRVVLEOHWKDWWKHSUHVHQFHRIVXFKVLPLODUEXWW\SRORJLFDOO\XQXVXDOZD\VRI FRQVWUXFWLQJWKHGXDOLQWKH1DPEXEUDQFKRIWKH &IWKLVTXRWHRQWKHVWDWXVRI0DNOHZIURP/HEROGHWDO µ7KHSHRSOHZKRVSHDNWKH 0DNOHZODQJXDJHVHHPWREHDVPDOOJURXS7KH\DOVRVHHPPXFKOHVVSURXGRIWKHLUODQJXDJH DQGFXOWXUHWKDQWKH0DULQGSHRSOHGR7KHDGXOWVLQ:HOEXWLFRPSODLQHGWRWKHVXUYH\WHDPWKDW WKHLUFKLOGUHQGRQRWVSHDNWKHLUODQJXDJHDQGVRPHWLPHVPDNHIXQRIWKHPIRUXVLQJLW7KH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages ± DUH RQO\ VSRNHQ E\ SHRSOHRI PLGGOHDJHRU DERYHDQG DUHVLJQL¿FDQWO\HQGDQJHUHG Marind is often said to be in better shape, but there have been no recent detailed studies of WKHODQJXDJHZKLFKFRXOGYHULI\WKLV$VWKHVHH[DPSOHVPDNHFOHDUGRFXPHQWDU\ZRUNRQ ODQJXDJHVRQWKH,QGRQHVLDQVLGHRI6RXWKHUQ1HZ*XLQHDLVDSDUWLFXODUO\XUJHQWSULRULW\ Beyond that, a closer study of the whole Southern New Guinea region will plainly lead WRPDQ\GLVFRYHULHV±RIDKRVWRIXQGHVFULEHGOLQJXLVWLFSKHQRPHQDRIWKHG\QDPLFV of village multilingualism and its effects on language change, of the forces that drove the expansion of Trans-New Guinea languages, of a complex process of relatively recent colonisation as the land was built up over the last few millennia, of contacts between Papuan and Australian languages across the Torres Strait, linked by an Australian language .DOD.DZDZ REFERENCES $OSKHU%DUU\*HRIIUH\2¶*UDG\ &ODLUH%RZHUQ:HVWHUQ7RUUHV6WUDLWODQJXDJH FODVVL¿FDWLRQ DQG GHYHORSPHQW ,Q &ODLUH %RZHUQ %HWKZ\Q (YDQV DQG /XLVD 0LFHOL HGV Morphology and language history: In honour of Harold Koch $PVWHUGDP -RKQ%HQMDPLQV children themselves reportedly do not speak the vernacular, although they are able to understand LW7KLVZDVDFRQFHUQWRWKHDGXOWVEHFDXVHWKH\ZHUHDIUDLGWKHLUODQJXDJHZRXOGGLHRXW WKH\ VDLGWKHLUODQJXDJHZRXOGQRWEHVSRNHQDQ\PRUHLQ\HDUV1HYHUWKHOHVVWKHVXUYH\WHDPGLG notice some older women who only spoke the vernacular language and were not able to speak ,QGRQHVLDQ¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages $UND:D\DQ&RQVWUXFWLYHQXPEHUV\VWHPVLQ0DURULDQGEH\RQG,Q0LULDP%XWW 7UDF\+ROORZD\.LQJ HGV Proceedings of the LFG11 conference&6/,3XEOLFDWLRQV KWWS FVOLSXEOLFDWLRQVVWDQIRUGHGX/)*SDSHUVOIJDUNDSGI $UND:D\DQ7KLVYROXPH3URMHFWLQJPRUSKRORJ\DQGDJUHHPHQWLQ0DURULDQLVRODWHRI 6RXWKHUQ1HZ*XLQHD $\UHV0DU\& 7KLVVLGHWKDWVLGH /RFDOLW\DQGH[RJDPRXVJURXSGH¿QLWLRQLQWKH Morehead area, southwestern Papua.&KLFDJR,/ 8QLYHUVLW\RI&KLFDJR3K'WKHVLV %OHYLQV-XOLHWWH $QGUHZ3DZOH\7\SRORJLFDOLPSOLFDWLRQVRI.DODPSUHGLFWDEOH YRZHOVPhonology &RUEHWW*UHYLOOH*Number&DPEULGJH &DPEULGJH8QLYHUVLW\3UHVV 'RQRKXH 0DUN 6\OODEOHV PRUDH DQG YRZHOV LQ .DQXP 6HPLQDU +DQGRXW $XVWUDOLDQ1DWLRQDO8QLYHUVLW\8QSXEOLVKHGPV 'UDEEH3 Spraakkunst van het Marind 6WXGLD,QVWLWXWL$QWKURSRV 0|GOLQJ 0LVVLHKXLV6W*DEULsO (YDQV 1LFKRODV Bininj Gun-wok: a pan-dialectal grammar of Gun-djeihmi, Kunwinjku and Kune 3DFL¿F/LQJXLVWLFV . &DQEHUUD 3DFL¿F/LQJXLVWLFV5HVHDUFK 6FKRRORI3DFL¿FDQG$VLDQ6WXGLHV$XVWUDOLDQ1DWLRQDO8QLYHUVLW\ (YDQV 1LFKRODV $XVWUDOLDQ ODQJXDJHV UHFRQVLGHUHG D UHYLHZ RI 'L[RQ Oceanic Linguistics (YDQV1LFKRODV 7ZRpusRQHPDNHVWKLUWHHQ VHQDU\QXPHUDOVLQWKH0RUHKHDG± 0DURUHJLRQLinguistic Typology (YDQV1LFKRODV)RUWKFRPLQJD9DOHQF\LQ1HQ,Q$QGUHM0DOFKXNRY0DUWLQ+DVSHOPDWK %HUQDUG&RPULH ,UHQ+DUWPDQQ HGV Valency Classes in the World’s Languages (YDQV1LFKRODV)RUWKFRPLQJE,QÀHFWLRQLQ1HQ,Q0DWWKHZ%DHUPDQ HG The Oxford +DQGERRNRI,QÀHFWLRQ (YDQV 1LFKRODV )RUWKFRPLQJ F 7KH ODQJXDJHV RI VRXWKHUQ 1HZ *XLQHD ,Q 3DOPHU Languages and linguistics of New Guinea. )RUG.HYLQ 'DQD2EHU $VNHWFKRI.DODZ.DZDZ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Multiplicity of Trans-Fly languages 3DZOH\$QGUHZ5REHUW$WWHQERURXJK-DFN*ROVRQ 5RELQ+LGH HGV Papuan pasts: Cultural, linguistic and biological histories of Papuan-speaking peoples. &DQEHUUD 3DFL¿F/LQJXLVWLFV 3DZOH\ $QGUHZ 5HFHQW UHVHDUFK RQ WKH KLVWRULFDO UHODWLRQVKLSV RI WKH 3DSXDQ languages, or, what does linguistics say about the prehistory of Melanesia? In Jonathan )ULHGODHQGHU HG Population genetics, linguistics and culture history in the southwest 3DFL¿F2[IRUG 2[IRUG8QLYHUVLW\3UHVV 3DZOH\$QGUHZ +DUDOG+DPPDUVWU|P)RUWKFRPLQJ7KHWUDQV1HZ*XLQHDIDPLO\,Q Palmer, Languages and linguistics of New Guinea. 3LSHU1LFN $VNHWFKJUDPPDURI0HU\DP0LU&DQEHUUD $180$WKHVLV 5D\6LGQH\+ 7KHODQJXDJHVRIWKH:HVWHUQ'LYLVLRQRI3DSXDJRAI 5D\6LGQH\+ A grammar of the Kiwai Language, Fly Delta, Papua. With a Kiwai vocabulary by the late Rev. E. Baster Riley3RUW0RUHVE\ (GZDUG*HRUJH%DNHU >7KLVVRXUFHLVFLWHGLQWKHOLWHUDWXUHZLWKYDULRXVGDWHV DQG DQG GHWHUPLQLQJLWVH[DFWGDWHRIDSSHDUDQFHLVSUREOHPDWLFVLQFHWKHÀ\OHDIFRQWDLQVQR GDWHZKLOHWKHSUHIDFHE\5D\EHDUVWKHGDWH ,VWLFNWR DVWKHFRPPRQHVW \HDUZLWKZKLFKLWLVFLWHG@ 5HHVLQN*HU5XWK6LQJHU 0LFKDHO'XQQ ([SODLQLQJWKHOLQJXLVWLFGLYHUVLW\RI 6DKXOXVLQJSRSXODWLRQPRGHOVPLoS Biology H 5RVV0DOFROP3URQRXQVDVDSUHOLPLQDU\GLDJQRVWLFIRUJURXSLQJ3DSXDQODQJXDJHV ,Q3DZOH\HWDOPapuan pasts 6LHJHO-HII7KHVLJQL¿FDQFHRIWKHQXPEHUVDQGLQWKH1DPDODQJXDJH 3DSXD 1HZ *XLQHD 3RZHUSRLQW 3UHVHQWDWLRQ )UHLEXUJ ,QVWLWXWH IRU $GYDQFHG 6WXGLHV 8QSXEOLVKHGPV 6XWHU(GJDU2EMHFW9HUEVLQ+XRQ3HQLQVXODODQJXDJHV8QSXEOLVKHGPV 9DQ%DDO-DQ Dema. Description and Analysis of Marind-Anim Culture (South New Guinea). 7KH+DJXH 0DUWLQXV1LMKRII 9RRUKRHYH&/ 6RPH QRWHV RQ WKH 6XNL*RJRGDOD VXEJURXS RI WKH &HQWUDO DQG 6RXWK1HZ*XLQHD3K\OXP,Q6WHSKHQ$:XUP 'RQDOG&/D\FRFN HGV 3DFL¿F linguistic studies in honour of Arthur Capell 3DFL¿F/LQJXLVWLFV& &DQEHUUD 3DFL¿F/LQJXLVWLFV$XVWUDOLDQ1DWLRQDO8QLYHUVLW\ 9ULHV /RXUHQV GH A short grammar of Inanwatan, an endangered language of the Bird’s Head of Papus, Indonesia 3DFL¿F /LQJXLVWLFV &DQEHUUD 3DFL¿F /LQJXLVWLFV$XVWUDOLDQ1DWLRQDO8QLYHUVLW\ :LOOLDPV)( Papuans of the Trans-Fly2[IRUG &ODUHQGRQ :XUP6WHSKHQ The Papuan languages of Oceania7ELQJHQ 1DUU ::) 7UDQV)O\ 7HDP &RQVHUYDWLRQ 6FHQDULRV IRU WKH 7UDQVÀ\ (FRUHJLRQ 1HZ *XLQHD 3RZHUSRLQW3UHVHQWDWLRQ%ULVEDQH8QSXEOLVKHGPV Nicholas Evans QLFKRODVHYDQV#DQXHGXDX MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY /DQJXDJH'RFXPHQWDWLRQ &RQVHUYDWLRQ6SHFLDO3XEOLFDWLRQ1R 'HFHPEHU Melanesian Languages on the Edge of Asia: Challenges for the 21st Century, HGE\1LFKRODV(YDQVDQG0DULDQ.ODPHU± KWWS QÀUFKDZDLLHGXOGFVS 6 KWWS KGOKDQGOHQHW Projecting morphology and agreement in Marori, an isolate of southern New Guinea I Wayan Arka Australian National University/ Universitas Udayana 7KLV SDSHU LV WKH ¿UVW GHWDLOHG LQYHVWLJDWLRQ RQ DJUHHPHQW LQ 0DURUL ,VRODWH 3DSXDQ 0HUDXNH,QGRQHVLD KLJKOLJKWLQJ LWV VLJQL¿FDQFH LQ the cross-linguistic understanding of NUM(BER) expression and in the XQL¿FDWLRQEDVHGWKHRU\RIDJUHHPHQW0DURULVKRZV3(56DQG180 DJUHHPHQW ZLWK GLVWULEXWHG H[SRQHQFH LQ '8$/7KH SDSHU SURSRVHV that DUAL is formed by two basic NUM features (SG, PL) each with LWVELQDU\YDOXHVDQGWKDW'8$/LV>6*3/@ XQPDUNHG 7KHQRYHO aspect of the analysis is the idea that the NUM feature is mapped onto DODQJXDJHVSHFL¿FVWUXFWXUHGVHPDQWLFVSDFHRI180$PRUSKHPHLV analysed as carrying a feature bundle, with the semantic spaces referred WRE\WKHLQGLYLGXDOIHDWXUHVSRVVLEO\RYHUODSSLQJZLWKHDFKRWKHU7KH proposed analysis can provide a natural explanation for NUMBER agreement in Marori and can be extended to account for unusual cases RI180DJUHHPHQWDQGH[SUHVVLRQLQRWKHUODQJXDJHV 1. INTRODUCTION. Marori1 is a Papuan language (isolate, Trans New Guinea (Ross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cc Creative Commons Attribution Non-Commercial No Derivatives Licence Morphology and agreement in Marori 151 0DURUL ODQJXDJH LV KLJKO\ HQGDQJHUHG 7KHUH DUH VHYHUDO ÀXHQW VSHDNHUV OHIW RXW of a total of IDPLOLHV RU SHRSOH 0DURUL SHRSOH W\SLFDOO\ KDYH PL[HG PDUULDJHV with Marind and non-Papuan Indonesians such as the Tanimbar people and currently the -DYDQHVH7KHVRFLROLQJXLVWLFVXUYH\FDUULHGRXWLQ 6RKQ/HEROG .ULHQV UHSRUWVWKHSUHFDULRXVQDWXUHRIWKHODQJXDJHZKLFK,IXUWKHUFRQ¿UPHGZKHQ,GLGP\ ¿HOGZRUNLQDQG 2. BASIC FACTS ON MARORI SYNTAX 2.1. MARORI CLAUSAL SYNTAX IN BRIEF. The basic clause structure in Marori is shown LQ¿JXUH7KHYHUEDOVWUXFWXUHW\SLFDOO\FRQVLVWVRIDOH[LFDOYHUE 9(5% DQGDOLJKWRU DX[LOLDU\ $8; YHUE7KH$8;YHUEFDQEHPRUSKRORJLFDOO\FRPSOH[FRQVLVWLQJRIDQ $8;URRWDQGSRVVLEO\RQHRUPRUHDI¿[HV7KLVLVGHSLFWHGLQ¿JXUH3 [NP* VERB AUX.VERB]CLAUSE FIGURE 1 Most of the Javanese people coming to west Papua were originally part of the transmigration pro- JUDPVSRQVRUHGE\WKHJRYHUQPHQW7KH\DUHQRZWKH¿UVWRUVHFRQGJHQHUDWLRQERUQLQ0HUDXNH DQGFDOOWKHPVHOYHV-DPHU -DZD0HUDXNH 3 'LIIHUHQWOH[LFDOYHUEVPD\WDNHGLIIHUHQW$8;YHUEVGHSHQGLQJRQOH[LFDOFODVVHV7KLVLVQRW GLVFXVVHGLQWKLVSDSHU MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 35()$8;522768)) (U) (A) (S) (S) FIGURE 2 $JUHHPHQWRQWKH DX[LOLDU\ YHUELVH[SUHVVHGE\SRUWPDQWHDXDI¿[HVZLWKWKHSUH¿[ VKRZLQJXQGHUJRHUDJUHHPHQWDQGWKHVXI¿[DFWRUVXEMHFWDJUHHPHQW7DEOHVKRZVIUHH SURQRXQVDQGWKHLUFRUUHVSRQGLQJDJUHHPHQWDI¿[HVRQWKHYHUE)RUVLPSOLFLW\RQO\WKH DFWRUVXI¿[SDUDGLJPLQWKH3DVW7HQVHLVJLYHQ FREE ACTOR SUFFIX PRONOUN UNDERGOER PREFIX (PAST) 1SG na i- -men ‘NonPL’ 1NonSG nie yar-( TABLE 1.)UHHSURQRXQV8QGHUJRHUSUH¿[HVDQGSDVWDFWRUVXI¿[HVLQ0DURUL Examples with agreement morphology on the AUX are given in (1)5 However, certain YHUEVRIKLJKIUHTXHQF\LQGDLO\OLIHHJµVLW¶DQGµEULQJ¶RIWHQKDYHLUUHJXODULQÀHFWLRQ RUWKH7163(56PRUSKRORJ\GLUHFWO\RQWKHYHUEVHJ-du ‘1SG(PRES)’6 on the verb ‘bring/take’ in 7 $EEUHYLDWLRQVXVHGLQWKHJORVVHV µ¿UVWVHFRQGWKLUGSHUVRQ¶$µDFWRU¶$%6µDEVROXWLYH¶ AUX ‘auxiliary’, DET ‘determiner’, DU ‘dual’, ERG ‘ergative’, F ‘female’, FUT ‘future’, M ‘male’, NonPL ‘non plural’, NonSG ‘non singular’,O ‘object’, NrPST ‘near past’, PERF µSHUIHFWLYH¶ 3/ µSOXUDO¶ 3266 µSRVVHVVLYH¶ 352* µSURJUHVVLYH¶ 5()/ µUHÀH[LYH¶ 6* ‘singular’, PRES ‘present’, PST ‘past’, U ‘undergoer’, 5 7KHUH LV QR VWDQGDUG RUWKRJUDSK\ IRU 0DURUL \HW 7KLV SDSHU IROORZV WKH ,QGRQHVLDQOLNH RUWKRJUDSK\FRPPRQO\XVHGE\P\0DURULFRQVXOWDQWVHJy represents the approximant /j/ and ngWKHYHODUQDVDOƾ&RQVRQDQWVZLWKSUHQDVDOVDUHZULWWHQZLWKPRUHWKDQRQHV\PEROHJmb, nd, and ngg%LODELDOIULFDWLYHVDUHZULWWHQDV f (voiceless) and v YRLFHG 6 7KHFRQFHSW35(6(17LQ0DURULFDQFRYHUWKHWLPH12: WRGD\ DQGRIWHQ\HVWHUGD\ 7 $X[LOLDULHVDQGOH[LFDOYHUEVDUHLQGHSHQGHQWZRUGVHDFKFDQKDYHWKHLURZQDI¿[HV:KHQWKH\ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 153 (1) D Ka ku=ndo-Ø. 6* UXQ $8;1RQ3/1367 ‘you (sg) will run’ E Nawa payung=i nde=ngge-ben. 6* XPEUHOOD 8 EX\ $8;1U367 µ,ERXJKWDQXPEUHOOD¶ ujif NH PH QD¿V ndon-du tamba yabah ngguo-f.8 ELUG 5(/ ZLVK 6* \HVWHUGD\ EULQJ6* 3(5)GLH $8;'8367 µ7KH WZR ELUGVWKDW,ZDQWHGWRWDNHZLWKPH\HVWHUGD\ZHUHDOUHDG\GHDG¶ 0DURUL VKRZV VSOLW LQWUDQVLWLYLW\ 7KH SDWLHQWLYH 6 DQG 3DWLHQW2EMHFW LV W\SLFDOO\ marked by =i These examples show the contrast where the patientive S argument na ‘I’ is marked by =i (3a) whereas the agentive S must not be marked by =i (3b (3) D Na=i patar yu-nggo-f. 6* 8 FROG 6*$8;367 µ,VXIIHUHGIURPFROG¶ E (¿ ramon (*=i) ku=ndo-f. WKDW ZRPDQ UXQ $8;1RQ3/367 µ6KHWKHZRPDQUDQRII¶ The split appears to be skewed: only patientive S of states as in (3a) is treated as object- OLNH3DWLHQWLYH6RIPRWLRQSUHGLFDWHVOLNHµIDOO¶UHFHLYHVVXI¿[DJUHHPHQW Nie yanadu purfam pa=saron-den kwi uyow ngge. 1RQ3/ WZRSHUVRQ VRRQ IDOO'835(6 WUHH DERYH IURP µ:HWZRDUHDERXWWRIDOORIIIURPWKHWUHH¶ The following examples show that na in (5a) functions as Subject appearing without =i In (5b), it functions as object; hence taking =i. (5) D Tamba=na Albert=i keswe=mi-men. DOUHDG\ 6* $OEHUW 8KLW 6*0$8;1RQ3/367 µ,DOUHDG\KLW$OEHUW¶ DSSHDUWRJHWKHUIRUPLQJSKRQRORJLFDOZRUGVWKH\DUHVHSDUDWHGE\D DQRWDWLRQFRQYHQWLRQDOO\ XVHGIRUFOLWLFV HJ ku=ndo-Ø in (1)/LNHZLVHWKHVDPHFRQYHQWLRQLVDSSOLHGWRVLPLODUFDVHV such as the free pronoun na as in tamba=naµDOUHDG\ 6*¶(6)RUWKHEHQH¿FLDU\SRVWSRVLWLRQ na which can becomes =n forming phonological words with other items as in na=n=du µ6* IRU 5()/¶DVLQ(7b) 8 A more precise gloss for ngguoZRXOGEHµ$8;1RQ6*1RQ3/¶ WKHWKUHHZD\GLVWLQFWLRQRI ngguµ$8;6*¶nggoµ$8;3/¶DQGngguoµ$8;'8¶VXJJHVWVWKDWYRZHO-u attached to the root ngg- is actually associated with ‘NonPL’ and -oZLWKµ1RQ6*¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori E (¿ purfam na=i kaswa=ri-ma-m. WKDW SHUVRQ 6* 8 KLW $8;1RQ3/367 µ7KHSHUVRQKLWPH¶ In a three-place (ditransitive) structure, the Goal/Recipient argument object is marked by =i 7KH YHUE LV LQÀHFWHG VKRZLQJ DJUHHPHQW ZLWK WKLV *RDO 13LQ DGGLWLRQ WR WKH DJUHHPHQWZLWKWKHDFWRU13 (6) D Tamba=na Robertus=i bosik nji=me-feri. DOUHDG\ 6* 5REHUWXV 8 SLJ 0JLYH $8;5367 µ,DOUHDG\JDYH5REHUWDSLJ DORQJWLPHDJR ¶ E Tamba=na Maria=i bosik njo=mo-fori. DOUHDG\ 6* 0DULD 8 SLJ )JLYH ) $8;5367 µ,DOUHDG\JDYH0DULDDSLJ DORQJWLPHDJR F Robertus/Maria na=i ERVLN L PR¿ 5REHUW0DULD 6*8 SLJ 6*JLYH $8;5367 µ5REHUWRU0DULDJDYHPHDSLJ¶ $EHQH¿FLDU\SDUWLFLSDQWLQDWKUHHSODFHSUHGLFDWHLVPDUNHGE\=na or =n7KHYHUE DJUHHVZLWKWKHWKHPHSDWLHQWQRWZLWKWKHEHQH¿FLDU\13 (7) D Maria ka=na di bosik eyew Ø-nda-Ø tanamba. 0DULD 6* IRU VRRQ SLJ VHH $8;1367 QRZ µ0DULDUHDGLO\KXQWVDSLJIRU\RXQRZ¶ E Nawa ¿V nandu dakai tawramon. 6* \HVWHUGD\ QD Q GX GDND L WDZ UDPRQ 6* IRU 5()/ ZDWHU 8 WDNH $8;1RQ3/367 µ,EDLOHGZDWHURXWIRUP\VHOI\HVWHUGD\¶ 2.2. AGREEMENT TYPES IN MARORI. Agreement in Marori is of two types: clausal SUHGLFDWHDUJXPHQWDJUHHPHQWDQGSKUDVDOQRXQGHWHUPLQHUDJUHHPHQW In the predicate-argument agreement, the core arguments (subject and object) agree in 3(56DQG180ZLWKHLWKHUWKH$8;RUWKHPDLQOH[LFDOYHUERUERWK7KH¿UVWDQGPRVW FRPPRQSDWWHUQLVWKHRQHZKHUHWKH$8;LVLQÀHFWHGDQGWKHOH[LFDOSUHGLFDWHUHPDLQV FRQVWDQWLQLWVIRUP,QWKHIROORZLQJH[DPSOHVWKH$8;LVLQÀHFWHG nadam, ndamon) whereas the lexical verb eyewLVQRWLQÀHFWHG (8) D Kie tamba Maria=na bosik eyew nadam. 16*3(5) 0 IRU SLJ VHH QQGDP 1RQ6*$8;1RQ3/367 µ E Nawa ¿V Maria=na bosik eyew ndamon. 6* \HVWHUGD\ 0DULD IRU SLJVHH QGDPRQ $8;1RQ3/367 µ,KXQWHGDSLJIRU0DULD¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 155 1RWHWKDWWKHDJUHHPHQWPRUSKRORJ\PD\KDYHGLVWULEXWHGH[SRQHQFHZLWKLQWKH$8; This is seen in (8a), where the second person dual actor past tense agreement in nadam is formed by the discontinuous formatives n- and -m, added to the auxiliary root nda FIWDEOH 9 ,QWKHVHFRQGW\SHWKHYHUELWVHOILVLQÀHFWHGWRVKRZDJUHHPHQW7KLVYHUELVW\SLFDOO\ DVVRFLDWHGZLWKDFWLYLW\RIKLJKIUHTXHQF\LQGDLO\OLIHVXFKDVµEULQJ¶DQGµVLW¶,QWKH following examples, the agreement morphology -duLVDI¿[HGWRWKHYHUEV (9) D Pa=na ka=na ujif nde-du sokodu. VRRQ 6* 6* IRU ELUG EULQJ6*35(6 RQH µ, ZLOO EULQJRQHELUGIRU\RX¶ E Nawa kursi uyowé kuye-du. 6* FKDLU RQWRS VLW6*35(6 µ,VLWRQDFKDLU¶ 7KHWKLUGW\SHLVLQÀHFWLRQRQERWKWKHOH[LFDOSUHGLFDWHDQGWKH$8;7KLVLVWKHFDVH ZLWK SUHGLFDWHV WKDW HQFRGH FHUWDLQ TXDOLWLHV VXFK DV µUHG¶ DQG µELJ¶ 7KHVH SUHGLFDWHV DUH LQÀHFWHG IRU 180 VKRZLQJ RSSRVLWLRQ RI 6* DQG 1RQ6* 7KH LQÀHFWLRQ PD\ EH PRUSKRORJLFDOO\UHJXODU HJpara ‘red’ Æ para-wonµUHG6*¶ para-ndeµUHG1RQ6*¶ RUVXSSOHWLYH HJsielµELJ6*¶ kofeµELJ1RQ6*¶monjunµVPDOO6*¶menindumµVPDOO 1RQ6*¶ Consider the following examples: (10) D (¿ nam pu para-won te. 6* 3266 KDLU UHG6* EH 1RQ3/ 35(6 µ+HUKLVKDLULVUHG¶ E Emde usindu nam pu para-nde te-re(re). 1RQ6* 3/ 3266KDLU UHG1RQ6* EH3/35(6 µ7KHLU 3/ KDLULVUHG¶ Sentence (10a)VKRZVVLQJXODUDJUHHPHQWZKHUHWKHVLQJXODUVXI¿[ -won must be used on the lexical predicate paraµUHG¶DQGWKHDX[LOLDU\VKRZVWKLUGSHUVRQ1RQ3/ LH PRUSKRORJ\6HQWHQFH(10b)LVWKHFRXQWHUSDUWVHQWHQFHWKDWVKRZVSOXUDODJUHHPHQW7KH NonSG -nde is used on paraµUHG¶DQGWKHVXI¿[-re(re) RQWKHDX[LOLDU\ In addition to the predicate-argument agreement just outlined, Marori shows agreement EHWZHHQWKHGHWHUPLQHUDQGWKHQRXQKHDGLQWKHQRXQSKUDVH7KHGHWHUPLQHULQ0DURUL VKRZVDQRSSRVLWLRQRI6*YV1RQ6* H¿µ'(76*¶YVemndeµ'(71RQ6*¶ (11) D H¿ UDPRQ VRNRGX ‘the (one) woman’ DET woman one 9 The form nadam is analysed as having an underlying form n-nda-m7KHIRUPnadam involves epenthesis vowel harmony with the consonant nd of the auxiliary nda becoming d. The prenasal SDUWSHUKDSVEHFRPHVWKHFRGDRIWKH¿UVWV\OODEOH nan.dam) which is then weakened and lost (na. dam) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 156 E emnde ramon yanadu ‘the two women’ '(71RQ6* ZRPDQ WZR F emnde ramon usindu ‘all of the women’ '(71RQ6* ZRPDQ DOO 3. PROJECTION ISSUES IN AGREEMENT AND PROPOSED ANALYSIS. Before discussing the issues posed by agreement in Marori and the proposed analysis, it is useful to have a brief UHYLHZRIWKHOH[LFDOLVWDSSURDFKWRDJUHHPHQW 3.1. A LEXICALIST THEORY OF AGREEMENT. :KLOHVSHFL¿FGHWDLOVDQGPHFKDQLVPV are different, all theories of agreement operate on the same principles: compatible or same features are allowed to pass through in the formation of larger syntactic structures ZKHUHDVLQFRPSDWLEOHRQHVDUHQRW,QDOH[LFDOLVWQRQGHULYDWLRQDOIUDPHZRUNRIJUDPPDU HJ /)* %UHVQDQ 'DOU\PSOH )DON DQG +36* 6DJ :DVRZ %HQGHU WKHPHFKDQLVPLVGRQHYLDXQL¿FDWLRQRIIHDWXUHV)HDWXUHVRIWKHVDPH RUFRPSDWLEOHYDOXHVZLOOVXFFHVVIXOO\XQLI\7KLVFDQEHLQIRUPDOO\UHSUHVHQWHGLQ¿JXUH IRUDODQJXDJHOLNH(QJOLVKWKDWUHTXLUHVVXEMHFWYHUEDJUHHPHQW7KH180IHDWXUHRI >1806*@FDUULHGE\68%-13ZLOOEHFRPH>68%->1806*@@,WWKHQXQL¿HVZLWKWKH VDPHIHDWXUHFDUULHGE\WKHYHUEDVVKRZQ¿JXUHE,IWKHYHUEKDVGLIIHUHQWLQFRPSDWLEOH YDOXH WKHQ WKH XQL¿FDWLRQ IDLOV 7KH IHDWXUH FODVKHV LQGLFDWHG E\ D VWDU LQ ¿JXUH F a. [SUBJNP VP]clause | | b. [SUBJ[NUM SG]] U [SUBJ[NUM SG]] = [SUBJ[NUM SG]] c. [SUBJ[NUM SG]] U [SUBJ[NUM PL]] = * FIGURE 3 The challenge is how to develop a linguistically motivated feature structure that can FDSWXUHWKHFRPSOH[SDWWHUQVRIDJUHHPHQWLQDSDUWLFXODUODQJXDJHDQGDFURVVODQJXDJHV I argue that NUM features and their structures must be mapped onto the semantic space of 180ZKLFKPXVWEHHVWDEOLVKHGRQDODQJXDJHVSHFL¿FEDVLV7KLVLVIXUWKHUGLVFXVVHGLQ sub-section EHORZ A theory of agreement must also be able to deal with intricacies of different types of DJUHHPHQWLQFOXGLQJLQGHWHUPLQDF\DVLQFDVHDJUHHPHQW 'DOU\PSOH.LQJ 6DGOHU DQGWKHGRPDLQVUHODWLRQVLQYROYHG7KHUHDUHDWOHDVWWKUHHRIWHQLQWHUUHODWHGNLQGVRI agreement: grammatical, semantic and pragmatic :HFKVOHU =ODWLF.DWKRO 3ROODUG 6DJ 10 10 3ROODUGDQG6DJ 3ROODUG 6DJ DOVRGLVFXVVHGKRQRUL¿FDJUHHPHQWHJLQ-DSDQHVHDQG .RUHDQ7KLVW\SHRIKRQRUL¿FDJUHHPHQWLVDOVRHQFRXQWHUHGLQ%DOLQHVHDQDO\VHGDVSUDJPDWLF MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 157 ,QJUDPPDWLFDODJUHHPHQW &21&25' WKHDJUHHLQJXQLWVVKDUHJUDPPDWLFDOIHDWXUHV 3XUHO\V\QWDFWLFIHDWXUHVHJ120FDVHDUHIHDWXUHVUHTXLUHGE\WKHJUDPPDUWRHQFRGH V\QWDFWLFUHODWLRQV)RUH[DPSOHJUDPPDWLFDODJUHHPHQWLVREVHUYHGLQWKH6HUER&URDWLDQ NP in where the determiner and adjective agree in case (in addition to gender and number) with the head noun: ov-a star-a knjig-a WKLV120)6* ROG120)6* ERRN ) 1206* :HFKVOHU =ODWLF In semantic agreement, the agreeing units share referential indices: PERS, NUM, *(1' 7KHVH IHDWXUHV DUH HVVHQWLDOO\ VHPDQWLF EHFDXVH WKH\ LQGLFDWH UHIHUHQWV LQ WKH H[WHUQDOZRUOG+RZHYHUWKH\DUHDOVRRIWHQJUDPPDWLFDOLVHGLQPDQ\ODQJXDJHVDQGDUH RIWHQWLHGWRJUDPPDWLFDOIHDWXUHV)RUH[DPSOHWKH\DUHRIWHQH[SUHVVHGE\DSRUWPDQWHDX PRUSKHPH 7KXV WKH 68%-9(5% DJUHHPHQW LQ (QJOLVK LV JUDPPDWLFDO DV ZHOO DV semantic, because we have cases like committee are/is … or where the noun committee FDQKDYHVLQJXODURUSOXUDOLQWHUSUHWDWLRQ7KHWZRLQWHUSUHWDWLRQVDOORZVLQJXODUDQGSOXUDO DJUHHPHQWRQWKHYHUE7KHSOXUDODJUHHPHQWVKRZVWKHDJUHHPHQWZLWKWKHSOXUDOUHIHUHQWV RIWKHVXEMHFWGHVSLWHWKHIRUPRIWKHVXEMHFWQRXQEHLQJVLQJXODU Note that agreement in English is also grammatical in the sense that the agreeing SUBJ LV REOLJDWRULO\ UHTXLUHG E\ WKH YHUE DQG WKDW WKH DJUHHPHQW YHUEDO PRUSKRORJ\ PDNHV UHIHUHQFHWRWKHV\QWDFWLFSURSHUW\RIVXEMHFWKRRG Pragmatic agreement, also called anaphoric agreement (see footnote 11), is a case of co- UHIHUHQWLDOHOHPHQWVZKLFKVKRZFRPSDWLEOHUHIHUHQWLDOSURSHUWLHV3UDJPDWLFDJUHHPHQWLV W\SLFDOO\QRWFRQVWUDLQHGE\FHUWDLQV\QWDFWLFGRPDLQV&DVHVVKRZLQJOHIWGLVORFDWLRQDV in English (13a)EHORZEHORQJVWRSUDJPDWLFDJUHHPHQW$JUHHPHQWRIWKHW\SHVKRZQLQ (13b) from Kambera (an Austronesian language of Sumba, Indonesia) where the free NP VXEMHFWLVRSWLRQDOO\SUHVHQWDOVREHORQJVWRDQDSKRULFDJUHHPHQW11 (13) D John, I like him very much. E (I Ama) na-kei-nja ri. ART father 3sN-buy-3pD vegetable (Klamer 1996) µ)DWKHUEX\VYHJHWDEOHVIRUWKHP¶ Predicate-argument agreement in Marori is, as we shall see in the next sub-section EDVLFDOO\VHPDQWLFSUDJPDWLFLQQDWXUH 3.2. THE NATURE OF AGREEMENT IN MARORI. Agreement in Marori is not grammatical, EXWVHPDQWLFSUDJPDWLFLQQDWXUH,WLVQRWJUDPPDWLFDOEHFDXVHWKHDJUHHPHQWIHDWXUHV 180DQG3(56 DUHHVVHQWLDOO\UHIHUHQWLDO KHQFHVHPDQWLF LQQDWXUH,Q0DURULWKHVH features are not grammaticalised to become part of an agreement system that makes reference DJUHHPHQWLQ$UND %UHVQDQ 0FKRPER GLVWLQJXLVKJUDPPDWLFDODQGDQDSKRULF DJUHHPHQW 11 The distribution of a resumptive pronoun which is a case of anaphoric agreement may be also FRQVWUDLQHG WR D FHUWDLQ GHJUHH E\ WHUPKRRGFRUHQHVV RI DUJXPHQWV )RU H[DPSOH UHVXPSWLYH SURQRXQVLQ%DOLQHVH $UND DUHUHVWULFWHGWRFRUHDUJXPHQWV 6XEMHFWDQG2EMHFW MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 158 WRV\QWDFWLFIXQFWLRQVDQGRUV\QWDFWLFPDUNLQJ,WVKRXOGEHQRWHGWKDWZKLOHDJUHHPHQW has been described in this paper to involve subject and object, these Subject and Object ODEHOVVKRXOGEHXQGHUVWRRGDVPDFUR VHPDQWLF UROHV0DURULKDVQRV\QWDFWLFVXEMHFW SLYRWRIWKHW\SHIRXQGLQ(QJOLVKRUFHUWDLQ$XVWURQHVLDQODQJXDJHVVXFKDV,QGRQHVLDQ *UDPPDWLFDODJUHHPHQWUHTXLUHVWKDWWKHDJUHHLQJ13EHREOLJDWRULO\SUHVHQW7KLVLV QRW WKH FDVH ZLWK 0DURUL7KH IUHH 13 WKDW WKH YHUE DJUHHV ZLWK LV RIWHQ GURSSHG LH RSWLRQDOO\ SUHVHQW 7KLV LV VKRZQ LQ WKH IROORZLQJ H[DPSOHV H[WUDFWHG IURP WKH )URJ storyLQ0DURUL6HQWHQFH F FRPHVZLWKQRIUHH$813V7KHDJUHHPHQWPRUSKRORJ\ on the verb, both of which are zero formatives, anaphorically refers to the NPs mentioned earlier in the texts (‘the dog’, ‘Thomas’, and ‘(the) frog’)13 D Koro Thomas ¿ njaj uyow … dog Thomas with bed top µ7KRPDVDQGWKHGRJZHUH VOHHSLQJ RQWKHEHG«¶ )URJ6WRU\B3DVNDOLV WKUHHIRXUOLQHVODWHUOLQH E Mar tok reruwo rowae kuya-maf. 1(* IURJ MDU LQVLGH %(1RQ3/367 µ7KHUHZDVQRIURJLQVLGHWKHMDU¶ F 0EH WDQDPED H\HZ QGD¿ 352* QRZ VHH $8;1RQ3/5367 µ 7KH\ZHUH WZR QRZORRNLQJIRU LW LHWKHIURJ ¶ Further evidence that agreement in Marori is semantic in nature comes from the fact that, when the agreeing NP is present, it is for a functional-semantic reason to create a VSHFL¿FUHIHUHQW7KLVLVWKHFDVHZLWK'8$/UHIHUHQFH7KXVWKH1RQ6*SURQRXQemnde µDJUHHV¶ZLWKWKH1RQ3/$FWRUVXI¿[-m in (15a) to create a dual referent, as the translation VKRZV:KHQWKHDFWRULV3/-imLVXVHG FIWDEOH JLYLQJULVHWRndim (15b) (15) D Emnde na=n bosik eyew nda-m. 1RQ6* 6* IRU SLJ VHH QGDP 3-AUX-1RQ3/367 µ7KH\ WZR KXQWHGDSLJIRUPH¶ E Emnde usindu Maria=na bosik eyew ndim. 1RQ6* DOO 0DULD IRU SLJ VHHQGDLP $8;3/367 µ7KH\ DOOPRUHWKDQWZR KXQWHGDSLJIRU0DULD¶ It should be noted that the formation of DUAL reference in Marori is achieved by This is the frog story (Frog, where are you? E\0HUFHU0D\HU 13 ,QÀHFWLRQ VKRZLQJ WHQVH LQ 0DURUL LV FRPSOH[7KHUH LV PRUH WKDQ RQH ZD\ RI GRLQJ LW DQG V\QFUHWLVPDGGVWRWKHFRPSOH[LW\3DVWWHQVHIRU$FWRUIRUH[DPSOHFDQEHH[SUHVVHGE\ DGGLQJWKHVXI¿[-f (PST) or ¿ W\SLFDOO\UHPRWHSDVW 5367 DVVHHQLQ F RUDGGLQJ±m µ1RQ3/¶DVLQ D VHHDOVRWDEOH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 159 combining NonSG and NonPL morphemes in phrasal syntax as well as in word-internal V\QWD[([DPSOH(15)LOOXVWUDWHVWKHIRUPDWLRQRI'8$/LQFODXVDOV\QWD[([DPSOH(8), repeated here as (16a)VKRZVWKHIRUPDWLRQRI'8$/ZLWKLQWKHYHUE LHWKHFRPELQDWLRQ RIWKH1RQ6*DFWRUSUH¿[n- and -m 1RWHWKDQWKHFRPELQDWLRQRIn- -im is used when the actor is plural, giving rise to nedim as seen in (16b), and no n- is used with -m when the actor is singular giving rise to ndam (16c)7KXV D IRUPDO DQG IXQFWLRQDO DQDO\VLV of agreement in Marori must take into account these phrasal and sublexical layers of DJUHHPHQW7KHDJUHHPHQWDFURVVWKHVHOD\HUVKDVWREHGHDOWZLWKLQDXQLIRUPZD\7KH issue will be further discussed in subsection EHORZ (16) D Kie tamba Maria=na bosik eyew nadam. 1RQ6* 3(5) 0DULD IRU SLJ VHHQQGDP 1RQ6*$8;1RQ3/367 µ E Kie usindu Maria=na bosik eyew nedim. 1RQ6* DOO 0DULD IRUSLJVHH QQGDLP 1RQ6*$8;3/367 µ F Ka Maria=na bosik eyew ndam. 6* 0DULD IRU SLJ VHH QGDP $8;1RQ3/367 µ 3.3. PROJECTING MORPHOLOGY 3.3.1. What is projection? The notion of projection is one of the central concepts in PRGHUQV\QWDFWLFWKHRULHV,WUHIHUVWRWKHPHFKDQLVPE\ZKLFKD VXE XQLWRIDVWUXFWXUH determines or constrains a larger (syntactic) structure which it is a part of, or a structure it LVUHODWHGWR7KXVRQHFDQWDONDERXW OH[LFDO FDWHJRULFDOSURMHFWLRQHJDYHUE 9 LQWKH OH[LFRQ LVSURMHFWHGWRYHUESKUDVH 93 LQV\QWD[,Q&KRPVN\DQWHUPVWKH(33 ([WHQGHG Projection Principle) is proposed to ensure that the verb which is projected to syntax must KDYHDQ13LQWKHVXEMHFWSRVLWLRQ &KRPVN\ ,QWKH/)*PRGHO 'DOU\PSOH WKHWHUPµSURMHFWLRQ¶UHIHUVWRPDSSLQJRUFRUUHVSRQGHQFHEHWZHHQOD\HUVRIVWUXFWXUHV Projection of morphology to syntax refers to how a morpheme in a sublexical structure determines or constraints the structure of phrasal or clausal syntax of which the word LVSDUW%\µVWUXFWXUH¶ZHPHDQ JUDPPDWLFDO VWUXFWXUHRIGLIIHUHQWNLQGV7KHUHOHYDQW ones for the purpose of the present discussion are semantic (predicate-)argument structure ZKHUH$YV3DUHUHOHYDQW ZRUGLQWHUQDOVWUXFWXUH ZKLFKDJUHHPHQWDI¿[HVDUHSDUWRI DQGSKUDVDODQGFODXVDOV\QWD[ ZKLFKWKHDJUHHLQJ13VDUHSDUWRI Of particular interest are the projection issues in relation to the agreement patterns SUHVHQWHG HDUOLHU $GRSWLQJ D WUDGLWLRQDO YLHZ ZKHUH PRUSKRORJ\ DQG V\QWD[ DUH WZR GLIIHUHQWEXWUHODWHGGRPDLQVRIJUDPPDUZHKDYHWKHIROORZLQJTXHVWLRQV KRZGRZH maintain the distinction while at the same time capture the idea that the same principle applies across boundary of morphology and syntax? Regarding NUM agreement, what can we learn from Marori in relation to the feature structure of NUM? What is the best analysis, and to what extent is the analysis applicable to other languages? MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 160 ,QZKDWIROORZV,DGGUHVVWKHVHTXHVWLRQV,SURSRVHDOH[LFDOO\EDVHGDQDO\VLVIRU180 DJUHHPHQWLQ0DURULZKHUH'8$/LVQRWSULPLWLYH,VNHWFKKRZWKHSURSRVHGDQDO\VLVFDQ EHH[WHQGHGWRDFFRXQWIRUFRPSOH[180V\VWHPVLQRWKHUODQJXDJHV 3.3.2. Proposed feature structure and claims. The points of the analysis are the IROORZLQJ)LUVWIROORZLQJ+DOH ,DGRSWWKHDQDO\VLVWKDW6*DQG3/IHDWXUHVDUH WKHPRVWEDVLF180IHDWXUHV(DFKKDVDELQDU\YDOXH DVVKRZQLQ¿JXUHD a. NUM ={[+/- SG], [+/-PL] } b. DUAL =[-SG,-PL] (where [-SG] is NSG and [-PL] is NPL) FIGURE 6HFRQGRQWKHEDVLVRI¿JXUHDWKH'8$/LQDWKUHHZD\180V\VWHPDVREVHUYHG in Marori (SG, DU, PL) is analysable as being formed out of these basic NUM features, QDPHO\>6*3/@7KLVLVVKRZQLQ¿JXUHE 7KLUGDVVHHQIURPIHDWXUHVSHFL¿FDWLRQLQ¿JXUHE'8$/LVXQPDUNHG,WLVIRUPHG RXWRIDFRPELQDWLRQRIWZRIHDWXUHVZLWKQHJDWLYHYDOXHV There is evidence from Marori WKDW'8$/LVLQGHHGHQFRGHGE\WZRXQGHUVSHFL¿HGPRUSKHPHVHJn- -m in (16) glossed DV1RQ6*DQG1RQ3/UHVSHFWLYHO\7KHUHLVDOVRHYLGHQFHIURP1HQ D3DSXDQODQJXDJH of southern New Guinea) where certain verbal stems expressing DUAL are unmarked and WKHIRUPDWLRQRI6*3/LVDFKLHYHGE\KDYLQJDGGLWLRQDOPDUNLQJRQWKHVHVWHPV )RXUWKZKLOHDVSHFL¿FQXPEHUPRUSKRORJ\VLJQDOVWKHSUHVHQFHRIDQXPEHUIHDWXUH I claim that the absence of number morphology associated with a form does not mean that WKHIRUPFRQWULEXWHVQRQXPEHUIHDWXUH:KDWQXPEHULQIRUPDWLRQLVFRQWULEXWHGE\WKH IRUPLVOH[LFDOO\GHWHUPLQHGZLWKLQWKHODUJHUV\VWHPRIWKHODQJXDJH)RUH[DPSOHWKH demonstrative this or thatLQ(QJOLVKFDQEHDQDO\VHGDVFDUU\LQJ>180>3/@@EHFDXVHD demonstrative is part of the nominal category in English where plural is morphologically PDUNHG+HQFHLQRXUDQDO\VLVthis/thatFDUULHV>3/@15 (and is compatible with a noun FDUU\LQJ>3/@VXFKDVchildren7KHGH¿QLWHDUWLFOHthe, however, does not enter into +DUOH\ 5LWWHU SURYLGHDQDQDO\VLVZKHUHGXDOLVXQLYHUVDOO\DVVRFLDWHGZLWKSRVLWLYH VSHFL¿FDWLRQRIERWK0LQLPDODQG*URXS VHPDQWLF IHDWXUHVURXJKO\FRUUHVSRQGLQJWRµVLQJXODU¶ DQGµSOXUDO¶ZLWKHYLGHQFHIRUH[DPSOHFRPLQJIURP+RSL,QWKLVODQJXDJHGXDOLVH[SUHVVHG E\ERWKVLQJXODUDQGSOXUDOIRUPV(YLGHQFHIURP3DSXDQODQJXDJHVDVGLVFXVVHGLQWKLVSDSHU however, shows that dual is expressed distributively by two NonSG and NonPL morphemes VXSSRUWLQJWKHDQDO\VLVWKDWWKHLU6*DQG3/IHDWXUHVFDUU\QHJDWLYHYDOXHV 15 2QHPLJKWZDQWWRDQDO\VHWKDW(QJOLVKVLQJXODUQRXQVDQGGHPRQVWUDWLYHVWKLVWKDWFDUU\>6*@ :KLOHWKLVLVLQWXLWLYHO\UHDVRQDEOHWKHUHDUHJRRGUHDVRQVZK\WKLVDQDO\VLVLVXQWHQDEOH,WZRXOG PHDQWKDWVLQJXODULVPRUSKRORJLFDOO\PDUNHGLQ(QJOLVK LHWKHUHLVDGHGLFDWHGPRUSKRORJ\WR PDUNVLQJXODUZKLFKLVQRWWKHFDVH 7KHWKLUGSHUVRQVLQJXODUSUHVHQWWHQVHVLVQRWVROHO\IRU QXPEHU ,QDGGLWLRQLWZRXOGOHDGWRDQXQZDQWHGRXWFRPHLQWKHXQL¿FDWLRQSURFHVVDOORZLQJ XQDFFHSWDEOHVWUXFWXUHZLWKIHDWXUHXQL¿FDWLRQRI>180>6*3/@LQ(QJOLVKVXFKDVLQ this children MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 161 QXPEHUFRQWUDVWLQ(QJOLVKDQGWKHUHIRUHFDUULHVQRQXPEHUIHDWXUH 7KH¿IWKNH\DQGQHZSURSRVDOLVWKHPRGHOOLQJDQGFRQFHSWLRQRI180V\VWHP,DUJXH WKDWWKH180V\VWHPPXVWEHXQGHUVWRRGDVUHÀHFWLQJODQJXDJHVSHFL¿FFDWHJRULVDWLRQV of the semantic space of NUM, and that NUM features are distinguished and structured on the basis of the corresponding structures of the relevant NUM spaces to which they are PDSSHGRQWR I therefore claim that feature operations to establish NUM referents are determined RUFRQVWUDLQHGE\WKHVHPDQWLFVSDFHRI180RIWKHODQJXDJH7KLVDOORZVXVWRSURYLGH D QDWXUDO H[SODQDWLRQ IRU FHUWDLQ FDVHV ZKLFK DSSHDU WR EH XQXVXDO HJ WKH FRGLQJ RI exhaustive set/plural or paucal using SG/DUAL morpheme in Nen (discussed in section EHORZ ,QWKHSURSRVHGDQDO\VLVGLIIHUHQWLQWHUSUHWDWLRQVRI3/ZKLFKPD\RUPD\QRW include the meaning of DUAL/TRIAL as in Larike (discussed in section ) also follow QDWXUDOO\ To begin with, the simplest model of pairing between (NUM) FORM and its FRUUHVSRQGLQJVHPDQWLFVSDFH 0($1,1* LVDUJXDEO\WKHRQHVKRZQLQ¿JXUHEHORZ (I will show later that the MEANING part is richly structured with possible overlapping VSDFHV 7KHOLQHZLWKDQDUURZDWWKHHQGDVVRFLDWHGZLWK3/85$/LVPHDQWWRFDSWXUHWKH LGHDWKDWSOXUDOLW\LVTXDQWLWDWLYHO\XQVSHFL¿HG,QFRQWUDVWVLQJXODULW\ LHEHLQJµRQH¶ LV TXDQWLWDWLYHO\VSHFL¿F KHQFHQRDUURZLVUHSUHVHQWHGDWWKHHQGRI180VSDFH16 (1). ‘one’ ‘two’ ‘three’ ‘four’ … (MEANING) SINGULAR PLURAL (FORM) FIGURE 5 Languages differ in the way the space between the two ends is divided and encoded OLQJXLVWLFDOO\7RFDSWXUHWKHGLIIHUHQFHVDQGVLPLODULWLHVLWLVQHFHVVDU\WRUHSUHVHQWWKH internal structure of the space explicitly, from which the abstraction of atomic NUM IHDWXUHV >6*@DQG>3/@ FDQEHSRVWXODWHG7RLOOXVWUDWHWKHSRLQWVWKH180V\VWHP LQ0DURULLVFRPSDUHGZLWKWKDWRI(QJOLVK(QJOLVKLVVLPSOHUDQGLVGLVFXVVHG¿UVW (QJOLVKKDVDWZRZD\180V\VWHPVKRZLQJ6* µRQH¶ YV3/ µPRUHWKDQRQH¶ RSSRVLWLRQ,WLVWKH3/85$/FDWHJRU\WKDWLVPRUSKRORJLFDOO\PDUNHGLQWKLVODQJXDJHRQ QRXQV7KDWLVWKHUHLVDGHGLFDWHG3/85$/PRUSKHPHLQ(QJOLVK*LYHQWKDW3/KDVD ELQDU\ YDOXHDQGWKDW6*LVDQDO\VDEOHDV>3/@WKHVLPSOHVWDQDO\VLVLVWKDW(QJOLVK 16 1RWHWKDWZHDUHWDONLQJDERXW180LQQDWXUDOODQJXDJHVHPDQWLFV,QPDWKHPDWLFDOVHQVHRQH FDQWDONDERXWVSDFHVEHORZRQHRU]HUR LHPLQXVHV LQZKLFKFDVHWKHUHVKRXOGEHDQDUURZ VSHFL¿HGIRUWKHOLQH)RUVLPSOLFLW\,DOVRLJQRUHWKHFRPSOLFDWLRQLQWKHFRQFHSWLRQRI180 in ‘mass’ nouns, where the FORM is SINGULAR but it does not refer to an individuated ‘one’ HQWLW\HJ(QJOLVKwater and air MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori HPSOR\VRQH180IHDWXUHRQO\QDPHO\WKH3/IHDWXUH7KHVFKHPD¿JXUHUHSUHVHQWVWKH PDSSLQJRIWKH3/IHDWXUHRQWRWKHVHPDQWLFVSDFHRI180LQ(QJOLVK$VVHHQWKHIHDWXUH >3/@FDUULHGE\D3/85$/IRUPPDSVRQWRD180VSDFHFRYHULQJµPRUHWKDQRQH¶ LQGLFDWHGE\DWKLFNOLQHLQWKHDUURZ$6,1*8/$5IRUPDQDO\VHGDVFDUU\LQJWKH>3/@ IHDWXUHZKLFKLVPDSSHGRQWRµRQH¶ The semantic space of NUM in English: PL -PL +PL ‘one’ two ‘three’ ‘four’ … SINGULAR PLURAL (FORM) FIGURE 6. The semantic space of NUM in English 0DURULKRZHYHUKDVDWKUHHZD\180V\VWHP 6*3/DQG'8$/*LYHQWKHIDFWWKDW both the SG and PL have their marking on the verb in Marori, we can say that Marori DFWLYDWHVERWK6*DQG3/IHDWXUHVLQLWVV\VWHP8QOLNHLQ(QJOLVKZKHUH6,1*/8$5LV >3/@0DURUL¶V6,1*8/$5LVLQGHHGDVVRFLDWHGZLWKWKH6*IHDWXUH7KXVWKHVHPDQWLF VSDFHRI180LQ0DURULFDQEHUHSUHVHQWHGLQ¿JXUH%RWK6*DQG3/IHDWXUHVZLWKWKHLU UHVSHFWLYHELQDU\YDOXHVDUHSUHVHQWLQ0DURULDQGPDSSHGRQWRWKH180VSDFH (1). The semantic space of NUM in Marori SG PL +S G -SG -PL +PL ‘one’ two ‘three’ ‘four’ … SINGULAR DUAL PLURAL FIGURE 77KHVHPDQWLFVSDFHRI180LQ0DURUL 7KH180VSDFHRI0DURULQHHGVVRPHH[SODQDWLRQ7KHIHDWXUHVDUHKLHUDUFKLFDOO\ VWUXFWXUHG7KH3/IHDWXUHLVHPEHGGHGLQWKH>6*@IHDWXUHYDOXH7KLVLVWRFDSWXUHWKH LQWXLWLRQWKDWWKH¿UVWKLJKRUGHUGLVWLQFWLRQLVEHWZHHQ6*YV1RQ6*:HKDYHVHHQWKDW there is morphological evidence for this in Marori: the distinction is expressed by two GLVWLQFWPRUSKHPHV ZLWKFRPSOH[LW\LQWKHYHUEDOPRUSKRORJ\DVVHHQLQWDEOH 7KHUH LVDOVRHYLGHQFHIURPRWKHUODQJXDJHVHJ/DULNH GLVFXVVHGLQVHFWLRQ below) where PL LVXQGHUVWRRGLQLWVEURDGVHQVHDVµQRQVLQJXODU¶ LHDOVRFRYHULQJµGXDO¶ LQDGGLWLRQWR MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 163 LWVQDUURZVHQVHRIµWKUHHRUPRUH¶ 7KH DQDO\VLV DV VKRZQ LQ ¿JXUH DOORZV XV WR DFFRXQW IRU IDFWV DERXW GLVWULEXWHG H[SRQHQWV LQ 180 H[SUHVVLRQV LQ 0DURUL DQG DOVR LQ RWKHU ODQJXDJHV :H FDQ straightforwardly account for the economical way of encoding of DUAL in Marori by PHDQVRIFRPELQLQJWKHDYDLODEOH1RQ6* >6*@ DQG1RQ3/ >3/@ PRUSKHPHV However, it should be noted that the three-way number distinction in Marori can also EH HQFRGHG E\ GLVWLQFW IRUPV W\SLFDOO\ IRU WKH ¿UVW SHUVRQ LQ WKH SUHVHQWIXWXUH WHQVH These forms are often associated with partially irregular lexically-determined paradigmatic SDWWHUQVHJdu, -den, -men, and -ru, -ren and -menIRUWKH¿UVWSHUVRQVLQJXODUGXDODQG plural categories in present and future tenses as shown in (17 DE )RUWKHSDVWWHQVHDV seen in (17c), -menLVXVHGIRUWKH¿UVWSHUVRQLUUHVSHFWLYHRIWKHQXPEHULQFRQWUDVWWRWKH VHFRQGDQGWKLUGSHUVRQV (17) D 7KHSDUDGLJPRIWKH DX[LOLDU\ YHUEµEHDWVLW¶LQWKHSUHVHQWWHQVHLQ0DURUL Singular kuyedu kami kuye Dual kuyeden kanermi kuye Plural minggemen kaminenggem minggri E 7KHSDUDGLJPRIWKH DX[LOLDU\ YHUEµEHDWVLW¶LQWKHIXWXUHWHQVHLQ0DURUL Singular miru kami mi Dual miren kanermi mi Plural minggemen kaminenggem minggem F 7KHSDUDGLJPRIWKH DX[LOLDU\ YHUEµEHDWVLW¶LQWKHSDVWWHQVHLQ0DURUL Singular kuyemen kuyem kuyem Dual kuyemen norowem kuyem Plural mingrimen minenggrim minggrim While the dual forms -den as in kuyeden and -ren as in miren encode number, they are DFWXDOO\SRUWPDQWHDXPRUSKHPHVWKDWDOVRHQFRGHVSHFL¿FSHUVRQDQGWHQVH LHSUHVHQW IXWXUH LQIRUPDWLRQ7KHUHIRUH WKH\ DUH LQ D VHQVH QRW UHDOO\ GHGLFDWHG '8$/ QXPEHU PRUSKHPHV *LYHQ WKH RYHUDOO V\VWHP RI QXPEHU LQ 0DURUL ZH FDQ VWLOO PDLQWDLQ WKH DQDO\VLVWKDWWKHUHLVQRQHHGWRKDYHD'8$/IHDWXUHLQWKLVODQJXDJH'8$/PRUSKHPHV such as -denZKLOHJORVVHGDV'8 $/ IRUVLPSOLFLW\FDQEHVSHFL¿HGDVFDUU\LQJ>6* 3/@DVSDUWRIIHDWXUHEXQGOHV>6*3/35(6@IHDWXUHV LHPHDQLQJD¿UVWSHUVRQ GXDOSUHVHQWWHQVHPRUSKHPH7KHDQDO\VLVDFFRXQWVIRUWKHIDFWWKDWWKHDX[LOLDU\LWLV DI¿[HGWR HJkuyeden) can enter into subject agreement with nie ‘1NonSG’ as in (18) EHFDXVH WKH VXEMHFW FDUULHV WKH VDPH IHDWXUH YDOXH >6*@ ZLWK ZKLFK LW FDQ XQLI\ 7KH PHFKDQLVPRIXQL¿FDWLRQLVIXUWKHUGLVFXVVHGLQVHFWLRQEHORZ (18) Nie purfam Jayapura di kuye-den. 1RQ6* SHUVRQ -D\DSXUD VRRQ EHDW'835(6 µZH WZR DUHLQ-D\DSXUDVRRQ¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori To conclude, there is good evidence to support the analysis that the three-way NUM V\VWHPLQ0DRULKDVWZREDVLF180IHDWXUHV6*DQG3/ZLWKELQDU\ YDOXHV7KH (1). IHDWXUHEXQGOHVLQ0DURULDUHUHSUHVHQWHGLQ¿JXUHNUM system in Marori SINGULAR DUAL PLURAL (NUMBER CATEGORIES) [SG +] [SG -] [PL +] (FEATURE BUNDLES) [PL-] [PL -] [SG -] SG NSG&NPL PL (MORPHOLOGICAL CODING) FIGURE 8180V\VWHPLQ0DURUL 'LVWULEXWHG180H[SRQHQFHDFURVVPRUSKRORJ\DQGV\QWD[Having discussed the feature structure, we are now ready to discuss the issue of distributed NUM exponence IXUWKHULQDSULQFLSOHGDQGSUHFLVHZD\7KLVFDQEHVWUDLJKWIRUZDUGO\GRQHZLWKLQWKH XQL¿FDWLRQEDVHGPRGHORIJUDPPDUDVGHVFULEHGLQ,QZKDWIROORZVZHGLVFXVVDQG H[HPSOLI\KRZ'8$/LVDUULYHGDWLQPRUSKRORJ\DQGV\QWD[ Consider example (19a)ZKHUHWKH$FWRUµ\RX¶LV'8$/,WV1RQ6*H[SRQHQWVFRPH from syntax (the free pronoun kieµ1RQ6*¶ DQGPRUSKRORJ\ WKHDI¿[HVn- and -m in the YHUE 7KHVHPRUSKHPHVFDUU\WKH180IHDWXUHZLWKFRPSDWLEOHYDOXHVZKLFKWKHQXQLI\ WRIRUP'8$/7KHXQL¿FDWLRQLVVKRZQLQ(19b) (19) D Kie tamba Maria-na bosik eyew nadam. 1RQ6* 3(5)0IRUSLJ VHDUFKQQGDP 1RQ6*$8;1RQ3/367 µ E >180>6*@@ 8>180>6*@@8 >180>3/@@ 1806* kie n- -m [ [ PL - @ @ 7KHIRUPDWLRQRI'8$/FDQWDNHSODFHLQWKHOH[LFRQDQGV\QWD[,Q(19), it is formed E\WKHXQL¿FDWLRQRIn- and -m when the verb nadamLVFUHDWHG:KHQWKHYHUEnadam combines in syntax with the free pronoun kie, the NUM information from these units IXUWKHUXQL¿HV7KHYHUEnadam is not acceptable if the actor is singular kaµ6*¶IRU which the verb ndam must be used as seen in (19c)7KHXQL¿FDWLRQIDLOVEHFDXVHka carries >180>6*@@IHDWXUHZKLFKLVLQFRPSDWLEOHZLWKWKDWFDUULHGE\n- -m (19d) F .D 0DULD QD ERVLN H\HZ QGDP QDGDP 6* 0DULD IRU SLJ VHH QGDP $8;1RQ3/367 µ G >180>6*@@ 8 >180>6*@@8>180>3/@@ 1806* ka n- -m [ PL - @@ In '8$/LVIRUPHGLQV\QWD[QRWLQPRUSKRORJ\8QOLNHLQnadam (19), there MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 165 LV QR 1RQ6* DI¿[ LQ WKH YHUE PRUSKRORJ\ RI ndam WR PDNH '8$/ 7KH YHUE ndam emerges from the lexicon searching out its Actor and Undergoer arguments, and the NUM LQIRUPDWLRQ >3/@ IURPWKHDFWRUVXI¿[mXQL¿HVZLWKWKH180LQIRUPDWLRQIURPWKH free pronoun emndeJLYLQJULVHWR'8$/7KHXQL¿FDWLRQLVVKRZQLQ E D Emnde na-n bosik eyew ndam. 3NonSG 1SG-for pig search Ø-nda-m 3-AUX-1RQ3/367 µ7KH\ WZR VHDUFKHGDSLJIRUPH¶ E >180>6*@@8 >180>3/@@ 1806* | | [ [ PL - @ @ emnde- [ -m@9(5% 7KH XQL¿FDWLRQEDVHG IHDWXUH DQDO\VLV RI $JUHHPHQW SUHVHQWHG DERYH VKRZV WKH IROORZLQJSRLQWV)LUVWLWDOORZVXVWRPDLQWDLQWKHWUDGLWLRQDOGLVWLQFWLRQRIPRUSKRORJ\ and syntax and at the same time also to capture the projection of morphology to syntax whereby referential information (in this case NUM values) can pass up across the ERXQGDU\ RI PRUSKRORJ\ DQG V\QWD[ 6HFRQG 180 DJUHHPHQW LV HVVHQWLDOO\ IHDWXUH value compatibility, which operates on the basis of the same principle irrespective of ZKHWKHULWWDNHVSODFHLQDFODXVHRUDZRUGOHYHO7KLUGZLWKIHDWXUHVEHLQJPDSSHGRQWKH semantic space of NUM, we can also capture the fact that agreement is more than simply FRPSDWLELOLW\RIIHDWXUHV*LYHQWKH180VSDFHRI¿JXUHWKHDJUHHPHQWLVIXQFWLRQDO EHFDXVHZKHQ>6*@DQG>3/@FRPELQHWKH\QDUURZGRZQWRVHOHFWWKH180VSDFHRI '8$/ 4. TYPOLOGICAL NOTES. A typological space of NUM is proposed in subsection 3/DQG6*DUHWKHEDVLF180IHDWXUHVZLWKELQDU\YDOXHV /DQJXDJHVYDU\ZLWK UHVSHFWWRZKHWKHURQHRUERWKRIWKHPDUHDFWLYDWHG,WKDVEHHQDUJXHGWKDWWKHIHDWXUH VWUXFWXUHLVKLHUDUFKLFDOZLWK3/EHLQJHPEHGGHGLQ>6*@DQGWKDW'8$/LVQHJDWLYHO\ GH¿QHGDV>6*3/@ 7KHTXHVWLRQQRZLVZKHWKHUWKHSURSRVHGDQDO\VLVRI'8$/LQ0DURULFDQEHH[WHQGHG to account for DUAL in other languages, possibly in those with richer NUM distinctions HJWULDORUSDXFDO 'LVFXVVLQJWKHVHLQGHSWKDFURVVODQJXDJHVLVEH\RQGWKHVFRSHRIWKH SUHVHQWSDSHU+RZHYHULQZKDWIROORZV,GLVFXVV'8$/LQWKUHHRWKHUODQJXDJHV 1HQ 3DSXDQ +RSL 8WR$]WHFDQ86 DQG/DULNH $XVWURQHVLDQ0DOXNX,QGRQHVLD 4.1. DUAL IN NEN. '8$/LQ1HQ (YDQV WKLVYROXPH LVXQPDUNHG 1RQ '8$/ 1' LVPDUNHGHJowab ‘talk (of two) Æ owab-ta ‘talk (of one, or three or PRUH¶ '8$/LQ1HQKRZHYHUPD\DOVREHPDUNHGHJaka-wµVHH'8¶YVaka-taµVHH1' 8QOLNHLQ0DURULWKH'8$/YVQRQ'8$/PDUNLQJLVV\VWHPDWLFLQ1HQ1HQDUJXDEO\ DFWLYDWHV'8$/DVDUHOHYDQW180IHDWXUHLQLWVJUDPPDU,PSRUWDQWO\WKHUHLVQRVSHFL¿F morphology for PL in Nen: it is expressed by means of a compositional strategy making XVHRIWKHDYDLODEOH XQGHUVSHFL¿HG 180PDUNHUV7KLVLVIXUWKHUGLVFXVVHGEHORZ /LNH0DURUL1HQDOVRVKRZVGLVWULEXWHGH[SRQHQFHIRUWKHIRUPDWLRQRIVSHFL¿F180 MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 166 UHIHUHQFH$EVROXWLYHIUHHSURQRXQVVKRZQR180GLVWLQFWLRQVEXW3(56GLVWLQFWLRQRQO\ %RXQGLQÀHFWLRQDODI¿[HVRQWKHYHUEVKRZWKH6*YV1RQ6*GLVWLQFWLRQV7KXVLQWKH IROORZLQJH[DPSOHVWKHVSHFL¿FUHIHUHQFHRIWKH¿UVWSHUVRQVLQJXODU D ¿UVWSHUVRQ plural E DQG¿UVWSHUVRQGXDO F LVGHWHUPLQHGE\WKHFRPELQDWLRQRIWKH¿UVWIUHH pronoun ynd XQVSHFL¿HGIRU180 DQGWKHDJUHHPHQWPRUSKRORJ\RQWKHYHUE ZKLFK supplies NUM information): D tog-am ynd w-aka-t-e (YDQV FKLOG(5* $%6 6*8VHH1RQ'8VJ$ µ7KHFKLOGVHHVPH¶ E tog-am ynd yn-aka-t-e FKLOG(5* $%6 1RQ6*8VHH1RQ'8VJ$ µ7KHFKLOGVHHVXV RUPRUH ¶ F ämbs är-äm ynd yn-akae-w-ng RQH PDQ(5* $%6 1RQ6*8VHH'8a6*$!'82 µ2QHPDQVHHVWKHWZRRIXV¶ On the basis of the available evidence, I propose that the structure of the semantic space RI180DQGUHODWHGIHDWXUHVLQ1HQLVVKRZQLQ¿JXUH FIGURE 97KHVHPDQWLFVSDFHRI180LQ1HQ 7KHDQDO\VLVDVGHSLFWHGLQ¿JXUH KDVWKHIROORZLQJDGYDQWDJHV)LUVW1HQDFWLYDWHV WKHIHDWXUH'8$/LQLWV180V\VWHP&UXFLDOO\WKH'8$/IHDWXUHLVVWUXFWXUHGDVSDUWRI >6*@7KHVSDFHRI'8$/µWZR¶LVUHIHUDEOHE\PHDQVRI>'8@ 1RWHWKDWERWKQHJDWLYH XQGHUVSHFL¿HG DQGSRVLWLYHQXPEHUYDOXHQXPEHUPD\EH DVVRFLDWHGZLWKRYHUWPRUSKHPHVLQ1HQ HJWKHQRQGXDOPRUSKHPH-t as in w-aka- t-e ‘1sgU-see’ D ,Q RWKHU ZRUGV ZH DGRSW DQ DQDO\VLV ZKHUH D QHJDWLYH YDOXH RI QXPEHU GRHV QRW QHFHVVDULO\ PHDQ WKDW LW LV PRUSKRORJLFDOO\ XQPDUNHG &RQYHUVHO\ D MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 167 morphologically simple form may carry a positive value of DUAL number feature lexically by default as is the case with owabµWDON RIWZR 7KHDQDO\VLVRIHPEHGGLQJ>'8@XQGHUWKH>6*@IHDWXUHLQWKHVWUXFWXUHGVHPDQWLF VSDFH¿QGVLWVHPSLULFDOVXSSRUWIURPWKHGLVWULEXWHGH[SRQHQFHIHDWXULQJWKLVODQJXDJH The presence of NonSG and DUAL morphemes to express DUAL as seen in in WKHSUHVHQFHRID'8$/PRUSKHPHVHHPVWREHUHGXQGDQWDW¿UVW+RZHYHUJLYHQWKH distributed exponence (where two exponents are needed to express DUAL), the two H[SRQHQWVDUHQDWXUDOO\WKRVHDVVRFLDWHGZLWKWKH'8$/VSDFHQDPHO\>'8@DQG>6*@ 7KHVSDFHRI'8$/LVDVSHFL¿FSRUWLRQRI180VSDFHRI>6*@ yä-trom-aran ۆPQ KRXVH 1RQ6*8EHHUHFWHG67$7 '8 µ7ZRKRXVHVDUHVWDQGLQJ¶ 6HFRQG WKH VSDFH RI >'8@ LV VSOLW 7KLV LV FRQVLVWHQW ZLWK WKH PHDQLQJ RI 1RQ '8$/ 1' DQGWKHGLVWULEXWLRQRI1'PRUSKHPHLQWKLVODQJXDJH)RUH[DPSOHWKH1' PRUSKHPHLVH[SHFWHGWREHXVHGIRUSOXUDOUHIHUHQFH7KLVLVLQGHHGWKHFDVH FI>6*@ 1RQ6* ZKLFKFRPELQHVZLWK>'8@ 1' LQH[DPSOH E 7KLUG LQ RXU DQDO\VLV WKH VSDFHV RI WZR FDWHJRULHV PD\ RYHUODS HJ WKH VSDFHV RI >6*@DQG>'8@LQ1HQ7KHFRQFHSWLRQRIRYHUODSSLQJVSDFHVLVLQIDFWVLJQL¿FDQWIRU VSHFL¿F180UHIHUHQFHDQGFRGLQJ7KXVWKHFRGLQJRID6*UHIHUHQWLQ1HQPDNHVXVH of the exponents signifying SG and NON-DUAL as seen in 7KLVLVH[SHFWHGRQWKH SURSRVHGVWUXFWXUHGVHPDQWLFVSDFHRI180LQGLVWULEXWHGPRUSKRORJ\ WURPQJU\ ۆPQ KRXVH 6*8EHHUHFWHG67$7 1RQ'8 µ$KRXVHLVVWDQGLQJ¶ Fourth, the analysis with conception of NUM involving overlapping spaces provides a natural account for what is otherwise a peculiar strategy of coding plural and exhaustive SOXUDOSDXFDOLQ1HQ $VVHHQLQ¿JXUH WKHVSDFHDVVRFLDWHGZLWK1RQ'8$/ >'8@ LVVSOLWLQWRWZR RQH RYHUODSVZLWKWKHVSDFHRI>6*@DQGWKHRWKHUZLWKWKHVSDFHRI>6*@,QWKHODWWHUFDVH LWLVHTXLYDOHQWWRWKHVSDFHRISOXUDO LHµWKUHHRUPRUH¶ ,QRWKHUZRUGVWKHVSDFHRI SOXUDOLVWKHSRUWLRQRIWKHVSDFHRI>6*@WKDWLV1RQ'8$/ >'8 6LQFHERWK>6*@DQG >'8@KDYHWKHLUUHVSHFWLYHFRGLQJPRUSKHPHVLWLVQRWVXUSULVLQJWKDW1HQGRHVQRWQHHG DVSHFLDOPDUNHUIRUSOXUDO%RWK1RQ6*DQG1RQ'8PRUSKHPHVDUHXVDEOHWRHQFRGH SOXUDODVH[HPSOL¿HGLQ EHORZ7KHLUXVHPHHWVWKHODQJXDJHVSHFL¿FUHTXLUHPHQWRI GLVWULEXWLYHH[SRQHQWVLQH[SUHVVLQJ180LQWKLVODQJXDJH lWURPQJU\ ۆPQ KRXVH 1RQ6*8EHHUHFWHG67$7 1RQ'8 µ7KUHHRUPRUHKRXVH V DUHVWDQGLQJ¶ Finally, the expression of what Evans calls the exhaustive set/universal ‘all’, which DOVR DSSHDUV WR EH XQXVXDO DW ¿UVW EXW LV LQ IDFW D QDWXUDO ZD\ RI H[SUHVVLQJ 180 LQ WKHSURSRVHGDQDO\VLV([KDXVWLYHVHWLVH[SUHVVHGE\PHDQVRIVLQJXODUPRUSKRORJ\LQ combination with dual morphology, as seen in 7KLVPLJKWEHHTXLYDOHQWWRµSDXFDO¶ LQRWKHUODQJXDJHV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 168 The expression of exhaustive set/all must refer to the space that is complementary to the space of the non-exhaustive plural expressed by the combination of non-singular and non- GXDOPRUSKHPHVDVH[HPSOL¿HGLQ $VVHHQLQ)LJXUH WKHVSDFHRIWKHQRQH[KDXVWLYH SOXUDOLVLQWKHULJKWHQGRI180VSDFH7KHVSDFHIRUWKHH[KDXVWLYHSOXUDO RUSDXFDO LV ORJLFDOO\WKHRQHLQWKHOHIWLQFOXGLQJWKDWRI>6*@ $JDLQ GXH WR WKH GLVWULEXWHG H[SRQHQFH UHTXLUHPHQW RI 180 H[SUHVVLRQ LQ 1HQ Nen needs no special morpheme to encode the exhaustive/paucal NUM because there are UHVRXUFHVDOUHDG\DYDLODEOHIRUWKLVQDPHO\WKHPRUSKHPHVVLJQLI\LQJ>6*@DQG>'8@ as seen in +RZHYHUZHKDYHWRQRWHWKHIDFWWKDWWKHFRPELQDWLRQRIWKHVHPRUSKHPHV is not compositional: the meaning has been ‘lexicalised’ as ‘exhaustive plural’ in contrast to ‘unlimited or general plural’ expressed by the combination of non-singular and non-dual PRUSKHPHVLQ1HQ .y-trom-aran ۆPQ KRXVH 6*8EHHUHFWHG67$7 '8 µ$OOWKHKRXVHVDUHVWDQGLQJ¶ 4.2. DUAL IN HOPI. The proposed analysis for Marori and Nen can be applied to DFFRXQWIRU+RSLGDWD,Q+RSL +DOH &RUEHWW , the combination of SG and PL morphemes give rise to DUAL interpretation as seen in F &RUEHWWH[SODLQVFDVHV H[HPSOL¿HGLQ F as ‘constructed’ numbers: dual is constructed from the number on the SURQRXQDQGWKDWRQWKHYHUE D Pam wari. WKDW6* UXQ3(5)96* µ+HVKHUDQ¶ E Puma yùutu. WKDW3/ UXQ3(5)93/ µ7KH\ SOXUDO UDQ¶ F Puma wari. WKDW3/ UXQ3(5)96* µ7KH\ WZR UDQ¶ The analysis of Hopi agreement in this paper is in the same spirit as the analysis VXJJHVWHGE\+DOH +HVXJJHVWHGWKDW'8$/LQWHUSUHWDWLRQFRXOGEHDFKLHYHGYLD LQWHUVHFWLRQRIWKHWZRELQDU\RSSRVLWLRQV >6*@>3/@ +RZHYHUWKHSUHFLVHGHWDLO RI+DOH¶VDQDO\VLVDVWRKRZWKHµLQWHUVHFWLRQ¶H[DFWO\ZRUNVUHPDLQVXQFOHDU,QWKLVSDSHU we present the analysis as a system of feature mapping onto a structured semantic space of 180DVVKRZQLQ¿JXUH7KHUHOHYDQW180IHDWXUHVDUHSURFHVVHGLQWKHVDPHZD\DV RWKHUJUDPPDWLFDOIHDWXUHVLQWKHJUDPPDU7KHXQL¿FDWLRQRI180IHDWXUHVLVH[SHFWHG to be constrained and/or functionally motivated by the possible reference to the semantic VSDFHRI180 MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 169 (1). The Semantic space of NUM in Hopi SG PL +S G -SG -PL +PL ‘one’ two ‘three’ ‘four’ … SINGULAR DUAL PLURAL FIGURE 10. The semantic space of NUM in Hopi /HW PH EH VSHFL¿F DERXW KRZ '8$/ LQ +RSL FDQ EH DUULYHG DW ,W LV HVVHQWLDOO\ LQ the same way as that in Marori, but with some constraints due to the mapping onto the VHPDQWLFVSDFHRI180)LUVWZHKDYHWRVSHFLI\KRZWKH180IHDWXUHLVFDUULHGE\WKH UHOHYDQWPRUSKHPHVLQ+RSL,WVKRXOGEHQRWHGWKDW+RSLLVXQOLNH0DURULLQWKDWLWKDV QRGHGLFDWHGXQGHUVSHFL¿HG1RQ6*1RQ3/PRUSKHPHVLHWKRVHFDUU\LQJ>6*@>3/@ IHDWXUHV3URQRPLQDORUYHUEDOPRUSKHPHVLQ+RSLJORVVHGDV6*3/FDQEHDQDO\VHGDV FDUU\LQJIHDWXUHEXQGOHVZLWKWKHLUYDOXHVDVVKRZQLQ¿JXUH a. PL morpheme b. SG morpheme FIGURE 11 The second point of the analysis is the implication the mapping of the NUM feature onto WKHVHPDQWLFVSDFHRI180KDVLQUHODWLRQWRWKHFRPELQDWRU\SURSHUW\RIWKHJUDPPDU 7KH PDSSLQJ FDQ EH ZLGH RU QDUURZ *LYHQ WKH VWUXFWXUHG VSDFH LQ ¿JXUH WKH PDSSLQJRI3/85$/PRUSKHPH ¿JXUHD IRUH[DPSOHPD\EHDVVRFLDWHGZLWKWKH VSDFHFRUUHVSRQGLQJWR>6*@ µZLGH¶LQFOXGLQJµWZR¶ 7KDWLVµSOXUDO¶PHDQVµPRUHWKDQ RQH¶$OWHUQDWLYHO\WKH3/85$/PRUSKHPHPHDQVµPRUHWKDQWZR¶ LHUHIHUULQJWRWKH VSDFH FRUUHVSRQGLQJ WR >3/@ µQDUURZ¶ H[FOXGLQJ µWZR¶ /LNHZLVH WKH 6,1*8/$5 PRUSKHPH ¿JXUHE FDQUHIHUWRWKHH[DFWQDUURZVSDFHRIµRQH¶GXHWRLWV>6*@IHDWXUH RUDOWHUQDWLYHO\WRDZLGHUVSDFHLQFOXGLQJµWZR¶GXHWRLWV>3/@ ,QXQL¿FDWLRQEDVHGJUDPPDUQRWKLQJWKHRUHWLFDOO\SUHYHQWVWKHXQL¿FDWLRQRI>3/@ FDUULHGE\WKHSOXUDOIRUPDQG>6*@FDUULHGE\WKHVLQJXODUIRUPEHFDXVHHDFKFDUULHV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 170 GLIIHUHQWDWWULEXWHVZLWKUHVSHFWLYHYDOXHV+RZHYHUJLYHQWKHFRQFHSWLRQWKDWHDFK180 IHDWXUHLVPDSSHGRQWRDSRUWLRQRIVHPDQWLFVSDFHRI¿JXUHWKHLQWHUSUHWDWLRQRIWKH FRPELQDWRU\SRVVLELOLW\RIPRUSKHPHVZLWK>6*@ZLWKWKDWRI>3/@LVFRQVWUDLQHG,QRQH interpretation, when the ‘narrow’ space is referred to, the two do not refer to a common 180VSDFH,QWKHRWKHULQWHUSUHWDWLRQZKHUHWKHZLGHVSDFHVDUHUHIHUUHGWR LHERWK IHDWXUHV>3/@DQG>6*@FDUULHGE\VLQJXODUDQGSOXUDOIRUPVUHIHUWRWKHFRPPRQVSDFH WKDW LQFOXGHV µWZR¶ WKHQ WKH '8$/ LQWHUSUHWDWLRQ LV DUULYHG DW 7KH FRQGLWLRQ RI WKH XQL¿FDWLRQWRDUULYHDW'8$/LQWHUSUHWDWLRQLQ+RSLFDQEHVKRZQLQ¿JXUH17 &RQGLWLRQ *LYHQWKH180VSDFHRI+RSLLQÀJXUHWKHVSDFHRI [+SG] is mutually exclusive with that of [+PL] (*i.e. [+SG]U[+PL]) Hence: PL U SG DU [PL +] [SG +] = [SG -] [SG -] [PL -] [PL -] FIGURE 12 ,QVKRUWXQGHUVSHFL¿HG180IHDWXUHVFDUULHGE\6*3/PRUSKHPHVDOORZIRUZLGH 180 VSDFH UHIHUHQWV7KHVH VHUYH DV UHVRXUFHV IRU FRPELQDWRU\ SXUSRVHV WR UHIHU WR D VSHFL¿F180UHIHUHQWVXFKDV'8$/7KXVODQJXDJHVVXFKDV+RSLGRQRWQHHGWRDKDYH DGHGLFDWHGPRUSKHPHIRU'8$/DV6*3/IRUPVDUHXVDEOHIRUWKLV 4.3. DUAL IN LARIKE. Larike, an Austronesian language of Maluku (Laidig & Laidig LVUHSRUWHGWRKDYHDIRXUZD\180V\VWHP 6*'875,$/DQG3/ 7KHIXOOVHWV RIWKHIRXUZD\180GLVWLQFWLRQDUHRQO\HQFRXQWHUHGLQ¿UVWVHFRQGDQGWKLUGSHUVRQ KXPDQSURQRPLQDOIRUPV7KHLQÀHFWLRQIRUWKHWKLUGSHUVRQQRQKXPDQLVGHIHFWLYH7KH VXEMHFWDQGREMHFWVHWVDUHVKRZQLQWDEOHDQGWDEOHUHVSHFWLYHO\ 1RQSURQRPLQDO IRUPVDUHQRWLQÀHFWHGIRU180 SG DUAL TRIAL PLURAL 1 EX au- aruai- aridu- ami- INC ituai- itidu- ite- ai- iruai- iridu- imi- 3 HUM mei matuai- matidu- mati NHUM i- - - iri- TABLE 2.6XEMHFWSUH¿[HV 17 :KLOHZHKDYHXQL¿FDWLRQRIIHDWXUHVLQ¿JXUHWKHLQWHUSUHWDWLRQRI^>6*@>3/@`DVGXDOLV DFWXDOO\DVVRFLDWHGZLWKWKHQRWLRQRILQWHUVHFWLRQLQWKHVHPDQWLFVSDFHRIQXPEHU MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 171 SG DUAL TRIAL PLURAL 1 EX -a/u -arua -aridu -ami INC -itua -itidu -ite -ne -irua -iridu -imi 3 HUM -ma -matua -matidu -mati NHUM -a - - -ri TABLE 3.2EMHFWVXI¿[HV 7KH180VSDFHLQ/DULNHFDQEHUHSUHVHQWHGLQ)LJXUHIRUWKHIROORZLQJUHDVRQV )LUVW'8$/DQG75,$/DUH180IHDWXUHVZLWKRXW>@YDOXHLHSULYDWLYH8QOLNHLQ Nen, there is no evidence in Larike that the system makes use of the opposition of DUAL YVQRQ'8$/'XDODQGWULDOLQ/DULNHZHUH KLVWRULFDOO\ GHULYHGIURPQXPHUDOµWZR¶DQG µWKUHH¶UHVSHFWLYHO\ /DLGLJ /DLGLJ 7KH\DUHWUXHGXDODQGWULDOIRUPVLQWKHVHQVH WKDWWKH\UHIHUWRH[DFWTXDQWLWLHVRIµWZR¶DQGµWKUHH¶DQGQHYHUXVHGWRUHIHUWRYDJXH QRWLRQRIVHYHUDODVLVDSDXFDORUOLPLWHGSOXUDOLQRWKHUODQJXDJHVVXFKDV [SG] [PL] +S G -SG DU TRIAL -PL +PL ‘one’ two ‘three’ ‘four’ … SINGULAR DUAL TRIAL PLURAL FIGURE 13. The Semantic space of NUM in Larike 6HFRQGWKHSOXUDOIRUPVLQ/DULNHPD\EHDOVRXVHGZKHQUHIHUULQJWRTXDQWLWLHVRI WZRRUWKUHH /DLGLJ /DLGLJ 7KLVLVWKHHYLGHQFHWKDWWKHSOXUDOIRUPVFRQWDLQ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori >6*3/@IHDWXUHVZKLFKDOORZWKHZLGHLQWHUSUHWDWLRQRI3/7KDWLVLWV>6*@IHDWXUH allows the use of the PL form to cover the wide semantic space that includes ‘two’, ‘three’, DQGµIRXURUPRUH¶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µSOXUDO¶EHFDXVHLWFRYHUVWKHVSDFHVRIµWZR¶DQGµWKUHH¶,WLVKRZHYHUOLPLWHG as it is contrasted with the space of PL (+PL) which, as indicated in the diagram, has no XSSHUOLPLWSRLQW,QVKRUWEHFDXVHRIWKHDEVHQFHRI'8$/75,$/IRUPIRUWKHWKLUG person non-human, the SG form is naturally extended to refer to this limited plural space EHFDXVHWKH6*IRUPFDUULHV>3/@IHDWXUHLQLW 5. CONCLUSIONS. This paper has provided an explicit analysis of how NUM morphology LV SURMHFWHG WR V\QWD[ LQ 0DURUL ,W LV SURSRVHG WKDW 180 IHDWXUHV EH HVWDEOLVKHG RQ ODQJXDJHVSHFL¿FVWUXFWXUHGVHPDQWLFVSDFHRI180DQGWKDWWKHUHDUHWZREDVLF180 IHDWXUHV QDPHO\ 6* DQG 3/ HDFK ZLWK ELQDU\ YDOXHV (DFK LV SRVVLEO\ LQGHSHQGHQWO\ PDSSHGRQWRWKHVHPDQWLFVSDFH,WLVDUJXHGWKDW'8$/FDQEHXQPDUNHGDQDO\VHGDV >6*3/@+RZHYHU'8$/FDQEHDOVRPDUNHGH[SUHVVHGE\DGHGLFDWHG'8$/IRUP 7KLVLVHQFRXQWHUHGLQ1HQDQG/DULNH It has also been demonstrated that the proposed analysis treats NUM morphemes as FDUU\LQJDEXQGOHRIIHDWXUHVZLWKHDFKRSHUDWLQJLQGHSHQGHQWO\7KLVSURYLGHVDQDWXUDO explanation for what appears to be unusual NUM agreement or expressions as found in 1HQDQG+RSL7KHSKHQRPHQDRIIDFXOWDWLYH3/85$/DVIRXQGLQ/DULNHFDQDOVREH DFFRXQWHGIRULQWKHSURSRVHGDQDO\VLV REFERENCES $UND,:D\DQBalinese morphosyntax: a lexical-functional approach. Canberra: 3DFL¿F/LQJXLVWLFV $UND,:D\DQ6SHHFKOHYHOVVRFLDOSUHGLFDWHVDQGSUDJPDWLFVWUXFWXUHLQ%DOLQHVH $OH[LFDODSSURDFKPragmatics %RHODDUV-DQ+0& The linguistic position of south-western New Guinea/HLGHQ %ULOO %UHVQDQ-RDQLexical functional syntax/RQGRQ %ODFNZHOO %UHVQDQ -RDQ 6 0FKRPER 7RSLF SURQRXQ DQG DJUHHPHQW LQ &KLFKHZD Language &KRPVN\1RDP Lectures on Government and Binding Theory'RUGUHFKW )RULV &RUEHWW*UHYLOOH*Number&DPEULGJH &DPEULGJH8QLYHUVLW\3UHVV 'DOU\PSOH0DU\Lexical Functional Grammar. 6\QWD[DQG6HPDQWLFV 6DQ 'LHJR $FDGHPLF3UHVV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Morphology and agreement in Marori 173 'UDEEH3HWHU Talen en Dialecten van Zuid-West Nieuw-Guinea 0LFUR%LEOLRWKHFD $QWKURSRV )UHLERXUJ $QWKURSRV,QVWLWXW 'UDEEH3HWHU Spraakkunst van het Marind(Studia Instituti Anthropos 11 0|GOLQJ 0LVVLHKXLV6W*DEULsO (YDQV 1LFKRODV &RPSOHPHQWDULW\ XQL¿FDWLRQ DQG QRQPRQRWRQLFLW\ %RXQG SURQRPLQDOV IUHH 13V DQG DUJXPHQW VWDWXV LQ D GRXEOHPDUNLQJ ODQJXDJH 1HQ 8QSXEOLVKHGPV (YDQV1LFKRODVDying words: Endangered languages and what they have to tell us 2[IRUG :LOH\%ODFNZHOO )DON I Wayan Arka ǁĂLJĂŶ͘ĂƌŬĂΛĂŶƵ͘ĞĚƵ͘ĂƵ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY /DQJXDJH'RFXPHQWDWLRQ &RQVHUYDWLRQ6SHFLDO3XEOLFDWLRQ1R 'HFHPEHU Melanesian Languages on the Edge of Asia: Challenges for the 21st Century, HGE\1LFKRODV(YDQVDQG0DULDQ.ODPHUSS± KWWS QÀUFKDZDLLHGXOGFVS 7 KWWS KGOKDQGOHQHW ‘Realis’ and ‘irrealis’ in Wogeo: A valid category? 0DWV([WHU Heinrich-Heine-Universität Düsseldorf Finite verb forms in Wogeo, an Austronesian language of New Guinea, DUHREOLJDWRULO\PDUNHGZLWKDSRUWPDQWHDXSUH¿[GHQRWLQJSHUVRQDQG number of the subject on the one hand, and a grammatical category WKDWLVFRQYHQWLRQDOO\JORVVHGLQWKHOLWHUDWXUHDVUHDOLV±LUUHDOLVRQWKH RWKHU ,Q VLPLODU ODQJXDJHV WKH ODWWHU FDWHJRU\ LV XVXDOO\ GHVFULEHG DV modal, with a certain range of meanings which is, in many cases, only YDJXHO\ GH¿QHG $ PRUH LQGHSWK LQYHVWLJDWLRQ RI WKH YHUEDO V\VWHP of Wogeo and the functional distribution of the respective categories VKRZVKRZHYHUWKDWWKHODQJXDJHLVTXLWHGLIIHUHQWIURPDSRVWXODWHG SURWRW\SLFDOUHDOLV±LUUHDOLVODQJXDJH&HQWUDODWWULEXWHVRIWKHVXSSRVHG UHDOLV±LUUHDOLVVHPDQWLFVDUHQRWUHDOL]HGE\WKHREOLJDWRU\SUH¿[HVEXW E\RWKHUPRUSKRV\QWDFWLFPHDQVZKLOHWKHSUH¿[HVDUHUHVWULFWHGWRRQO\ DVPDOOSDUWRIWKHDVVXPHGUHDOLV±LUUHDOLVGRPDLQ 1. INTRODUCTION.1 In the linguistic literature, ‘realis’ and ‘irrealis’ have most often been discussed under the more general heading of mood and modality7KHVHLQWXUQDUHWHUPV which are almost universally used in linguistics (with or without difference in meaning), \HW D VDWLVIDFWRU\ GH¿QLWLRQ LV ODUJHO\ D PDWWHU RI RQJRLQJ GHEDWH 7KH SUREOHP ZLWK PDQ\H[LVWLQJGH¿QLWLRQVLVWKDWWKH\DUHHLWKHUWRRYDJXHDQGOHDYHWRRPXFKWRLPSOLFLW assumptions, as is often the case in purely descriptive contexts; or, if they attempt to be H[SOLFLWWKH\IUHTXHQWO\UHVRUWWRGLVMXQFWLYHFKDUDFWHUL]DWLRQVLQYROYLQJVWDWHPHQWVOLNH 1 I wish to thank the speakers of Wogeo, above all my main consultants, Conny Tarere, Michael Ganem and the late Albert Kulbobo, for welcoming me and sharing their knowledge of the ODQJXDJHZLWKPH,DOVRWKDQN$VWULG$QGHUVRQIRULQWURGXFLQJPHWRWKH:RJHRZRUOGDQG the Research Council of Norway as well as the Institute for Comparative Research in Human &XOWXUH2VOR1RUZD\IRUIXQGLQJWKH¿HOGZRUNWKDWWKLVSDSHULVEDVHGRQ7KDQNVDUHDOVR GXH WR WKH SDUWLFLSDQWV DW WKH:RUNVKRSRQ WKH /DQJXDJHV RI 3DSXD )HEUXDU\ ± 0DQRNZDUL,QGRQHVLDIRUKHOSIXOGLVFXVVLRQVDQGIHHGEDFN)LQDOO\,DPYHU\JUDWHIXOWR-RKDQ van der Auwera, Marian Klamer, Daniel Kölligan, and an anonymous reviewer for their valuable FRPPHQWVRQDQHDUOLHUYHUVLRQRIWKLVSDSHU cc Creative Commons Attribution Non-Commercial No Derivatives Licence ‘Realis’ and ‘irrealis’ in Wogeo 175 “A is X or<or=´$IXOOGLVFXVVLRQRIWKHWHUPVPRRGDQGRUPRGDOLW\LVZHOOEH\RQG WKHVFRSHRIWKLVSDSHU KRZHYHUDZRUNLQJGH¿QLWLRQLVQHHGHGWRLQYHVWLJDWHWKHLVVXHRI UHDOLV±LUUHDOLVLQDPHDQLQJIXOZD\ ,Q WKH IROORZLQJ VHFWLRQ WKHUHIRUH VXFK ZRUNLQJ GH¿QLWLRQV DUH GLVFXVVHG DQG WKH SRVLWLRQRIUHDOLV±LUUHDOLVZLWKUHVSHFWWRWKHFDWHJRU\RIPRRG RUPRGDOLW\ LVGLVFXVVHG 7KHQ D EULHI UHYLHZ RI SURSRVHG µUHDOLV±LUUHDOLV¶ FDWHJRULHV DFURVV ODQJXDJHV LV JLYHQ DQGWKHFRPSDUDELOLW\RIWKRVHFDWHJRULHVLVGLVFXVVHG)LQDOO\DQRYHUYLHZRIWKHYHUEDO morphosyntax of Wogeo is given and the usefulness of the realis–irrealis terminology is UHDVVHVVHGLQWKHOLJKWRIWKHHYLGHQFHWKDWFDQEHJDLQHGIURPWKH:RJHRGDWD 2. TERMINOLOGICAL ISSUES. $V D ¿UVW VWHS DV REVHUYHG E\ &ULVWRIDUR LW LV important to distinguish between the semantic (or conceptual) domain we are dealing with, on the one hand, and any grammatical categories that realizeWKDWGRPDLQRQWKHRWKHU)RU the former, the term modality is often used, whereas the term mood is commonly reserved IRUWKHODWWHU7KHGLVWLQFWLRQEHWZHHQVHPDQWLFGRPDLQDQGJUDPPDWLFDOFDWHJRU\ZLOOEH WDNHQDVIXQGDPHQWDOLQWKHGLVFXVVLRQWKDWIROORZV 3DOPHU GH¿QHVPRGDOLW\DVEHLQJ³FRQFHUQHGZLWKWKHVWDWXVRIWKHSURSRVLWLRQ WKDW GHVFULEHV WKH HYHQW´7KLV LV DQ H[DPSOH RI ZKDW KDV EHHQ UHIHUUHG WR DERYH DV D YDJXHGH¿QLWLRQVLQFHLWLVOHIWLPSOLFLWZKDWH[DFWO\LVPHDQWE\concerned with and, especially, the status of the proposition ± status in relation to what? Somewhat more H[SOLFLWLVWKHGH¿QLWLRQJLYHQE\3RUWQHU ZKRVXJJHVWVWKDW³PRGDOLW\LVWKH linguistic phenomenon whereby grammar allows one to say things about, or on the basis RIVLWXDWLRQVZKLFKQHHGQRWEHUHDO´$V3RUWQHUKLPVHOISRLQWVRXWLWLVQRWLPPHGLDWHO\ REYLRXVKRZWRGH¿QHWKHWHUPreal \HWWKHGH¿QLWLRQLVPRUHXVHIXOLQSUDFWLFHWKDQ 3DOPHU¶V Further differences can be found in the ways in which different researchers subdivide WKHPRGDOVHPDQWLFGRPDLQ*LYyQ HJYLHZVWKHGLYLVLRQEHWZHHQpresuppositions and assertions as primary; assertions are then divided into realis and irrealis; and realis DVVHUWLRQVDUHFODVVL¿HGDVpositive or negative3DOPHU RQWKHRWKHUKDQGWDNHVD more traditional position, distinguishing propositional modality (subdivided into epistemic YVevidential) from event modality (subdivided into deonticYVdynamic )LQDOO\%\EHH (1998) distinguishes four subdomains: agent-oriented, speaker-oriented, epistemic3 and subordinatingPRGDOLW\7KHPRVWVWULNLQJZD\LQZKLFK%\EHH¶VDSSURDFKGLIIHUVIURPWKH former two, though, is that she argues that the supposed subdomains of modality are really four independent semantic domains, the connection between which is mainly diachronic, QRWV\QFKURQLF7KHVXEGLYLVLRQVZLWKLQWKHGRPDLQRIPRGDOLW\WKDW*LYyQ3DOPHUDQG %\EHHSURSRVHDUHVXPPDUL]HGLQWDEOH ,Q3DOPHU¶VWHUPLQRORJ\G\QDPLFPRGDOLW\VXEVXPHVDELOLW\DQGZLOOLQJQHVV 3 Agent-oriented modality (in Bybee’s terms) includes, but need not be restricted to: obligation, permission, volition, ability; speaker-oriented: imperative, permissive; epistemic: uncertainty, SRVVLELOLW\SUREDELOLW\ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo 176 *LYyQ Palmer Bybee Presupposition Propositional modality: Agent-oriented modality Realis assertion: Epistemic Speaker-oriented modality Positive Evidential Epistemic modality Negative Event modality: Subordinating modality Irrealis assertion Deontic Dynamic TABLE 1.6XEGLYLVLRQVRIPRGDOLW\DFFRUGLQJWR*LYyQ 3DOPHU DQG Bybee (1998) $ GLIIHUHQW DSSURDFK LV WDNHQ E\ YDQ GHU$XZHUD 3OXQJLDQ 7KH\ FKRRVH WR restrict the use of the term modal to those categories whose functions can be described by reference to the concepts of possibility and necessity, explicitly excluding categories OLNHYROLWLRQHYLGHQWLDOLW\HWFIURPWKHUHDOPRIPRGDOLW\7KHFODVVL¿FDWLRQRIYDQGHU $XZHUD 3OXQJLDQLVVXPPDUL]HGLQWDEOH Possibility Non-epistemic Participant-external Epistemic Participant-internal Non-deontic Deontic Non-deontic Deontic Participant-internal Participant-external Epistemic Non-epistemic Necessity TABLE 2. Subdivisions of modality according to van der Auwera & Plungian (1998) 2EYLRXVO\*LYyQ3DOPHU%\EHHDQGYDQGHU$XZHUD 3OXQJLDQVXEGLYLGHWKHVHPDQWLF GRPDLQRIPRGDOLW\RQWKHEDVLVRIGLIIHUHQWFULWHULD7KHVHVKRXOGWKHUHIRUHEHVHHQDV complementary approaches which can very well be applied independently to arrive at cross- FXWWLQJFODVVL¿FDWLRQV7KHTXHVWLRQWKDWSRVHVLWVHOILVWKHQZKLFKRIWKHVWUDWHJLHV LIDQ\ LV RUDUH PRVWIUXLWIXOLQVROYLQJWKHUHDOLV±LUUHDOLVLVVXHZHDUHFXUUHQWO\FRQFHUQHGZLWK )RUUHDVRQVZKLFKZLOOEHFRPHFOHDULQVHFWLRQVDQGEHORZ,ZLOODGRSWWKHUHVWULFWLYH approach of van der Auwera & Plungian (1998) as a working hypothesis for the domain RIPRGDOLW\ $VZLOOEHFRPHFOHDULQVHFWLRQWKHVHPDQWLFGRPDLQWKDWDSXWDWLYHUHDOLV±LUUHDOLV domain has been claimed to subsume overlaps to a large degree with what different authors assume to be within the realm of modality, plus other areas that would not traditionally be MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo 177 YLHZHGDVPRGDOVXFKDVHJIXWXUH WHQVH RUKDELWXDO DVSHFW ,WLVWKHUHIRUHLQVWUXFWLYH DVDVWDUWLQJSRLQWWRORRNDWGLIIHUHQWSURSRVDOVDVWRZKDWUHDOLV±LUUHDOLVUHDOO\LV0DXUL 6DQVz SURYLGHDYHU\JRRGRYHUYLHZRIWKHFXUUHQWGHEDWH7KHPDLQSRVLWLRQVWKDW are relevant to the present discussion can, according to them, be summarized as follows: ,UUHDOLVLVDNLQGRIµPHJDPRGDOLW\¶VXEVXPLQJDQXPEHURIPRGDOVXEGRPDLQV 5HDOLV±LUUHDOLVLVWKHVDPHDVPRGDOLW\ 5HDOLVDQGLUUHDOLVDUHWKHPVHOYHVPRGDOFDWHJRULHV Realis and irrealis are the values of a category ‘reality status’ which is independent of PRGDOLW\ ,IWKHODVWSRVLWLRQDGYRFDWHGHJE\(OOLRWW LVFRUUHFWLWVKRXOGEHSRVVLEOHWR LGHQWLI\WKHVHPDQWLFFRQWHQWWKDWLVH[SUHVVHGE\VXFKDFDWHJRU\3LHWUDQGUHD argues in a top-down approach in favor of a category of ‘reality status’ as distinct from PRGDOLW\)RUKHULUUHDOLVVWDWHVRIDIIDLUVDUHnon-actualized, meaning they are “presented DVQRWJURXQGHGLQSHUFHLYDEOHUHDOLW\´ The task of identifying the meaning expressed by ‘reality status’ is taken up in a YHU\GLIIHUHQWZD\E\GH+DDQ ,QKLVERWWRPXSW\SRORJLFDOVWXG\KHVHWVRXWWR investigate the claim that there is a prototypical semantic core that can be assigned to those FDVHVWKDWKDYHEHHQDQDO\]HGDVLQVWDQFHVRIUHDOLV±LUUHDOLV+LVFRQFOXVLRQKRZHYHULV negative: Many alternative core meanings can be found, none of which can convincingly EHDUJXHGWRKDYHSULRULW\RYHUWKHRWKHUV7KXVLWLVFRPSOHWHO\RSHQZKDWVKRXOGEHWKH FRUHDQGZKDWVKRXOGEHWKHSHULSKHU\RIWKHFDWHJRU\µUHDOLW\VWDWXV¶7KHUHIRUHGH+DDQ DUJXHVLWFDQQRWDWSUHVHQWEHVKRZQWREHDW\SRORJLFDOO\YDOLGFDWHJRU\ 3. PREVIOUS TYPOLOGICAL STUDIES. Having been sensitized to the complexity of the issues involving modality and reality status as well as the relationship between the two, we are now in a position to give a concise overview of previous typological studies relating WRWKHLVVXHRIWKHHOXVLYHµUHDOLV±LUUHDOLV¶FDWHJRU\LQYDULRXVODQJXDJHVODQJXDJHIDPLOLHV DQGJHRJUDSKLFDODUHDV:HZLOOIRFXVRQWKUHHVWXGLHV %XJHQKDJHQ (OOLRWW DQGYDQ*LMQ *LSSHU Bugenhagen’s (1993) paper is particularly interesting in the present context because it investigates the semantics of what is called ‘irrealis’ in seven Austronesian languages of 1HZ*XLQHD7KHODQJXDJHVLQKLVVDPSOHDUHWKHUHIRUHERWKJHQHWLFDOO\DQGJHRJUDSKLFDOO\ FRPSDUDEOHWR:RJHR2QWKHEDVLVRIKLVGDWDEDVHKHLGHQWL¿HVZKDWFDQEHGHVFULEHGDV a prototypical semantic core for the realis and irrealis categories (for the given language family and area): prototypical realis semantics is associated with positive polarity, non- future tense, perfective aspect and declarative speech acts,5 while irrealis semantics is associated with future tense, hypothetical conditional clauses, counterfactual conditional In Bugenhagen (1993), as almost everywhere else (including this paper), irrealis is taken to be the category in need of explanation, with realisOHIWDVWKHXQPDUNHGPHPEHURIWKHGLFKRWRP\7KH relationship between the two terms is thus fundamentally asymmetrical 5 $VOLJKWO\GLIIHUHQWFRUHPHDQLQJIRUUHDOLVLVDVVXPHGE\YDQGHU$XZHUD 'HYRV QDPHO\D³PDLQFODXVHDI¿UPDWLYHGHFODUDWLYHUHIHUULQJWRWKHSUHVHQWWLPHVSKHUH´ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo 178 FODXVHV FRPSOHPHQWV RI µZDQW¶ DQG QHJDWLYH SXUSRVH FODXVHV µOHVW¶ %XJHQKDJHQ¶V SURWRW\SLFDOXVHVRIUHDOLVDQGLUUHDOLVDUHVXPPDUL]HGLQWDEOH Realis Irrealis Positive polarity Future tense Non-future tense Hypothetical conditional clauses Perfective aspect Counterfactual conditional clauses Declarative speech acts Complements of ‘want’ Negative purpose clauses (‘lest’) TABLE 3. Prototypical uses of realis and irrealis in Austronesian languages of New Guinea according to Bugenhagen (1993) The characterization of the (supposed) irrealis semantic domain by means of a number of notions reminds us of Bybee’s (1998) view of the domain as a set of notions linked by SDUWLDOVLPLODULWLHV IDPLO\UHVHPEODQFHV DVGLVFXVVHGDERYH7KLVYLHZLVDXJPHQWHGE\ Bugenhagen (again, for his data set only) by explicitly postulating a semantic focal area within the broader domain where the languages are largely in agreement, and more peripheral DUHDVZKHUHLQGLYLGXDOODQJXDJHVVKRZVSHFL¿FSDWWHUQV /RRNLQJDW%XJHQKDJHQ¶VOLVW one would have to state more precisely that it represents several interconnected focal areas UDWKHUWKDQRQHDVSURSRVHGE\GH+DDQ %XJHQKDJHQH[SOLFLWO\SRLQWVRXWKRZHYHU that despite the relatedness and close proximity of the languages, “no two of them exhibit DFRPSOHWHO\LGHQWLFDOUDQJHRIXVHVIRUWKHLULUUHDOLVIRUPV´ :HVKDOOVHHEHORZ ZKHWKHU:RJHR¿WV%XJHQKDJHQ¶VJHQHUDOL]DWLRQV (OOLRWWWRRLQYHVWLJDWHVDQXPEHURIODQJXDJHVZLWKDQDOOHJHGUHDOLV±LUUHDOLVGLVWLQFWLRQ ZLWKWKHDLPWR³DUULYHLQGXFWLYHO\DWDW\SRORJLFDOGHVFULSWLRQRIWKLVFDWHJRU\´ The number of languages included in her database (16) is slightly larger than the number of languages investigated by Bugenhagen, and she uses a different sampling strategy, with ODQJXDJHVGUDZQIURPZLGHO\GLIIHUHQWIDPLOLHVDQGJHRJUDSKLFDODUHDV Elliott arrives at a result which is completely different from Bugenhagen’s (1993): She argues for a grammatical category reality status (the term originating in Whorf 1938) with the values realis and irrealis, and she claims that it is in fact possible to identify D FRPPRQ VHPDQWLF FRPSRQHQW LQ DOO XVHV RI WKH FDWHJRU\ )RU (OOLRWW WKH FRPPRQ semantic core of irrealis is that “irrealis events or states are perceived as being located LQDQDOWHUQDWLYHK\SRWKHWLFDORULPDJLQHGZRUOGEXWQRWWKHUHDOZRUOG´ 7KH semantic area thus covered by ‘irrealis’ is, however, extremely broad and includes potential HYHQWVFRQGLWLRQDOVHYHQWVTXDOL¿HGE\PRGDOLW\DQGFRPPDQGV DGGLWLRQDOO\QHJDWLRQV KDELWXDOVDQGLQWHUURJDWLYHVPD\DOVREHVXEVXPHGE\µLUUHDOLV¶ I see two problems in Elliott’s approach: First, the distinction (if any) between modality RQWKHRQHKDQGDQGKHUµUHDOLW\VWDWXV¶RQWKHRWKHULVQRWGH¿QHGV\VWHPDWLFDOO\ DQGVHFRQG WKHODUJHFURVVOLQJXLVWLFGLIIHUHQFHVLQWKHVHPDQWLFVRIµLUUHDOLV¶DUHOHIWXQH[SODLQHG 9DQ*LMQ *LSSHU XVHDWKLUGDSSURDFKSURYLGLQJDQLQGHSWKDQDO\VLVRIWKH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo 179 UHDOLV±LUUHDOLV V\VWHP RI D VLQJOH ODQJXDJH FRXQWHUIDFWXDO SRVVLEOH>í6&@ SRVVLEOH>6&@ IDFWXDO>í76@ IDFWXDO>76@ 9DQ*LMQ *LSSHUWKXVLQWURGXFHWKHLGHDRIDQHPSLULFDOO\EDVHGLPSOLFDWLRQDOKLHUDUFK\ DQGVXEKLHUDUFKLHV LQWRWKHGLVFXVVLRQ8QIRUWXQDWHO\KRZHYHUDVZHZLOOVHHEHORZ :RJHRFRQVWLWXWHVDFOHDUFRXQWHUH[DPSOHWRWKHJHQHUDOL]DWLRQH[SUHVVHGLQWKDWKLHUDUFK\ It seems likely that the data base that van Gijn & Gipper base their proposal on is much too VPDOOWRDGHTXDWHO\FDSWXUHDSKHQRPHQRQDVFRPSOH[DVWKHRQHXQGHUGLVFXVVLRQKHUH ,QP\YLHZZKDWYDQ*LMQ *LSSHU¶V DSSURDFKGRHVQRWDGHTXDWHO\H[SODLQ is the fundamental asymmetry between the alleged endpoints of the continuum (on the RQHKDQGµUHDOLV¶DVDFURVVOLQJXLVWLFDOO\IDLUO\ZHOOGH¿QHGFDWHJRU\FRYHULQJDUDWKHU narrow semantic area; and on the other hand, ‘irrealis’ as an extremely wide, vague, and fuzzy category with large cross-linguistic variation and no clearly discernible semantic FRUH 0RUHRYHUµIDFWXDOLW\¶LVXVXDOO\ LIQRWDOZD\V QRWWKHonly semantic component of the relevant grammatical categories; therefore, the supposed continuum may be better GHVFULEHGDVWKHUHVXOWRIFURVVFODVVL¿FDWLRQE\GLIIHUHQWLQGHSHQGHQWFDWHJRULHV 4. REALIS AND IRREALIS IN WOGEO. We will now turn to Wogeo and the formal and VHPDQWLFSURSHUWLHVRILWVµUHDOLV±LUUHDOLV¶PRUSKRORJLFDOFDWHJRU\:RJHRLVDQ$XVWURQHVLDQ ODQJXDJHVSRNHQE\DWPRVW DQGSUREDEO\OHVVWKDQ SHRSOHRQ9RNHRDQG.RLO,VODQGV RIIWKHQRUWKFRDVWRI1HZ*XLQHD3UHYLRXVDQWKURSRORJLFDOVWXGLHVRQ:RJHRLQFOXGH +RJELQ DQG$QGHUVRQ ([WHU LVDQDQDO\VLVRIWKHSKRQRORJ\ RIWKHODQJXDJHDQG$QGHUVRQ ([WHU LVDFROOHFWLRQRIWUDGLWLRQDO:RJHRWH[WV IRUWKHVSHHFKFRPPXQLW\DVZHOODVDPDLQO\DQWKURSRORJLFDODFDGHPLFDXGLHQFH([WHU VWLOOZRUNLQSURJUHVVLVLQWHQGHGWREHDFRPSUHKHQVLYHJUDPPDWLFDOGHVFULSWLRQ 7KHGDWDSUHVHQWHGKHUHDUHEDVHGRQP\RZQ¿HOGZRUNFRQGXFWHGLQ DQG )LQLWHYHUEVLQ:RJHR LHDOOYHUEIRUPVH[FHSWYHUEDOQRXQVJHUXQGV DUHPDUNHG ZLWKDQREOLJDWRU\SRUWPDQWHDXSUH¿[WKDWGHQRWHVWKHSHUVRQDQGQXPEHURIWKHVXEMHFW DVZHOODVUHDOLVRULUUHDOLV67KDWPHDQVWKDWQRQHRIWKHYDOXHVRIWKHGLFKRWRPRXVUHDOLV± LUUHDOLVFDWHJRU\LVIRUPDOO\XQPDUNHGLQ:RJHR,WDOVRPHDQVWKDWHYHU\VHQWHQFHZLWKD 6 ,PSHUDWLYHDQGSURKLELWLYHIRUPVDUHWKHRQO\H[FHSWLRQVWRWKLVJHQHUDOL]DWLRQ VHHEHORZ ±7R facilitate the discussion below, I will continue to use the terms realis and irrealis for the time EHLQJ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo 180 verbal predicate in Wogeo is marked either as realis or as irrealis; there are no unmarked VHQWHQFHV DQGE\WKHVDPHWRNHQQRXQPDUNHGHYHQWV 1RRWKHUSDUWRIWKHYHUELQ:RJHR DSDUWIURPWKHVWHP LVIRUPDOO\REOLJDWRU\7KXVLWLVIDLUWRVD\WKDWLQDOOUHVSHFWVWKH :RJHRYHUEDOV\VWHPLVRUJDQL]HGDURXQGWKHUHDOLV±LUUHDOLVFDWHJRU\ $VFDQEHVHHQIURPWKHWHPSODWHLQWDEOHVORWVíDQGí RSWLRQDO DQGVORWí REOLJDWRU\ DOOFRQWDLQLQIRUPDWLRQUHODWHGWRWHQVHDVSHFWDQGRUPRRG 6ORWíFRQWDLQV the counterfactual SUH¿[ VORW í FRQWDLQV WKH future, tentative, proximal imperfective and distal imperfective SUH¿[HV DQG VORW í FRQWDLQV WKH person/number/realis–irrealis SRUWPDQWHDXSUH¿[HV7 CNTF TAM PNM INCH CAUS IPFV Stem IPFV DIR APPL P N BEN P N (RDP) (RDP) í í í í í í 0 1 3 5 6 7 8 TABLE 4. Schematic morphological structure of the verb in Wogeo (obligatory slots are bold; slots that show higher internal coherence are shaded grey) 7DEOHJLYHVDQRYHUYLHZRYHUWKH310SUH¿[HV VORWí LQ:RJHR$VFDQEHVHHQWKHUH are four number categories (singular, plural, dual, paucal); tildes indicate synonymous IRUPV,QVSHFWLRQRIWKHSDUDGLJPLPPHGLDWHO\VKRZVWKDWLWLVTXLWHµPHVV\¶ 7KHUHDUH PDQ\ KRPRQ\PRXV IRUPV HJ PLRLS and 1PLIRR, 1DURLS and 1PAURLS) and partly KRPRQ\PRXVIRUPV HJSGRLSDQGSGIRR) without a clearly discernible pattern (although FRQVSLFXRXVO\WKHGLVWLQFWLRQEHWZHHQUHDOLVDQGLUUHDOLVLVQHXWUDOL]HGLQWKHSOXUDO 1RW surprisingly, corresponding realis and irrealis forms appear to be diachronically related; V\QFKURQLFDOO\KRZHYHUWKHWZRFDWHJRULHVFDQQRWEHUHGXFHGWRDVLPSOHUDQDO\VLV The table only shows the so-called plainUHDOLV±LUUHDOLVSDUDGLJP LHZLWKVORWVí DQG í UHPDLQLQJ HPSW\ ,I WKH FRPSOHWH 310 SDUDGLJPV RI DOO FRPSOH[ FDWHJRULHV are taken into account, an extremely complex picture emerges, which includes multiple complicating factors such as vowel assimilation; idiosyncratic fusions, vowel changes, and YRZHOGHOHWLRQV DQGHYHQPRUHFRPSOH[SDWWHUQVRIV\QRQ\P\DQGKRPRQ\P\)RUWKH point made in the present paper, therefore, this morphophonological and morphological FRPSOH[LW\ZLOOQRWEHGHDOWZLWKIXUWKHU 7 $EEUHYLDWLRQV XVHG LQ WKLV SDSHU $ µDVSHFW¶ APPL µDSSOLFDWLYH¶ BEN µEHQHIDFWLYH¶ CAUS µFDXVDWLYH¶ CNTF µFRXQWHUIDFWXDO¶ DIR µGLUHFWLRQDO¶ DIST µGLVWDO¶ DU µGXDO¶ FOC µIRFXV¶ FUT µIXWXUH¶ INCH µLQFKRDWLYH¶ IPFV µLPSHUIHFWLYH¶ IRR µLUUHDOLV¶ 0 µPRRG¶ 1 µQXPEHU¶ NEG µQHJDWLYH¶ NMLZ µQRPLQDOL]HU¶ 3 µSHUVRQ¶ PAU µSDXFDO¶ PL µSOXUDO¶ PROH µSURKLELWLYH¶ PROX µSUR[LPDO¶ RDP µUHGXSOLFDWLRQ¶ RECP µUHFLSURFDO¶ RLS µUHDOLV¶ SG µVLQJXODU¶ 7 µWHQVH¶ TENT µWHQWDWLYH¶ TOP µWRSLF¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo 181 Person/number (Plain) realis (Plain) irrealis 1SG o-lako go-lako SG go-lako ~ ko-lako go-lako 3SG e-lako de-lako 1PL ta-lako ta-lako PL ka-lako ka-lako 3PL da-lako da-lako 1DU to-lako ~ te-lako tog-lako ~ teg-lako DU kad-lako ~ kod-lako kad-lako ~ kod-lako 3DU do-lako ~ de-lako dog-lako ~ deg-lako 1PAU to-lako ~ te-lako tog-lako ~ teg-lako PAU koto-lako koto-lako 3PAU doto-lako doto-lako TABLE 5.7KH310SUH¿[HVLQ SODLQ UHDOLVDQGLUUHDOLVIRUPVRI:RJHRlako ‘go’ 6ORWVRWKHUWKDQííDQGíLQWDEOH QDPHO\VORWVííDQG FRQWDLQ7$0UHODWHG LQIRUPDWLRQWRREXWLWLVDUJXHGKHUHWKDWWKHDIRUHPHQWLRQHGVORWV LHVORWVííDQG í IRUPDXQLWRIWKHLURZQ)RUPDOO\WKH\DUHDXQLWEHFDXVHWKH\GLVSOD\PRUSKRORJLFDO idiosyncrasies between each other, such as fusion, vowel assimilation, and a number of RWKHU LUUHJXODULWLHV )XQFWLRQDOO\ WKH\ DUH D XQLW LQ VKRZLQJ D QXPEHU RI FRPELQDWRU\ LQWHUGHSHQGHQFHV REOLJDWRU\RSWLRQDODQGH[FOXGHGFRPELQDWLRQV 7KHVDPHGRHVQRW DSSO\WRWKHRWKHUVORWVZKHUHDJJOXWLQDWLRQDQGDODUJHGHJUHHRIFRPELQDELOLW\SUHGRPLQDWH 7KHUHVXOWLQJFRPELQDWLRQVRIVORWVííDQGíIRUPFRPSOH[7$0FDWHJRULHV8 which are given convenient summary labels (which I will call complex-category ODEHOV KHUH Those TAM combinations that are well-formed, along with their complex-category labels, DUHVKRZQLQWDEOH:KHUHPRUHWKDQRQHIRUPLVJLYHQIRUDQ\FRPSOH[FDWHJRU\WKRVH IRUPVDUHV\QRQ\PRXV9 8 µ&RPSOH[¶VKRXOGKHUHEHWDNHQWRPHDQIRUPDOO\QRWVHPDQWLFDOO\FRPSOH[ 9 Note that the so-called tentative forms express the meaning ‘to try it with X-ing’ (or ‘to X and see ZKDWKDSSHQV¶ QRWµWRWU\WR;¶±$VZLOOEHFRPHREYLRXVIURPDFORVHULQVSHFWLRQRIWDEOH the tentative and counterfactualPDUNHUVDUHKRPRQ\PRXV7ZROLQHVRIDUJXPHQWDUHSXWIRUZDUG here to justify their analysis as different morphemes: (1) Forms such as s-o-lako ‘I try it with going’ (tentative) and s-o-lako ‘I would have gone’ (counterfactual) show a contrast in meaning WKDW,FRQVLGHUIXQGDPHQWDOHQRXJKWRH[FOXGHDQDQDO\VLVZLWKDVLQJOHSRO\VHPRXVPRUSKHPH 7KH GHVFULSWLRQ RI WKH GLVWULEXWLRQDO IDFWV LV VLPSOL¿HG LI RQH DVVXPHV WKDW WKH WHQWDWLYH morpheme is in slot í DORQJZLWKWKHIXWXUHPRUSKHPH ZKLOHWKHFRXQWHUIDFWXDOPRUSKHPHLV LQVORWí FIWDEOH 7KHWHQWDWLYHDQGIXWXUHPDUNHUV EHLQJLQWKHVDPHVORW VKRZLGHQWLFDO PRUSKRSKRQRORJLFDO EHKDYLRU LQ HYHU\ GHWDLO WKH FRXQWHUIDFWXDO PDUNHU FDQ WKHQ EH SUH¿[HG WR WKH IXWXUH 310 FRPSOH[7KH FRQVSLFXRXV QRQFRPELQDELOLW\ RI WKH FRXQWHUIDFWXDO DQG WHQWDWLYHPDUNHUV FIWDEOH PLJKWKDYHSKRQRORJLFDOUHDVRQV KDSORORJ\OHDGLQJWRDFKDQJHRI *se-s-o-lako tabo > s-o-lako tabo), thus rendering the negative tentative form homonymous to the MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo Complex category Example Range of meanings (Plain) realis o-lako ‘I go’, ‘I went’ 1SG.RLS-go (Plain) irrealis go-lako ‘I must go’, ‘I want to go’, ‘I will 1SG.IRR-go go (now)’ Future m-o-lako ‘I will go’, ‘I can go’, ‘I may go’ FUT-1SG.RLS-go mo-go-lako FUT-1SG.IRR-go Tentative s-o-lako ‘I try it with going’ TENT-1SG.RLS-go so-go-lako TENT-1SG.IRR-go Counterfactual s-o-lako ‘I would have gone’ CNTF-1SG.RLS-go Proximal imperfective k-o-lako ‘I am going (nearby)’, ‘I was PROX.IPFV-1SG.RLS-go going (nearby)’ Distal imperfective o-lako ‘I am going (further away)’, DIST.IPFV;1SG.RLS-go ‘I was going (further away)’ TABLE 6. Complex TAM categories encoded on Wogeo lako ‘go’ As mentioned above, imperatives and prohibitives are exceptions to the pattern illustrated LQWDEOH7KHLPSHUDWLYHLVIRUPHGE\WKHEDUHVWHPZLWKRXWWKHRWKHUZLVHREOLJDWRU\310 SUH¿[HV WKHSURKLELWLYHLVIRUPHGE\DFRPELQDWLRQRIDYHUEDOQRXQDQGDIUHHJUDPPDWLFDO PRUSKHPH 7KH IRUPDWLRQ RI LPSHUDWLYHV DQG SURKLELWLYHV LV VXPPDUL]HG LQ WDEOH Complex category Example Range of meanings Imperative lako ‘Go!’ go Tentative imperative se-lako ‘Try it with going!’ TENT-go Prohibitive lako~lako dol ‘Don’t go!’ go~NMLZ PROH TABLE 7. Imperative and prohibitive forms of Wogeo lako ‘go’ QHJDWLYH SODLQ UHDOLVIRUP MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo 183 Negations in Wogeo (with the exception of prohibitives) are formed analytically by a combination of the counterfactual SUH¿[ D realis 310 SUH¿[ DQG WKH QHJDWRU tabo Table 8 presents the negative forms of the corresponding non-negative forms found in WDEOH106HYHUDOLQWHUHVWLQJIDFWVFDQEHQRWHG ¿UVWO\WKHREOLJDWRU\FRPELQDWLRQRIWKH FRXQWHUIDFWXDOZLWKWKHUHDOLVLVXQXVXDODQGVXUSULVLQJ6HFRQGO\LQWKHRQO\FRPSOH[ FDWHJRU\ZKHUHUHDOLVDQGLUUHDOLVSUH¿[HVFDQEHXVHGLQWHUFKDQJHDEO\LQWKHQRQQHJDWLYH form (namely the future), the presence of the counterfactual plus negator precludes the XVHRIWKHLUUHDOLVSUH¿[ WKHRWKHUQRQQHJDWLYHFDWHJRU\FRPSDWLEOHZLWKERWKUHDOLVand LUUHDOLVSUH¿[HVWKHWHQWDWLYHGRHVQRWKDYHDVSHFL¿FQHJDWLYHIRUPDVH[SODLQHGDERYH $QGWKLUGO\WKHUHLVRQHFDWHJRU\ WKHIXWXUH ZKHUHWKHFRXQWHUIDFWXDOSUH¿[LVRSWLRQDO Corresponding complex category Example Range of meanings (Plain) realis s-o-lako tabo ‘I do not go’, ‘I did not go’ CNTF-1SG.RLS-go NEG (Plain) irrealis >1RQHJDWLYHIRUPH[LVWV@ — Future se-m-o-lako tabo ‘I will not go’, ‘I cannot CNTF-FUT-1SG.RLS-go NEG go’, ‘I may not go’ m-o-lako tabo FUT-1SG.RLS-go NEG Not possible: *se-mo-go-lako tabo CNTF-FUT-1SG.IRR-go NEG *mo-go-lako tabo FUT-1SG.IRR-go NEG Tentative >1RQHJDWLYHIRUPH[LVWV@ — Counterfactual s-o-lako tabo ‘I would not have gone’ CNTF-1SG.RLS-go NEG Proximal se-k-o-lako tabo ‘I am not going (nearby)’, ‘I imperfective CNTF-PROX.IPFV-1SG.RLS-go NEG was not going (nearby)’ Distal imperfective >1RQHJDWLYHIRUPH[LVWV@ — TABLE 8. Negation of complex TAM categories encoded on Wogeo lako ‘go’ 10 7KUHHRIWKHFDWHJRULHVLQWDEOHKDYHQRVSHFL¿FQHJDWLYHIRUP (plain) irrealis, tentative, and distal imperfective7RH[SUHVVWKHPHDQLQJRIDQHJDWLYH SODLQ LUUHDOLVWKHprohibitive is used FIWDEOH ZKLOHWKHPHDQLQJVRIQHJDWLYHWHQWDWLYHDQGQHJDWLYHGLVWDOLPSHUIHFWLYHDUHERWK expressed by the negative (plain) realis MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo This brief exposition of the verbal morphology of Wogeo shows that in the majority of forms, the language employs a system where the realis and irrealis morphemes co-occur with other grammatical markers in complex categories, forming a joint system in Palmer’s ± WHUPLQRORJ\%XWZKLOHWKHUHDOLVDQGLUUHDOLVPRUSKHPHVcan also occur independently in the so-called (plain) realis and (plain) irrealis categories, the more SHULSKHUDO PDUNHUV VXFK DV IXWXUH HWF DUH REOLJDWRULO\ ERXQG WR WKH UHDOLV DQG LUUHDOLV PRUSKHPHVDQGFDQQRWRFFXUZLWKRXWWKHODWWHU 7RVXPXS WKHVRFDOOHGµUHDOLV¶SUH¿[HVDUHLQYROYHGLQWKHIRUPDWLRQRIWKHIROORZLQJ complex morphological categories in Wogeo: (plain) realis, counterfactual, proximal imperfective, distal imperfective, future and tentative LQWKHODWWHUWZRRSWLRQDOO\±WKH\ DUHDOWHUQDWLYHO\IRUPHGZLWKWKHµLUUHDOLV¶SUH¿[HVZLWKRXWFKDQJHLQPHDQLQJ 7KHVR FDOOHG µLUUHDOLV¶SUH¿[HV RQ WKH RWKHU KDQG DUH XVHG LQ WKH IRUPDWLRQ RI WKH IROORZLQJ categories: (plain) irrealis, future and tentative (again, in the latter two, their use is optional DQGDOWHUQDWHVZLWKWKHµUHDOLV¶SUH¿[HV 6HHQIURPWKHRSSRVLWHSHUVSHFWLYHWKHIROORZLQJ complex categories are formed exclusively ZLWK WKH µUHDOLV¶ SUH¿[HV (plain) realis, counterfactual, proximal imperfective and distal imperfective,WLVWKXVRQO\WKH(plain) irrealisWKDWLVIRUPHGH[FOXVLYHO\DQGREOLJDWRULO\ZLWKWKHµLUUHDOLV¶SUH¿[HV Having looked at the formal distribution of the ‘realis/irrealis’ morphemes in Wogeo, we will now turn to the range of meanings that is associated with each of the respective IRUPV11 First, the ‘realis’ morphemes are associated with the following meanings: General: a) Present, past (obligatorily) b) Counterfactual; proximal imperfective; distal imperfective (obligatorily, but always in combination with the respective markers) c) Future, ability, permission; tentative (optionally; always with the respective markers) 6SHFL¿FV\QWDFWLFFRQVWUXFWLRQV a) Negations (obligatorily) b) Protasis and apodosis of simple conditional clauses (obligatorily) c) Protasis of counterfactual conditional clauses (obligatorily; always with the counterfactual marker) d) Protasis and apodosis of hypothetical conditional clauses, apodosis of counterfactual conditional clauses (optionally; always with the future marker) 11 ‘Associated with’ is a deliberately vague term: while the attribution of certain meanings to individual morphemes is straightforward in the case of the (plain) realis and irrealis categories, it is not at all clear what the contribution of the respective morphemes is in the case of the complex FDWHJRULHV ,Q VRPH WKH µUHDOLVLUUHDOLV¶ SUH¿[HV PD\ FRQWULEXWH WR WKH UHVXOWLQJ JUDPPDWLFDO meaning, while in others, they may merely be compatible V\QFKURQLFDOO\ ZLWKWKRVHPHDQLQJV 7KLVTXHVWLRQLVQRWWULYLDODQGEH\RQGWKHVFRSHRIWKLVSDSHU MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo 185 The semantic associations of the ‘irrealis’ morphemes, on the other hand, are as follows: General: a) Obligation, volition, immediate future (obligatorily) b) Future, ability, permission; tentative (optionally; always with the respective markers) 6SHFL¿FV\QWDFWLFFRQVWUXFWLRQV a) Complements of ‘want’ (obligatorily) b) Protasis and apodosis of hypothetical conditional clauses, apodosis of counterfactual conditional clauses (optionally; always with the future marker) Some typical examples will serve as illustrations of the kinds of contexts in which the YDULRXVIRUPVRFFXU([DPSOH VKRZVWKHXVHRIWKH SODLQ UHDOLVIRUPLQWKLVFDVH expressing past tense7KLVLVDSURWRW\SLFDOH[DPSOHLQWKHVHQVHRI%XJHQKDJHQ LQ that it illustrates the use of a realis form to express positive polarity and non-future tense LQDGHFODUDWLYHVSHHFKDFW (1) (Plain) realis va, ilo-g e-la-muta~muta-k-iko I inside-1SG 3SG.RLS-INCHEHWLUHGRI~IMPV-APPL-2SG µ0H,EHFDPHWLUHGRI\RX¶ 7XUQLQJWRWKHµLUUHDOLV¶SUH¿[ZHFDQREVHUYHWKDWLQ RQHRIWKHFRUHPHDQLQJVRI (plain) irrealis in Wogeo, obligationLVH[SUHVVHG 3ODLQ LUUHDOLV iko go-la-boalé va na o-taval=te you 2SG.IRR-INCH-tell.3SG I FOC 1SG.RLS-die=TOP µ $QRWKHUW\SLFDOFRQVWUXFWLRQVSHFL¿FXVHRIWKH SODLQ LUUHDOLVLVVKRZQLQ QDPHO\DV DFRPSOHPHQWRIµZDQW¶/LNHWKHH[DPSOHJLYHQLQ WKLVXVHLVH[FOXVLYHWRWKHLUUHDOLV (3) (Plain) irrealis as complement of ‘want’ do-boré dog-va gon-iak, vaine boe ramata 3DU.RLS-want 3DU.IRR-RECP play-APPL.PL woman and man du-rú ma they-DU FOC µ7KH\ZDQWHGWRVOHHSZLWKHDFKRWKHUWKDWZRPDQDQGPDQ¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo 186 ,QWKHH[DPSOHVZHKDYHVHHQVRIDUWKHUHZDVDELXQLTXHUHODWLRQVKLSEHWZHHQWKHIRUPDO PDUNHUVDQGWKHPHDQLQJVWKH\H[SUHVVHG([DPSOHV DQG LQFRQWUDVWVKRZWKH LQGLVFULPLQDWHXVHRIWKHµUHDOLV¶DQGµLUUHDOLV¶SUH¿[HVLQFRPELQDWLRQZLWKWKHIXWXUHSUH¿[ )XWXUH IRUPHGIURPWKHUHDOLVEDVH vavá iko va m-u-kila-k-an-iko udemtaregá name.3SG you I FUT-1SG.RLS-call-APPL.3SG-BEN-2SG 8GHPWDUHJi µ,WVQDPHZKLFK,ZLOOFDOOLWIRU\RXLV8GHPWDGHJi¶ (5) Future (formed from the irrealis base) va kat va mo-go-jale-k oageva I canoe I FUT-1SG.IRR-JRGRZQ-APPL.3SG 9RNHR µ,ZLOOEULQJP\FDQRHGRZQWR9RNHR¶ The somewhat unexpected exclusive association of the counterfactual with the ‘realis’ SUH¿[HVLVLOOXVWUDWHGLQ ZKHUHLWLVXVHGLQWKHSURWDVLVRIDFRXQWHUIDFWXDOFRQGLWLRQDO (6) Counterfactual s-e-vá iko sa-k-lako, katé mo-la-moet CNTF-3SG.RLS-happen you CNTF-2SG.RLS-go thus FUT.2SG.RLS-INCH-disappear µ,I\RXKDGJRQH\RXZRXOGKDYHEHHQORVW¶ ([DPSOH ¿QDOO\LOOXVWUDWHVZKDWLVE\IDUWKHPRVWFRPPRQXVHRIWKHFRXQWHUIDFWXDO category in Wogeo, namely as the negated counterpart of the (plain) realis category (the VRFDOOHGµQHJDWHGUHDOLV¶ $VLQ DQG DERYHWKLVIRUPDQGIXQFWLRQLVH[FOXVLYHO\ DVVRFLDWHGZLWKWKHµUHDOLV¶SUH¿[ (7) Counterfactual as negated counterpart of (plain) realis natú e-ot taumdabí, e-ot, e-t-dom~doma, FKLOG3SG 3SG.RLS-come afternoon 3SG.RLS-come 3SG.RLS-INCH-look~IPFV[3PL] tabo tiná s-i-mia tabo but mother.3SG CNTF-3SG.RLS-stay NEG ‘Her son came in the afternoon, he came, looked around, but his mother ZDVQRWWKHUH¶ 6XPPLQJXSVHYHUDOREVHUYDWLRQVVXJJHVWWKHPVHOYHV:KDWVHHPVWREHHVSHFLDOO\ LQWHUHVWLQJ LV WKDW YDQ *LMQ *LSSHU¶V LPSOLFDWLRQDO KLHUDUFK\ LV QRW YDOLG IRU :RJHRVLQFHFRXQWHUIDFWXDOV±FUXFLDOWRWKHLUFODLP±DUHDOZD\VIRUPHGIURPWKHrealis EDVH QRW WKH LUUHDOLV EDVH 7KDW H[FOXVLYH DVVRFLDWLRQ RI WKH FRXQWHUIDFWXDO VHPDQWLFV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo 187 ZLWK WKH µUHDOLV¶ SUH¿[ LQ :RJHR LV DOVR RQH RI WKH WZR PDLQ GLVFUHSDQFLHV EHWZHHQ Bugenhagen’s (1993) generalizations and the Wogeo data, the other one being the fact that his list in fact does not include what can be said to constitute the semantic core of the (plain) irrealis morphological category in Wogeo: obligation and volition2WKHUWKDQWKRVH two (rather substantial) discrepancies, however, the functional range of the ‘realis’ and ‘irrealis’ morphemes in Wogeo can be described as largely consistent with Bugenhagen’s UHVXOWV 7REHVXUHVXFKDSXUHO\QHJDWLYHFKDUDFWHUL]DWLRQRIWKHFDWHJRU\LVQRWVDWLVIDFWRU\ As could be observed in the description of the semantic range covered by forms involving WKH µLUUHDOLV¶SUH¿[ LQ:RJHR HLWKHU DORQH RU LQ FRPELQDWLRQ ZLWK RWKHU SUH¿[HV WKDW range is largely coextensive with the domain of non-epistemic necessity in the sense of van der Auwera & Plungian (1998): irrealis in Wogeo can be said to express non-epistemic QHFHVVLW\:RJHRLVWKHUHIRUHDUJXDEO\DJRRGH[DPSOHRIDmood-prominent language in WKHVHQVHRI%KDW As we have observed above, Wogeo is not untypical in showing such ‘aberrations’ IURP D VXSSRVHG SURWRW\SLFDO UHDOLV±LUUHDOLV V\VWHP 2Q WKH FRQWUDU\ MXGJLQJ IURP WKH W\SRORJLFDOVWXGLHVDYDLODEOH:RJHRVHHPVWRUHSUHVHQWWKHUXOHUDWKHUWKDQWKHH[FHSWLRQ What can one do with such a situation? Two basic possibilities readily present themselves, QHLWKHURIZKLFKLQP\YLHZLVGHVLUDEOH2QHSRVVLELOLW\ZRXOGEHWRVD\WKDWLI:RJHR GRHVQRW¿WWKHH[SHFWHG RUSUHGLFWHG SDWWHUQWKHQLWIROORZVWKDWWKH:RJHRFDWHJRU\LV QRWDQLQVWDQFHRIWKDWSDWWHUQLQWKH¿UVWSODFH6XFKDQDSSURDFKPLJKWPDNHVHQVHLIRQH KDVJRRGDSULRULUHDVRQVWRDVVXPHWKDWWKHSUHGLFWHGFDWHJRU\LVLQGHHGYDOLGDQGXVHIXO The main problem that I see with that approach, however, is that a common semantic denominator can usually be ‘constructed’ for any subdomain of modality (in fact, that is what constitutes the semantic basis for the observed pattern of ‘family resemblances’ ZLWKLQWKHGRPDLQ 6RLI:RJHRLVQRWDJRRGH[DPSOHRIWKHVXSSRVHGFDWHJRU\±ZKLFK of the many other observed types of systems should be taken as a better example? 7KHVHFRQGSRVVLELOLW\ZRXOGEHWRPDNHWKHFODLPPRUHJHQHUDO+RZHYHUWKDWPD\ not be a very helpful suggestion when it comes to characterizing individual grammatical V\VWHPV3UHFLVHO\DV%\EHH SRLQWVRXW VXFKDFRQFHSWLVWRREURDGWREHRISUDFWLFDO descriptive use because it glosses over, and fails to explain, the very large differences that H[LVWEHWZHHQLQGLYLGXDOODQJXDJHVLQWKLVUHVSHFW The solution to the problem that I propose is that, as Bybee (1998) suggests, a ODQJXDJHVSHFL¿FQDUURZHUFDWHJRU\PLJKWEHPRUHKHOSIXOKHUHWKDQWKHZLGHFDWHJRU\ realis–irrealis; and what applies to Wogeo would likewise apply to other languages, WRR 2EVHUYHG GLIIHUHQFHV EHWZHHQ ODQJXDJHV DUH WKHQ EHVW XQGHUVWRRG DV GLDFKURQLF relations of grammaticalization within the semantic domain of modality, and between that GRPDLQDQGLWVQHLJKERULQJGRPDLQV7KHWHUPVrealis and irrealis may still be useful for comparative and historical purposes, where precisely such grammaticalization processes DQGVHPDQWLFVKLIWVQHHGWREHFDSWXUHG±NHHSLQJLQPLQGWKDWLQWKDWXVDJHWKH\DUHQR PRUHVSHFL¿F UDWKHUHYHQOHVVVSHFL¿F WKDQWKHWHUPVmodal and non-modalWKHPVHOYHV Note, however, that volitionZRXOGKDYHWREHH[SOLFLWO\LQFOXGHGHJDVDVSHFLDOFDVHRIYDQGHU Auwera & Plungian’s (1998) participant-internal necessity MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo 188 As for Wogeo, the language seems to be in the middle of a grammaticalization process, with the original ‘realis/irrealis’ markers on the way to being semantically bleached, while WKHSDUWO\IXVHGPRUSKV FRPELQDWLRQVRIVORWVííDQGíLQWDEOH DUHRQWKHZD\WR EHFRPLQJQHZSRUWPDQWHDXPRUSKV2QWKHRWKHUKDQGLQWKHPDMRULW\RIFDVHVWKHROG ‘realis/irrealis’ markers are still more or less formally and/or functionally transparent in the formation of parallel sets of what I have called complex categories FIWDEOH 13 5. CONCLUSION. In this paper, I have tried to assess the conceptual relevance of the terms UHDOLV±LUUHDOLVWKHLUUHODWLRQVKLSZLWKWKHGRPDLQRIPRGDOLW\ LWVHOIDFRQWURYHUVLDODUHD DQGWKHLUDSSURSULDWHQHVVDVGHVFULSWLYHJUDPPDWLFDOWHUPV ,WZDVVKRZQWKDWODQJXDJHVWKDWKDYHEHHQFODLPHGWRPDNHXVHRIDUHDOLV±LUUHDOLV category show extremely large variation in the semantic content of that category; indeed, QRWHYHQDSURWRW\SLFDOFRUHPHDQLQJFDQEHLGHQWL¿HGFURVVOLQJXLVWLFDOO\1HLWKHUDWRS down nor a bottom-up (typological) approach has, in my view, so far been able to provide FRQYLQFLQJHYLGHQFHWKDWWKHUHLVLQGHHGDQHHGWRSRVWXODWHVXFKDFDWHJRU\ It is of course conceivable that something like non-factuality is a valid concept in the minds of speakers, and that all the partial resemblances and diachronic developments WKDWFDQEHVHHQLQWKHGDWDDUHDFWXDOO\JURXQGHGLQVXFKDFRQFHSW+RZHYHU,VHHD GDQJHURIFLUFXODULW\LQWKHDQDO\VLVKHUH LWLVHTXDOO\SRVVLEOHWKDWSDUDOOHORYHUODSSLQJ and interacting diachronic developments of neighboring (but in principle independent) domains could create the illusionRIDQXQGHUO\LQJµVXSHUFDWHJRU\¶OLNHUHDOLW\VWDWXV'RHV a putative concept of reality status bring about the observable facts, or do the observable facts (which really arise through independent developments) look as though they instantiate some concept? Different typological studies were assessed that try to characterize realis–irrealis either DVDZHOOGH¿QHG \HWDEVWUDFW FDWHJRU\DVDFDWHJRU\ZLWKDSURWRW\SLFDOFRUHDQGIX]]\ ERXQGDULHVRUDVDQLPSOLFDWLRQDOKLHUDUFK\+RZHYHULWKDVEHHQDUJXHGLQWKLVSDSHU that all those attempts fail to solve the basic problem: namely, that the supposed category LV HLWKHU WRR YDJXH VR WKDW SUDFWLFDOO\DQ\ ODQJXDJH PD\ ¿W LQ LW WRR QDUURZ VR WKDW ODQJXDJHVSHFL¿FLGLRV\QFUDVLHVRXWZHLJKDQ\JHQHUDOL]DWLRQV RUWRRODQJXDJHVSHFL¿F (so that the category itself becomes arbitrary, and not comparable from a typological point RIYLHZ 'DWDIURP:RJHRZDVSUHVHQWHGWRLOOXVWUDWHWKLVSRLQW 7DNLQJ LQWR DFFRXQW WKH WKHRUHWLFDO GLI¿FXOWLHV ZLWK WKH FRQFHSW reality status, the ODFNRIXQHTXLYRFDOOLQJXLVWLFHYLGHQFHLQIDYRURILWDQGWKHIDFWVWKDWFDQEHOHDUQHG from Wogeo, my view is that it is probably wisest at this point to side with Bybee (1998) DQGGH+DDQ ,DJUHHZLWKWKHPLQVD\LQJWKDWXQWLOHYLGHQFHWRWKHFRQWUDU\LV presented, what we are dealing with is not one large, highly abstract domain but rather PDQ\VPDOOHULQGHSHQGHQWGRPDLQV7KHFRQQHFWLRQEHWZHHQWKRVHVPDOOHUGRPDLQVLV mainly diachronic via common paths of grammaticalization (van der Auwera & Plungian SynchronicallyWKHGRPDLQVDUHFKDUDFWHUL]HGPDLQO\E\SDUWLDOUHVHPEODQFHV As to the nature of the smaller domains that, as a whole, take the place of ‘reality 13 3UDFWLFDOO\ WKLV FUHDWHV WKH SUREOHP RI JORVVLQJ PRUSKV OLNH LQ ([DPSOHV ± WKDW DUH DUJXDEO\LQVRPHFRQWH[WVVHPDQWLFDOO\HPSW\EXWQRWLQRWKHUVOLNHWKHµUHDOLVLUUHDOLV¶SUH¿[HV LQWKHFRPSOH[PRUSKRORJLFDOFDWHJRULHVRI:RJHR MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo 189 VWDWXV¶LWLVSUREDEO\EHVWWRVWLFNWRIDLUO\ZHOOGH¿QHGGRPDLQVOLNHWKH UDWKHUUHGXFHG GRPDLQRIPRGDOLW\DVGH¿QHGE\YDQGHU$XZHUD 3OXQJLDQ DORQJVLGHGRPDLQV OLNHHYLGHQWLDOLW\LOORFXWLRQDU\IRUFHSRODULW\HWF,WLVWKHODQJXDJHVSHFL¿FLQWHUDFWLRQ between them that accounts for the type of ‘reality status’ system characteristic of any JLYHQODQJXDJH Finally, it was suggested that realis–irrealis may nevertheless sometimes be useful as a pair of terms to capture certain formal diachronic processes and relationships within DQGEHWZHHQODQJXDJHV HJLQWKHKLVWRULFDOFRPSDUDWLYHVWXG\RI$XVWURQHVLDQRU1HZ Guinea area languages), but that different terms that more accurately capture the semantics RIDJLYHQODQJXDJHVSHFL¿FFDWHJRU\PD\EHPRUHKHOSIXOLQPDQ\LIQRWPRVWGHVFULSWLYH FRQWH[WV REFERENCES $QGHUVRQ$VWULGLandscapes of relations and belonging: Body, place and politics in Wogeo, Papua New Guinea (Person, Space and Memory in the Contemporary 3DFL¿F 1HZ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY ‘Realis’ and ‘irrealis’ in Wogeo 190 *LYyQ7Syntax: An introductionQGHGQ$PVWHUGDP %HQMDPLQV +DDQ)HUGLQDQGGH,UUHDOLV )DFWRU¿FWLRQ"Language Sciences± +RJELQ ,DQ The island of menstruating men: Religion in Wogeo, New Guinea 6FUDQWRQ &KDQGOHU +RJELQ ,DQ The leaders and the led: Social control in Wogeo, New Guinea 0HOERXUQH 0HOERXUQH8QLYHUVLW\3UHVV 0DXUL &DWHULQD $QGUHD 6DQVz :KDW GR ODQJXDJHV HQFRGH ZKHQ WKH\ HQFRGH reality status? Language Sciences ± 3DOPHU)5Mood and modalityQGHGQ &DPEULGJH7H[WERRNVLQ/LQJXLVWLFV &DPEULGJH &DPEULGJH8QLYHUVLW\3UHVV 3LHWUDQGUHD3DROD7KHFRQFHSWXDOVWUXFWXUHRILUUHDOLW\ $IRFXVRQQRQH[FOXVLRQ RIIDFWXDOLW\DVDFRQFHSWXDODQGDOLQJXLVWLFFDWHJRU\Language Sciences± 3RUWQHU 3DXO Modality 2[IRUG 6XUYH\V LQ 6HPDQWLFV DQG 3UDJPDWLFV 2[IRUG 2[IRUG8QLYHUVLW\3UHVV van der Auwera, Johan (see Auwera) van Gijn, Rik (see Gijn) :KRUI%/ 6RPHYHUEDOFDWHJRULHVLQ+RSLLanguage± Mats Exter H[WHU#SKLOXQLGXHVVHOGRUIGH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY /DQJXDJH'RFXPHQWDWLRQ &RQVHUYDWLRQ6SHFLDO3XEOLFDWLRQ1R 'HFHPEHU Melanesian Languages on the Edge of Asia: Challenges for the 21st Century, HGE\1LFKRODV(YDQVDQG0DULDQ.ODPHUSS ± KWWS QÀUFKDZDLLHGXOGFVS 8 KWWS KGOKDQGOHQHW From mountain talk to hidden talk: Continuity and change in Awiakay registers Darja Hoenigman The Australian National University When the Awiakay of East Sepik Province in Papua New Guinea left their village or bush camps and went to the mountains, they used a different linguistic register, ‘mountain talk’, in which several lexical LWHPVDUHUHSODFHGE\WKHLUDYRLGDQFHWHUPV,QWKLVZD\WKH$ZLDND\ would prevent mountain spirits from sending sickness or dense fog LQ ZKLFK WKH\ ZRXOG JHW ORVW RQ WKHLU MRXUQH\V 2YHU WKH ODVW GHFDGH SHRSOH¶VWULSVWRWKHPRXQWDLQKDYHEHFRPHPRUHIUHTXHQWGXHWRWKH HDJOHZRRG EXVLQHVV +RZHYHU &KULVWLDQLW\ FDXVHG D GHFOLQH LQ WKH XVHRIµPRXQWDLQWDON¶ cc Creative Commons Attribution Non-Commercial No Derivatives Licence Mountain talk to hidden talk: Awiakay registers 192 1. INTRODUCTION. This paper will look at two linguistic registers practiced by the $ZLDND\SHRSOH1 One of these, which we can refer to as ‘mountain talk’, was originally XVHGZKHQWUDYHOOLQJWRWKHPRXQWDLQVEXWLVQRZPRUHRUOHVVREVROHWH7KHRWKHUQHZO\ developed register, which we can name ‘hidden talk’, is used when Awiakay people travel WRWRZQ%RWKDUHUHIHUUHGWRDVND\PHQGDµGLIIHUHQWODQJXDJH¶RUND\PRPEDµGLIIHUHQW WDON¶E\WKH$ZLDND\,ZLOOH[SORUHWKHZD\VLQZKLFKµKLGGHQWDON¶FDQEHYLHZHGDVD FRQWLQXDWLRQRIµPRXQWDLQWDON¶ It is not uncommon for languages of the New Guinea Highlands to have special linguistic UHJLVWHUVFKDUDFWHULVHGE\OH[LFDOVXEVWLWXWLRQVDQGXVHGLQSDUWLFXODUVRFLDOFRQWH[WV,Q Kewa, for example, the use of a special speech variety is associated with notions of high PRXQWDLQVEHLQJLQKDELWHGE\ZLOGGRJVDQGVSLULWVIURPZKRPRQHPXVWSURWHFWRQHVHOI Similarly, Huli use a special vocabulary when travelling through country inhabited by GHPRQV )UDQNOLQ 2WKHUµKLGGHQODQJXDJHV¶DUHXVHGLQULWXDOO\UHVWULFWHGFRQWH[WV ZKLOHKXQWLQJ 7HOHIROWUDSSLQJUDWV LELG RURQSDQGDQXVKDUYHVWLQJH[SHGLWLRQVZKHQ FRRNLQJDQGHDWLQJFDVVRZDU\ 3DZOH\ HWF+RZHYHUVRPHRIWKHVHUHJLVWHUVKDYH GHFOLQHG )UDQNOLQ 6WHIDQLZ While several authors have looked into lexical substitution registers, few have attempted WRWUDFHWKHGLDFKURQLFFKDQJHV7KLVSDSHUZLOOVKRZKRZWKHXVHRIDUHJLVWHULVDGDSWHGWR QHZVRFLRHFRQRPLFFLUFXPVWDQFHV7KHH[DPSOHRIµKLGGHQWDON¶SURYLGHVXVZLWKWKHUDUH RSSRUWXQLW\RIDQDO\VLQJWKLVSURFHVVZKLOHLWXQIROGV 2. THE AWIAKAY AND THEIR LANGUAGE. Awiakay is a Papuan language spoken by 300 people living in Kanjimei village in the East Sepik Province of Papua New Guinea VHHPDS The village itself is located on the Konmei River, which is a tributary of the .DUDZDULZKLOHWKHPDMRUSDUWRIWKH$ZLDND\ODQGVWUHWFKHVVRXWKLQWRWKHPRXQWDLQV 1 7KLV SDSHU ZDV RULJLQDOO\ SUHVHQWHG DW :RUNVKRS RQ WKH /DQJXDJHV RI 3DSXD Melanesian Languages on the Edge of Asia: Past Present and FutureLQ0DQRNZDUL,QGRQHVLD)HEUXDU\ ,ZRXOGOLNHWRWKDQN1LFN(YDQVIRULQYLWLQJPHWRSDUWLFLSDWHDWWKHFRQIHUHQFHDQGIRU VXJJHVWLRQVRQKRZWRLPSURYHWKHSDSHU)RUYDOXDEOHFRPPHQWVRQHDUOLHUGUDIWV,ZLVKWRWKDQN &KULVWLDQ '|KOHU %HWKZ\Q (YDQV$QGUHZ 3DZOH\$ODQ 5XPVH\ /LOD 6DQ 5RTXH DQG %RUXW 7HOEDQ7KHDFFRPSDQ\LQJ¿OPVZRXOGEHEXWPHUHFXWVZLWKRXWWKHH[SHUWLVHDQGDUWLVWLFH\HRI Gary Kildea who was generous with his time and patience in helping me edit the footage and sub- WLWOHWKHHGLWHGVHJPHQWVTenkyu tumas, Masta G! I am also grateful to the two referees, Rupert 6WDVFKDQG/RXUHQV'H9ULHVIRUWKHLUGHWDLOHGUHYLHZVRIWKHSDSHUDQGKHOSIXOVXJJHVWLRQVIRU IXUWKHUZRUNRQWKLVVXEMHFW0\JUHDWHVWGHEWKRZHYHUOLHVZLWKWKH$ZLDND\SHRSOHIRUVKDULQJ WKHLUOLIHZRUOGZLWKPH Awiakay is not only how the speakers refer to their language, but it s also used by the inhabitants RI.DQMLPHLWRUHIHUWRWKHPVHOYHV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 193 MAP 1.DQMLPHL:HZDNURXWH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 194 7KH$ZLDND\HFRQRP\UHPDLQVODUJHO\DVXEVLVWHQFHRQH3HRSOHVXSSOHPHQWWKHLUVDJR GLHWE\KXQWLQJ¿VKLQJDQGJDWKHULQJ*DUGHQLQJLVRIPLQRULPSRUWDQFH3 FIGURE 17LNLQMDRZDVKLQJVDJR 1RZDGD\VDOO$ZLDND\DGXOWVDUHELOLQJXDOLQ7RN3LVLQDQG$ZLDND\0XOWLOLQJXDOLVP in other local languages is less common, but it occurs in the few families where one spouse is from the neighbouring Asangamut villageAmong adult Awiakay, the use of 7RN3LVLQLVFRQ¿QHGPDLQO\WRVLWXDWLRQVZKHUHLWIXQFWLRQVDVDODQJXDJHRIDXWKRULW\ &RGHVZLWFKLQJEHWZHHQ7RN3LVLQDQG$ZLDND\RFFXUVLQSXEOLFVSHHFKHVLQTXDUUHOVDQG in other situations where a speaker (of any gender and age) wants to take an authoritative SRVLWLRQLQWKHFRPPXQLFDWLYHDFW$OOFKLOGUHQDUHÀXHQWLQ$ZLDND\EXWDFTXLUH7RN3LVLQ DWDYHU\HDUO\VWDJH7KH\DUHDGGUHVVHGSULPDULO\LQ$ZLDND\ZKLOH7RN3LVLQLVXVHGIRU VFROGLQJ :RUGVIURP7RN 3LVLQ ± SDUWLFXODUO\ RQHV GHQRWLQJ LWHPV DQG FRQFHSWV ZKLFK KDYH HQWHUHGWKHYLOODJHIURPWKHRXWVLGH±GRHQWHU$ZLDND\DQGDUHXVHGLQHYHU\GD\VSHHFK Many of them are nativised, that is, adapted to the rules of Awiakay phonology and PRUSKRORJ\0RUHRYHU7RN3LVLQYHUEVZKLFKDUHERUURZHGLQWR$ZLDND\DFTXLUHDVSHFLDO VXI¿[-bapo-, which is attached to the borrowing and precedes the normal Awiakay verb HQGLQJ FI+RHQLJPDQ For example, Tok Pisin verb ‘buy’ gets adapted by 3 There is both linguistic and cultural evidence that gardening has been adopted relatively recently +RHQLJPDQ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 195 adding the ‘Awiakayser’ -bapo-, DVZHOODV$ZLDND\WHQVHQXPEHUDQGSHUVRQHQGLQJV (1) baim Æ baim-bapo-pali-k (TP) buy (TP) buy-LA-PRES-1SG As we shall see, Awiakay words have also been coined for many of these borrowings, EXWDUHRQO\XVHGLQVSHFL¿FVLWXDWLRQV 3. AWIAKAY ‘MOUNTAIN TALK’. The Awiakay employ four basic terms to describe their landscape: DQGDƾ ‘swamp’, mip µÀRRGSODLQ¶palakayµÀDWJURXQG¶DQGSRQGRƾ, which GHQRWHVODQGRIVLJQL¿FDQWO\KLJKHUHOHYDWLRQWKDQLWVVXUURXQGLQJVDQGFDQEHWUDQVODWHGDV µPRXQWDLQ¶7KH$ZLDND\FRQVLGHUWKHLUPRXQWDLQRXVODQGWREHJLQDERYHaPDERYHVHD OHYHOZLWKWKHKLJKHVWPRXQWDLQRQ$ZLDND\ODQG,QMDLƾEHLQJPDERYHVHDOHYHO FIGURE 2. ,QMDLƾ 1RQVWDQGDUGDEEUHYLDWLRQVWREHIRXQGLQWKLVSDSHU 73 IRU7RN3LVLQ /$IRUORDQDGDSWDWLRQ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 196 FIGURE 3. Drawing Awiakay mountains and creeks Before the commercial eaglewood5 trade, which started in the Awiakay region just EHIRUHWKH$ZLDND\ZHQWWRWKHPRXQWDLQVPDLQO\LQVHDUFKRINDQXƾLVD ‘blackpalm’, ZKLFKWKH\XVHGIRUPDNLQJERZVRUIRUVKRUWKXQWLQJWULSV6XFKWULSVZHUHUHVWULFWHGWR some mountains only, as others were perceived to be heavily populated by both HQGHPEDƾ ‘mountain spirits’ and WDƾJLDµVSLULWVRIWKHGHDG¶WKHODWWHUEHLQJSDUWLFXODUO\PDOHYROHQW On these trips people used to employ a different linguistic register, which I refer to as µPRXQWDLQWDON¶LQZKLFKFHUWDLQOH[LFDOLWHPVDUHUHSODFHGE\DYRLGDQFHWHUPV7KHUHLV DP\WKDERXWDPDQZKREHFDPHORVWLQWKHPRXQWDLQVDQGPHWDIHPDOHPRXQWDLQVSLULW This spirit hid him and taught him the avoidance terms for animals, plants and foods which SHRSOHVKRXOGXVHZKLOHLQWKHPRXQWDLQV6RPHRIWKHVHOH[LFDOSURKLELWLRQVDQGWKHLU UHSODFHPHQWVDUHPHQWLRQHGLQWKHP\WK7RGD\$ZLDND\SHRSOHUHPHPEHUQRPRUHWKDQ VRPHDYRLGDQFHWHUPV 5 Eaglewood (Gyrinops ledermannii) or Tok Pisin garu, from Indonesian gaharu, is known for its fragrant resin, which Awiakay call is-kamia OLWHUDOO\µWUHHPHDW¶µZRRGHQPHDW¶ ,WLVSURGXFHG DVWKHWUHH¶VUHVSRQVHWRDQLQMXU\DQGLVWKXVIRXQGRQO\LQDVPDOOSHUFHQWDJHRIHDJOHZRRGWUHHV This black resinous wood is highly sought after by traders because of its commercial value, and is sold to Middle Eastern countries and Japan “for religious, medical, ceremonial and domestic DFWLYLWLHVE\$VLDQ%XGGKLVWVDQG0RVOHPV´ *XQQHWDO MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 197 Ordinary Awiakay Mountain Awiakay aisia ͚ĞĞů͛ ŶŽƚĞƌŵƐŚŽƵůĚďĞƵƐĞĚĂƚĂůů ĂLJżŐǁĂż͚ŇLJŝŶŐĨŽdž͛ apuria ͚ƚLJƉĞŽĨďĞĞ͛ kamdok ͚ĐůŽƵĚ͛ kandukya ͚ǁŚŝƚĞ͛ ŬĂǁŝŶ͚ŵŽƵŶƚĂŝŶďŝƌĚʹ ƟŹĞƉĂǁŝĂŬĂLJ͚ƌĞĚďŝƌĚ͛ ƐƉŝƌŝƚŽĨĂĚĞĂĚŵĂŶ͛ ƚƵŵĂŶũŝżŐŽLJ͚ƚŚĞŚĂŝƌLJŽŶĞ͛ kayma ͚ĐĂƐƐŽǁĂƌLJ͛ OR kondamin panba ͚ƚǁŽůĞŐƐ͛ ŬŽżŐŽŶŽż ͚Ă ƚĂůů ƚLJƉĞ ŽĨ ŐŝŶŐĞƌ (Alpinia sp͘Ϳ͛ ŝƐŬĂżŐĂ͚ƚƌĞĞůĞĂĨ͛ ;сŶĂŵĞŽĨĂŵŽƵŶƚĂŝŶƐƉŝƌŝƚͿ momok (ƚĂǁĂͿ ͚ƐƉŝŶĞ ŽĨ Ă ƚLJƉĞ ŽĨĐĂŶĞǁŚŝĐŚƚŚĞǁŝĂŬĂLJƵƐĞŝŶ ĐŝƌĐƵůĂƌƌŽŽĨďƵŝůĚŝŶŐ͛ injam kanja ͚ĐĂŶĞƚŽŽƚŚ͛ ;сŶĂŵĞŽĨĂŵŽƵŶƚĂŝŶƐƉŝƌŝƚͿ ŵƵżŐƵŵĂ‘ƚĞƌŵŝƚĞƐ͛ŶĞƐƚ͛ nam tapuka ͚ŽůĚǁŽŵĂŶ͛ ĂŶĚĂżŐĂŵŐŽLJŬŽůŽŬŽƚ tao ͚ƐĂŐŽƐƉŝŶĞƐ͛ ͚ƐŽŵĞƚŚŝŶŐďĞůŽŶŐŝŶŐƚŽƐǁĂŵƉ͛ OR ŶŽƚĞƌŵƐŚŽƵůĚďĞƵƐĞĚĂƚĂůů tay ͚ƐĂŐŽ͛ kandukya kolokotay ͚ǁŚŝƚĞĨŽŽĚ͛ ƵŵďƵż͚ƐůŝƚĚƌƵŵ͛OR ͚ŐĂƌĂŵƵƚ ƚƌĞĞ ;sŝƚĞdž ĐŽŶĨŽƐƐƵƐͿ͛ ŶŽƚĞƌŵƐŚŽƵůĚďĞƵƐĞĚĂƚĂůů ĨƌŽŵǁŚŝĐŚƐůŝƚĚƌƵŵƐĂƌĞŵĂĚĞ ĞŵǁŝŬŽůŽŬŽůĂLJ yaki ͚ƚŽďĂĐĐŽ͛ ͚ƚŚĞƐŵŽŬŝŶŐƚŚŝŶŐ͛ yambuka ͚ůĞĂǀĞƐ ŽĨ Ă ƚLJƉĞ ŽĨ ŝƐŬĂżŐĂ͚ƚƌĞĞůĞĂĨ͛ ŐŝŶŐĞƌ͛ OR ŶŽƚĞƌŵƐŚŽƵůĚďĞƵƐĞĚĂƚĂůů ;сŶĂŵĞŽĨĂŵŽƵŶƚĂŝŶƐƉŝƌŝƚͿ ŶĂŵĞ ŽĨ ĂŶLJ ĮƐŚ ĨŽƵŶĚ ŝŶ ƚŚĞ ŶŽƚĞƌŵƐŚŽƵůĚďĞƵƐĞĚĂƚĂůů ƵƉƉĞƌƉĂƌƚƐŽĨǁŝĂŬĂLJĐƌĞĞŬƐ TABLE 1. Avoidance terms in mountain Awiakay By using this linguistic register, the Awiakay would satisfy the demands of mountain spirits and prevent them from carrying out malevolent acts, such as sending sickness or MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 198 GHQVHIRJLQZKLFKWKH\ZRXOGJHWORVW 2YHUWKHODVWGHFDGHSHRSOH¶VWULSVWRWKHPRXQWDLQVKDYHEHFRPHPRUHIUHTXHQWGXH WRWKHFRPPHUFLDOHDJOHZRRGWUDGH This wood grows mainly at altitudes between 70 and P2Q$ZLDND\ODQGWKLVLVWKHUHJLRQVRXWKRI.DQMLPHLZKHUHWKHODQGVWDUWVULVLQJ LQWRWKHPRXQWDLQV VHHPDS As people spent more time in the mountains, one might expect that ’mountain talk’ ZRXOG WKULYH DW OHDVW , GLG +RZHYHU D &DWKROLF FKDULVPDWLF PRYHPHQW ZKLFK WKH Awiakay accepted in 1995, demanded of people that they radically cut their traditions and EUHDNWKHLUUHODWLRQVKLSVZLWKWKHVSLULWV 7HOEDQDE ±DQGWKHUHIRUHZLWKWKHLU ODQG6 Sickness (and even death) caused by not thoroughly implementing the expected SUDFWLFHVZRXOGQRZQRORQJHUEHLQÀLFWHGE\WKHVSLULWVIRUnot following their demands, EXWUDWKHUE\*RGIRUIROORZLQJWKHP,QRUGHUWRSURWHFWWKHPVHOYHVIURP*RG¶VDQJHU people were now forced to abandon the very practices which used to protect them from VSLULWVZKLFKKDVPHDQWDJUDGXDOGHFOLQHLQWKHXVHRIµPRXQWDLQWDON¶ FIGURE 4. Searching for eaglewood at Umbim 4. AWIAKAY ‘HIDDEN TALK’. While ‘mountain talk’ has declined, a new linguistic register UDWKHUVLPLODULQLWVIRUPDQGIXQFWLRQKDVVSUXQJXSLQDGLIIHUHQWVHWWLQJKay menda 6 7KH$ZLDND\JUDGXDOO\DFFHSWHG&KULVWLDQLW\LQWKH V&DWKROLFPLVVLRQDULHVZKRRFFDVLRQDOO\ visited the village put most effort into uprooting initiation rites, while many of the customary SUDFWLFHVFRQQHFWHGZLWKVSLULWVFRQWLQXHGWRFRH[LVWZLWKWKHQRPLQDO&KULVWLDQLW\ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 199 ‘different language’, or what I will refer to as ‘hidden talk’ is used when the Awiakay go to WKHWRZQWRVHOOWKHHDJOHZRRGZKLFKWKH\KDYHKDUYHVWHGDQGWREX\JRRGV :HZDNLVWKHSURYLQFLDOFDSLWDODQGKDVLQUHFHQW\HDUVEHFRPHIUHTXHQWO\YLVLWHGE\ SHRSOHIURPUHPRWHDUHDVZKRFRPHWRVHOOHDJOHZRRGDQGVPDOOTXDQWLWLHVRIJROG,QWKH early years of the eaglewood trade Indonesian buyers would themselves travel around the SURYLQFHWREX\WKHDURPDWLFZRRG+RZHYHUWKLVEHFDPHGDQJHURXVDVLWZDVNQRZQWKDW they were carrying huge amounts of money, and they were often robbed (reportedly two of WKHPZHUHHYHQNLOOHGLQHDUO\LQDQDWWDFNRQWKH6HSLN5LYHU :KHQWKHLQLWLDOERRP in the eaglewood trade declined, these foreign buyers did not earn enough to be willing to WDNHWKHULVNVDQGVRWKH\JDYHXSWKHLU¿HOGWULSV2QWKHRWKHUKDQGSHRSOHZKRVHODQGLV rich in eaglewood earned enough to buy outboard motors and started travelling to Wewak themselves, to sell their eaglewood and also to buy goods which can only be obtained in WRZQ7 This increase in visitors with money who are not used to town has coincided with DQLQFUHDVHLQFULPH,WLVQRWXQFRPPRQIRUYLVLWRUVWRWRZQWREHUREEHGRIDOOWKHLU SRVVHVVLRQV Being aware of these dangers, the Awiakay people try to be extremely cautious when in Wewak, and have (among other things, such as carrying cassowary bone daggers) started SUDFWLVLQJµKLGGHQWDON¶+LGGHQWDONLVDUHJLVWHURIOH[LFDOVXEVWLWXWLRQLQZKLFKDOO7RN 3LVLQERUURZLQJVZKLFKDUHXVHGLQHYHU\GD\$ZLDND\LQWKHYLOODJH±DQGPD\WKHUHIRUH EHXQGHUVWRRGE\RXWVLGHUV±DUHUHSODFHGE\QHZO\FRLQHG$ZLDND\WHUPV&RLQLQJQHZ words for newly introduced items and concepts is a common practice in many languages RI 1HZ *XLQHD %XW ZKDW PDNHV WKLV SKHQRPHQRQ GLIIHUHQW IURP VLPLODU SURFHVVHV LQ other languages is the special function that the Awiakay attribute to these newly-coined expressions in their vernacular (namely concealing the meaning of commonly used Tok Pisin borrowings), and the special social setting in which this is done (not in the village, EXWZKHQJRLQJWRWKHWRZQ ,QRWKHUZRUGVLWLVLPSRUWDQWWRQRWHWKDW7RN3LVLQWHUPV ±SDUWLFXODUO\RQHVGHQRWLQJLWHPVDQGFRQFHSWVZKLFKKDYHHQWHUHGWKHYLOODJHIURPWKH RXWVLGH±doHQWHUHYHU\GD\$ZLDND\DVLWLVXVHGLQWKHYLOODJH(YHQWKRXJKPDQ\RIWKHVH borrowings have been nativised by adapting them to the rules of Awiakay phonology and morphology, they are nonetheless parts of the Awiakay language (which is not spoken by people from other villages and even less so by anyone in the town) that could be understood E\RWKHUSHRSOH$ZLDND\ZRUGVKDYHWKHUHIRUHEHHQFRLQHGIRUPDQ\RIWKHVHORDQVEXW WKH\DUHSULPDULO\XVHGLQVLWXDWLRQVUHTXLULQJkay menda, while Tok Pisin terms continue WREHXVHGLQ.DQMLPHL8 To illustrate how kay mendaZRUNVFRQVLGHUKRZSHRSOHUHIHUWRDJDOORQGUXP,Q Tok Pisin this is fotifo$WUDGLWLRQDOLWHPZLWKWKHPRVWVLPLODUIXQFWLRQWRDGUXPZDVD EXFNHWPDGHIURPDODUJHEDPERRXVHGIRUFDUU\LQJZDWHU,Q$ZLDND\LWLVFDOOHGyomoy, 7 Out of six outboard motors in Kanjimei four were bought with people’s earnings from selling HDJOHZRRG 8 In certain situations, when the Awiakay people want to conceal their talk from visitors from other villages, they will resort to kay menda even at home in Kanjimei, but typically this register is used ZKHQHYHUWKH\JRWRWRZQ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 200 and people adopted this term to replace fotifo+RZHYHUZKLOHDWKRPH RQWKHLUODQG WKH Awiakay would continually use fotifo, but leaving their land, particularly when going to town, they would call it yomoyZKHQVSHDNLQJ$ZLDND\WRHDFKRWKHU 6RIDU,KDYHFROOHFWHGDERXW$ZLDND\FUHDWLRQVXVHGLQkay menda in place of Tok 3LVLQWHUPV VHH$SSHQGL[ :HFDQGLYLGHWKHVHWHUPVLQWR¿YHJURXSVDFFRUGLQJWRWKH way in which they were created: ±WHUPVZKLFKGHQRWHREMHFWVZLWKVLPLODUIXQFWLRQV wallet: (TP) hanpaus Æ kundambiµFRFRQXW¿EUHIRUVWRULQJWREDFFR¶ ±WHUPVZKLFKGHQRWHREMHFWVVLPLODULQIRUP WKH\ORRNVLPLODU petrol: (TP) petrol Æ yom ‘water’ balloon: (TP) balun Æ mumba ‘bladder’ gold: (TP) gol ÆNLƾLP‘sand’ ±GHVFULSWLYHWHUPV store: (TP) stoa Æ kolokot yawa ‘things house’ bra: (TP) susu kalabus Æ isik ulakaplakay ‘(something that) covers breasts’ ±OH[LFDOFDOTXHV toilet: (TP) haus pekpek ÆHQHƾ\DZD ‘shit house’ ±DEVXUGO\LQFRQJUXRXVWHUPV DZRUGGHQRWLQJVRPHWKLQJWKDWSHRSOH¿QGGLVJXVWLQJLV XVHGIRUVRPHWKLQJWKH\¿QGGHOLFLRXVRQWKHEDVLVRISK\VLFDOUHVHPEODQFH noodles: (TP) nudols Æ kundam enga ‘earthworm shit’ chocolate cream: (TP) soklit krim Æ HQHƾPROD ‘diarrhoea/rotting shit’ tinned (mushroom) sauce: (TP) (?) Æ PHƾJZDN‘vomit’ Although coining new words for newly introduced items and concepts is a common practice in many languages of New Guinea, it is the special function that the Awiakay KDYHDWWULEXWHGWRXVLQJWKHVHQHZO\FRLQHGH[SUHVVLRQVLQWKHLUYHUQDFXODU LHFRQFHDOLQJ the meaning of commonly used Tok Pisin borrowings) and the special social setting in ZKLFKWKLVLVGRQH LHQRWLQWKHYLOODJHEXWZKHQJRLQJWRWKHWRZQ ZKLFKPDNHVWKLV SKHQRPHQRQGLIIHUHQWIURPVLPLODUSURFHVVHVLQRWKHUODQJXDJHV Some of the examples of how kay menda is used can be drawn from a video recorded HDJOHZRRGVHOOLQJWULSWR:HZDNLQ6HSWHPEHU ,QVSLWHRIDIRUWQLJKWZLWKRXWDQ\UDLQLQWKHPRXQWDLQVZKLFKZRXOG¿OOXSWKHFUHHNV that would send water to the Karawari river, the Kangrimei passage was still navigable, which saved Desmon Asuk, Dicson Tumak, Sailus Kaim, Justin Pupi and me a whole DIWHUQRRQ¶V MRXUQH\ GRZQ WKH .DUDZDUL 5LYHU$SDUW IURP VKRUWHQLQJ WKH ORQJ MRXUQH\ XVLQJWKLVVKRUWFXWDOVRPHDQVWKDWWKH$ZLDND\FDQVDYHDERXW¿YHJDOORQVRISHWURO DQG a bit more on the way back when going upriver), trade for food with the people from MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 201 Karawari-speaking Kaiwaria and Masandanai villages along the channel and overnight in RQHRIWKHLUFDPSV.DQJULPHLZDVYHU\ORZWKRXJKVRZHKDGWRWXUQRIIWKHPRWRUVRDV QRWWRKLWWKHEUDQFKHVDQGWUHHWUXQNVO\LQJDWWKHERWWRP:KLOHSDGGOLQJ$VXNZKRLV more experienced in travelling to town, started a conversation in which he repeated for the \RXQJHUER\VDQGPHKRZZHVKRXOGEHKDYHZKHQZHFRPHWR:HZDN9 KAY MENDA – FILM 1 transcript ƐƵŬ͗ 1RƾRPJXVDQGDDƾNDNSHNHƾJR\HQGXƾRSLDQJRPEHPJR\ ϭ ROXNXQMDWDƾDQDNDSDƾJXPEHPiskamia salimEDSRƾJDSHNHƾERS When you meet somebody in the town, don’t tell them that we came down to sell tree meat>HDJOHZRRG@ $XQGD\DPHƾJDSHNHƾ7DƾDQSDƾJRPEHP :HMXVWFDPHGRZQIRUDWULS7HOOWKHPWKDW 9 ,QWKHWUDQVFULSWOH[LFDOVXEVWLWXWHVDUHEROGHGDQGXQGHUOLQHG,QWUDQVODWLRQRULJLQDO$ZLDND\ meanings are underlined, while their ‘hidden’, kay mendaPHDQLQJVIROORZLQVTXDUHEUDFNHWV HJiskamia (lexical substitute) is translated tree meat>HDJOHZRRG@7RN3LVLQH[SUHVVLRQVDUH blue and underlined like this. In the transcripts the translations are more faithful to the original $ZLDND\WH[WZKLOHWKH\KDGWREHVOLJKWO\PRGL¿HG VKRUWHQHG IRUVXEWLWOLQJSXUSRVHV7KH WUDQVFULSWVFRPH¿UVWVRWKDWWKHUHDGHUFDQEHFRPHIDPLOLDUZLWKWKHPHDQLQJRIDYRLGDQFHWHUPV EHIRUHZDWFKLQJWKH¿OPVZKLFKDUHDQLQWHJUDOSDUWRIWKLVSDSHU,QWKHVXEWLWOHVOH[LFDOVXEVWL- WXWHVDUHLQ\HOORZ,QRUGHUWRLOOXVWUDWHKRZkay menda works, I have glossed Awiakay terms using their original Awiakay meanings, rather than their kay mendaPHDQLQJVHJ,KDYHJORVVHG ikakapan as ‘carving’ (ordinary Awiakay) instead of ‘writing’ (kay menda MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 202 ƐƵŬ͗ .RƾJRWPD\DQDQGDDND\DxDƾJXQD\HODƾDQ$QDND \DxDƾJXQD\ .RƾJRWPD\>'DUMD@ZLOOQRWWHOOWKHPHLWKHU6KHZRQ¶WWHOO WƵƉŝ͗ An opepon. 6KHNQRZV ĂƌũĂ͗ 1Lƾ QLƾDQGDRSHSDOLN ,,NQRZ ƐƵŬ͗ 0P$QGDRSHSRQ$ND\DxDƾJXQD\ 0KP>DJUHHV@6KHNQRZV6KHZRQ¶WWHOO ĂƌũĂ͗ Andoposa opepalik. ϭϱ ,NQRZWKDWYHU\ZHOO ƐƵŬ͗ (ODNWRNDƾJXPJR\NXQMDNRORNRWDHODND\SD\SPDƾJDHSDOXƾJR\ tok. This is the only thing we get stones>PRQH\@IRU ĂƌũĂ͗ Yo. 7KHUHDGHULVQRZLQYLWHGWRZDWFKD¿OPH[FHUSWIURPRXUWULSWRWKHWRZQZKLFKLV available at KWWS \RXWXEHW/]/&SZ]$Z ϭ͗ϮϮ͘ The Awiakay are afraid of being held by rascals and robbed, so eaglewood selling trips DUHDOZD\VSHUPHDWHGZLWKVHFUHF\1RRQHHYHUGLVFXVVHVWKHLUEXVLQHVVZLWKSHRSOHZKRP they meet on the river or in the camps where they overnight, let alone with anyone in the WRZQ:HZDNLVSHUFHLYHGWREHDGDQJHURXVSODFH\HWRQHZKHUHWKH$ZLDND\FDQJHWDOO WKHJRRGVWKH\GHVLUH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 203 meaning of ‘pickpockets’ OLQH &DOOLQJQDPHVRIGDQJHURXVHQWLWLHVLVRIWHQDYRLGHG10 VRDOWKRXJKWKH¿YHRIXVDUHDORQHLQWKHFDQRHRQWKH.DQJULPHLDQGWKHUHLVQRGDQJHURI anyone else overhearing our conversation, he chooses to use kay mendaWHUPVIRUUDVFDOV As it started getting dark we decided to spend the night at Kambatiman, a Masandanai FDPSLQWKHPLGGOHRIWKH.DQJULPHLSDVVDJHDERXWKDOIZD\WR$QJRUDP VHHPDS :H ZHUHQRWDORQHWKHUH±DIDPLO\IURPDQHDUE\.DLZDULDYLOODJHVWD\HGLQDQRWKHUVKHOWHU That is why Tumak and I used kay menda to replace Tok Pisin terms which could reveal 7XPDN¶VSODQVLQWRZQ KAY MENDA – FILM 2 transcript ĂƌũĂ͗ Tumak, o! ϭ +H\7XPDN Amba... amba im... amba momba ikakapan? :KDWZKDWDUH\RXGRZKDWDUH\RXcarving>ZULWLQJ@" dƵŵĂŬ͗ Amba ...? Ey! :KDW"2K>looks up in surprise@ Paypmanda ... paypmanda George sakay mamgoy bal emba tike mimbia ikakapalik. Stones>PRQH\@stones ,DPcarving>ZULWLQJ@ the name of the stones>DPRXQWRIPRQH\@*HRUJH¶V>ZLIH@JDYHPHWRtake >EX\@DEDOO ĂƌũĂ͗ Yo. M-m. Kaykay olukunja givim paypmanda. Yo. 5 Now please watch KWWS \RXWXEH',6LM$6Z Ϭ͗ϯϳ When asked what he was doing, Tumak was taken by surprise when he saw me with a YLGHRFDPHUD$VKHZDVDZDUHRIWKHSUHVHQFHRISHRSOHIURPDQRWKHUYLOODJH D.DLZDULD woman was sitting in a nearby wind house and approached when we started speaking), and possibly reminded by myself using a ‘hidden’ term for writing, he knew that he had to replace the expressions like ‘money’ and ‘buy’, with their kay mendaWHUPV+RZHYHUZH FRXOGKHDUKLPKHVLWDWLQJFDUHIXOO\WKLQNLQJKRZWRIRUPXODWHKLVVHQWHQFHV Arriving in Angoram in the early afternoon of the following day and storing the canoe ZLWKWKH,PDQPHULSHRSOHZHPDQDJHGWR¿QGDQHDUO\µEDFNORDG¶WUXFNWKDWZDVJRLQJWR 10 (YHQ LQ WKH YLOODJH SHRSOH ZRXOG RIWHQ DYRLG XVLQJ D ZRUG IRU HJ D KDUPIXO VSLULW ,QVWHDG of saying QXƾJXP ‘gigantic python’ when describing a picture where a speaker believed this creature was threatening a man depicted in the drawing, he would say kalak ambam ‘this what’ or kolokolayµWKLVWKLQJ¶ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 204 WDNHXVWR:HZDN7KHPDMRUSDUWRIWKHMRXUQH\ZDVRYHUDQGDOWKRXJKWKHULGHDORQJWKH dirt road to Wewak is a rough one, the men who were in charge of the canoe could now WDNHVRPHUHVW7KHUHZHUHRQO\DFRXSOHRIRWKHUSHRSOHVLWWLQJRQWKHWUXFNDQGZDLWLQJ with us, and the leisurely conversation that took place did not involve anything that would demand secrecy, but the closer to Wewak the Awiakay get, the more urge they feel to speak DPRQJWKHPVHOYHVLQDZD\WKDWRWKHUSHRSOHGRQRWXQGHUVWDQGWKHP11 KAY MENDA – FILM 3 transcript ƐƵŬ͗ Mawia tok. ϭ *UHDW ^ĂŝůƵƐ͗ 3HNHSLDƾSHNHSLDƾ\DDPEXƾ :HFDPHGRZQULYHUDQGZHDUHJRLQJQRZ .XPDSPDQGDƾNRORƾ7RP We’re in the coconut shell/DWHU ƐƵŬ͗ Aka kumap manda. Now please watch KWWS \RXWXEHUY=D&=3.F Ϭ͗ϰϴ ,QGLYLGXDOV¿UVWEHFRPHDFTXDLQWHGZLWKkay menda in the village, but only put it into SUDFWLFHZKHQWUDYHOOLQJWRWRZQ(YHU\WULSWRWKHWRZQLVWKHUHIRUHDWUDLQLQJIRUWKHER\V ZKRDUHQRW\HWIXOO\FRPSHWHQWLQWKLVUHJLVWHU7KH\DUHWDXJKWkay menda by the more H[SHULHQFHGPHQ7KHVHDOVRFRUUHFWWKHER\VZKHQWKH\PDNHPLVWDNHV)LOPVKRZVKRZ $VXNFRUUHFWV6DLOXVZKRFDOOVFDUµFRFRQXWVKHOO¶LQVWHDGRIµWXUWOHVKHOO¶ 11 $IWHURQHRIWKHLQWHUQDOYLOODJH¿JKWV3XSL¶VEURWKHU1DPD\VDLG ³Gutpela taimEXƾDNDDPEDOXƾ TaunXƾDQLQDQJR\WRNSXNXQLQDƾ´µ:HQHYHUJHWWRJHWKHUDWJRRGWLPHV:KHQZHJRWRWKH WRZQZHWKLQNRIHDFKRWKHU¶>,QWKHYLOODJHZHWHQGWRTXDUUHODQG¿JKW%XWZKHQZHJRWRWKH WRZQZHVWLFNWRJHWKHUDVRQHDQGWDNHFDUHRIHDFKRWKHU@ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 205 When they are in Wewak, the Awiakay normally overnight with people in Masandanai FDPSDW.ULD .UHHUPDUNHW DVHWWOHPHQWRIWKH.DUDZDULVSHDNLQJFRPPXQLWLHV$OWKRXJK WKH\DUHRQIULHQGO\WHUPVZLWK.DUDZDULSHRSOH DOEHLWQRWWKHLUZDQWRNV WKH\¿QGLW YHU\LPSRUWDQWWRFRQFHDOWKHLUEXVLQHVVDQGSODQVIURPWKHP7KHER\ZHVHHVLWWLQJDQG ZULWLQJLQ¿OPDIWHUZHDUULYHLQ:HZDNLVD0DVDQGDQDLER\ZKRJRHVWRVFKRROLQWRZQ WKHRWKHUVDUH$ZLDND\GLVFXVVLQJWKHLUSODQVIRUWKHIROORZLQJGD\ KAY MENDA – FILM 4 transcript ƐƵŬ͗ 1RƾDPEDNRORNRWDPDHenamin nan? ϭ What will you take>EX\@ ^ĂŝůƵƐ͗ 1LƾDQLQDNR\DPEDHƾJDQLQDN ,ZLOOJRZKDWJRDQGtake>EX\@ ƐƵŬ͗ Pisikanda, pisikanda kakaym. 4XLFNO\>FRPHRQ@WHOOPH ^ĂŝůƵƐ͗ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 206 ^ĂŝůƵƐ͗ Amba tok yom HƾDPERQJR\. $QGZKDWHOVHVKH>'DUMD@ZLOOtake>EX\@ water>SHWURO@ Now please watch KWWS \RXWXEH,1Q%K*$S7[F ϭ͗ϱϮ Asuk asks Sailus what he is going to buy, and Sailus hesitates with his answer, not NQRZLQJZKDWWRFDOOµWKHWKLQJIRUVWUDLQLQJVDJRÀRXU¶LQkay menda+HDYRLGVXVLQJWKH 7RN3LVLQWHUPE\FDOOLQJLWµZKDWIRUVWUDLQLQJ¶ OLQH µZKDW¶VWDQGLQJIRUµWKDWWKLQJ¶ FI IQ %RWK3XSLDQG6DLOXVDUHVHDUFKLQJIRUWKHULJKWWHUP OLQHVDQG ZKHQ7XPDN gives up and calls a Tok Pisin word streina ‘strainer’ (line 8) at the same time when Asuk VD\VWKDWWKH\FDQQRWUHPHPEHU7XPDNLVLQVWDQWO\UHSULPDQGHGE\6DLOXVZKRFDOOVKLP ‘fucker’ (line 10), which makes them all laugh, but Asuk defends him by saying that there ZDVQRRWKHUZD\RIWHOOLQJWKLV OLQH DVKHKLPVHOIDVWKHPRVWFRPSHWHQWVSHDNHURI kay mendaDQGWKHOHDGHURIWKLVWULSWR:HZDNFDQQRWWKLQNRIDVXLWDEOHDYRLGDQFHWHUP The conversation continues by Asuk telling what he will buy the next day and turns to how WKH\DUHDOOJRLQJWRJRZLWKPHWREX\SHWURODQGEULQJLWWR0DVDQGDQDLFDPS6DLOXVLV VWXFNDJDLQZKHQKHZDQWVWRVD\WKDW,¿UVWQHHGWRJRWRWKHEDQNWRWDNHRXWP\PRQH\ %\FDOOLQJWKHEDQNDµVSLULWKRXVH¶ OLQH KHLVFRQIXVHGDJDLQDQGXVHVamba ‘what’ HYHQZKHQKHZDQWVWRUHIHUWRPRQH\$VXNKHOSVKLPRXWE\UHPLQGLQJKLPRIERWKWHUPV SD\SPDƾJD‘stones’ for money and SD\SPDƾJD\DZDµVWRQHKRXVH¶IRUEDQN OLQHVDQG 6DLOXVFRUUHFWVKLPVHOIE\XVLQJDQDOWHUQDWLYHWHUPIRUPRQH\paypmanda, saying that when I take out my money and buy all the goods, we will load it all onto a truck (for $QJRUDP MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 207 In situations like this kay menda becomes a kind of a mind game which all participants HQMR\HYHQWKRXJKLWVSULPDU\SXUSRVHLVWRPDNHWKH$ZLDND\IHHOVDIHUZKLOHLQWRZQ 7KHQH[WGD\ZHZHQWVKRSSLQJ$V$VXNZDQWHGWRSXWVRPHRIWKHPRQH\KHHDUQHGZLWK eaglewood in the bank, he and Pupi went to do this business, while Sailus, Tumak and I ZHQWWRWKHVKRSV KAY MENDA – FILM 5 transcript dƵŵĂŬ͗ Wakon. SkulXƾSDND\DPHQDNSRNR\bag kalakiay enapok. ϭ 6RPDQ\,I,FRXOGtake>EX\@WKLVEDJ,FRXOGFDUU\LWWRVFKRRO ^ĂŝůƵƐ͗ Aka anda. SkulXƾSDND\DPHQDNSRNR\bag kalakiay enapok. 7UXH,I,FRXOGtake>EX\@WKLVEDJ,FRXOGFDUU\LWWRVFKRRO dƵŵĂŬ͗ Aka kolokot. Paypmanda tonaypeke wakon aka kiay enapok. What a thing! If I had lots of stones>PRQH\@, I would take>EX\@ PDQ\ ^ĂŝůƵƐ͗ Kandikak. $QGDƾJXƾ\DND\DPEODND\$QGDƾJXƾ +HUH[Something for@wandering around in swamps>JXPERRWV@ )RUVZDPSV $QGDƾJXƾ\DND\DPEODND\... 5 [Something for@wandering around in swamps>JXPERRWV@ dƵŵĂŬ͗ Bag, o! Oh, bag! Mawiakay kalak. 7KLVLVDJUHDWRQH ^ĂŝůƵƐ͗ Emay kalak yambongoy, poka pukulakana pokoy anda kaykay wakakanaype. If this sorcerer>UDVFDO@ keeps tailing us, I will bash his face till he VFUHDPV ^ĂŝůƵƐ͗ Amba pia kandikakay? Is this a piece of something? dƵŵĂŬ͗ .RORNRWPXƾD\DPEOD ϭϬ >6HH@WKH\DUHZDQGHULQJDURXQGDQGORRNLQJDWWKLQJV $SLDƾVDND\DPEDSLVLSPJR\tawel pisip. 7KLVLVOLNH$SLDƾ¶VZKDWOLNHDWRZHO ^ĂŝůƵƐ͗ Amgam? Wakon. +RZPXFK"$ORW dƵŵĂŬ͗ 3RNRQXƾSDVLSODND\. 6RPHWKLQJWRFOHDQ\RXUIDFHZLWK>WRZHO@ .XƾDQMDDPJR\WDƾDQLP" Is it for children? ^ĂŝůƵƐ͗ 7RPNHOHHORNLD\RSRƾDQDN ϭϱ ,¶OOORRNDWWKDWODWHU MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 208 dƵŵĂŬ͗ $ƾJH\D /HW¶VJR ^ĂŝůƵƐ͗ $ƾJH /HW¶VJR dƵŵĂŬ Emay nanday okokaim yambon. (to A sorcerer>UDVFDO@LVIROORZLQJ\RX ŵĞͿ͗ Emay nanday okokaim yambon, yo kon yambon. A sorcerer>UDVFDO@LVIROORZLQJ\RXKH¶VZDONLQJMXVWWKHUH 1LƾPDHPDHDQLƾ ϮϬ /HWPHJR¿UVW ^ĂŝůƵƐ͗ Aka ... aka mokoinay. Tawa pokombakanak. +HZRQ¶WKHZRQ¶WWRXFK\RX,¶OOEUHDNKLVERQHV Now please watch KWWS \RXWXEHUEO,M(($ Ϯ͗ϰϭ :KLOHZDQGHULQJDURXQGWKHVWRUHDQGORRNLQJDWDUWLFOHVVXFKDVJXPERRWV±ZKLFKKH does not know what to call in kay menda, and therefore uses a descriptive term ‘something IRUZDQGHULQJDURXQGLQVZDPSV¶ OLQHVDQG ±6DLOXVJRWDIHHOLQJWKDWWKHPHQEHKLQG KLPZHUHQRWMXVWH\HVKRSSLQJ,QOLQHKHERDVWVKRZKHZLOOEDVKWKHUDVFDO¶VIDFH ZKLFKLVDWWKHVDPHWLPHDZDUQLQJIRU7XPDNDQGPHWREHFDUHIXO)LQGLQJDVPDOOWRZHO KHZRQGHUVZKDWLWLVZKLOH7XPDN¶VDWWHQWLRQLVZLWKWKHDOOHJHGSLFNSRFNHWV+HWKHQ answers Sailus, attempting to remember an avoidance term for towel, but in the end uses the Tok Pisin word tawel/DWHUKHFRUUHFWVKLPVHOIXVLQJDGHVFULSWLYHWHUPµVRPHWKLQJ WRFOHDQ\RXUIDFHZLWK¶ OLQH ,ISHRSOHGRQRWNQRZRUGRQRWUHPHPEHUDQDOUHDG\ established kay menda term, they often try to create one on the spot, and in such cases they ZRXOGIUHTXHQWO\UHVRUWWRGHVFULSWLRQ+RZHYHU7XPDNLVDOHUWDQGDQ[LRXVEHFDXVHRI WKHDOOHJHGUDVFDOVDQGKHVXJJHVWVWKDWZHOHDYH+HFDOPO\ZDUQVPHWKDWD SRWHQWLDO UDVFDOLVIROORZLQJPHDQGVXJJHVWVWKDWKHJRHVDKHDG OLQH %HLQJQHUYRXVKLPVHOI Sailus boasts again, assuring me that he can protect me if somebody wanted to rob me OLQH +DYLQJH[SHULHQFHGVRPHWURXEOHVWKHPVHOYHVDQGKHDULQJVWRULHVDERXWSHRSOH EHLQJDWWDFNHGDQGUREEHGWKH$ZLDND\DUHDOZD\VWHQVHZKHQLQWRZQ0DQ\RIWKHP particularly young boys, release this uneasiness by boasting how mean they will turn if DQ\RQHGDUHVDWWDFNWKHP:KLOHWKLVFDQEHDPHDQLQJOHVVHYHQMRFXODUHYHU\GD\SUDFWLFH LQWKHYLOODJH WKRXJKDOVRHPSOR\HGGXULQJ¿JKWV LWEHFRPHVDPHDQVRIUHDVVXULQJRQH DQRWKHUZKHQLQWRZQ :HZHUHMXVWDERXWWROHDYHWKHVWRUHZKHQ$VXN DQG3XSL ZKR KDG¿QLVKHG WKHLU EXVLQHVV DQG ZHUH DOUHDG\ ORRNLQJ IRU XV FDPH LQ$V D JURXS RI ¿YH ZH ZHUH D OHVV attractive target for the robbers or pickpockets, so we stayed there to take a look at tapes ZLWKSRSXODUPXVLFWRUFKHVDQGNQLYHV7XPDNDQG,ZHUHORRNLQJIRUDODPSIRUKLVpap µPDWHUQDOXQFOH¶ DQGP\µIDWKHU¶ $\PDNDQ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 209 KAY MENDA – FILM 6 transcript ƐŚŽƉ Bilong disla bateri save stap insait. Narapela kain em i stap ϭ ĂƐƐŝƐƚĂŶƚ͗ long hapsait long narapela glas. 7KHVHRQHVKDYHEDWWHULHVLQVLGH7KHRWKHURQHVDUHRYHUWKHUH Yu minim wanemplakain? What kind do you want? dƵŵĂŬ͗ Narapela. Glas tasol em i go daun olsem... Disla em nogat? $QRWKHUNLQGZKHUHWKHODPSIROGVGRZQ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 210 ĂƌũĂ͗ Kondamin isapasa? By two sticks>%\.LQD@? dƵŵĂŬ͗ Kondamin isapasa ... ϮϬ By two sticks>%\.LQD@ Kay... kay \DZDƾZDNDQMLƾDQLQDƾ $QRWKHUZH¶OOJRDQGVHDUFKIRULWLQDQRWKHUhouse>VWRUH@ OK, kay ya$QJRUDPZDNDQMLQDƾ OK, we’ll go and search for it in another h[ouse@LQ$QJRUDP Now please watch KWWS \RXWXEHW1,[Z2Y ]B ϭ͗ϮϬ As it turns out that they do not have the kind of lamps that Aymakan asked for, Tumak suggests that we search in another store, for which he uses a shortened version of kay menda term, yawa ‘house’, instead of kolokot yawaµKRXVHRIWKLQJV¶ OLQH 3HRSOHWHQG to shorten words in ordinary Awiakay all the time, and this practice is sometimes applied to kay mendaDVZHOO Tumak then looks at other lamps and torches they sell in this store and says that the SULFHVZHQWXSVLQFHKHZDVODVWLQWKHWRZQDIHZPRQWKVDJR)RUWKHSULFHJRLQJXS he uses the verb wam-, which originally means ‘go/come up’ in the sense ‘in the upriver GLUHFWLRQ¶RUµXSWRWKHKRXVH¶ EXWQRWµXSWRWKHPRXQWDLQ¶ $V$\PDNDQH[SHFWHGWKDWWKH ODPSKHZDQWHGZRXOGFRVWDURXQG.,DVNZKHWKHUWKHSULFHZDVGRXEOHGDQG7XPDN FRQ¿UPVWKDWLWZHQWXSE\µWZRVWLFNV¶RQHWUHHVWLFNHTXDOOLQJ.7KLVWHUPFRPHV IURPWKHFRORQLDOGD\VLQ31*ZKHQWKH¿UVWPRQH\ZDVLQWURGXFHGWRWKH$ZLDND\DQG WKH\GHYLVHGWKHLURZQQDPLQJV\VWHPIRUWKHFRLQVDQGQRWHV 0RVW UHJLVWHUV DUH QRW ³VRFLRORJLFDOO\ KRPRJHQHRXV IRUPDWLRQV´ $JKD ZKLFKPHDQVWKDWQRWHYHU\RQHLVHTXDOO\FRPSHWHQWLQWKHPDQG$ZLDND\µKLGGHQWDON¶LV QRH[FHSWLRQ:KLOHHYHU\$ZLDND\SHUVRQFDQVSHDNDWOHDVWDOLWWOHELWRIkay menda, the PRVWFRPSHWHQWVSHDNHUVDUHWKHPHQZKRWUDYHOPRVWIUHTXHQWO\WRWRZQ+RZHYHUZH FRXOGVHHWKDWHYHQLQWKLVJURXSWKHOHYHORIÀXHQF\YDULHVDQGGHSHQGVRQVHYHUDOIDFWRUV not excluding an individual speaker’s skills such as cunning, which is an essential part of µKLGGHQWDON¶ At the moment kay menda is still in the making and we can witness its on-going GHYHORSPHQW:RPHQZKR QRUPDOO\ VWD\ LQ WKH YLOODJH GR QRW KDYH PDQ\ FKDQFHV RI using kay menda LQSUDFWLFH KRZHYHUPDQ\RIWKHPWDNHDQDFWLYHSDUWLQFUHDWLQJLW:LWK DKXJHLQÀX[RIPDWHULDOJRRGVIURP,QGRQHVLDVKRSVLQWRZQDUHIXOORILWHPVSUHYLRXVO\ XQNQRZQWRWKH$ZLDND\ZKLFKPHDQVWKDWWKH\ERUURZWHUPVIRUWKHPIURP7RN3LVLQ When such an item is brought to the village, its form and function is eagerly studied and GLVFXVVHGDQGVRRQHURUODWHUVRPHERG\FRPHVXSZLWKDQ$ZLDND\WHUPIRULW,WWDNHV VRPHWLPHEHIRUHWKHVSHDNHUVDGRSWVXFKDWHUPRUFUHDWHDQHZRQHZKLFKWKH\¿QGPRUH DSSURSULDWH The usage of a number of terms varies, and one can either (a) use the same DYRLGDQFHWHUPIRUVHYHUDOGLIIHUHQW7RN3LVLQH[SUHVVLRQVHJ 73 kemera ‘camera’, (TP) +DYLQJWKHDELOLW\WRVNLOIXOO\GHFHLYHRWKHUSHRSOHLVKLJKO\YDOXHGE\WKH$ZLDND\ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 211 skrin ‘television screen’, (TP) gras ‘mirror’ are all referred to as memek ‘lightning’ in kay menda, or (b) use different kay menda H[SUHVVLRQVIRUWKHVDPHWKLQJHJmap kulamba yomba ‘water from ground hole’ or payp kulamba yomba ‘water from stone shelter’ for µVKRZHU¶7KHODWWHUXVXDOO\KDSSHQVZKHQDWHUPKDVQRWEHHQDGRSWHGE\DOOVSHDNHUV 5. CONTINUITY AND CHANGE IN AWIAKAY LINGUISTIC REGISTERS. I would argue that kay menda as it is spoken today in town is not a completely new register, but a continuation of the kay menda ZKLFKXVHGWREHVSRNHQLQWKHPRXQWDLQV7KHre appear to be many IXQFWLRQDODQGVRFLDODVZHOODVVRPHVWUXFWXUDOOLQJXLVWLFVLPLODULWLHVLQWKHLUXVH7KHUH is, however, no overlap in vocabulary, as ‘mountain talk’ used to ‘hide’ the meaning of Awiakay terms denoting people’s immediate environment, while ‘hidden talk’ creates DYRLGDQFHWHUPVIRU7RN3LVLQERUURZLQJVGHQRWLQJUHFHQWO\LQWURGXFHGLWHPVDQGFRQFHSWV Parallels between the two ͚ŵŽƵŶƚĂŝŶƚĂůŬ͛ ͚ŚŝĚĚĞŶƚĂůŬ͛ ǀĂƌŝĞƟĞƐŽĨkay menda ƉĂƐƚ͖ ƉƌĞƐĞŶƚ͖ ŶŽǁĂĚĂLJƐŽďƐŽůĞƚĞ ƌĞŐŝƐƚĞƌŝŶƚŚĞŵĂŬŝŶŐ ƵƐĞĚŝŶƵŶĨĂŵŝůŝĂƌƚĞƌƌŝƚŽƌLJ͖ ŵŽƵŶƚĂŝŶƐ;ŝŶŚĂďŝƚĞĚďLJƐƉŝƌŝƚƐͿ tĞǁĂŬ͕ĂůůƐƚŽƉƐŽŶƚŚĞǁĂLJ ĨĂƌĨƌŽŵƚŚĞǀŝůůĂŐĞŽƌĐĂŵƉƐ ƚŚĞƌĞǁŚĞƌĞƚŚĞǁŝĂŬĂLJ ĞŶĐŽƵŶƚĞƌƵŶĨĂŵŝůŝĂƌƉĞŽƉůĞ ƉĞŽƉůĞŐŽƚŚĞƌĞƚŽŐĞƚ ŬĂŶƵżŝƐĂ ;ǁŽŽĚĨŽƌďŽǁƐͿ͕ ƐĞůůŝŶŐĞĂŐůĞǁŽŽĚ͕ ƐŽŵĞƚŚŝŶŐƚŚĞLJŶĞĞĚ ŚƵŶƟŶŐ;ŶŽǁĂĚĂLJƐŚĂƌǀĞƐƟŶŐ ďƵLJŝŶŐŐŽŽĚƐ ĞĂŐůĞǁŽŽĚͿ ĚĂŶŐĞƌŽƵƐĞŶƟƟĞƐ ĞŶĚĞŵďĂż ‘ŵŽƵŶƚĂŝŶƐƉŝƌŝƚƐ͛ ĞŵĂLJ,͚ĂƐƐĂƵůƚƐŽƌĐĞƌĞƌ͛с ƌĂƐĐĂůƐ͕ƉŝĐŬƉŽĐŬĞƚƐ ƉŽƐƐŝďůĞĚĂŶŐĞƌƐ ƐŝĐŬŶĞƐƐĐĂƵƐĞĚďLJƐƉŝƌŝƚƐ͕ ƌŽďďĞƌLJ͕ƚŚĞŌ͕ƉŚLJƐŝĐĂůŝŶũƵƌLJ ŐĞƫŶŐůŽƐƚ͕ĚĞĂƚŚ ƉƌĞǀĞŶƟŽŶŽĨĚĂŶŐĞƌƐ ƉŽƐƐŝďůĞďLJŝŵƉůĞŵĞŶƟŶŐƚŚĞ ƉŽƐƐŝďůĞďLJƵƐŝŶŐŬĂLJŵĞŶĚĂ ĞdžƉĞĐƚĞĚƉƌĂĐƟĐĞƐ͕ŝ͘Ğ͘ƵƐŝŶŐ ĂŶĚƉƌĂLJŝŶŐ ŬĂLJŵĞŶĚĂ͕ŶŽǁĂĚĂLJƐƉƌĂLJŝŶŐ ƉĞƌƐŽŶƐǁŚŽĞŶŐĂŐĞŝŶ ŵĞŶ ŵĞŶ ƌĞůĞǀĂŶƚƐŽĐŝĂůƉƌĂĐƟĐĞƐ ;ŐŽŝŶŐƚŽŵŽƵŶƚĂŝŶƐͬƚŽǁŶͿ ĂŶĚĂƌĞƉƌŽĮĐŝĞŶƚŝŶƚŚŝƐ ƌĞŐŝƐƚĞƌ ŽƚŚĞƌƐĨĂŵŝůŝĂƌǁŝƚŚƚŚĞ ǁŽŵĞŶ͕ƚĞĞŶĂŐĞƌƐ ǁŽŵĞŶ͕ƚĞĞŶĂŐĞƌƐ ƌĞŐŝƐƚĞƌ ĐƌĞĂƚĞĚďLJ ͚ŵŽƵŶƚĂŝŶƐƉŝƌŝƚ͛ƚĂƵŐŚƚƉĞŽƉůĞ ĂůůǁŝĂŬĂLJ͖ŝŶƚŚĞŵĂŬŝŶŐ ŚŽǁƚŽƉƌŽƚĞĐƚƚŚĞŵƐĞůǀĞƐ TABLE 2. Parallels between the two varieties of kay menda For instance, both varieties of kay menda employ descriptive terms for their substitutes, HJ µFDVVRZDU\¶ EHFRPHV WXPDQMLƾJH ‘the hairy one’ in mountain talk, while ‘store’ becomes kolokot yawaµWKLQJVKRXVH¶LQKLGGHQWDON%RWKYDULHWLHVDUHXVHGZKHQSHRSOH venture into the ‘unknown’ territory, far away from the village or camps in order to get MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 212 VRPHWKLQJWKH\QHHG7KHPRXQWDLQVZKLFKDUHQRWHPSW\EXWDUH±MXVWOLNHWKHUHVWRI $ZLDND\ODQG±LQKDELWHGE\VSLULWVDUHDSODFHZKHUHPHQJRKXQWLQJJHWEODFNSDOP for their bows and nowadays harvest eaglewood, while the town, with all the unfamiliar people they meet, is the place where the Awiakay sell their eaglewood and buy the goods WKH\QHHG,QERWKVHWWLQJVWKH\PD\HQFRXQWHUGDQJHURXVHQWLWLHV±HQGHPEDƾ ‘mountain VSLULWV¶RUWKHUDVFDOVDQGSLFNSRFNHWVLQWKHWRZQ±ZKLFKPD\GDPDJHWKHPRUWKHLU SRVVHVVLRQV,QERWKFDVHVWKHGDQJHUVFDQEHSUHYHQWHGE\XVLQJkay menda, just that due to the changed relationship with spirits ‘mountain talk’ is nowadays replaced by praying, ZKLOHLQWKHWRZQSUD\HULVRQO\VXSSOHPHQWDU\WRµKLGGHQWDON¶,QERWKFRQWH[WVLWLVPHQ who venture to these faraway places and use kay menda there, while women, even if they DFFRPSDQ\WKHLUKXVEDQGVRUEURWKHUVVWD\EHKLQG±HLWKHULQEXVKFDPSVZDLWLQJIRU the men to return from the mountain, or in Angoram, waiting for the men to return from :HZDN,QERWKFDVHVZRPHQDQGWHHQDJHUVDUHQHYHUWKHOHVVIDPLOLDUZLWKWKHUHJLVWHU While ‘mountain talk’ is seen as a gift a spirit gave to the people to protect themselves, it must originally have been a fairly conscious creation, in which people chose to modify FHUWDLQHOHPHQWVRIRUGLQDU\$ZLDND\LQRUGHUWRDUULYHDWDGLIIHUHQWFRGH FI3DZOH\ RQ.DODPµSDQGDQXVODQJXDJH¶ ‘Hidden talk’, however, is being continually DQGDFWLYHO\FUHDWHGE\DOO$ZLDND\ In some socio-linguistic contexts the introduction of new commercial trades leads to increased exposure to and use of regional languages and a decline of local languages͘ However, in this instance it has also created circumstances in which the local language has GHYHORSHGDQHZGLPHQVLRQThe eaglewood trade seems to have re-strengthened people’s UHODWLRQVKLSZLWKWKHLUODQGZKLFKKDGRWKHUZLVHEHHQZHDNHQHG FIGURE 5. Continuity and change in the varieties of kay menda MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 213 By interpreting their environment through the same cosmology, with their actions being VWURQJO\ LQÀXHQFHG E\ WKH &DWKROLF FKDULVPDWLF PRYHPHQW 7HOEDQ E DQG E\ WKH rules and changes it brought, the Awiakay have transferred the same practice (namely a lexical substitution register) with a similar function (hiding the meaning in order to protect WKHPVHOYHVIURPEHLQJKDUPHG IURPWKHLUPRXQWDLQVWRDGLIIHUHQWVRFLDOVHWWLQJRIWRZQ REFERENCES $JKD$VLI5HJLVWHU,Q$OHVVDQGUR'XUDQWL HG Key terms in language and culture, 2[IRUG %ODFNZHOO $JKD$VLI5HJLVWHUVRI/DQJXDJH,Q$OHVVDQGUR'XUDQWL HG A companion to linguistic anthropology&DPEULGJH &DPEULGJH8QLYHUVLW\3UHVV $JKD$VLILanguage and social relations.&DPEULGJH 1HZ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 214 3DZOH\ $QGUHZ .DODP 3DQGDQXV /DQJXDJH $Q ROG 1HZ *XLQHD H[SHULPHQW LQ ODQJXDJH HQJLQHHULQJ ,Q7RP 'XWWRQ 0DOFROP 5RVV 'DUUHO7U\RQ HGV The language game: papers in memory of Donald C. Laycock 3DFL¿F/LQJXLVWLFV& .&DQEHUUD 3DFL¿F/LQJXLVWLFV 6WDVFK5XSHUW-RNLQJDYRLGDQFH D.RURZDLSUDJPDWLFVRIEHLQJWZRAmerican Ethnologist 6WDVFK 5XSHUW 5HIHUHQWZUHFNLQJ LQ .RURZDL $ 1HZ *XLQHD DEXVH UHJLVWHU DV HWKQRVHPLRWLFSURWHVWLanguage in Society 7HOEDQ%RUXW D7KH3RHWLFV RI WKH&URFRGLOH &KDQJLQJ &XOWXUDO3HUVSHFWLYHVLQ $PERQZDULOceania 7HOEDQ%RUXWE0RGL¿FDWLRQRISHUFHSWLRQLQD6HSLNFRPPXQLW\3DSHUSUHVHQWHG DWWKHWK&RQIHUHQFHRIWKH(XURSHDQ6RFLHW\IRU2FHDQLVWV9HURQD-XO\ 7HOEDQ%RUXW $VWUXJJOHZLWKVSLULWV +LHUDUFK\ULWXDOVDQGFKDULVPDWLFPRYHPHQW LQD6HSLNFRPPXQLW\3DPHOD-6WHZDUW $QGUHZ6WUDWKHUQ HGV Religious and ritual change: Cosmologies and histories 5LWXDO6WXGLHV0RQRJUDSK6HULHV 'XUKDP1& &DUROLQD$FDGHPLF3UHVV 7HOEDQ%RUXW 'DQLHOD9iYURYi3ODFHVDQGVSLULWVLQD6HSLNVRFLHW\The Asia 3DFL¿F-RXUQDORI$QWKURSRORJ\ Darja Hoenigman GDUMDKRHQLJPDQ#DQXHGXDX MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 215 Appendix: GLOSSARY OF KAY MENDA TERMS English Tok Pisin loan kay menda gloss bucket made of a big 44 gallon drum fotifo yomoy bamboo airplane balus QDLPWDQGRȄJD eagle-canoe SD\SPDȄJD amount of money hamas moni name of money mimbia axe tamiok mundum stone axe bag bek yambam grass basket ball bal SDSXND\PDȄJD orange tree fruit ball bal yupim wild pandanus ball balloon balun mumba bladder bank benk SD\SPDȄJD\DZD house of stones basin bikpela dis yakaopay earthen dish (large) water mixed with bark beer bia o yomba ashes (traditionally made salt) dish made of the soft part of the Arecoid big dish, boat dis PRQGDȄ palm (Rhopaloblaste sp.) petiole big sturdy bag renbo bag yambam basket book, paper, anything for buk, niuspepa NDVDȄJD dry banana leaf reading [something that] bra susu kalabus/bra isik ulakaplakay covers breasts bullet bulit tasia tamanda spirit arrow bush knife, busnaip PDOD\HȄJD\D sago machete machete buy baim e- take camera kemera television screen skrin memek lightning mirror gras candle kendol yandom endia tree sap cap kep NRSRQXȄWLD head skin chewing gum big boy / P.K. kamba endia breadfruit sap chocolate cream soklit krim HQHȄPROD diarrhoea /rotting shit cigarette (bought Spia, Pal Mal kandukya yakia white [man] cigarette in town) clothes klos kolokot things things ol samtink bikpela (save) computer NDQGHȄROXNXQMD big man / person man MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 216 English Tok Pisin loan kay menda gloss cup, mug kap palendem coconut shell carnivorous plant cup, mug kap ZDXQD (Nepenthes ampullaria sp.) eaglewood garu is kamia tree meat Eucharist yukarist pamben a kind of a nut ÀUHOLJKWHU masis pat VWLFNIRUPDNLQJÀUH ÀVKLQJKRRN huk tao sago thorn ÀVKLQJQHW net HZH\ net made of bark rope earthen ‘frying pan’ frying pan epay RUÁDWVWRQHXVHGIRU cooking sago gaol kalabus ZDQGD\\DZD chickens’ house glasses, sunglasses, (ai) glas, QRNRPJXQXȄWLD eyelids diving goggles gogols gold gol NLȄLP sand guitar gita WDVLDSXQMLPED spirit hand drum DQGDȄJXȄ\DND [something with which gum boots gam but yamblakay to] walk in the swamps gun gan WDVLDNDQXȄJD spirit bow tree species having buttress roots (buttress roots can be kicked gun gan \DPEXȄNXQGD with the heel or struck with an ax or other tool to make a gun-like booming sound) hard biscuits biskit tasia taya spirit sago house with a tin roof / haus kapa WDVLD\DZD spirit-house town house instant noodles nudols kundam en HȄ JD earthworm shit iron post ain makam main post in a house K10 ten kina isapasa stick knife naip \RPED\ NDSD\D bamboo (small knife) lamp lem yambat sago stem torch saveman/ learned man QRNRPJDSDZL red-eyed savemeri lighter / torch / masis / tos / lem tasia yamba VSLULWÀUH lamp bamboo / wooden loudspeaker spika WHSXȄ ‘loudspeaker’ marble marbol LPDȄPDȄJD tree nut medicines marasin tasia pamyamba spirit ginger MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 217 English Tok Pisin loan kay menda gloss SD\SPDȄJD money mani stone payp manda mosquito net taunam DLȄ basket for sleeping necklace neklis WRNRPERQRȄWLD neck skin oat/nut/dried fruit ?? koña taya honey(comb)-sago bar oil of native tree WRPEDWDVLD Campnosperma oil / gear oil wel / giawel tombaya brevipetiolata (TP wel diwai) outboard motor moto WDVLDPRQDȄJD spirit-paddle beetle family outboard motor moto ZDRD\PD Rhinchophoridae (sago beetle) paint (for grass/ leaves for producing pen NXQDNXPEXȄ sago) paint for kuna pencil pensil NDZD\WLxLSODND\ paint drawing pencil / biro pensil / bairo yambao ember petrol / kerosene / petrol / kerosin / yom water beer / soft drinks bia / sop drink pillow pilo tasia kumunda spirit wooden pillow plate pleit tane earthen plate policeman polis tam dog pot DȄJDV earthen pot powder milk Sunshine LVLN \RPED breast (milk) prais em i go SD\SPDȄJD price went up stones went upriver antap ZDPRȄDQ radio redio HPXȄNXQGD buttress roots radio redio \RPJRȄPDQGD turtle shell car kar rascal, bandit sanguma emay assault sorcerer seeds of Arecoid palm rice rais NDXȄZDZD\D Rhopaloblaste sp. rope rop DZDP vine/ thorns of a vine nail nil rubber gloves glav NRORQRȄWLD hand skin salt sol tasia oua spirit ‘salt’ halved coconut shell serving spoon kumu spun ipikapa IRUSUHVVLQJVDJRÁRXU bamboo tongs (for serving tongs ?? kula holding hot items or ritual use) MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Mountain talk to hidden talk: Awiakay registers 218 English Tok Pisin loan kay menda gloss big dish dish made of the soft part of ship sip PRQGDȄNDQGHȄJH the Arecoid palm (Rhopaloblaste sp.) petiole shoes su panben tia leg skin shovel sawel VLȄJD\DȄ ?? map kulamba water from ground shower sawa \RPEDSD\S hole / kulamba yomba ~ stone shelter soap sop tomba tree (oil) body spray bodi sprei soap sop \RPHȄD\ water spit soap sop yom karay water foam spoon made of coconut spoon spun kap shell store stoa NRORNRW\DZD house of things string string pipisimba pandanus string sugar suga sugar cane/spirit sugar LPDWWDVLDLPDWD lollies, candies loli cane sunglesses sanglas WHPQRNRPJD sun-eye tabernacle tabernakol yao house spirit JDUDPXW (TP for telephone telipon WDVLDXPEXȄJD ‘slit-drum’) tin roof kapa ZDNȄJD sago thatch shingles tinned (mushroom) ?? PHȄJZDN vomit sauce toilet toilet HQHȄ\DZD shit house trousers trausis kumbayn tia tree bark skin ZDVLSLWLD [!ZDLSLD ‘part trousers trausis string bag skin of a torn string bag’] T-shirt singlis RPXQXȄWLD body skin umbrella ambrela D\ȄJZDȄWLD VNLQRIDÁ\LQJIR[ umbrella ambrela embum grass hood FRFRQXWÀEUHIRU wallet hanpaus kundambi storing tobacco watch hanwas WHPPDȄJD sun/time ‘fruit’ write raitim ikak- carve MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY /DQJXDJH'RFXPHQWDWLRQ &RQVHUYDWLRQ6SHFLDO3XEOLFDWLRQ1R 'HFHPEHU Melanesian languages on the edge of Asia: Challenges for the 21st century, HGE\1LFKRODV(YDQVDQG0DULDQ.ODPHUSS ± KWWS QÀUFKDZDLLHGXOGFVS 9 KWWS KGOKDQGOHQHW Cross-cultural differences in representations and routines for H[DFWQXPEHU Michael C. Frank Stanford University The relationship between language and thought has been a focus of SHUVLVWHQW LQWHUHVW DQG FRQWURYHUV\ LQ FRJQLWLYH VFLHQFH$OWKRXJK debates about this issue have occurred in many domains, number is an ideal case study of this relationship because the details (and even the existence) of exact numeral systems vary widely across languages DQGFXOWXUHV,QWKLVDUWLFOH,GHVFULEHKRZFURVVOLQJXLVWLFDQGFURVV cultural diversity—in Amazonia, Melanesia, and around the world— JLYHVXVLQVLJKWLQWRKRZV\VWHPVIRUUHSUHVHQWLQJH[DFWTXDQWLWLHV DIIHFWVSHDNHUV¶QXPHULFDOFRJQLWLRQ7KLVERG\RIHYLGHQFHVXSSRUWV the perspective that numerals provide representations for storing and PDQLSXODWLQJTXDQWLW\LQIRUPDWLRQ,QDGGLWLRQWKHGLIIHULQJVWUXFWXUH RITXDQWLW\UHSUHVHQWDWLRQVDFURVVFXOWXUHVFDQOHDGWRWKHLQYHQWLRQ of widely varied routines for numerical tasks like enumeration and DULWKPHWLF 1. INTRODUCTION.1 The relationship between language and thought is one of the most IDVFLQDWLQJ²DQG WKH PRVW FRQWURYHUVLDO²WRSLFV LQ FRJQLWLYH VFLHQFH 3RVHG E\ :KRUI WKHTXHVWLRQRIZKHWKHUFURVVOLQJXLVWLFGLIIHUHQFHVOHDGWRGLIIHUHQFHVLQFRJQLWLRQ KDVEHHQVWXGLHGH[WHQVLYHO\DFURVVDZLGHUDQJHRIGRPDLQV5HFHQWZRUNRQWKLVTXHVWLRQ KDVFRPHIURPFRORUSHUFHSWLRQ .D\%HUOLQ0DI¿ 0HUUL¿HOG:LQDZHUHWDO 5REHUVRQ +HQOH\ QDYLJDWLRQDQGVSDWLDOODQJXDJH +HUPHU 6SHONH /HYLQVRQ.LWD+DXQ 5DVFK WKHRU\RIPLQG 3\HUV 6HQJKDV JHQGHU 1 7KDQNV WR 0DUN 'RQRKXH IRU HQFRXUDJLQJ PH WR YLVLW 0DQRNZDUL ,QGRQHVLD , JUDWHIXOO\ acknowledge all of my collaborators in the work reported here, including Ted Gibson, Evelina )HGRUHQNR5HEHFFD6D[H'DQ(YHUHWWDQG'DYLG%DUQHU7KDQNVDOVRWR6XVDQ&DUH\DQG/HUD %RURGLWVN\ IRU YDOXDEOH GLVFXVVLRQ RI WKH WKHRUHWLFDO LGHDV SUHVHQWHG KHUH )LQDOO\ WKDQNV WR David Barner, Nick Evans, Ev Fedorenko, Ted Gibson, and two anonymous reviewers for giving FRPPHQWVRQDSUHYLRXVYHUVLRQRIWKLVPDQXVFULSW cc Creative Commons Attribution Non-Commercial No Derivatives Licence Cross-cultural differences ~ exact number %RURGLWVN\6FKPLGW 3KLOOLSV HYHQWSHUFHSWLRQ 3DSDIUDJRX+XOEHUW 7UXHVZHOO )DXVH\ %RURGLWVN\ REMHFWLQGLYLGXDWLRQ /XF\ %DUQHU/L 6QHGHNHU FDWHJRUL]DWLRQ /XS\DQ5DNLVRQ 0F&OHOODQG DQGPDQ\RWKHUV The term “number” is generally ambiguous between grammatical markings like singular/plural DQG QXPHUDOV WKDW GHVFULEH WKH H[DFW FDUGLQDOLW\ RI VHWV +HUH , ZLOO DYRLG WKH FXPEHUVRPH language necessary to disambiguate in every instance and use the terms “numbers” and “numerical cognition” under the assumption that these terms refer to numerals representing the exact cardinalities of large sets and the broad range of cognitive operations that are carried out ZLWKVXFKVHWVUHVSHFWLYHO\ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number )HGRUHQNR/DL6D[H *LEVRQ WKDWQXPHULFDOUHSUHVHQWDWLRQVDUHFXOWXUDODUWLIDFWV WKDWDUHXVHGIRUWKHRQOLQHHQFRGLQJRITXDQWLW\LQIRUPDWLRQ7KHIRUPRIDOLQJXLVWLFRU FXOWXUDOUHSUHVHQWDWLRQRIQXPEHUDQGWKHHI¿FLHQF\RIWKHURXWLQHVIRUPDQLSXODWLQJWKLV representation each affect what computations are possible using this representation; the online availability of this representation (in the moment a computation is desired) is a SUHUHTXLVLWHIRUSHUIRUPLQJWKHFRPSXWDWLRQ2QHYHUVLRQRIWKLVYLHZZDV¿UVWDUWLFXODWHG E\.D\DQG.HPSWRQ DQGLWDQGLWVYDULDQWVDUHFXUUHQWO\H[SHULHQFLQJDUHVXUJHQFH LQ FRJQLWLYH VFLHQFH 'HVVDOHJQ /DQGDX *HQWQHU :LHVH VHH HJ )UDQNHWDO IRUPRUHGHWDLOHGGLVFXVVLRQ $ VHFRQGDU\ JRDO RI WKLV UHYLHZ LV WR DUJXH IRU DQ DSSURDFK ZKHUHE\ ¿HOGZRUNHUV VXSSOHPHQWVWDQGDUGHOLFLWDWLRQWHFKQLTXHVZLWKSV\FKRORJLFDOH[SHULPHQWDWLRQWKDWWHVWVWKH FRJQLWLYHFRQVHTXHQFHVRIGLIIHUHQWQXPHULFDOUHSUHVHQWDWLRQVDQGURXWLQHV%HFDXVHRIWKH immense linguistic and cultural diversity in regions like Amazonia and Melanesia and the relative isolation of these populations, investigation of numerical systems in these regions’ indigenous cultures provides especially rich evidence regarding the range of variation LQQXPEHUV\VWHPV0HODQHVLDLQSDUWLFXODULVOLNHO\WRKDUERXUWKHJUHDWHVWGLYHUVLW\RI QXPEHUV\VWHPVLQWKHZRUOG /HDQ (WKQRJUDSKLFREVHUYDWLRQDQGSV\FKRORJLFDO observation can play complementary roles in characterizing this diversity, providing both QDWXUDOLVWLFREVHUYDWLRQVDQGSUHFLVHDQGJHQHUDOL]DEOHPHDVXUHPHQWV$QGJLYHQWKHUDSLG GHFUHDVHVLQOLQJXLVWLFGLYHUVLW\LQWKHVHUHJLRQV (YDQV D LWLVHVSHFLDOO\LPSRUWDQW to document not only the facts of languages in Amazonia and Melanesia, but also the SV\FKRORJLFDOFRQVHTXHQFHVRIWKHVHODQJXDJHVIRUWKHLUVSHDNHUV 2. REPRESENTATIONS AND ROUTINES FOR NUMBER. The past twenty years have seen an explosion of interest in representations of exact number as an example of an important, XQLTXHO\KXPDQFRQFHSW\HWRQHWKDWLVEXLOWRXWRISULPLWLYHFRPSRQHQWVWKDWFDQHDFKEH REVHUYHGLQLQIDQWVDQGPHPEHUVRIRWKHUVSHFLHV 'HKDHQH &DUH\ 2QWKHRQH hand, numbers are a key part of every modern society: they facilitate a huge set of human EHKDYLRUVIURPFRPSOH[IHDWVRIHQJLQHHULQJWRHFRQRPLFH[FKDQJHVXVLQJFXUUHQF\2Q WKHRWKHUUHSUHVHQWDWLRQVRITXDQWLW\LQIRUPDWLRQFDQEHREVHUYHGLQLQIDQWVPRQNH\V¿VK DQGDKRVWRIRWKHUFUHDWXUHV *DOOLVWHO ;X 6SHONH+DXVHUHWDO 7KXV in the domain of number, cognitive scientists can ask how basic cognitive abilities can be combined into a sophisticated conceptual system and, in particular, what role language SOD\VLQWKLVFRPELQDWLRQ 7KHEDVLFFRJQLWLYHV\VWHPVWKDWSURYLGHQRQYHUEDOUHSUHVHQWDWLRQVRITXDQWLW\DUH QRZZHOOHVWDEOLVKHG )HLJHQVRQ'HKDHQH 6SHONH 7KH¿UVWLVDV\VWHPWKDWFDQ track the location and identity of up to three or four objects at a time, likely based in visual DWWHQWLRQRUWUDFNLQJ7KHVHFRQGLVWKHDSSUR[LPDWHQXPEHUV\VWHP $16 ZKLFKFDQ represent the approximate magnitude of sets of objects but not the identities of individuals ZLWKLQWKHVHVHWV'HVSLWHWKHSUHVHQFHRIERWKRIWKHVHV\VWHPVLQSUHOLQJXLVWLFLQIDQWV OHDUQLQJKRZWRXVHOLQJXLVWLFQXPHUDOVLVDSURWUDFWHGSURFHVV,QW\SLFDOO\GHYHORSLQJ English-speaking children, the time period from learning the meaning of “one” to mastering WKHXVHRIQXPEHUZRUGVXSWR³WHQ´FDQODVWD\HDURUPRUH :\QQ Despite consensus about the basic facts, the role of language is contested in both this GHYHORSPHQWDOSURJUHVVLRQDQGLWVHQGUHVXOW2QWKH³ERRWVWUDSSLQJ´DFFRXQWOHDUQLQJ WKHPHDQLQJVRIQXPHUDOVLQWKHFRXQWOLVWLVDUHVXOWRI¿UVWPDSSLQJQXPEHUZRUGVIURP MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number “one” up to “three” or “four” onto small number representations, and then performing an LQGXFWLYHVWHSWKDWUHFRJQL]HVWKHSDUDOOHOEHWZHHQWKHVHTXHQWLDOUHODWLRQVKLSEHWZHHQWKH ZRUGVLQWKHFRXQWOLVWDQGWKHVHTXHQWLDOUHODWLRQVKLSLQKHUHQWLQWKHLUGH¿QLWLRQV7KH VSHFL¿FVRIODQJXDJH²ERWKLQWKHVWUXFWXUHRIWKHFRXQWOLVWDQGLQWKHXVHRIQXPEHU QDPHVDVSODFHKROGHUVIRUFRQFHSWV²SOD\DQHVVHQWLDOUROHLQWKLVDFFRXQW &DUH\ 3LDQWDGRVL7HQHQEDXP *RRGPDQ ,QFRQWUDVWWKH³PDSSLQJ´YLHZVXJJHVWVWKDW ZRUGVOLNH³IRXU´RU³VHYHQ´DUHGH¿QHGLQWHUPVRILQQDWHQXPEHUFRQFHSWVDQGLGHQWL¿HG HLWKHUQRLVLO\XVLQJWKH$16RUSUHFLVHO\XVLQJDFRXQWURXWLQH2QWKLVNLQGRIDFFRXQW language plays a peripheral role: it does not help to create new concepts, it simply helps to name and recognize pre-existing concepts by using enumeration routines like counting *HOPDQ *DOOLVWHO One broad area of agreement between these views, however, is the distinction between numerical representations and numerical routines, and the importance of their interaction LQDOORZLQJWKHLUXVHUVWRVWRUHDQGPDQLSXODWHH[DFWTXDQWLWLHV *HOPDQ %XWWHUZRUWK &DUH\ %\QXPHULFDOUHSUHVHQWDWLRQ,PHDQKHUHDVHWRIV\PEROVXVHGIRUWKH WDVNRIUHSUHVHQWLQJH[DFWTXDQWLWLHV7KHFKRLFHRIDUHSUHVHQWDWLRQRIQXPEHULQFOXGHV the medium of representation (linguistic, like a count list; externalized, like a counting stick; or even supported by visual imagery, like a mental abacus representation) and the LQWHUQDO VWUXFWXUH RI WKHVH UHSUHVHQWDWLRQV HJ WKDW (QJOLVK VSHDNHUV VD\ ³QLQHW\QLQH´ WRPHDQ ZKLOH)UHQFKVSHDNHUVVD\³TXDWUHYLQJWGL[QHXI´ %\QXPHULFDOURXWLQH,PHDQDQDOJRULWKPWKDWLVFRPPRQO\XVHGWROHYHUDJHVXFK DUHSUHVHQWDWLRQLQDSDUWLFXODUQXPHULFDOWDVN([DPSOHVRIURXWLQHVUDQJHIURPVLPSOH enumeration to the complex sets of steps that schoolchildren are taught to follow in order WRSHUIRUPDGGLWLRQRUGLYLVLRQRIODUJHTXDQWLWLHV 3. NUMERICAL ABILITIES WITHOUT REPRESENTATIONS OF EXACT NUMBER. What is numerical cognition like in the absence of linguistic numerals in a language?3 Are there DQ\URXWLQHVIRUPDQLSXODWLRQRIH[DFWTXDQWLW\WKDWDUHSRVVLEOHLQWKHDEVHQFHRIH[DFW numerical representations? This section reviews recent work with the Mundurukú and 3LUDKm WZR LQGLJHQRXV JURXSV LQ %UD]LO WKDW H[SORUHV WKH FRJQLWLYH FRQVHTXHQFHV RI VSHDNLQJDODQJXDJHZLWKOLPLWHGRUQRYRFDEXODU\IRUH[DFWTXDQWLWLHV 3.1. MEASURING NUMBER VOCABULARY. *RUGRQ FODLPHG WKDW 3LUDKm KDG D FRXQWLQJV\VWHPFRQVLVWLQJRIZRUGVIRUWKHTXDQWLWLHV hói DQG hoí) as well as a word for “many” (aibaagi He reported data from only a single elicitation (in which a speaker 3 7KHTXHVWLRQRIZKDWLWPHDQVWRKDYHH[DFWQXPHUDOVLQDODQJXDJHLVDPELJXRXV DQLQGLYLGXDO speaker can in principle have access to a particular, idiosyncratic mapping between symbols DQGTXDQWLWLHV RUDPDSSLQJFDQEHFRQYHQWLRQDOL]HGDQGDYDLODEOHWRPDQ\RUDOOVSHDNHUVRI D ODQJXDJH$OWKRXJK WKHUH are cases of idiosyncratic or heterogeneous number systems (for SUHOLPLQDU\GDWDRQWKLVLVVXHVHHHJ)UDQN +RQH\PDQ WKHH[DPSOHVGLVFXVVHGKHUHDOO show relatively broad consensus across speakers, shown via experimental procedures used with a VDPSOHRILQGLYLGXDOVIURPWKHFRPPXQLW\ Here and throughout the article I will use the Arabic numerals as a shorthand for the expression ³WKH TXDQWLW\ 1´ UHJDUGOHVV RI ZKHWKHU WKH TXDQWLW\ LV ODUJH RU VPDOO UDWKHU WKDQ IROORZLQJ standard W\SRJUDSKLFDOFRQYHQWLRQV ³RQH´YV GHSHQGLQJRQTXDQWLW\,ZLOOTXRWHQXPEHUV OLNH³VHYHQ´WRUHIHUWRDZRUGIRUDTXDQWLW\ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number used the “two” word hoí WRUHIHUWRWKHTXDQWLWLHVDQG 7KHVHGDWDZHUHEURDGO\LQ accordance with a description of Pirahã as a “one, two, many” language, a type found in RWKHUQRQLQGXVWULDOL]HGVRFLHWLHV 0HQQLQJHU +DPPDUVWU|P ,QWKHLUZRUNRQ0XQGXUXN~3LFD/HPHU,]DUGDQG'HKDHQH SHUIRUPHGD VWUXFWXUHGHOLFLWDWLRQH[SHULPHQW7KH\SUHVHQWHGVHWVRI±GRWVLQUDQGRPRUGHUWRDGXOWV DQGFKLOGUHQDQGDVNHGKRZPDQ\GRWVZHUHSUHVHQWLQHDFKVHW0XQGXUXN~SDUWLFLSDQWV UHVSRQGHGFRQVLVWHQWO\ZLWKDVHWRIFRQYHQWLRQDOL]HGWHUPVIRUWKHTXDQWLWLHV±7KHVH WHUPVZHUHXVHGE\SDUWLFLSDQWVLQQHDUO\DOOFDVHV)RUSDUWLFLSDQWVXVHGDFRQYHQWLRQDO WHUPDOPRVWDVRIWHQEXWRFFDVLRQDOO\XVHGWKHVDPHWHUPWRUHIHUWRDQG)RURI SDUWLFLSDQWVXVHGDWHUPPHDQLQJ³RQHKDQG´ZKLOHRIRWKHUVXVHGDYDJXHUWHUPWKDW 3LFDDQGFROOHDJXHVWUDQVODWHGDV³VRPHQRWPDQ\´DQGWKDWZDVXVHGIRURWKHUTXDQWLWLHV ±DVZHOO$ERYHRQO\WKLVODWWHUWHUPDQGDWHUPPHDQLQJ³PDQ\´ZHUHXVHGZLWK DQ\ IUHTXHQF\7KLV H[SHULPHQW JLYHV HYLGHQFH WKDW 0XQGXUXN~ GRHV KDYH VRPH H[DFW numerals, but lacks a recursive number naming system and exact number vocabulary for ODUJHTXDQWLWLHV5 )ROORZLQJRQ*RUGRQ RXURZQZRUNUHYHDOHGDGLIIHUHQWYLHZRI3LUDKmTXDQWLW\ YRFDEXODU\XVLQJDVWUXFWXUHGHOLFLWDWLRQWDVNOLNH3LFDHWDO :HVKRZHGSDUWLFLSDQWV sets of objects and asked “how much/many are there?”, increasing the cardinalities of the VHWIURP±DQGWKHQGHFUHDVLQJIURP± RUYLFHYHUVD :HIRXQGWKDWWKHTXDQWLWLHV for which our participants used particular words changed depending on the context of the HOLFLWDWLRQ LQFUHDVLQJYVGHFUHDVLQJ ,QSDUWLFXODUDOWKRXJKSDUWLFLSDQWVXVHGhói only IRULQWKHLQFUHDVLQJFRQWH[WWKH\XVHGLWIRUXSWRREMHFWVLQWKHGHFUHDVLQJHOLFLWDWLRQ This context effect strongly suggests that hói LVQRWDZRUGIRU2QRXUYLHZWKHPRVW likely conclusion from these data is that it is a relative term like “few,” “fewer,” or even ³VPDOO´$QRWKHU SRVVLEOH SRVLWLRQ KRZHYHU LV WKDW hói is polysemous between “one” and “a few”; this view is of course logically possible, but provides no account of why or XQGHUZKDWFRQGLWLRQVDQH[DFWPHDQLQJZRXOGEHDYDLODEOH7KHWKUHHZRUGVGRFXPHQWHG E\*RUGRQDUHFRQ¿UPHGE\VHYHUDOQRQQDWLYH3LUDKmVSHDNHUVWREHWKHRQO\ZRUGVIRU TXDQWLWLHVOHDGLQJXVWRFRQFOXGHWKDW3LUDKmVHHPVWRKDYHQR XQDPELJXRXV ZRUGVIRU H[DFWQXPEHUV QRWHYHQDZRUGIRU 7KH$PD]RQLDQ ¿QGLQJV VXJJHVW WKDW UHSUHVHQWDWLRQV RI H[DFW TXDQWLWLHV DUH QRW D OLQJXLVWLFXQLYHUVDO,QDGGLWLRQWKH\UDLVHWKHLQWULJXLQJTXHVWLRQRIZKHWKHUDQ\RWKHU ODQJXDJHVZLWKRXWQXPHUDOVKDYHEHHQPLVFODVVL¿HGDV³RQHWZRPDQ\´ODQJXDJHVGXH WR WKH DEVHQFH RI H[SHULPHQWDO GDWD6 In order to determine the semantics of possible numerals, single-participant elicitations should be replaced with structured elicitations and QXPHUDO FRPSUHKHQVLRQ WDVNV :\QQ (YHQ GDWD IRU D KDQGIXO RI SDUWLFLSDQWV LQ 5 Note that IRU GHYHORSPHQWDO UHVHDUFKHUV WKH JROG VWDQGDUG IRU FKLOGUHQ KDYLQJ DFTXLUHG WKH meaning of a numeral for 7 is success in comprehension-based tasks like “give a number” (Wynn /H&RUUHHWDO&RQGU\ 6SHONH ,QWKH³JLYHDQXPEHU´WDVNSDUWLFLSDQWVDUH VLPSO\DVNHGWR³JLYHPH1REMHFWV´DQGWKHFDUGLQDOLW\RIWKHVHWWKH\JLYHLVUHSRUWHG1HLWKHU the Mundurukú nor the Pirahã have been tested on such a task, so more work remains to be done WRSUREHWKHPHDQLQJVRIWKHDWWHVWHGYRFDEXODU\LWHPV 6 +DPPDUVWU|P JLYHV D OLVW RI RWKHU ODQJXDJHV WKDW KDYH VXFK V\VWHPV DQG QRWHV WKLV possibility, though Pirahã may be the only one of these that lacks any singular-plural marking as ZHOO MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number VXFKWDVNVFDQEHLQIRUPDWLYHDQGFDQSURYLGHDQLQH[SHQVLYHVXSSOHPHQWWRFXUUHQW¿HOG PHWKRGV 3.2 CONSEQUENCES OF LIMITED NUMBER VOCABULARY. In contrast with linguistic representations of number, which vary across societies, a large body of evidence shows that an approximate number sense (ANS) is available to all human beings as well as members RIRWKHUVSHFLHV7KLVDSSUR[LPDWHVHQVHOHDGVXVWREHDEOHWRPDNHHVWLPDWHVRIDVHW¶V TXDQWLWLHVZLWKRXWXVLQJDQHQXPHUDWLRQURXWLQH The ANS has been characterized extensively in human and non-human animals (for UHYLHZVHH)HLJHQVRQHWDO*DOOLVWHO (VWLPDWHVRITXDQWLW\PDGHE\WKH$16 IROORZ:HEHU¶VODZ HJ:KDOHQ*DOOLVWHO *HOPDQ ;X ZKLFKVWDWHVWKDW the probability of a correct response in a discrimination task is related to the magnitude RIWKHVWLPXOXVEHLQJGLVFULPLQDWHG:HEHU¶VODZOHDGVWRWKHSUHGLFWLRQRIWKHUHODWLRQ ı c in participants’ data, where and ıare the mean and standard deviation of the magnitude estimates (across trials or participants) and c is a constant holding across a UDQJHRIPDJQLWXGHV7KHWHUPc is often referred to as the FRHI¿FLHQWRIYDULDWLRQRU&29 $FRQVWDQW&29LPSOLHVWKDWWKHODUJHUWKHTXDQWLW\EHLQJHVWLPDWHGWKHODUJHUWKHDYHUDJH HUURULQWXUQVLJQDOLQJWKDWWKH$16LVEHLQJXVHG ,Q 3LFD HW DO¶V VWXG\ 0XQGXUXN~ SDUWLFLSDQWV DQG )UHQFK FRQWUROV SHUIRUPHG FRPSDULVRQDGGLWLRQDQGVXEWUDFWLRQWDVNV:KHQSDUWLFLSDQWVZHUHDVNHGWRFKRRVHWKH larger of two large sets of dots (and were not given enough time to count), both groups SHUIRUPHG VLPLODUO\ VKRZLQJ D FRQVWDQW &29FRQVLVWHQW ZLWK :HEHU¶V ODZ +RZHYHU ZKHQ SDUWLFLSDQWV ZHUH DVNHG WR JLYH WKH UHVXOWLQJ TXDQWLW\ LQ D VXEWUDFWLRQ SDUDGLJP ZKHUHREMHFWVZHUH¿UVWDGGHGWRDQGWKHQVXEWUDFWHGIURPDQRSDTXHFRQWDLQHU)UHQFK participants performed nearly perfectly, while the Mundurukú made errors that were again FRQVLVWHQWZLWKWKHRSHUDWLRQRIWKH$16&UXFLDOO\WKHGHVLJQRIWKLVWDVNUHTXLUHGRQO\ UHVSRQVHVLQWKHUDQJHZKHUHWKH0XQGXUXN~FRXOGKDYHUHVSRQGHGYHUEDOO\ TXDQWLWLHV ± UXOLQJ RXW WKH H[SODQDWLRQ WKDW WKH\ FRXOG QRW LQGLFDWH WKH FRUUHFW UHVSRQVH HYHQ WKRXJKWKH\NQHZLW /LNHWKH0XQGXUXN~WKH3LUDKmDOVRUHOLHGRQWKH$16WRSHUIRUPQXPHULFDOWDVNV *RUGRQ SHUIRUPHG D UDQJH RI PDWFKLQJ WDVNV GHVLJQHG WR SUREH WKH DELOLW\ RI SDUWLFLSDQWVWRVWRUHDQGPDQLSXODWHH[DFWTXDQWLWLHV,QWKHVLPSOHVWWDVNSDUWLFLSDQWVZHUH DVNHGWRSURGXFHD±PDWFKEHWZHHQWZRVHWVE\VHOHFWLQJWKHFRUUHFWTXDQWLW\RIREMHFWV WRDOLJQZLWKDWDUJHWVHW,QPRUHGLI¿FXOWWDVNVWKHWDUJHWVHWZDVSUHVHQWHGLQDFOXVWHURU ZDVSUHVHQWHGRQO\EULHÀ\DQGSDUWLFLSDQWVZHUHDJDLQDVNHGWRSURGXFHDWDUJHWVHWRIWKH VDPHFDUGLQDOLW\3DUWLFLSDQWVPDGHHUURUVLQDOOWDVNVHYHQWKH±PDWFKWDVNDOWKRXJK their errors were larger in those tasks where the target set was presented for a short period RIWLPH:KHQ*RUGRQFRQVROLGDWHGGDWDDFURVVDOOWDVNVWKHSDWWHUQRIUHVSRQVHVDJDLQ VKRZHGDFRQVWDQW&29/LNHWKH0XQGXUXN~UHVXOWVWKHVH¿QGLQJVVXJJHVWWKDWDQDORJ HVWLPDWLRQXVLQJWKH$16LVWKHGHIDXOWVWUDWHJ\LQVLWXDWLRQVZKHUHQRFRXQWOLVWLVDYDLODEOH %RWKVHWVRIUHVXOWVOHIWRSHQDQLPSRUWDQWTXHVWLRQKRZHYHU GLG0XQGXUXN~DQG 3LUDKmSDUWLFLSDQWVXQGHUVWDQGWKDWODUJHTXDQWLWLHVcould be exact, even if they did not NQRZKRZWRH[SUHVVRUPDQLSXODWHWKHP")RUH[DPSOH*RUGRQ¶V±PDWFKLQJWDVNZDV WKHVLPSOHVWWDVNLQHLWKHUDVVHVVPHQW\HW3LUDKmVWLOOPDGHHUURUV:HUHWKHVHHUURUVGXH WRFRQIXVLRQDERXWZKDWZDVEHLQJDVNHGRUGLI¿FXOWLHVLQFRPSOHWLQJWKHWDVNRUZHUH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number WKH\LQVWHDGGXHWRDPRUHIXQGDPHQWDOFRQFHSWXDOGLIIHUHQFH"2QWKH¿UVWLQWHUSUHWDWLRQ WKH 3LUDKm PDGH HUURUV LQ PDWFKLQJ XS ODUJHU TXDQWLWLHV RI REMHFWV HLWKHU EHFDXVH WKH\ did not understand that an exact response was called for (even though they could have produced such a response) or because they made manual errors in alignment even though WKH\XQGHUVWRRGZKDWZDVEHLQJDVNHGRIWKHP2QWKHVHFRQGLQWHUSUHWDWLRQKRZHYHUWKH 3LUDKmGLGQRWXQGHUVWDQGWKDWDFRUUHFWUHVSRQVHUHTXLUHGPDWFKLQJH[DFWO\EHFDXVHWKH\ GLGQRWHYHQKDYHDYDLODEOHDFRQFHSWRIH[DFWHTXLYDOHQFH 7KHDFWXDOFRPSXWDWLRQDOGHPDQGVIRUVXFFHVVLQWKH±PDWFKLQJWDVNDUHTXLWHORZ In order to succeed, it is only necessary to match individuals until there are no more left WRPDWFK7KLVWDVNFDQEHDFFRPSOLVKHGZLWKRXWHYHUUHSUHVHQWLQJWKHWRWDOTXDQWLW\VR VXFFHVVLQWKHWDVNGRHVQRWGHPRQVWUDWHWKHH[LVWHQFHRIH[DFWTXDQWLW\UHSUHVHQWDWLRQV$ ±PDWFKRIH[DFWO\LWHPVFDQEHSHUIRUPHGZLWKRXWHYHUPHQWDOO\UHSUHVHQWLQJ2Q WKHRWKHUKDQGDWUXHIDLOXUHLQWKHWDVN²DQLQDELOLW\WRVHOHFWWKH±PDWFKLQJDOJRULWKP even with appropriate training and unlimited time—would suggest that the Pirahã truly did QRWWKLQNLQWHUPVRIH[DFWHTXLYDOHQFHRUH[DFWPDWFKHV On a recent visit to the Pirahã, my collaborators and I replicated a number of Gordon’s WDVNVZLWKDODUJHUVDPSOHRISDUWLFLSDQWV 1 DVRSSRVHGWR1 LQWKHSUHYLRXVVWXG\ In order to ensure task understanding, we included a systematic training phase in which we GHPRQVWUDWHGZKDWWKHFRUUHFWUHVSRQVHZRXOGEHIRURQHWULDOZLWKDVPDOOTXDQWLW\DQG WKHQJDYHFRUUHFWLYHIHHGEDFNRQDQRWKHUVHWRIVPDOOTXDQWLW\WULDOVXQWLOSDUWLFLSDQWVZHUH SHUIRUPLQJFRQVLVWHQWO\ )UDQNHWDO ,QWKHPRUHGLI¿FXOWPDWFKLQJWDVNVZHIRXQG precisely the pattern of ANS usage that Gordon documented, with errors increasing along ZLWKWKHTXDQWLW\RIREMHFWVEHLQJHVWLPDWHG VHH¿JXUHIRUDQH[DPSOHRIWKHWHVWLQJ HQYLURQPHQW 2XU UHVXOWV GLIIHUHG IURP *RUGRQ¶V LQ WKH ± PDWFKLQJ WDVN KRZHYHU There, only one participant made any errors and the rest performed perfectly, suggesting WKDW WKLV WDVN ZDV TXDOLWDWLYHO\ GLIIHUHQW IURP WKH RWKHUV 'HVSLWH QRW KDYLQJ OLQJXLVWLF UHSUHVHQWDWLRQVRIH[DFWTXDQWLWLHVDYDLODEOHWRWKHPWKLVJURXSRI3LUDKmXQGHUVWRRGWKDW DQH[DFWUHVSRQVHZDVUHTXLUHG7KLVUHVXOWVKRZVWKDWRXUSDUWLFLSDQWVPDGHWKHDSSURSULDWH generalization from a few training examples with small numbers: that every target item should be matched with exactly RQHLWHPQRWWKDWWKHWZRVHWVVKRXOGPDWFKDSSUR[LPDWHO\ That they made this generalization consistently across individuals strongly suggests that WKHQRWLRQRIDQH[DFWUDWKHUWKDQDSSUR[LPDWH±PDWFKZDVDYDLODEOHWRWKHP WKRXJK DJDLQQRWWKHUHSUHVHQWDWLRQRIDSDUWLFXODUH[DFWTXDQWLW\OLNH 2QH¿QDOGDWDVHWEHDUVRQWKLVTXHVWLRQKRZHYHU(YHUHWWDQG0DGRUD FRQGXFWHG DUHSOLFDWLRQRIRXUSUHYLRXVZRUNZLWKDQRWKHUJURXSRI3LUDKmIURPDGLIIHUHQWYLOODJH Although they again replicated the pattern of ANS usage on more complex matching tasks, they found results congruent with Gordon’s: their participants made systematic errors RQWKH±PDWFKLQJWDVNV(YHUHWWDQG0DGRUDDUJXHGWKDWWKHVXFFHVVRIWKHSDUWLFXODU SDUWLFLSDQWVLQRXUH[SHULPHQWVZDVGXHWRH[SRVXUHWKDWPHPEHUVRIWKLVYLOODJHKDG WRLQQRYDWHGQXPEHUZRUGVDQGQXPHULFDOSURFHGXUHV$SSDUHQWO\0DGRUDKDGFRQGXFWHG numerical training sessions with the members of this village; nevertheless, our elicitation WDVNVVKRZHGQRHYLGHQFHIRUNQRZOHGJHRILQQRYDWHGQXPEHUZRUGV7KLVFODLPEULQJV up an interesting possibility: could it be that exposure to some representations of exact number— even without the long-term adoption of these representations—facilitates the FRQVWUXFWLRQRID±PDWFKVWUDWHJ\"$OWKRXJKWKHFXUUHQWGDWDGRQRWSURYLGHHQRXJK MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number information to evaluate this claim, perhaps it can be assessed via future developmental or FURVVFXOWXUDOZRUN FIGURE 1.$3LUDKmSDUWLFLSDQWLQ)UDQNHWDO LQWKHRUWKRJRQDOPDWFK FRQGLWLRQ7KHH[SHULPHQWHUVKDYHSODFHGVSRROVRIWKUHDGDQGWKHSDUWLFLSDQW KDVPDWFKHGWKHPZLWK EDOORRQV To summarize, evidence from the Pirahã and Mundurukú demonstrates that in cultures ZLWKRXW UHSUHVHQWDWLRQV RI ODUJH H[DFW TXDQWLWLHV LQGLYLGXDOV DUH QRW DEOH WR UHPHPEHU RU PDQLSXODWH VXFK TXDQWLWLHV H[DFWO\ VXJJHVWLQJ D FRQQHFWLRQ EHWZHHQ OLQJXLVWLF UHSUHVHQWDWLRQVDQGWKHDELOLW\WRFUHDWHURXWLQHVIRUPDQLSXODWLQJH[DFWQXPEHU,QVWHDGRI UHPHPEHULQJH[DFWTXDQWLWLHVERWKJURXSVXVHGDQHVWLPDWLRQVWUDWHJ\ZKLFKDOORZHGIRU DSSUR[LPDWHO\FRUUHFWUHVSRQVHVHYHQLQUHODWLYHO\GLI¿FXOWWDVNV1HYHUWKHOHVVHYLGHQFH IURPWKH3LUDKmVXJJHVWVWKDWLWLVSRVVLEOHWRFUHDWHDQGXVHDURXWLQHIRUH[DFW±PDWFK HYHQZLWKRXWDQXQDPELJXRXVOLQJXLVWLFUHSUHVHQWDWLRQRI 4. DISTINGUISHING COGNITION FROM CULTURAL EXPOSURE IN NUMBER REPRESENTATION. 7KHHYLGHQFHDERYHVXJJHVWVWKDWURXWLQHVIRUVWRULQJDQGPDQLSXODWLQJH[DFWTXDQWLWLHV correlate with the cultural presence of linguistic representations of number, but the precise QDWXUHRIWKLVFRUUHODWLRQLVXQNQRZQ2QHSRVVLELOLW\LVWKDWODQJXDJHIRUQXPEHUFRXOG simply co-occur with cultural routines for number, rather than being a causal factor in the FRJQLWLRQ RI LQGLYLGXDO VSHDNHUV 2Q WKLV NLQG RI DFFRXQW ODQJXDJH IRU QXPEHU ZRXOG GHYHORSDORQJVLGHDVHWRI SRVVLEO\QRQYHUEDO URXWLQHVIRUPDQLSXODWLQJH[DFWTXDQWLWLHV VSULQJLQJIURPWKHVDPHEDVLFFXOWXUDOQHHGV6SHDNHUVZRXOGOHDUQQXPEHUZRUGVEXW WKH\ZRXOGDOVROHDUQDOJRULWKPVIRUGRLQJPDWFKLQJWDVNVIRUFKXQNLQJODUJHTXDQWLWLHV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number LQWRVHWVRIVPDOOHUTXDQWLWLHVDQGIRUWDOO\LQJWRNHHSWUDFNRITXDQWLWLHVRYHUWLPH)RU example, the use of an abacus would constitute a parallel, non-linguistic routine that could VXSSRUWQXPHULFDOFDOFXODWLRQ VHHEHORZIRUPRUHGHWDLOV 2QWKHRWKHUKDQGDQRWKHU possibility is that language for number could be necessary in the moment for the precise PDQLSXODWLRQRIH[DFWTXDQWLWLHV WKDWLVODQJXDJHFRXOGEHDQHFHVVDU\FRQVWLWXHQWLQWKHVH URXWLQHV OLNHLQWKHFDVHRIYHUEDODULWKPHWLFEXWXQOLNHLQWKHFDVHRIDQDEDFXV 5HFHQWVWXGLHVKDYHEHJXQWRGLIIHUHQWLDWHEHWZHHQWKHVHWZRDFFRXQWV)LUVWZRUN with signers in Nicaragua has investigated the numerical abilities of individuals in a highly numerate culture who nonetheless have limited representations of exact number and limited URXWLQHVIRUPDQLSXODWLQJWKHVHUHSUHVHQWDWLRQV6HFRQGSV\FKRSK\VLFDOH[SHULPHQWDWLRQ with verbal interference tasks has begun to manipulate the online availability of linguistic UHSUHVHQWDWLRQVRIH[DFWQXPEHULQKLJKO\QXPHUDWHHGXFDWHGDGXOWV7KHVHWZRVHWVRI VWXGLHVDUHUHYLHZHGEHORZ 4.1. CULTURAL EXPOSURE ALONE DOES NOT SCAFFOLD EXACT NUMBER. Nicaraguan Sign Language (NSL) is a new sign language created over the last 30 years as specialized schools have brought together the community of deaf individuals in Nicaragua (Senghas, .LWD 2]\XUHN $VWKH1LFDUDJXDQGHDIFRPPXQLW\KDVJURZQDQGWKHDJHDWZKLFK children are exposed has become younger, NSL has evolved into a fully-featured, highly grammaticized language that includes number words, complex spatial language (Senghas &RSSROD DQG VRSKLVWLFDWHG FRQVWUXFWLRQV IRU UHSRUWLQJ WKH WKRXJKWV RI RWKHUV 3\HUV 6HQJKDV Since NSL speakers live in a numerate community, playing gambling games and using PRQH\WKH\KDYHDPSOHRSSRUWXQLWLHVWRDFTXLUHQXPHULFDOURXWLQHV1HYHUWKHOHVVQXPEHU signs in NSL underwent rapid standardization in the early 1990s, transforming from iconic ¿QJHUVLJQV²ZLWKDQXPEHURI¿QJHUVFRUUHVSRQGLQJWRWKHTXDQWLW\EHLQJLQGLFDWHG²WRD VHWRIVLPSOHURQHKDQGHGVLJQVWKDWDUHOHVVLFRQLF7KLVFKDQJHKDVFUHDWHGDSRSXODWLRQ of speakers with a range of experience with numbers signs: there are older adults who did not learn either system as children; younger adults who learned the iconic system but have since learned the second system; and adolescents who learned the second system during FKLOGKRRG )ODKHUW\ 6HQJKDV %\NHHSLQJFXOWXUDOH[SRVXUHUHODWLYHO\FRQVWDQW but varying linguistic representation, the case of NSL thus presents an opportunity to test ZKHWKHUFXOWXUDOH[SRVXUHWRQXPHULFDOURXWLQHVLVVXI¿FLHQWIRUDFFXUDWHSHUIRUPDQFHRI numerical tasks or whether it is necessary to have linguistic representations in order to DFTXLUHRUFDUU\RXWQXPHULFDOURXWLQHV )ODKHUW\DQG 6HQJKDV WHVWHG 16/ VSHDNHUV DFURVV WKH IXOO UDQJH RI DJHV RQ DVHWRIWDVNVWKDWLQFOXGHGPDWFKLQJWDVNVOLNHWKRVHXVHGE\*RUGRQ DVZHOODV WDVNVUHTXLULQJWDSSLQJRXWTXDQWLWLHVFRXQWLQJDQGVHOHFWLQJVHWVXVLQJQXPEHUZRUGVDQG WUDQVODWLQJEHWZHHQPRQHWDU\QRWHVDQGFRLQV$FURVVDOOWDVNVWKHJURXSWKDWPDGHIDUDQG away the most errors were the older adults that had not fully mastered even the iconic count OLVWV,QGLYLGXDOVZKRKDGPDVWHUHGHLWKHUFRXQWOLVWPDGHVPDOOEXWV\VWHPDWLFHUURUV² indicating that they were not perfectly accurate in using their count routine in challenging VLWXDWLRQV²EXWWKHSHUIRUPDQFHRIROGHUDGXOWVZKRFRXOGQRWFRXQWGLIIHUHGVLJQL¿FDQWO\ IURPHYHQWKDWRIWKHROGHUDGXOWVZKRKDGEHHQDEOHWRPDVWHUWKHLFRQLFFRXQWURXWLQH In addition, as with the Pirahã, all NSL participants—even the non-counters— VXFFHHGHGLQJUDVSLQJWKHVLPSOHVW±PDWFKLQJWDVNV:KHQPDWFKLQJWDVNVEHFDPHPRUH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number complex and the stimuli being matched were presented ephemerally (via tapping, or via SXWWLQJLWHPVRQHE\RQHLQWRDQRSDTXHFXS DFFXUDF\ZDVFRQVLGHUDEO\ORZHUIRUWKHQRQ FRXQWHUV7KHQRQFRXQWHUVNQHZWKDWWKHUHZDVVRPHWKLQJWKH\GLGQRWNQRZKRZHYHU² WKH\H[SUHVVHGXQFHUWDLQW\DERXWODUJHUTXDQWLWLHVDQGKDGGHYHORSHGKHXULVWLFVWUDWHJLHV IRUPDNLQJFKDQJHLQWKHPRQHWDU\WDVNV7KH\NQHZWKDWDQH[DFWDQVZHUZDVUHTXLUHG EXWGLGQRWNQRZKRZWRFDOFXODWHWKDWDQVZHU7KXVOLNHWKH3LUDKm16/VSHDNHUVZLWKRXW DFRXQWURXWLQHZHUHDEOHWRVHOHFWDQH[DFWTXDQWLW\PDWFKLQJVWUDWHJ\HYHQLQWKHDEVHQFH RIDUHOLDEOHPHWKRGIRUPHQWDOO\UHSUHVHQWLQJLQGLYLGXDOTXDQWLWLHV Although many deaf children in Nicaragua are now given opportunities to learn NSL, there are still some individuals who have not had access to the broader deaf community and have instead built up more idiosyncratic sign systems for communicating with their IDPLOLHV DQG PRUH LPPHGLDWH FRPPXQLW\ ³+RPHVLJQ´ V\VWHPV RI WKLV VRUW DQG WKHLU relationship to conventional language have been studied extensively, in the US and around WKH ZRUOG *ROGLQ0HDGRZ 0\ODQGHU 5HFHQW ZRUN E\ 6SDHSHQ &RSSROD 6SHONH&DUH\DQG*ROGLQ0HDGRZ LQYHVWLJDWHVQXPHULFDOFRJQLWLRQLQ1LFDUDJXDQ KRPHVLJQHUV&RQJUXHQWZLWKWKHZRUNZLWK16/VSHDNHUV6SDHSHQDQGFROOHDJXHVIRXQG that homesigners, who could not produce a consistent count list or perform matching tasks, ZHUHVWLOODEOHWRFRPSDUHPRQHWDU\GHQRPLQDWLRQVZLWKKLJKDFFXUDF\ In addition, although they could not produce a correct ordering of number signs, the KRPHVLJQHUVGLGVWLOONQRZZRUGVIRUH[DFWTXDQWLWLHV7KLVNQRZOHGJHDOORZHG6SDHSHQ DQGFROOHDJXHVWRSHUIRUPDQLPSRUWDQWH[DFWQXPHURVLW\UHFRJQLWLRQWDVN,QWKLVWDVNWKH homesigners were told that some exact number of objects were in a box, and then the array LQWKHER[ZDVWUDQVIRUPHG HLWKHUYLDDFKDQJHLQWKHQXPEHURIREMHFWVRUQRW :KHQWKH WUDQVIRUPDWLRQGLGQRWFKDQJHWKHTXDQWLW\LQWKHER[WKHKRPHVLJQHUVDOPRVWDOZD\VXVHG WKHVDPHJHVWXUHDVWKHH[SHULPHQWHU ZKHQWKHWUDQVIRUPDWLRQGLGFKDQJHWKHTXDQWLW\ WKH\QHYHUXVHGWKHVDPHJHVWXUH5XOLQJRXWDSUDJPDWLFH[SODQDWLRQIRUWKLVEHKDYLRU HJDSSO\LQJWKHSULQFLSOHRIFRQWUDVW &ODUN QHDUO\DOOSDUWLFLSDQWVXVHGJHVWXUHV that matched the direction of the transformation, for example signaling a larger number WKDQWKHRULJLQDOJHVWXUHZKHQDQREMHFWKDGEHHQDGGHGWRWKHVHW7KLVWDVNJLYHVFOHDU evidence that the homesigners understood that each set had an exact numerical value, even LIWKH\GLGQRWKDYHDQHUURUOHVVURXWLQHIRU¿QGLQJWKDWYDOXH Although both NSL users and homesigners grew up in a highly numerate culture, this fact alone did not create the concepts and routines necessary to succeed in complex exact QXPEHUWDVNV,QDGGLWLRQVXSSRUWLQJWKH3LUDKm±PDWFKLQJUHVXOWVWKH1LFDUDJXDQGDWD suggest that neither number words nor a count routine are necessary to understand the idea WKDWDVHWKDVDQH[DFWTXDQWLW\HYHQLIWKDWTXDQWLW\FDQQRWEHQDPHGRUVWRUHGLQPHPRU\ While the Nicaraguan data implicate linguistic representations (rather than cultural H[SRVXUHWRURXWLQHV DVSOD\LQJDFDXVDOUROHLQWKHDELOLW\WRPDQLSXODWHH[DFWTXDQWLWLHV LWLVDVHSDUDWHTXHVWLRQZKHWKHUWKLVUROHLVonline,QRWKHUZRUGVIRUDQLQGLYLGXDOZLWKD OLIHWLPHRISUDFWLFHUHSUHVHQWLQJH[DFWTXDQWLWLHVGRHVUHSUHVHQWLQJDTXDQWLW\OLNHUHTXLUH the use of language in the moment such that if linguistic resources were not available at WKDWPRPHQWWKLVWDVNZRXOGEHFRPHPXFKPRUHGLI¿FXOWRULPSRVVLEOH"7RDQVZHUWKLV TXHVWLRQZHWXUQWRSV\FKRSK\VLFDOWDVNVSHUIRUPHGZLWKQXPHUDWH(QJOLVKVSHDNHUV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number 4.2. NUMBER WORDS MUST BE AVAILABLE ONLINE FOR ENUMERATION. 9HUEDO interference methods have been used widely for testing the online dependence of various WDVNVRQODQJXDJH 1HZWRQ GH9LOOLHUV:LQDZHUHWDO+HUPHU9D]TXH] 6SHONH .DWVQHOVRQ 9HUEDOLQWHUIHUHQFHUHIHUVWRDFODVVRIH[SHULPHQWDOSDUDGLJPV in which participants are asked to perform a task while simultaneously occupying their verbal system by performing a separate verbal task, such as repeating a word like “the,” repeating strings of numbers, or “shadowing” (repeating words after immediately after KHDULQJWKHPVSRNHQRQDUHFRUGLQJ $VDFRQWUROIRUWKHJHQHUDOL]HGGXDOWDVNFRVWRI SHUIRUPLQJWZRWDVNVDWRQFH 3DVKOHU SHUIRUPDQFHLQWKHWDUJHWWDVNXQGHUYHUEDO interference is often compared to performance in the target task paired with a non-verbal WDVNOLNHVKDGRZLQJDFODSSHGSDWWHUQ $KDQGIXORIVWXGLHVKDYHXVHGYHUEDOLQWHUIHUHQFHWRPHDVXUHQXPHULFDOEHKDYLRU +RZHYHU PRVW KDYH GRQH VR XVLQJ QXPEHU WDVNV WKDW ZHUH WKHPVHOYHV YHUEDO )RU example, Logie and Baddeley (1987) found that rapid repetition of “the” caused more HUURUVLQFRXQWLQJWKDQHLWKHUOLVWHQLQJWRVSHHFKRUWDSSLQJD¿QJHUVXJJHVWLQJWKDWDFWLYH VSHHFKSURGXFWLRQLQWHUIHUHGZLWKXVHRIWKHVDPHV\VWHPWRFRXQW$PRUHUHFHQWVWXG\ E\&RUGHV*HOPDQ*DOOLVWHODQG:KDOHQ VKRZHGDQ$UDELFQXPHUDODQGDVNHG participants to press a key that number of times while either repeating “the” or counting YHU\TXLFNO\7KH\IRXQGWKDWSDUWLFLSDQWVXQGHUYHUEDOVXSSUHVVLRQ VKRZHG DFRQVWDQW FRHI¿FLHQW RI YDULDWLRQ²LQGLFDWLQJ XVH RI WKH $16²ZKLOH WKRVH ZKR ZHUH FRXQWLQJ VKRZHGDGHFUHDVLQJ&29 SHUKDSVFDXVHGE\WKHELQRPLDOHUURUVLPSOLHGE\VNLSSLQJ QXPEHUVLQWKHFRXQWOLVW 7KHVHWZRVWXGLHVJLYHHYLGHQFHWKDWODQJXDJHLQWHUIHUHQFHGRHV cause participants to make errors when aspects of the task are linguistic, but leaves open WKHSRVVLELOLW\RIEHWWHUSHUIRUPDQFHLQSXUHO\QRQOLQJXLVWLFWDVNV In order to test this possibility, my colleagues and I conducted a series of experiments where we replicated the matching tasks used with the Pirahã, performing these tasks with a group of English speakers who were simultaneously shadowing complex texts (Frank et DO 7KLVSDUDGLJPKDGWKHEHQH¿WRIXVLQJDSXUHO\QRQYHUEDOPHDVXUHRIQXPEHU knowledge and of providing data that could be compared directly to those collected during ¿HOGZRUNZLWKWKH3LUDKm2XUUHVXOWVVXJJHVWHGVWURQJSDUDOOHOVEHWZHHQWKHSHUIRUPDQFH of the English speakers—who did not have number language available in the moment— DQGWKDWRIWKH3LUDKm²ZKRKDGQHYHUNQRZQZRUGVIRUQXPEHUV/LNHWKH3LUDKm DQG Nicaraguan populations), the English speakers under verbal interference were able to do WKH±PDWFKLQJWDVNZLWKUHODWLYHO\IHZHUURUV,QDGGLWLRQWKH(QJOLVKVSHDNHUVOLNHWKH RWKHUSRSXODWLRQVVKRZHGHYLGHQFHRIUHO\LQJRQWKH$16LQWKHKDUGHVWPDWFKLQJWDVNV Followup experiments using matched verbal and spatial memory interference tasks showed WKDWWKLVSDWWHUQZDVVSHFL¿FWRODQJXDJHLQWHUIHUHQFH +RZHYHU WKH (QJOLVK VSHDNHUV DOVR VKRZHG VRPH GLIIHUHQFHV IURP WKH 3LUDKm ,Q WKH PHGLXPGLI¿FXOW\ PDWFKLQJ WDVNV ZKHUH WKHUH ZHUH YLVXDO FXHV HJ PDWFKLQJ WKH TXDQWLWLHVRIWZRRUWKRJRQDOOLQHV WKH\PDGHHUURUVEXWWKHLURYHUDOOSHUIRUPDQFHGLG QRWVKRZWKHVLJQDWXUHRIWKH$16 DFRQVWDQWUHODWLRQVKLSEHWZHHQWKHTXDQWLWLHVEHLQJ PDWFKHGDQGWKHPDJQLWXGHVRIWKHHUURUVLQHVWLPDWLRQ ,QVWHDGWKHPDJQLWXGHRIWKHHUURUV LQFUHDVHGZLWKUHVSHFWWRWKHTXDQWLW\EHLQJPDWFKHG:HSRVLWHGWKDWWKHLUHUURUVUHVXOWHG from the use of ad hoc matching routines like making correspondences between sub-groups RIREMHFWV7KLVVDPHSDWWHUQRILQFUHDVLQJHUURUVZDVREVHUYHGLQWKH1LFDUDJXDQVLJQHUV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number who did know a count list, indicating that this pattern of data may generally result from the DSSOLFDWLRQRIIDOOLEOHURXWLQHV More broadly, the picture that emerges from the evidence so far suggests an online, FDXVDOUROHIRUODQJXDJHLQWKHUHSUHVHQWDWLRQRIQXPEHULQIRUPDWLRQ(YLGHQFHJDWKHUHG WKURXJKSV\FKROLQJXLVWLF¿HOGZRUNLQFRPELQDWLRQZLWKODERUDWRU\FRQWUROWDVNVVXJJHVWV WKDW UHSUHVHQWLQJ UHTXLUHV KDYLQJ VRPH LQWHUQDO V\PERO OLNH ³VHYHQ´ DYDLODEOH LQ WKH PRPHQW7KLVSDWWHUQRIHYLGHQFHVKRXOGQRWVXJJHVWWKDWWKHUHLVQRUROHIRUFXOWXUDOQHHGV LQWKHFUHDWLRQRIQXPHULFDOURXWLQHVKRZHYHU7KHQH[WVHFWLRQJLYHVVHYHUDOHWKQRJUDSKLF H[DPSOHVRILQWHUDFWLRQVEHWZHHQFXOWXUHDQGQXPHULFDOUHSUHVHQWDWLRQVDQGURXWLQHV 5. NUMERICAL REPRESENTATIONS AND ROUTINES CAN BE SHAPED BY CULTURE. A common perspective on English numerals—even from sophisticated, numerate adults—is WKDWWKH\DUHWUDQVSDUHQWOLQJXLVWLFWRROVWKDWGRQRWUHÀHFWDQLGLRV\QFUDWLFHYROXWLRQDU\ SURFHVVGULYHQE\VSHFL¿FFXOWXUDOQHHGV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number WZLVWWKHEDVHUHSUHVHQWDWLRQOHDGVWRDUHLQWHUSUHWDWLRQRI¿QJHUFRXQWLQJURXWLQHV WKHVH URXWLQHVEHFRPHEDVHDVZHOOXVLQJWKHZULVWDVD¿QDOORFDWLRQDQGUHFRQVWUXLQJWKH ¿QJHUFRXQWDVDFRXQWRI³DWWDFKPHQWSRLQWV´ ¿QJHUMRLQWVDQGZULVWMRLQW (YDQV E ,QWKLVZD\¿QJHUFRXQWLQJLVUHLQWHUSUHWHGZLWKUHVSHFWWRWKHEDVHUHSUHVHQWDWLRQWKDW KDVHYROYHG RUEHHQLQYHQWHG WRVXSSRUWDQLPSRUWDQWFXOWXUDOURXWLQH Third, changes to a numerical representation can also be motivated directly through FKDQJHVLQFXOWXUDOURXWLQHV$VSHFL¿FH[DPSOHRIWKLVNLQGRIFRPHVIURP6D[H He documented that speakers of Oksapmin, a language spoken in the West Sepik province of Papua New Guinea, used a body-count system (a common type in the region (Lean 7KLV FRXQW V\VWHP ZDV EDVH H[WHQGLQJ IURP RQH KDQG DORQJ WKH DUP RYHU WKHKHDGDQGWKURXJKWKHRWKHUDUPWRWKHRWKHUKDQG+RZHYHUZKHQXVHUVRIWKLVFRXQW system had limited experience with manipulating money, they made systematic errors in VLPSOHDGGLWLRQSUREOHPV HJ 6D[HIRXQGWKDWDOWKRXJKWKH\FRXOGFRXQWRXWWKH ¿UVWDGGHQG WKHVHLQH[SHULHQFHGXVHUVKDGQRWGHYHORSHGDFRUUHVSRQGHQFHVWUDWHJ\VR WKDWWKH\FRXOGNHHSWUDFNRIWKHQXPEHURIERG\SDUWVLQWKHVHFRQGDGGHQG Users more experienced with money manipulation had developed a number of strategies to circumvent this problem, however, including counting both the second addend and the sum of the addends in parallel, and splitting the body in two and using the second DUPWRWUDFNWKHVHFRQGDGGHQG7KHERG\VSOLWWLQJVWUDWHJ\ZDVPRVWVXFFHVVIXO LWZDV used by the participants that were most experienced with money manipulation, but also UHTXLUHGWKHPRVWDGDSWDWLRQRIWKHH[LVWLQJUHSUHVHQWDWLRQ7RXVHLW6D[H¶VSDUWLFLSDQWV KDGWRUHYHUVHWKHFRXQWOLVWVRWKDWLWFRXOGEHLQLWLDWHGIURPHLWKHUDUP6D[H¶VVWXG\ beautifully demonstrates how cultural pressures can lead to the creation of new routines for DULWKPHWLFDQGFDQLQWXUQOHDGWRFKDQJHVLQWKHEDVHUHSUHVHQWDWLRQ Body count systems also suggest how the choice of a base—or more generally the design RIDQXPEHUUHSUHVHQWDWLRQ²FDQLQWHUIHUHZLWKWKHGHYHORSPHQWRIHI¿FLHQWURXWLQHV,QWKH FDVHRI2NVDSPLQWKHEDVHZDVVRKLJKWKDWWKHHQXPHUDWLRQURXWLQHUHTXLUHGERWKKDQGV DQGKHQFHFRXOGQRWEHHDVLO\XVHGWRFUHDWHWZRVHSDUDWHEXIIHUVIRUDGGLWLRQ7KLVH[DPSOH is minor, however, compared with counting systems like one reported to be used by some VSHDNHUVRI2QH7KLVV\VWHPGHVFULEHGE\'RQRKXH LVLQSULQFLSOHUHFXUVLYHDQG LQ¿QLWHEXWLQSUDFWLFHVRFXPEHUVRPHWKDWLWLVUDUHO\XVHGWRFRXQWTXDQWLWLHVODUJHUWKDQD KDQGIXO2QHKDVLQGLYLGXDOOH[LFDOLWHPVIRUDQGEXWDOORZVVSHFL¿FFRQYHQWLRQDOL]HG FRPELQDWLRQVRIWKHVHZRUGVXSWR7KHLUFRXQWOLVWDGPLWVWKHIROORZLQJFRPELQDWLRQV DQG EXW QRW IRU H[DPSOH $OWKRXJK WKLV V\VWHP FRXOG EH XVHG WR H[SUHVV RU HYHQ LW TXLFNO\ EHFRPHV LPSUDFWLFDO IRU ODUJHUTXDQWLWLHV7KLVV\VWHPPD\HYHQEHDUHFHQWLQQRYDWLRQDQGKHQFHLQGLFDWLYHRID FRPPXQLW\ZKRVHXVHRIQXPEHUVLVLQÀX[ &URZWKHU 7KHVH H[DPSOHV JLYH D ÀDYRU IRU WKH ZD\V LQ ZKLFK WKH YDVW UDQJH RI QXPEHU representations and routines in the world’s languages can be shaped by their cultural FRQWH[W1HYHUWKHOHVVXQGHUVWDQGLQJWKHVSHFL¿FFRJQLWLYHFRQVHTXHQFHVRIWKLVYDULDWLRQ ZLOOUHTXLUHVLJQL¿FDQWH[SHULPHQWDO¿HOGZRUNDQGWKHIRUPRIWKHUHODWLRQVKLSEHWZHHQ SDUWLFXODUQXPHULFDOUHSUHVHQWDWLRQVDQGWKHURXWLQHVWKH\VXSSRUWLVPRVWO\XQNQRZQ7KH ODVWVHFWLRQRIWKLVUHYLHZJLYHVVRPHHYLGHQFHRQWKLVTXHVWLRQE\H[SORULQJDFDVHVWXG\ of a number representation that licenses a very different set of routines—in a different PHGLXP²IURPWKHRWKHUVZHKDYHUHYLHZHG PHQWDODEDFXV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number 6. NON-LINGUISTIC REPRESENTATIONS: EFFECTS OF REPRESENTATION STRUCTURE ON ROUTINE. This review began by asking about the relationship between language and thought and explored this relationship through the diversity in number representations across the ZRUOG¶VODQJXDJHV%XWLQIDFWWKHUHLVDZLGHYDULHW\RIQRQOLQJXLVWLFUHSUHVHQWDWLRQVXVHG DFURVVGLIIHUHQWFXOWXUHV$VLGHIURP¿QJHUFRXQWV\VWHPV0HQQLQJHU GHVFULEHVWKH near-universal use in the ancient world of tally sticks and knot-based systems to keep track RIODUJHTXDQWLWLHV,QPDQ\FDVHVWDOO\EDVHGV\VWHPVHYROYHGLQWRWKHXVHRIFRXQWLQJ ERDUGV²GHYLFHVWKDWDOORZHGIRUWKHJURXSLQJRIWDOO\REMHFWVOLNHSHEEOHV8QOLNHPRVW tally systems, counting boards incorporated the use of place value (in which a particular SRVLWLRQLQDQRWDWLRQVWDQGVIRUWKHRUGHURIPDJQLWXGHRIV\PEROVLQWKDWSRVLWLRQHJ 101 3 DPDMRULQQRYDWLRQWKDWDOORZHGWKHPWREHXVHGÀH[LEO\IRUD ZLGHYDULHW\RIUHFRUGNHHSLQJ The modern Soroban abacus (primarily used in Japan and China) likely evolved from 5RPDQFRXQWLQJERDUGV/LNHWKHVHFRXQWLQJERDUGVWKH6RUREDQDEDFXVXVHVDEDVH UHSUHVHQWDWLRQZLWKSODFHYDOXHVPDSSHGWRLQGLYLGXDODEDFXVFROXPQV7KHDEDFXV DQG some of the most sophisticated counting boards) incorporates a subsidiary base-5 as well, however: on a standard Soroban each column has a single bead on top that represents 5 units in that place value, and four beads on the bottom that each represent 1 in that place YDOXH &RPELQDWLRQV RI WKHVH EHDGV DOORZ WKH TXDQWLWLHV ± WR EH UHSUHVHQWHG XVLQJ D PD[LPXPRI¿YHEHDGV Like tally sticks and other enumeration devices, counting boards and abacuses are H[WHUQDOGHYLFHVWKDWDOORZWKHLUXVHUVWRHQXPHUDWHODUJHH[DFWTXDQWLWLHVDQGUHWDLQWKHP SUHFLVHO\RYHUORQJSHULRGVRIWLPH+RZHYHUWKHDEDFXVDOORZVXVHUVWRJREH\RQGWKH VLPSOH HQXPHUDWLRQWDVN E\ DOORZLQJ WKH GHYHORSPHQWDQG XVH RI HI¿FLHQW URXWLQHV IRU DULWKPHWLFFRPSXWDWLRQ8VLQJWKHEDVHZLWKLQFROXPQUHSUHVHQWDWLRQDQGEDVHSODFH value system, large computations can be broken into many small steps consisting of the addition of numbers below 5 and a corresponding set of “carry” operations (in which WKHSDUWVRIDUHVXOWJUHDWHUWKDQ DUHWUDQVIHUUHGWRWKHQH[WKLJKHVWSODFHYDOXH :LWK SUDFWLFHDEDFXVFDOFXODWLRQVFDQEHFRPHURXWLQL]HGDQGKLJKO\DFFXUDWH,QDKHDGWRKHDG competition in post-war Japan, a skilled abacus operator out-computed a calculator user .RMLPD &UXFLDOO\DEDFXVDGGLWLRQRSHUDWHVYLDDURXWLQHXVLQJVHWRIPHPRUL]HG RSHUDWLRQVWKDWDUHGLIIHUHQWIURPWKHFRPPRQO\XVHGEDVHDGGLWLRQRSHUDWLRQV Although abacus is an external computation aid, experienced abacus users can learn to internalize the abacus representation and make computations by manipulating beads on a PHQWDOLPDJHRIDQDEDFXV7KLVWHFKQLTXHNQRZQDVPHQWDODEDFXV 0$ LVZLGHO\WDXJKW in Japan and has been the focus of recent interest in math supplementary education programs LQ0DOD\VLD,QGLD&KLQDDQGDQXPEHURIRWKHUFRXQWULHVLQ$VLDDQGWKH0LGGOH(DVW Studies of MA have suggested that users do truly represent a mental abacus using visual LPDJHU\ +DWDQR +DWDQR 2VDZD 6WLJOHU )RUH[DPSOHWKH\PDNHRII by-5 errors far more than would be expected in standard linguistic calculation, indicating WKDWWKH\DUHLQDGYHUWHQWO\³GURSSLQJ´WKHEHDGIURPWKHLUPHQWDOUHSUHVHQWDWLRQ0$ users also seem to be able to compute while performing linguistic distractor tasks (Hatano DQGQHXURLPDJLQJVWXGLHVFRQ¿UPWKDW0$DFWLYLW\LQGXFHVDFWLYLW\LQFRUWLFDODUHDV UHODWHGWRYLVXRVSDWLDOZRUNLQJPHPRU\ 7DQDNDHWDO&KHQHWDO 0$LVDOVR MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number KLJKO\HIIHFWLYHDVDQDULWKPHWLFPHWKRG D0$XVHUWRRNWRSKRQRUVLQWKH:RUOG&XS RI0HQWDO&RPSXWDWLRQ Our own recent work investigated how it is that MA representations are possible, given WKHDWWHVWHGOLPLWVRQQXPHULFDOUHSUHVHQWDWLRQLQWKHYLVXDOV\VWHP )UDQN %DUQHU Neither of the two systems traditionally implicated in visual number processing (object SURFHVVLQJIRUVPDOOQXPEHUVXSWRDSSUR[LPDWHUHSUHVHQWDWLRQVDERYHWKDW ZRXOGEH DEOHWRUHSUHVHQWDQXPEHUOLNH RQWKHDEDFXVVLQFHWKLVZRXOGUHTXLUHUHSUHVHQWLQJWKH H[DFWSRVLWLRQVRI EHDGV'HVSLWHWKLV0$XVHUVDUHDEOHWRGRLPSUHVVLYHFRPSXWDWLRQV ZLWK IDU ODUJHU QXPEHUV )RU D YLVXDO FRPSDULVRQ RI DEDFXV FRPSXWDWLRQ DQG 0$ VHH ¿JXUH FIGURE 2. OHIW $FKLOGSHUIRUPLQJDSK\VLFDODEDFXVFRPSXWDWLRQ ULJKW 7KHVDPHFKLOG SHUIRUPLQJDPHQWDODEDFXVFRPSXWDWLRQ :HWHVWHGDODUJHJURXSRIFKLOGUHQ DJHV±\HDUV LQ*XMDUDW,QGLDZKRZHUH HQUROOHGLQ0$DIWHUVFKRROSURJUDPV:HDVNHGWKHVHFKLOGUHQWRGRWZRVWDQGDUG0$ WDVNV²DGGLWLRQRITXDQWLWLHVDQGWUDQVODWLRQRIDEDFXVFRQ¿JXUDWLRQVLQWR$UDELFQXPHUDOV² ZKLOHZHYDULHGWKHGLI¿FXOW\RIWKHWDVNV,QERWKWDVNVZHIRXQGWKDWWKHOLPLWDWLRQVRQ performance came from the number of columns on the abacus representation—in other ZRUGV WKH PD[LPXP SODFH YDOXH²UDWKHU WKDQ RQ RWKHU IHDWXUHV RI D JLYHQ WDVN )RU example, many participants were able to add between 7 and 9 two-digit addends together LQXQGHUVHFRQGVEXWYHU\IHZZHUHDEOHWRDGGIRXUGLJLWDGGHQGVLQWKHVDPHWLPH SHULRG,QFRQWUDVWWKHQXPEHURIEHDGVQHFHVVDU\WRPDNHDUHSUHVHQWDWLRQ HJZKHWKHUD FROXPQVKRZHGDQXPEHUOLNHZLWKEHDGVRU ZLWK¿YHEHDGV LQWKHVHSUREOHPVGLG QRWVHHPWRDIIHFWSHUIRUPDQFHRQFHWKHQXPEHURIFROXPQVZDVFRQWUROOHG To test the dependence of MA computations on language, we asked MA experts to SHUIRUPYHUEDOLQWHUIHUHQFHWDVNVDVWKH\GLGPHQWDOFRPSXWDWLRQV:KLOHWKHLUSHUIRUPDQFH was impaired slightly by verbal shadowing, they were if anything more impaired by VLPSO\WDSSLQJWKHLU¿QJHUVGXULQJWKHFRPSXWDWLRQ SUHVXPDEO\GXHWRWKHUHOLDQFHRIWKH FRPSXWDWLRQRQWKHDFFRPSDQ\LQJJHVWXUHVVHH¿JXUH ,QFRQWUDVWDJURXSRI$PHULFDQ college students—who used linguistic calculation strategies to do mental arithmetic—were highly impaired by verbal interference but experienced no interference from tapping their ¿QJHUV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number The interference data suggest that MA is a fundamentally visual representation, while performance in the addition experiments described above suggest that MA representations DUHFROXPQEDVHG:HK\SRWKHVL]HWKDWHDFKFROXPQLQWKHPHQWDODEDFXVLVPDSSHGWRD separate object representation in visual working memory, though the substructure of how HDFKFROXPQLVUHSUHVHQWHGLVVWLOOXQNQRZQ&RQVLVWHQWZLWKWKLVK\SRWKHVLVZHIRXQGWKDW novice abacus users showed some of the same signatures of column-based organization, suggesting that this non-linguistic format for number was adapted to the general visual FDSDFLWLHVRILWVXVHUVUDWKHUWKDQEHLQJWKHUHVXOWRIH[WHQVLYHSUDFWLFH7DNHQWRJHWKHU these data paint a picture of MA as a visual alternative to linguistic number representations that relies on the distinct structure of visual working memory, rather than phonological ZRUNLQJPHPRU\ DVLQODQJXDJHEDVHGWHFKQLTXHVIRUPHQWDODULWKPHWLF The example of MA goes beyond external physical representations of number like FRXQWLQJERDUGVDQGJLYHVVWURQJHYLGHQFHWKDWWKHPHQWDOUHSUHVHQWDWLRQRIH[DFWTXDQWLW\ LVSRVVLEOHLQPHGLXPVRWKHUWKDQODQJXDJH$OWKRXJKVRPHDXWKRUVKDYHVSHFXODWHGWKDW ODQJXDJHDQGH[DFWQXPEHUUHO\RQWKHVDPHFRPSXWDWLRQDOVXEVWUDWH +DXVHUHWDO WKHIDFLOLW\DQGÀH[LELOLW\LQFRPSXWDWLRQVKRZQE\0$XVHUVVXJJHVWVDGLIIHUHQWYLHZ Representations of exact number can be constructed using a variety of different resources— OLQJXLVWLFRUYLVXDO,QDGGLWLRQWKHVSHFL¿FRUJDQL]DWLRQRIWKHDEDFXV0$UHSUHVHQWDWLRQ is tailored to allow computations to be decomposed into many simple operations that can EHSUDFWLFHGLQGHSHQGHQWO\$OWKRXJKWKH0$DGGLWLRQURXWLQHUHTXLUHVPRUHVWHSVWKDQWKH PRVWFRPPRQYHUEDODOJRULWKPLWLVDOVRPRUHDFFXUDWHEHFDXVHLWQHYHUUHTXLUHVVWRULQJ SDUWLDOVXPV 7. CONCLUSIONS. %HOOHUDQG%HQGHU ZULWHWKDW³WKHUHPD\EHQRRWKHUGRPDLQ LQ WKH ¿HOG RI FRJQLWLYH VFLHQFHV ZKHUH LW LV VR REYLRXV WKDW ODQJXDJH LH WKH YHUEDO QXPHUDWLRQV\VWHP DIIHFWVFRJQLWLRQ LHPHQWDODULWKPHWLF ´7KHGDWDUHYLHZHGKHUH DUH FRQVLVWHQW ZLWK WKLV FRQWHQWLRQ KRZ D ODQJXDJH UHSUHVHQWV ODUJH H[DFW TXDQWLWLHV GUDPDWLFDOO\LQÀXHQFHVKRZLWVVSHDNHUVDUHDEOHWRVWRUHDQGPDQLSXODWHWKHP)RUWKLV UHDVRQ QXPEHU UHSUHVHQWDWLRQ SUHVHQWV DQ LPSRUWDQW FDVH WR JR EH\RQG WKH ¿UVW RUGHU TXHVWLRQVRIWKH:KRU¿DQGHEDWH²³GRHVODQJXDJHLQÀXHQFHWKRXJKW´²DQGDVNGHWDLOHG TXHVWLRQVDERXWKRZODQJXDJHSDUWLFLSDWHVLQFRQVWUXFWLQJUHSUHVHQWDWLRQVRIH[DFWQXPEHU DQGURXWLQHVIRUPDQLSXODWLQJTXDQWLWLHV,QYHVWLJDWLRQVRIWKHULFKQHVVRIFURVVFXOWXUDO YDULDWLRQLQ QXPEHU V\VWHPV VXJJHVW WKDWWKHUHDUHPDMRUEHKDYLRUDOFRQVHTXHQFHVWKDW correspond to what number words a language has and how those words are structured LQWRDFRXQWOLVW0RUHJHQHUDOO\WKHIRUPRIDQXPHULFDOUHSUHVHQWDWLRQ OLQJXLVWLFRUQRW VWUXFWXUHVWKHNLQGVRIURXWLQHVIRUHQXPHUDWLRQDQGDULWKPHWLFWKDWFDQEHSHUIRUPHG The data that lead to this conclusion could not have been gathered by the standard PHWKRGVRIFRJQLWLYHSV\FKRORJ\QRUE\WKHVWDQGDUGPHWKRGVRI¿HOGOLQJXLVWLFV0DQ\ RIWKHUHVXOWVFLWHGKHUHFRPHIURPFDUHIXOO\FRQWUROOHGVWXGLHVSHUIRUPHGLQWKH¿HOGZLWK populations that possess culturally, linguistically, or cognitively interesting numerical UHSUHVHQWDWLRQV 7KLV JHQHUDOL]DWLRQ VXJJHVWV WKH EHQH¿WV RI SV\FKROLQJXLVWLF ¿HOGZRUN WKDW FRPELQHV H[SHULPHQWDO GHVLJQ ZLWK FURVVFXOWXUDO RU FURVVOLQJXLVWLF SRSXODWLRQV 6XFK¿HOGZRUNLVHVSHFLDOO\LPSRUWDQWLQWKHVWXG\RIWKHGLYHUVHODQJXDJHVRI0HODQHVLD VLQFHRSSRUWXQLWLHVWRVWXG\WKHVHODQJXDJHVDUHTXLFNO\GLVDSSHDULQJ)XWXUH¿HOGZRUN² RQQXPEHUDQGLQRWKHUGRPDLQV²VKRXOGWDNHDGYDQWDJHRIWKHVHWHFKQLTXHVWRSUHVHQWD IXOOHUSLFWXUHRIWKHUHODWLRQVEHWZHHQODQJXDJHFXOWXUHDQGFRJQLWLRQ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number REFERENCES %DUQHU'3/L -6QHGHNHU:RUGVDVZLQGRZVWRWKRXJKW 7KHFDVHRIREMHFW UHSUHVHQWDWLRQCurrent Directions in Psychological Science ± %HOOHU 6 $ %HQGHU 7KH OLPLWV RI FRXQWLQJ 1XPHULFDO FRJQLWLRQ EHWZHHQ HYROXWLRQDQGFXOWXUHScience ± %RURGLWVN\//6FKPLGW :3KLOOLSV6H[V\QWD[DQGVHPDQWLFV,Q'*HQWQHU 6*ROGLQ0HDGRZ HGV Language in mind: Advances in the study of language and thought± &DPEULGJH0$ 0,73UHVV &DUH\6 The origin of concepts1HZ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number )UDQN 0 & ' / (YHUHWW( )HGRUHQNR ( *LEVRQ 1XPEHU DV D FRJQLWLYH WHFKQRORJ\ (YLGHQFHIURP3LUDKmODQJXDJHDQGFRJQLWLRQCognition ± )UDQN 0 & ( )HGRUHQNR 3 /DL 5 6D[H ( *LEVRQ 9HUEDO LQWHUIHUHQFH VXSSUHVVHVH[DFWQXPHULFDOUHSUHVHQWDWLRQCognitive Psychology± )UDQN0& 7+RQH\PDQ1XPEHUNQRZOHGJHLQD¿QLWHFRXQWLQJV\VWHP,Q Proceedings of the 33rd Annual Meeting of the Cognitive Science Society. Austin, TX: &RJQLWLYH6FLHQFH6RFLHW\ *DOOLVWHO& The organization of learning&DPEULGJH0$ 0,73UHVV *HOPDQ5 %%XWWHUZRUWK1XPEHUDQGODQJXDJH +RZDUHWKH\UHODWHG"Trends in Cognitive Sciences ± *HOPDQ5 &5*DOOLVWHO The child’s understanding of number&DPEULGJH0$ +DUYDUG8QLYHUVLW\3UHVV *HQWQHU' :K\ZH¶UHVRVPDUW,Q*HQWQHU *ROGLQ0HDGRZLanguage in mind, ± *HQWQHU' 6*ROGLQ0HDGRZ:KLWKHUZKRUI,Q*HQWQHU *ROGLQ0HDGRZ Language in mind, ± *HQWQHU' 6*ROGLQ0HDGRZ HGV Language in mind: Advances in the study of language and cognition. &DPEULGJH0$ 0,73UHVV *ROGLQ0HDGRZ6 &0\ODQGHU *HVWXUDOFRPPXQLFDWLRQLQGHDIFKLOGUHQ 7KH HIIHFWVDQGQRQHIIHFWVRISDUHQWDOLQSXWRQHDUO\ODQJXDJHGHYHORSPHQWMonographs of the Society for Research in Child Development ± *RUGRQ31XPHULFDOFRJQLWLRQZLWKRXWZRUGV (YLGHQFHIURP$PD]RQLDScience ± *XPSHU]-- 6&/HYLQVRQ Rethinking linguistic relativity&DPEULGJH8. &DPEULGJH8QLYHUVLW\3UHVV +DPPDUVWU|P+ :KHQFHWKH.DQXPEDVH±QXPHUDOV\VWHP"Linguistic Typology ± +DPPDUVWU|P + 5DULWLHV LQ QXPHUDO V\VWHPV ,Q -DQ :RKOJHPXWK 0LFKDHO &\VRXZ HGV Rethinking universals: How rarities affect linguistic theory (Empirical $SSURDFKHVWR/DQJXDJH7\SRORJ\ ±. %HUOLQ 'H*UX\WHU +DWDQR* 3HUIRUPDQFHRIH[SHUWDEDFXVRSHUDWRUVCognition ± +DWDQR* .2VDZD 'LJLWPHPRU\RIJUDQGH[SHUWVLQDEDFXVGHULYHGPHQWDO FDOFXODWLRQCognition ± ± +DXVHU0)7VDR3*DUFLD (6SHONH(YROXWLRQDU\IRXQGDWLRQVRIQXPEHU VSRQWDQHRXV UHSUHVHQWDWLRQ RI QXPHULFDO PDJQLWXGHV E\ FRWWRQWRS WDPDULQV Proceedings of the Royal Society of London 6HULHV% %LRORJLFDO6FLHQFHV +HUPHU / ( 6SHONH $ JHRPHWULF SURFHVV IRU VSDWLDO UHRULHQWDWLRQ LQ \RXQJ FKLOGUHQNature± +HUPHU9D]TXH]/(66SHONH $6.DWVQHOVRQ 6RXUFHVRIÀH[LELOLW\LQKXPDQ FRJQLWLRQ 'XDOWDVNVWXGLHVRIVSDFHDQGODQJXDJHCognitive Psychology ± .D\3%%HUOLQ/0DI¿ :50HUUL¿HOGThe world color survey3DOR$OWR &$ &6/,3UHVV .D\3 :.HPSWRQ :KDWLVWKH6DSLU:KRUIK\SRWKHVLV"American Anthropologist ± MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number .RMLPD 7 The Japanese abacus: Its use and theory 7RN\R &KDUOHV ( 7XWWOH &RPSDQ\ /HDQ* &RXQWLQJV\VWHPVRI3DSXD1HZ*XLQHDDQG2FHDQLD/DH 3DSXD1HZ *XLQHD8QLYHUVLW\RI7HFKQRORJ\GRFWRUDOGLVVHUWDWLRQ /H&RUUH0*9DQGH:DOOH(0%UDQQRQ 6&DUH\5HYLVLWLQJWKHFRPSHWHQFH SHUIRUPDQFHGHEDWHLQWKHDFTXLVLWLRQRIWKHFRXQWLQJSULQFLSOHVCognitive Psychology ± /HYLQVRQ 6 & 6 .LWD ' % 0 +DXQ % + 5DVFK 5HWXUQLQJ WKH WDEOHV /DQJXDJHDIIHFWVVSDWLDOUHDVRQLQJ Cognition± /L3 /*OHLWPDQ7XUQLQJWKHWDEOHV /DQJXDJHDQGVSDWLDOUHDVRQLQJCognition ± /RJLH5 $%DGGHOH\ &RJQLWLYHSURFHVVHVLQFRXQWLQJJournal of Experimental Psychology: Learning, Memory, and Cognition ± /XF\ - $ Language diversity and thought: A reformulation of the linguistic relativity hypothesis. &DPEULGJH &DPEULGJH8QLYHUVLW\3UHVV /XS\DQ*'+5DNLVRQ -/0F&OHOODQG/DQJXDJHLVQRWMXVWIRUWDONLQJ 5HGXQGDQW ODEHOV IDFLOLWDWH OHDUQLQJ RI QRYHO FDWHJRULHV Psychological Science ± 0HQQLQJHU. Number words and number symbols: A cultural history of numbers &DPEULGJH0$ 0,73UHVV 1HZWRQ$0 -*GH9LOOLHUV7KLQNLQJZKLOHWDONLQJ $GXOWVIDLOQRQYHUEDO IDOVHEHOLHIUHDVRQLQJPsychological Science± 3DSDIUDJRX$-+XOEHUW -7UXHVZHOO'RHVODQJXDJHJXLGHHYHQWSHUFHSWLRQ" HYLGHQFHIURPH\HPRYHPHQWVCognition ± 3DVKOHU+ 'XDOWDVNLQWHUIHUHQFHLQVLPSOHWDVNV 'DWDDQGWKHRU\Psychological Bulletin± 3LDQWDGRVL 6 - 7HQHQEDXP 1 *RRGPDQ %RRWVWUDSSLQJ LQ D ODQJXDJH RI WKRXJKW $IRUPDOPRGHORIQXPHULFDOFRQFHSWOHDUQLQJCognition ± 3LFD3&/HPHU9,]DUG 6'HKDHQH([DFWDQGDSSUR[LPDWHDULWKPHWLFLQDQ $PD]RQLDQLQGLJHQHJURXSScience ± 3LQNHU6 The language instinct1HZ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Cross-cultural differences ~ exact number 6WLJOHU -: 0HQWDO DEDFXV WKH HIIHFW RI DEDFXV WUDLQLQJ RQ &KLQHVH FKLOGUHQ¶V PHQWDOFDOFXODWLRQCognitive Psychology ± 7DQDND 6 & 0LFKLPDWD7 .DPLQDJD 0 +RQGD 1 6DGDWR 6XSHULRU GLJLW PHPRU\ RI DEDFXV H[SHUWV DQ HYHQWUHODWHG IXQFWLRQDO 05, VWXG\ Neuroreport :KDOHQ - &5 *DOOLVWHO 5 *HOPDQ 1RQYHUEDO FRXQWLQJ LQ KXPDQV 7KH psychophysics of number representation. Psychological Science± :KRUI%/ Language, thought, and reality&DPEULGJH0$ 0,73UHVV :LHVH+7KHFRHYROXWLRQRIQXPEHUFRQFHSWVDQGFRXQWLQJZRUGVLingua ± :LQDZHU-1:LWWKRIW0&)UDQN/:X$5:DGH /%RURGLWVN\5XVVLDQ EOXHVUHYHDOHIIHFWVRIODQJXDJHRQFRORUGLVFULPLQDWLRQProceedings of the National Academy of Sciences ± :\QQ. &KLOGUHQ¶VXQGHUVWDQGLQJRIFRXQWLQJCognition± ;X)7KHUROHRIODQJXDJHLQDFTXLULQJREMHFWFRQFHSWVLQLQIDQF\Cognition ± ;X) (6SHONH/DUJHQXPEHUGLVFULPLQDWLRQLQPRQWKROGLQIDQWVCognition %±% 0LFKDHO&)UDQN PFIUDQN#VWDQIRUGHGX MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY /DQJXDJH'RFXPHQWDWLRQ &RQVHUYDWLRQ6SHFLDO3XEOLFDWLRQ1R 'HFHPEHU Melanesian Languages on the Edge of Asia: Challenges for the 21st Century, HGE\1LFKRODV(YDQVDQG0DULDQ.ODPHUSS ± KWWS QÀUFKDZDLLHGXOGFVS 10 KWWS KGOKDQGOHQHW Keeping records of language diversity LQ0HODQHVLD 7KH3DFL¿FDQG5HJLRQDO Archive for Digital Sources in Endangered Cultures (PARADISEC) Nicholas Thieberger University of Melbourne Linda Barwick University of Sydney At the turn of this century, a group of Australian linguistic and musicological researchers recognised that a number of small collections RI XQLTXH DQG RIWHQ LUUHSODFHDEOH ¿HOG UHFRUGLQJV PDLQO\ IURP WKH 0HODQHVLDQDQGEURDGHU3DFL¿FUHJLRQVZHUHQRWEHLQJSURSHUO\KRXVHG and that there was no institution in the region with the capacity to take UHVSRQVLELOLW\ IRU WKHP 7KH UHFRUGLQJV ZHUH QRW KHOG LQ DSSURSULDWH FRQGLWLRQVDQGVRZHUHGHWHULRUDWLQJDQGLQQHHGRIGLJLWLVDWLRQ)XUWKHU there was no catalog of their contents or their location so their existence ZDVRQO\NQRZQWRDIHZSHRSOHW\SLFDOO\FROOHDJXHVRIWKHFROOHFWRU 7KHVHSUDFWLWLRQHUVGHVLJQHGWKH3DFL¿FDQG5HJLRQDO$UFKLYHIRU'LJLWDO Sources in Endangered Cultures (PARADISEC), a digital archive based on internationally accepted standards (Dublin Core/Open Archives Initiative metadata, International Asociation of Sound Archives audio standards and so on) and obtained funding to build an audio digitisation VXLWHLQ7KLVLVDQHZFRQFHSWLRQRIDGDWDUHSRVLWRU\EXLOWLQWR ZRUNÀRZV DQG UHVHDUFK PHWKRGV RI SDUWLFXODU GLVFLSOLQHV UHVSHFWLQJ GRPDLQVSHFL¿FHWKLFDOFRQFHUQVDQGUHVHDUFKSULRULWLHVEXWUHFRJQLVLQJ WKHQHHGWRDGKHUHWREURDGHULQWHUQDWLRQDOVWDQGDUGV7KLVSDSHURXWOLQHV the way in which researchers involved in documenting languages of Melanesia can use PARADISEC to make valuable recordings available ERWKWRWKHUHVHDUFKFRPPXQLW\DQGWRWKHVRXUFHFRPPXQLWLHV 1. INTRODUCTION. At the turn of this century, a group of Australian linguistic and PXVLFRORJLFDO UHVHDUFKHUV UHFRJQLVHG WKDW D QXPEHU RI VPDOO FROOHFWLRQV RI XQLTXH DQG cc Creative Commons Attribution Non-Commercial No Derivatives Licence Keeping Records of Language Diversity: PARADISEC 240 RIWHQLUUHSODFHDEOH¿HOGUHFRUGLQJVPDLQO\IURPWKH0HODQHVLDQDQG3DFL¿FUHJLRQVZHUH not being properly housed and that there was no institution in Australia which would take UHVSRQVLELOLW\IRUWKHP7KHUHFRUGLQJVZHUHQRWKHOGLQDSSURSULDWHFRQGLWLRQVDQGVRZHUH GHWHULRUDWLQJDQGLQQHHGRIGLJLWLVDWLRQ)XUWKHUWKHUHZDVQRFDWDORJRIWKHLUFRQWHQWVRU their location so their existence was only known to a few people, typically colleagues of the FROOHFWRU7KHVHUHVHDUFKHUVGHVLJQHGWKH3DFL¿FDQG5HJLRQDO$UFKLYHIRU'LJLWDO6RXUFHV in Endangered Cultures (PARADISEC), a digital archive based on internationally accepted standards and obtained Australian Research Council Infrastructure funding to develop DQDXGLRGLJLWLVDWLRQVXLWHLQ7KLVLVDQHZFRQFHSWLRQRIDGDWDUHSRVLWRU\EXLOW LQWRZRUNÀRZVDQGUHVHDUFKPHWKRGVRISDUWLFXODUGLVFLSOLQHVUHVSHFWLQJGRPDLQVSHFL¿F ethical concerns and research priorities, but recognising the need to adhere to broader LQWHUQDWLRQDOVWDQGDUGV 2. BACKGROUND. Researchers (in particular linguists, musicologists and anthropologists) working with speakers of small languages (those with few speakers) typically conduct ¿HOGZRUNWROHDUQKRZDVSHFWVRIWKHVHVRFLHWLHVIXQFWLRQKRZWKHODQJXDJHVDUHVWUXFWXUHG or how musicological knowledge is constituted, in addition to recording life stories, HWKQRELRORJLFDODQGRWKHULQIRUPDWLRQ7\SLFDOO\WKHVHDUHPLQRULW\HQGDQJHUHGODQJXDJHV IRUZKLFKQRSULRUGRFXPHQWDWLRQH[LVWV7KLVLVYLWDOO\LPSRUWDQWZRUNZKLFKRIWHQUHFRUGV language structures and knowledge of the culture and physical environment that would RWKHUZLVHEHORVW VHHHJ(YDQV 0DI¿+DUULVRQ :KLOHLWLVW\SLFDO for the interpretation and analysis of this data to be published eventually, the raw data is UDUHO\PDGHDYDLODEOH7KHGDWD²WDSHV¿HOGQRWHVSKRWRJUDSKVDQGYLGHR²DUHRIWHQQRW properly described, catalogued, or made accessible, especially in the absence of a dedicated UHSRVLWRU\7KLVPHDQVWKDWHQRUPRXVDPRXQWVRIGDWDRIWHQWKHRQO\LQIRUPDWLRQZHKDYH on disappearing languages, remain inaccessible both to the language community itself, and WRRQJRLQJOLQJXLVWLFUHVHDUFK The data that we create as part of our research endeavour should be reusable, both by ourselves and by others, and, in particular by the speakers and the general community with DQLQWHUHVWLQWKHQDWXUHRIOLQJXLVWLFGLYHUVLW\LQ0HODQHVLD%HVLGHWKHLPSHUDWLYHWRHQVXUH there are good records of these languages this is also because any claims that we make based on that data must themselves be replicable and testable by others, and because the effort of creating the data should not be duplicated later by others, and provide a foundation WKDWFDQEHEXLOWRQ,QRUGHUWREHPDGHDFFHVVLEOHWKHGDWDUHFRUGHGE\UHVHDUFKHUVPXVW be properly collated and indexed for public presentation and archiving (see Himmelmann :RRGEXU\ +RZHYHUXQWLOUHFHQWO\WKHUHKDVEHHQQRVLPSOHPHDQVIRU GRLQJWKLVDQGDFFHVVWRSK\VLFDODQDORJUHFRUGVFDQEHGLI¿FXOWLIQRWYLUWXDOO\LPSRVVLEOH ZKHQWKH\DUHVWRUHGLQDVLQJOHORFDWLRQ This issue is being faced by scholars in many disciplines, and is being addressed under the rubric of cyberinfrastructure (1DWLRQDO 6FLHQFH )RXQGDWLRQ 86 ) or ehumanities—how to build on existing knowledge and how to add new data that is being created in the course of various research projects so that the broader research community FDQEHQH¿WIURPLW7KLVLVDOOWKHPRUHLPSRUWDQWZKHQDOLQJXLVWPDNHVWKHRQO\UHFRUGLQJV IRUDQHQGDQJHUHGODQJXDJH±RQHWKDWPD\QRORQJHUEHVSRNHQLQWKHQHDUIXWXUH$XVWUDOLD and its immediate neighbours are home to a third of the world’s languages, most of which PD\QHYHUEHUHFRUGHG0DQ\RIWKHVHODQJXDJHVFRXOGLQFOXGHFRPSOHWHO\QRYHOVWUXFWXUHV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Keeping Records of Language Diversity: PARADISEC 241 RUZD\VRIYLHZLQJWKHZRUOGEXWHDFKRIWKHPUHÀHFWVWKHKLVWRU\RIWKHLUVSHDNHUVDQG LVZRUWK\RIGHWDLOHGUHFRUGLQJ0HODQHVLDLQSDUWLFXODULVDPRQJWKHPRVWOLQJXLVWLFDOO\ GLYHUVH UHJLRQV VHH +DPPDUVWU|P 1RUGKRII WKLV LVVVXH ZLWK 9DQXDWX KDYLQJ WKH KLJKHVWGHQVLW\RIODQJXDJHVSHUSHUVRQRIDQ\FRXQWU\ 6LJQL¿FDQW UHVRXUFHV DUH QRZ EHLQJ GHYRWHG WR UHFRUGLQJ HQGDQJHUHG ODQJXDJHV LQ Europe (the Documentation of Endangered Languages project administered by the Max Planck Institute, Nijmegen) and the UK (Endangered Languages Documentation Project) DQGLQWKH86$ WKHMRLQW16)1(+SURJUDPWLWOHG'RFXPHQWLQJ(QGDQJHUHG/DQJXDJHV )XUWKHUPRUHWKHUHDUHPDQ\ORFDOLQLWLDWLYHVIRUUHFRUGLQJRUDOWUDGLWLRQOLNHWKH¿HOGZRUNHU SURJUDPPHDWWKH9DQXDWX.DOMRUDO6HQWDRUWKHFROOHFWLRQVEHLQJPDGHE\WKH$JHQFHGH Développement de la Culture Kanak (ADCK) or the Academy for Kanak Languages in 1HZ&DOHGRQLD,IWKHGDWDDULVLQJIURPDOORIWKLVHIIRUWLVQRWSURSHUO\VDIHJXDUGHGLQ our region it will represent a loss of cultural information, not to mention an enormous ZDVWHRIHIIRUWDQGPRQH\0DQ\UHFRUGLQJVDUHQRWGHVFULEHGVXI¿FLHQWO\WRDOORZWKHLU contents to be discovered, and often there is little thought given to the methods involved in managing large multimedia datasets, which are especially vulnerable because they are in digital formats that are at risk (either due to lack of suitable digital data preservation and management infrastructure, or because of format obsolescence in a fast-changing digital PHGLDHQYLURQPHQW 7RRPXFKGDWDLVVWRUHGLQZD\VWKDWPDNHLWKDUGWRDFFHVVIRUWKH UHVHDUFKFRPPXQLW\OHWDORQHWKHEURDGHUFRPPXQLW\6RPHUHVHDUFKJURXSVGHYHORSWKHLU own computational solutions which, admittedly, serve their needs well but which renders WKHJURXSDQGWKHLUGDWDLVRODWHGIURPWKHUHVWRIWKHVFLHQWL¿FFRPPXQLW\7KHGHYHORSPHQW of a new methodology, which includes the adaptation or development of new tools, must be grounded in application of that methodology to real data (%LUGDQG6LPRQV 7KHUH are too many examples of ‘proofs of concept’ which set out directions for further work but ZKLFKDUHQRWLPPHGLDWHO\DSSOLFDEOHWRDQ\UHDOZRUOGSUREOHP 3. TECHNOLOGY GAP (THE DIGITAL DIVIDE) AND MULTIMEDIA. It is a concern to some that we use increasingly technological methods for recording traditional practices, while the cultures in which they are embedded and the people who practise them have OLWWOH DFFHVV WR WKH EHQH¿WV RIIHUHG E\ WKHVH WHFKQRORJLHV +RZ DSSURSULDWH LV LW WR XVH high technology, such as digital multimedia, with languages from villages that have no electricity? Of course, there is nothing new about the gap between the resources available to the researcher and those available to the researched, this is the colonial essence of any UHVHDUFK SURMHFW UXQ E\ D ¿UVWZRUOG OLQJXLVW 6XJJHVWLQJ WKDW D YLGHR UHFRUGHU LV PRUH FRORQLDOWKDQKDQGZULWWHQQRWHV VHHIRUH[DPSOH$LNKHQYDOG LJQRUHVWKHH[WUDFWLYH nature of both forms of recording, and, more importantly, ignores the need for researchers WRPDNHWKHULFKHVWSRVVLEOHUHFRUGIRUUHXVHE\WKHVSHDNHUFRPPXQLW\:HVKRXOGWKLQN in terms of what technology is appropriate for the task, and, in the case of recording oral tradition as the basis for both linguistic research and for heritage purposes, it is clear that ZHPXVWXVHPHWKRGVEDVHGLQGLJLWDOWHFKQRORJLHV %RZGHQDQG+DMHN EHFDXVH DQDORJUHFRUGLQJIRUPDWVDQGHTXLSPHQWDUHDOOEXWREVROHWH 6FKOOHU The realisation that we can use multimedia data to enrich our understanding of performance is not something recent, and indeed goes back to the days of phonograph UHFRUGLQJVDVWKLVTXRWHIURP0DOLQRZVNLDERXWKLV¿HOGZRUNLQWKH7UREULDQG,VODQGV illustrates: MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Keeping Records of Language Diversity: PARADISEC 242 If I could, by a good phonographic record, counterfeit the living voice of 7RNXOXEDNLNL >«@,VKRXOGFHUWDLQO\EHEHWWHUDEOHWRWUDQVODWHWKHWH[WLQ WKHVHQVHRILPSDUWLQJWRLWLWVIXOOFXOWXUDOÀDYRXUDQGVLJQL¿FDQFH$JDLQ if by cinematographic picture I could reproduce the facial expression, the ERGLO\DWWLWXGHWKHVLJQL¿FDQWJHVWXUHVWKLVZRXOGDGGDQRWKHUFRQWH[WXDO GLPHQVLRQ 0DOLQRZVNL While the technology to record and play back performances has been available since the late 1800s, it was rarely used by linguists until the second half of the twentieth century, DQG HYHQ WKHQ DQDORJ UHFRUGLQJV ZHUH GLI¿FXOW WR FUHDWH LQ WKH ¿HOG DQG ODWHU DQG WR DFFHVV,WLVRQO\ZLWKWKHDGYHQWRIGLJLWDOPHGLDWKDWZHVHHWKHGHYHORSPHQWRILQVWDQW access to time points within large media corpora and the associated (but still painfully slow) realization among linguists that they can create reusable corpora in which their DQDO\VLV FDQ EH HPEHGGHG 7KLHEHUJHU ,W LV FULWLFDO WKDW D GLVWLQFWLRQ LV FOHDUO\ PDGHEHWZHHQDUFKLYDOIRUPVRIWKHPHGLD KHOGLQKLJKUHVROXWLRQ¿OHVVXFKDVELW 96 kilohertz uncompressed audio, which are described in a catalog, and given persistent location and naming) and delivery or access forms of multimedia (which will be of lower resolution and often compressed for delivery via appropriate formats, such as the web or PRELOHSKRQHV 0XOWLPHGLDSUHVHQWDWLRQVDUHVHGXFWLYHLQWKHLUDELOLW\WRUHODWHSDUWVRI FROOHFWLRQVOLQNLQJWH[WVWRPHGLDRULPDJHVDQGPHGLDWRGLFWLRQDULHV:HKDYHKRZHYHU seen enough examples of multimedia packages that are costly, contain relatively small DPRXQWVRILQIRUPDWLRQDQGEHFRPHXQSOD\DEOHDIWHUDIHZ\HDUV 4. ACCESS TO DIGITAL DATA IN THE REGION. :LOOLDPV LQDUHSRUWRQWKHVWDWXV RIGLJLWDOFRPPXQLW\VHUYLFHVLQWKH3DFL¿FQRWHGWKDW >L@QIRUPDWLRQ RQ KDUGZDUH UHVRXUFHV >«@ VKRZV WKDW ZKLOH DOO OLEUDULHV archives and museums that responded have access to at least a computer, WKH VLWXDWLRQ LV EOHDN ([FHSW IRU OLEUDULHV LQ WKH 5HSXEOLF RI 3DODX DQG presumably in the Micronesian region) and university libraries and centres LQWKH8QLYHUVLW\RIWKH6RXWK3DFL¿FQHWZRUN)LML,QVWLWXWHRI7HFKQRORJ\ Fiji School of Medicine, National University of Samoa and University of 3DSXD1HZ*XLQHDWKHFRPSXWHUVDUHXVHGE\VWDIIIRUZRUNRSHUDWLRQV,Q the Library Service of Fiji, there is no computer for public use, with only one FRPSXWHULQWKHOLEUDU\7KH6XYD3XEOLF/LEUDU\LVLQDEHWWHUVLWXDWLRQ7KH 1LXH3XEOLF6FKRRO/LEUDU\7XYDOX&XOWXUH2I¿FHDQGWKH6DPRD1DWLRQDO Archives also do not have computer access for students or members of the FRPPXQLW\ It is clear from reports such as this (and from our own observation) that there is still a ORQJZD\WRJRLQWKHSURYLVLRQRIGLJLWDOLQIRUPDWLRQLQVPDOO3DFL¿F,VODQGFRPPXQLWLHV Nevertheless, in the decade since Williams’s report there have been unexpected advances LQ DFFHVV WR GLJLWDO UHVRXUFHV LQ HYHQ TXLWH UHPRWH DUHDV RI WKH 3DFL¿F 0RELOH SKRQH technology has been taken up with enthusiasm, and has coverage in many previously unconnected locations, allowing remote use of both telephony and the ‘mobile web’ (See MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Keeping Records of Language Diversity: PARADISEC 243 3LFWXUH 7KH:RUOG%DQNµ5XUDO&RPPXQLFDWLRQ3URMHFW¶ :RUOG%DQN LQ31* DLPVWRVLJQL¿FDQWO\LQFUHDVHWKHQXPEHURILQWHUQHWXVHUVWKHUHIURPWKHHVWLPDWHGFXUUHQW PRVWO\EDVHGLQ3RUW0RUHVE\DQGWRLQFUHDVHFRYHUDJHLQUXUDOGLVWULFWFHQWHUV We can expect to see mobile phones taking over functions of portable computers in remote locations and so should also plan on building access to cultural collections using WKHVH WHFKQRORJLHV 7KH GHYHORSPHQW RI PRELOH SKRQH GLFWLRQDULHV RI VPDOO ODQJXDJHV based on common formats of lexical databases (see, for example, the PARADISEC project Wunderkammer) can already provide online or local access to electronic dictionaries with VRXQGDQGLPDJHV6LPLODUO\QHZPHWKRGVRIVWUHDPLQJGLJLWDOPHGLDDOORZIRUHI¿FLHQW GHOLYHU\RIHWKQRJUDSKLFUHFRUGLQJVRYHUORZEDQGZLGWKLQFOXGLQJPRELOHSKRQHV7KH PARADISEC project EOPAS streams audio or video recordings of stories over the internet WRJHWKHUZLWKWH[W VHHWKHGLVFXVVLRQEHORZ XVLQJ+70/DQGRSHQVRXUFHPHGLD+70/ is an emerging web standard that allows streaming of multimedia within the stamdard web page, thus obviating the need for users to install additional software or plugins (Pfeiffer $OORIWKLVLQGLFDWHVWKDWFUHDWLQJSURSHUIRUPVRIUHFRUGLQJVLPDJHVDQGVRRQWKDW conform to accepted archival standards will allow them to be transformed into delivery IRUPDWVDSSURSULDWHWRWKHFRQWH[WLQZKLFKWKH\DUHWREHXVHG FIGURE 1. Publicity billboard for internet access via mobile phones (Port 9LOD-XQH 3KRWRE\1LFN7KLHEHUJHU MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Keeping Records of Language Diversity: PARADISEC 244 5. ETHICS OF INFORMATION PROVISION. ,QDGGLWLRQWRWKHTXHVWLRQRIHTXLWDEOHDFFHVV to the kind of cultural information that is now becoming commonplace on the internet, there is the more complex issue of the sensitivity of archival records being reintroduced in QHZFRQWH[WV5HFRUGLQJVPDGHLQWKH VPD\WDNHRQDFRQVLGHUDEO\GLIIHUHQWPHDQLQJ when used today, especially if there are land disputes that otherwise rely on oral accounts UHPHPEHUHG E\ WKH FXUUHQW JHQHUDWLRQ 7KH DUFKLYDO UHFRUG FDQ DVVXPH DQ DXWKRULW\ ZKHWKHUMXVWL¿HGRUQRW WKDWPD\EHDGYDQWDJHRXVWRVRPHLQWKHSUHVHQWGLVSXWHEXW GHWULPHQWDOWRRWKHUV:KLOHWKRVHUXQQLQJDQDUFKLYHFDQEHDZDUHWKDWVXFKSUREOHPVPD\ arise, it is impossible for them to know such details for all of the locations from which the DUFKLYHVWRUHVPDWHULDO In most societies there is some kind of protocol in place for access to certain kinds of LQIRUPDWLRQ 1RW HYHU\RQH FDQ UHDG WKH UHFRUGV RI FRPSDQ\ PHHWLQJV IRU H[DPSOH RU RI VHFUHW JRYHUQPHQWEXVLQHVV ,Q VPDOOHUVRFLHWLHVVXFK SURWRFROVPD\LQFOXGHDFFHVV WRVRQJVRUVWRULHVWKDWUHODWHWRWKH¿UVWFUHDWLRQRIWKHODQGRUWRWKHWUDYHOVRIDQFHVWUDO beings: see for example Lindstrom (1990) on what he terms ‘the economy of knowledge’ in 7DQQDVRXWKHUQ9DQXDWX7KHSURYLVLRQRIVXFKLQIRUPDWLRQIURPDQDUFKLYHPD\VXEYHUW the very power structures that promote the ongoing use of traditional languages and clearly WKLVLVDSRWHQWLDOO\GLI¿FXOWVLWXDWLRQIRUDQDUFKLYLVWWR¿QGWKHPVHOYHVLQ7KH(QGDQJHUHG /DQJXDJHV$UFKLYHDW62$6KDVEHHQZRUNLQJRQDV\VWHPIRUDOORZLQJPRUH¿QHJUDLQHG DFFHVVFRQGLWLRQVWREHVSHFL¿HGLQFOXGLQJIRUH[DPSOHWKHDELOLW\IRUSHRSOHRWKHUWKDQ GHSRVLWRUVWRGHWHUPLQHZKRFDQDFFHVVWKHUHFRUGLQJVRIWKHPVHOYHVVSHDNLQJ+RZHYHU our present focus has been on preservation of the records we have located and we consider it more important that the material be stored for later reuse than that the safer option (that WKHUHEHQRDUFKLYDOUHFRUG EHDGRSWHG 6. IMPLEMENTATION OF PARADISEC. In the initial phase of the PARADISEC project ZH HVWDEOLVKHG D VWHHULQJ FRPPLWWHHZLWK UHSUHVHQWDWLYHV RI HDFK RI WKH SDUWQHU universities (initially Sydney University, the University of Melbourne, ANU, and later 81( 7KHGLUHFWRURIWKHSURMHFWLV/LQGD%DUZLFNDWWKH8QLYHUVLW\RI6\GQH\ With invaluable technical support from both the National Library of Australia and the National Film and Sound Archive and with funds from the Australian Research Council we bought a Quadriga digitisation suite and employed an audio engineer and administrative DVVLVWDQWEDVHGDWWKH8QLYHUVLW\RI6\GQH\:HDOVREXLOWDYDFXXPFKDPEHUDQGORZ WHPSHUDWXUHRYHQWRDOORZXVWRWUHDWPRXOG\WDSHVWKDWUHTXLUHGVSHFLDOFDUHEHIRUHEHLQJ SOD\DEOH7DSHVVWRUHGDWWKH$18ZHUHLGHQWL¿HGDQGORFDWHGDQGWKHQSHUPLVVLRQZDV VRXJKWIURPWKHFROOHFWRUVRUWKHLUDJHQWVWRGLJLWLVHDQGDFFHVVLRQWKHPLQWRWKHFROOHFWLRQ ,QWKH¿UVW\HDURIIXQGLQJZHKDGWRFRPHXSZLWKRXWFRPHVWKDWZRXOGMXVWLI\IXUWKHU IXQGLQJJUDQWVDQGZHDLPHGIRUKRXUVRIGLJLWL]HGWDSHVLQWKDW¿UVW\HDU ZHDFKLHYHG WKLVJRDOLQWHQPRQWKV :HZURWHDFDWDORJGDWDEDVHLQ)LOHPDNHU3URDZDUHWKDWLWZRXOG provide us with an immediately usable tool that would ultimately have to be converted WRDQRQOLQHGDWDEDVH7KLVGDWDEDVHDOORZHGXVWRUH¿QHGDWDHQWU\IRUPVDQGFRQWUROOHG YRFDEXODULHVZLWKRXWUHO\LQJRQDSURJUDPPHU7KLV¿UVWFDWDORJZRUNHGZHOODQGH[SRUWHG WRWKH;0/¿OHVUHTXLUHGIRULQFOXVLRQDVKHDGHUVLQ%URDGFDVW:DYH)RUPDW %:) ¿OHV and also exported to a static repository for Open Archives Initiative harvesting via the 2SHQ/DQJXDJH$UFKLYHV&RPPXQLW\KDUYHVWHU )LOHVJHQHUDWHGE\WKLVV\VWHP DW NK]ELW DUHODUJHDURXQG*ESHUPLQXWH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Keeping Records of Language Diversity: PARADISEC 245 VLGHRIDFDVVHWWHDQGVRUHTXLUHGHGLFDWHGVWRUDJHIDFLOLWLHV:HHVWDEOLVKHGDWDSHEDFNXS V\VWHPZKLFKUDQSHULRGLFDOO\WRFRS\¿OHVIURPWKHKDUGGLVNWRVWRUDJHWDSHVEXWZHUH fortunate when the Australian Partnership for Advanced Computing (APAC) designated 3$5$',6(& D µ3URMHFW RI 1DWLRQDO 6LJQL¿FDQFH¶ DOORZLQJ XV WR XVH WKHLU PDVV GDWD VWRUDJHV\VWHPZLWKFRQVLGHUDEOHVWRUDJHVSDFHSURYLGHGWRVXSSRUWRXUZRUN7KH\IXUWKHU provided programming support by writing specialized software (called ‘Babble’) which SURYLGHVZHHNO\PRQWKO\DQGTXDUWHUO\UHSRUWVRQWKHVWDWHRIWKHFROOHFWLRQDVZHOODV QLJKWO\TXHU\LQJWKHVHUYHULQ6\GQH\DQGFRS\LQJ¿OHVWKDWDUHUHDG\IRUDUFKLYLQJ Data is organized by collector, but also by the internal logic of the collections (the same collector working on two different languages will have two collections, or a collection of YLGHRPD\EHGLVWLQFWIURPDFROOHFWLRQRIVWLOOLPDJHV 7KHFROOHFWLRQOHYHODOVRVSHHGVXS DXVHU¶VW\SLQJLQWRWKHFDWDORJDVFRPPRQ¿HOGVIURPWKHFROOHFWLRQOHYHOFDQEHLQKHULWHG GRZQWRWKHLWHPOHYHO2XUQDPLQJFRQYHQWLRQLVUDWKHUVLPSOH µ&ROOHFWLRQ,'¶µ,WHP,'¶ µ)LOH,'¶¶H[WHQVLRQ¶ DQGLWDOVRSURYLGHVWKHKLHUDUFKLFDO¿OHVWUXFWXUHLQWRZKLFK¿OHVDUH placed and stored on the server (with directories corresponding to the collection level and VXEGLUHFWRULHVFRUUHVSRQGLQJWRWKHLWHPOHYHO 6XEVHTXHQWO\ DQG ZLWK IXQGV IURP WKH $XVWUDOLDQ 5HVHDUFK &RXQFLO /LQNDJH ,QIUDVWUXFWXUH(TXLSPHQWDQG)DFLOLWLHV /,() SURJUDPPHZHEXLOWGLJLWLVDWLRQVXLWHVLQ Melbourne and Canberra, allowing us to preserve important heritage tape collections such DVWKRVHVKRZQLQWDEOHE\QRPHDQVDQH[KDXVWLYHOLVW:LWKRXWDGHGLFDWHGLQIUDVWUXFWXUH WRGHVFULEHPDQDJHDQGVWRUHWKLVPDWHULDOLWZRXOGVLPSO\EHORVW Mark Durie (Acehnese, Indonesia) &LQG\6FKQHLGHU $SPD9DQXDWX %DUU\$OSKHU &DSH TABLE 1. Examples of collections from Australia and its the region that have either been digitised by PARADISEC or accessioned as digital data by PARADISEC Now that many researchers are recording directly to digital formats, we provide advice DQGJXLGDQFHRQVXLWDEOHIRUPDWVDQGZRUNÀRZVWRIDFLOLWDWHLQJHVWLRQLQWRWKHUHSRVLWRU\ 2QUHWXUQIURP¿HOGZRUNGHSRVLWLQJLQ3$5$',6(&SURYLGHVDPHDQVRIVHFXUHEDFNXS RIUHVHDUFKHUV¶RWKHUZLVHYXOQHUDEOHGLJLWDOPHGLD¿OHV:HVWLOOKDYHDQHHGIRUGLJLWL]DWLRQ RI ROGHU DQDORJ FROOHFWLRQV D PXFK VORZHU SURFHVV WR SURGXFH D KLJK TXDOLW\ GLJLWDO SUHVHUYDWLRQPDVWHU¿OHIRUDUFKLYLQJ International Association of Sound and Audiovisual $UFKLYHV ,$6$ 7. LICENSING USE OF ITEMS IN THE COLLECTION. The primary aim of the project to date KDVEHHQRQSUHVHUYDWLRQRIXQLTXHFXOWXUDOUHFRUGV,QFOXGLQJDOLFHQFHRULQIRUPDWLRQ about how each item can be used, is critical to the establishment of a properly curated FROOHFWLRQEHFDXVHZLWKRXWLWWKHUHLVQRZD\RISURYLGLQJDFFHVV(DFKGHSRVLWRUPXVW¿OO MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Keeping Records of Language Diversity: PARADISEC 246 RXWDGHSRVLWIRUPVSHFLI\LQJDQ\FRQGLWLRQVWKDWPD\DSSO\WRWKHPDWHULDO:HSURYLGHD default set of access conditions which any user must agree to prior to being given access to data, and depositors can choose to allow this set of conditions to govern their collection, RUWRGHWHUPLQHWKHLURZQFRQGLWLRQV:HDUHSUHVHQWO\LQYHVWLJDWLQJWKHXVHRI&UHDWLYH Commons licences as a less restrictive and more standardised form of agreement (Newman , 6HHJHU 8. DELIVERY OF ARCHIVAL MATERIAL, PAGE IMAGES AND DYNAMIC MEDIA. We provide material from the collection to those authorized to receive it, typically in the form of GRZQORDGDEOH ¿OHV KRZHYHU ZH KDYH DOVR ZRUNHG RQ VSHFL¿F PHWKRGV IRU WKH RQOLQH GHOLYHU\RIWZRNLQGVRIPDWHULDO±SDJHLPDJHVDQGWLPHFRGHGPHGLD:HPDGHDYDLODEOH LPDJHVRISDJHVRI¿HOGQRWHV VHH¿JXUH IURPWKUHHGHFHDVHGUHVHDUFKHUVXVLQJ the Heritage Document Management System with a digital camera rig that we took to the KRPHRIWKHHVWDWH¶VH[HFXWRURUWRWKHRI¿FHLQZKLFKWKHSDSHUVZHUHVWRUHG7KHVHQRWHV from deceased researchers would otherwise have only been available in a single physical ORFDWLRQ$VZHGRQRWKDYHWKHUHVRXUFHVWRNH\ERDUGDOORIWKHVHPDQXVFULSWVWKHLPDJHV DUHVWRUHGLQWKHFROOHFWLRQZLWKVXI¿FLHQWFRQWH[WXDOPHWDGDWDWRPDNHWKHPGLVFRYHUDEOH RQWKHZHE$VQRWHGHDUOLHUWKHDUFKLYDOYHUVLRQRIHDFKLPDJHLVVWRUHGVHSDUDWHO\IURP WKHUHSUHVHQWDWLRQDOYHUVLRQ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Keeping Records of Language Diversity: PARADISEC 247 FIGURE 2.3DJHLPDJHVIURPWKH:XUPFROOHFWLRQRIRQOLQHPDQXVFULSWVVKRZLQJ¿QGLQJ DLGVIURPWKHKLJKHVWOHYHO WRSOHIW WRWKHLWHPOHYHO WRSULJKW DQG¿QDOO\WKHLPDJH LWVHOI ERWWRP KWWS SDUDGLVHFRUJDX¿HOGQRWHV6$:6$:KWP 9. THE ETHNOER ONLINE PRESENTATION AND ANNOTATION SYSTEM (EOPAS). While building a method for working with our own data we consider it important to create generalisable models and structures for others to use, and to engage in discussions and WUDLQLQJVHVVLRQVERWKLQRUGHUWRUH¿QHRXUPHWKRGRORJLHVDQGWRLPSDUWQHZLGHDV$Q example of such development is our work on the online presentation of interlinear glossed text together with recorded media (EOPAS), allowing material from any language to be KHDUG LQ FRQFHUW ZLWK LWV WUDQVFULSW DQG WUDQVODWLRQ 6FKURHWHU DQG7KLHEHUJHU $ number of tools for annotating language data have been produced recently and it is clear that more are envisaged now that several large projects are engaging with these issues in WKH86$8.*HUPDQ\DQGWKH1HWKHUODQGV$QQRWDWLRQLVDEDVLFWDVNWKDWLVXQGHUWDNHQ following recording, and now it is typically carried out with time-alignment, meaning MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Keeping Records of Language Diversity: PARADISEC 248 WKDWWKHWH[WKDVUHIHUHQFHVWRWLPHSRLQWVZLWKLQWKHPHGLD¿OH XVLQJVRIWZDUHVXFKDV Elan or Transcriber) and can take several forms, the most common of which, for linguists, LVLQWHUOLQHDUWH[W7KHVHWH[WVDUHDQDO\VHGDQGSDUVHGE\DJORVVLQJWRROWKDWSURGXFHV parallel lines of text, word translation and grammatical information, together with a free WUDQVODWLRQ 7KHVH WH[WV DUH WKHQ LQSXW LQWR (23$6 D VFKHPDEDVHG ;0/ V\VWHP IRU making explicit the relationship between parts of interlinear texts together with links to WKHVRXUFHPHGLD VHH¿JXUH ZKLFKDOORZVVHDUFKLQJDQGFRQFRUGDQFLQJOLQNHGGLUHFWO\ WR WKH PHGLD (23$6 LV SRUWDEOH WKH VRXUFH FRGH LV IUHHO\ DYDLODEOH DOORZLQJ RWKHU LQLWLDWLYHVWRFDSLWDOLVHRQWKHZRUNDQGSRWHQWLDOO\GHYHORSLWLQGLIIHUHQWGLUHFWLRQV7KH ultimate aim of this approach is to allow new perspectives on the data itself, provided E\FRQWH[WXDOLVHGDFFHVVWRSULPDU\GDWDDQGWKHQWRDOORZQHZUHVHDUFKTXHVWLRQVWREH asked, and richer answers to be provided, all in a fraction of the time that it would have WDNHQZLWKDQDORJGDWD FIGURE 3.([DPSOHRIDYLGHRFOLSZLWKWLPHDOLJQHGWH[WDVSUHVHQWHGLQ(23$6 MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Keeping Records of Language Diversity: PARADISEC 249 10. CURRENT STATUS OF THE PARADISEC COLLECTION. &XUUHQWO\ ODWH 3$5$',6(&FRQWDLQVLWHPVPDGHXSRI¿OHVWRWDOLQJ7%ZLWKMXVWRYHU KRXUVRIDXGLRGDWD'LJLWDOYLGHRDOUHDG\PDNHVXSDQLQFUHDVLQJO\VLJQL¿FDQWSDUW RIWKHFROOHFWLRQ:HKROGGDWDUHSUHVHQWLQJODQJXDJHVIURPFRXQWULHV VHHH[DPSOHV of the kinds of collections in table 1) which is organized into 163 collections, some 85 RIZKLFKUHSUHVHQWQHZ¿HOGZRUNHUVZKRKDYHGHSRVLWHGPDWHULDORQWKHLUUHWXUQIURP ¿HOGZRUN DQGRQHGXULQJWKHFRXUVHRIKHU¿HOGZRUN WKXVSURYLGLQJDFLWDEOHIRUPRI WKHLUGDWDIRUWKHLURZQUHVHDUFK7KLVPHDQVWKDWLQWKHLUGLVVHUWDWLRQVDQGSXEOLFDWLRQVWKH\ can refer precisely to the relevant linguistic data through citing the timecodes associated ZLWK WKH SHUVLVWHQW LGHQWL¿HU ZHE ORFDWLRQ RI WKHLU UHFRUGLQJV LQ WKH 3$5$',6(& FROOHFWLRQ&LWDWLRQRISULPDU\GDWDLVDFULWLFDOVWHSLQFRQGXFWLQJQHZUHVHDUFKEDVHGRQ WKDWGDWD7KHUHPDLQLQJFROOHFWLRQVDUHGLJLWLVHGIURPUHFRUGLQJVPDGHVLQFHWKH V 7KHSURYLVLRQRIWKLVVHUYLFHUHTXLUHVRQJRLQJVXSSRUWDQGQHJRWLDWLRQZLWKGHSRVLWRUVDQG we have found that a key to establishing the collection has been the depositors’ perception RIWKHEHQH¿WDFFUXLQJWRWKHPDQGWRWKHLUGDWDLQKDYLQJLWZHOOGHVFULEHG,QDGGLWLRQ there are collections we know about and would dearly love to digitise but we do not have WKHUHVRXUFHVWRGRWKLVZRUN7KHVHLQFOXGHODUJHDXGLRWDSHFROOHFWLRQVDWUDGLRVWDWLRQV DURXQGWKH3DFL¿FPDQ\LQORFDOODQJXDJHVDQGFROOHFWLRQVLQUHJLRQDOFXOWXUDOFHQWUHV WKDWGRQRWKDYHDQ\ORFDOHTXLSPHQWWRGLJLWL]HWKHLUFROOHFWLRQV)XUWKHUZHDUHUHJXODUO\ DSSURDFKHGE\IRUPHUFRORQLDOSDWURORI¿FHUVRUPLVVLRQDULHVZKRKDYHUHFRUGLQJVQRWHV RUSKRWRJUDSKVWKDWWKH\ZDQWWRSUHVHUYH Arthur Capell V3DFL¿FDQG31* WDSHVDQGDUFKLYHER[HVRI ¿HOGQRWHV Tom Dutton VRQZDUGV31* WDSHV William Foley V31*WDSHV John Harris 1960s, Kiwai, PNG, 75 tapes Don Laycock 1960s, PNG, 98 tapes Al Schütz VRQZDUGV9DQXDWXVL[WDSHV Stephen Wurm V6RORPRQ,VODQGVWDSHV aWDSHVDQGWUDQVFULSWV ¿HOGQRWHV %HUW9RRUKRHYH West Papua, 180 tapes TABLE 2. Example collections that have been digitized, described or curated by the 3$5$',6(&SURMHFW :HKDYHSXEOLVKHGRQRXUZHEVLWHDGHWDLOHGGHVFULSWLRQRIRXUZRUNÀRZGHYHORSHGRYHU seven years of operation, that describes the various processes involved in locating tapes and then assessing, accessioning, digitising and describing them, managing the resulting data DQGPHWDGDWDDQGWKHUHWXUQRIRULJLQDOWDSHV3$5$',6(&KDVEHHQFLWHGDVDQH[HPSODU\ system for audiovisual archiving using digital mass storage systems by the International $VVRFLDWLRQRI6RXQGDQG$XGLRYLVXDO$UFKLYHVDQGLQZRQWKH9LFWRULDQ(UHVHDUFK 6WUDWHJLF,QLWLDWLYHSUL]HIRUKXPDQLWLHVHUHVHDUFK MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Keeping Records of Language Diversity: PARADISEC 250 2QFHZHEXLOWWKHLQIUDVWUXFWXUHIRUDUHVHDUFKUHSRVLWRU\LQFOXGLQJWKHFDWDORJ¿OH system and naming conventions, it has been taken up by those researchers who are aware RIWKHQHHGWRGHVFULEHDQGSUHVHUYHWKHLUUHVHDUFKPDWHULDO2IWHQLWLVRQO\LQWKHSURFHVV RIGHSRVLWLQJZLWK3$5$',6(&WKDWDFROOHFWLRQLV¿UVWGHVFULEHGLQDV\VWHPDWLFZD\ ±RQHWKDWWKHQDOORZVWKHGHVFULSWLRQWREHVHDUFKHGE\2SHQ$UFKLYHV,QLWLDWLYHVHDUFK HQJLQHV DQG DOVR JRRJOH (YHU\ HLJKW KRXUV WKH 3$5$',6(&FDWDORJ LV TXHULHG E\ D service run by the Open Language Archives Community (OLAC) and any new or edited FDWDORJ HQWULHV DUH FRSLHG DQG PDGH DYDLODEOH WR WKHLU DJJUHJDWHG VHDUFK PHFKDQLVP Similarly, because the catalog complies with relevant standards, the Australian National Data Service (ANDS) has been able to incorporate our 163 collections into its national VHDUFKPHFKDQLVP7KHTXDOLW\RIWKHPHWDGDWDZHSURYLGHHQVXUHVWKDWWDUJHWHGVHDUFKHV E\ODQJXDJHQDPHFDQEHUHVROYHGZLWKRXWORFDWLQJVLPLODUEXWLUUHOHYDQWIRUPV 11. REGIONAL LINKS AND TRAINING. While the initial focus for our collection was the region around Australia (as suggested by the name we chose at the outset of the project), LWKDVEHFRPHFOHDUWKDWZHQHHGWRDFFHSWPDWHULDOWKDWKDVQRRWKHUSODFHWREHDUFKLYHG Typically, this means supporting Australian researchers whose research is outside of Australia, with the geographic spread of material we house now extending from India, LQWR&KLQDDQGDFURVVWR5DSDQXL (DVWHU,VODQG :LWKOLPLWHGUHVRXUFHV3$5$',6(& KDV QHYHUWKHOHVV HVWDEOLVKHG ZRUNLQJ UHODWLRQVKLSV ZLWK FXOWXUDO FHQWUHV LQ WKH 3DFL¿F UHJLRQ HJ WKH 9DQXDWX.DOMRUDO 6HQWD RU WKH ,QVWLWXWH RI 31* 6WXGLHV ZKLFK KDYH involved providing CD copies of relevant material and, in the case of the University of 1HZ&DOHGRQLDFOHDQLQJDQGGLJLWLVLQJROGUHHOWRUHHOWDSHVLQ'UHKX$VHULRXVFRQFHUQ for many such agencies in the region (as observed in Williams’ report, above) is the lack of FRQWLQXLW\LQIXQGLQJDQGLQVWDI¿QJZLWKWKHSRWHQWLDOUHVXOWWKDWFROOHFWLRQVHVWDEOLVKHG DQGFXUDWHGRYHUWLPHPD\EHDWULVN:HZRXOGOLNHWREHDEOHWRGLJLWL]HWKHPDQ\KRXUV RIWDSHVKHOGRIWHQLQOHVVWKDQLGHDOFRQGLWLRQVLQFRXQWULHVRIWKHUHJLRQ:HKDYHEHJXQ DQRFFDVLRQDOPDVVEDFNXSRIVLJQL¿FDQWFROOHFWLRQVRIGLJLWDOPDWHULDOIURPWKH9DQXDWX .DOMRUDO6HQWDDQGZRXOGOLNHWRH[WHQGWKLVDVDVHUYLFHWRRWKHUDJHQFLHV We regularly offer training workshops in linguistic research methods, including the use RIDSSURSULDWHWRROVDQGUHFRUGLQJPHWKRGVDQGLQGDWDPDQDJHPHQWIRUHWKQRJUDSKLF¿HOG PDWHULDO7KLVLVH[WUHPHO\LPSRUWDQWDVWKHPRUHLQIRUPHGWKHUHVHDUFKFRPPXQLW\FDQ become about the need for reuse of primary data, the more likely they are to be creating well-formed data that needs no extra handling by PARADISEC to be accessioned into the FROOHFWLRQ6XFKWUDLQLQJKDVEHHQRIIHUHGDWFRPPXQLW\,QGLJHQRXVODQJXDJHFHQWUHVDV ZHOODVLQDFDGHPLFVHWWLQJV :HFRRSHUDWHLQWZRIXUWKHULQLWLDWLYHVIRUGLVVHPLQDWLQJLQIRUPDWLRQ7KH¿UVWLVDEORJ (Endangered Languages and Cultures) and the second a resource website with FAQs and a PDLOLQJOLVW WKH5HVRXUFH1HWZRUNIRU/LQJXLVWLF'LYHUVLW\ %HFDXVHRIWKHUDSLGFKDQJHV in methods for recording, transcribing, and analysing human performance no one can keep FRPSOHWHO\XSWRGDWHVRWKHVHZHEEDVHGUHVRXUFHVDUHZLGHO\TXRWHGDQGDSSUHFLDWHGE\ WKHFRPPXQLW\RIUHVHDUFKHUV MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Keeping Records of Language Diversity: PARADISEC 251 12. THE FUTURE OF THE COLLECTION. As the value of data curation becomes clearer and the use of the collection increases, we will see more theoretical work based on SURSHUO\FXUDWHGDUFKLYDOPDWHULDO:HKDYHDOUHDG\VHHQOLQJXLVWVUHWULHYLQJZKDWDUHQRZ historical language records for use in comparison with current usage and for analysis of ODQJXDJH FKDQJH 6HUHQGLSLWRXV GLVFRYHULHV LQ WKH FROOHFWLRQV KDYH LQFOXGHG WKH GUDPD VSHFLDOLVW'LDQD/RRVHU¿QGLQJDSHUIRUPDQFHRI$OEHUW7RUR¶V UDGLRVHULDO6XJDU Cane Days, a historical drama about the ‘blackbirding’ days of indentured Kanak labour LQWKH4XHHQVODQGFDQH¿HOGV:KLOHGLVFUHWHVHFWLRQVRI7RUR¶VSOD\KDGEHHQSXEOLVKHG in local literary anthologies and magazines in the early 1980s, no complete script of the SOD\ZDVDYDLODEOH7RP'XWWRQKDGUHFRUGHGWKHFRPSOHWH¿YHSDUWSHUIRUPDQFHWDNHQ in Port Moresby in the 1970s, as well as an interview with Toro about the inspiration for, DQGJHQHVLVRIWKHSOD\7KHVHXQLTXHVRXQG¿OHVDOORZHG/RRVHUQRWRQO\WROLVWHQWRWKH RULJLQDOUDGLRSOD\LQSHUIRUPDQFHEXWWRFUHDWHDYHUEDWLPWUDQVFULSWIURPWKHUHFRUGLQJ PARADISEC is a project ahead of its time and so suffers from a lack of vision among IXQGLQJDJHQFLHV,WLVWUXO\FROODERUDWLYHPXOWLLQVWLWXWLRQDODQGFURVVGLVFLSOLQDU\ZKLFK GHVSLWHIUHTXHQWIXQGLQJDJHQF\UKHWRULFWRWKHFRQWUDU\ZHLJKVDJDLQVWLWEHLQJVXSSRUWHG WKURXJKQRUPDOUHVHDUFKIXQGLQJVRXUFHV We would like to extend the streaming server we have established to allow delivery of DQ\DFFHVVLEOHPDWHULDOLQWKHFROOHFWLRQ:HDUHDOVRLQWKHSURFHVVRIGHYHORSLQJDQDFFHVV V\VWHPZLWKDXWKHQWLFDWLRQDQGDXWKRUL]DWLRQRIXVHUV PARADISEC is part of several international networks of similar projects (DELAMAN or OLAC, cited above), but is a leading exponent of linguistic data curation even among WKDW¿HOG$XVWUDOLDQJRYHUQPHQWPRYHVWRHVWDEOLVKDQDWLRQDOGLJLWDOGDWDVHUYLFH DV\VWHP of repositories hosting digital data in the way that PARADISEC has done) are still in their HDUO\VWDJHVEXWZHDUHFRQ¿GHQWWKDW3$5$',6(&ZLOOEHFRPHSDUWRIVXFKDVHUYLFH ZLWKLQ WKH QH[W GHFDGH 2XU XQLTXH FROOHFWLRQ QHHGV WR EH VDIHO\ VKHSKHUGHG WKURXJK the intervening period, identifying more collections in need of digitisation, accessioning them, and providing the infrastructure for current researchers and postgraduate students to GHVFULEHDQGSUHVHUYHWKHLU¿HOGUHFRUGLQJV:HQHHGWRFRQWLQXDOO\SURYLGHWUDLQLQJDQG advice for researchers in order that their outputs can be accessioned with minimal extra KDQGOLQJ5HVHDUFKWKDWLVFRQGXFWHGZLWKRXWDQDZDUHQHVVRIDSSURSULDWHGDWDVWUXFWXUHV and formats will result in poor outputs that need to be converted, often with considerable HIIRUWWRPDNHWKHPDUFKLYDEOH,WLVXQOLNHO\WKDWWKLVDUGXRXVFRQYHUVLRQHIIRUWZLOOEH UHVRXUFHGDQGVRZHULVNORVLQJSULPDU\UHVHDUFKGDWD 13. CONCLUSION. PARADISEC is a practice-based archive, arising from a community of practice who recognised that it was part of our professional responsibility to ensure that WKHUHFRUGVZHFUHDWHDUHSURSHUO\FXUDWHGLQWRWKHIXWXUH7KLVLVDQHZFRQFHSWLRQRID GDWDUHSRVLWRU\DFFHVVLRQLQJSULPDU\UHVHDUFKLQWKHFRXUVHRI¿HOGZRUNRUVKRUWO\DIWHU DQGEXLOGLQJPHWKRGVDQGWRROVWRIDFLOLWDWHLWVGHSRVLWDQGFXUDWRQ,WLVXQLTXHLQLWVOLQNV RQWKHRQHKDQGWR¿HOGZRUNHUVDQGWRVSHDNHUVRI,QGLJHQRXVODQJXDJHVDQGRQWKHRWKHU KDQGWRWKHFXWWLQJHGJHWHFKQRORJLHVRI:HEDQG+70/ PARADISEC has been active in locating records of small languages and making WKHP DYDLODEOH IRU ORQJWHUP DFFHVV :H KDYH EHHQ SDUWLFXODUO\ DZDUH RI WKH QHHGV RI VPDOOODQJXDJHFRPPXQLWLHVHVSHFLDOO\WKRVHLQ31*DQGLVODQG0HODQHVLD,QZH MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Keeping Records of Language Diversity: PARADISEC 252 have collaborated with the Solomon Islands Museum and Archives to apply for funding WR GLJLWLVH WKHLU DXGLR FROOHFWLRQV 6LPLODU FROOHFWLRQV RI DXGLR ¿OP DQG YLGHR H[LVW LQ DJHQFLHVDFURVVWKH3DFL¿FDQGDUHLQQHHGRIXUJHQWDWWHQWLRQ2XUQHZFDWDORJZLOOPDNH streaming media available for viewing on a variety of platforms, including mobile phones, DQG WKLV VKRXOG DOORZ GHOLYHU\ RI WKHVH XQLTXH UHVRXUFHV WR WKHLU VRXUFH FRPPXQLWLHV PARADISC is keen to attract more funding, so as to locate and digitise more material, and SURYLGHWUDLQLQJWRVSHDNHUVWRFUHDWHWKHLURZQUHFRUGVQRZ:HFRXOGDOVRLQFUHDVHWKH representation of languages in our EOPAS system to provide online samples of as many ODQJXDJHVRIWKHUHJLRQDVSRVVLEOH7KHUHLVPXFKPRUHWREHGRQHEXWWKHZRUNGRQHE\ 3$5$',6(&ZLOODOORZIXWXUHZRUNWRJURZRQJRRGIRXQGDWLRQV REFERENCES $LNKHQYDOG$OH[DQGUD/LQJXLVWLF¿HOGZRUN VHWWLQJWKHVFHQHSprachtypologie und Universalienforschung %LUG 6WHYHQ *DU\ 6LPRQV 6HYHQ GLPHQVLRQV RI SRUWDELOLW\ IRU ODQJXDJH GRFXPHQWDWLRQDQGGHVFULSWLRQLanguage %RZGHQ-RKQ -RKQ+DMHN:KHQEHVWSUDFWLFHLVQ¶WQHFHVVDULO\WKHEHVWWKLQJWR GR GHDOLQJZLWKFDSDFLW\OLPLWVLQDGHYHORSLQJFRXQWU\,Q%DUZLFN/LQGD 1LFKRODV 7KLHEHUJHU HGV 6XVWDLQDEOH GDWD IURP GLJLWDO ¿HOGZRUN 6\GQH\ 6\GQH\ 8QLYHUVLW\3UHVV (YDQV1LFKRODV Dying words: Endangered languages and what they have to tell us 0DOGRQ0$ :LOH\%ODFNZHOO +DUULVRQ.'DYLGWhen languages die: The extinction of the world’s languages and the erosion of human knowledge1HZ MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Keeping Records of Language Diversity: PARADISEC 253 6FKURHWHU 5RQDOG DQG 1LFKRODV 7KLHEHUJHU (23$6 WKH (WKQR(5 RQOLQH UHSUHVHQWDWLRQRILQWHUOLQHDUWH[W,Q%DUZLFN/LQGDDQG1LFKRODV7KLHEHUJHU HGV 6XVWDLQDEOHGDWDIURPGLJLWDO¿HOGZRUN 6\GQH\ 6\GQH\8QLYHUVLW\3UHVV 6FKOOHU'LHWULFK6DIHJXDUGLQJWKHGRFXPHQWDU\KHULWDJHRIFXOWXUDODQGOLQJXLVWLF GLYHUVLW\Language Archive Newsletter 6HHJHU$QWKRQ\1HZWHFKQRORJ\UHTXLUHVQHZFROODERUDWLRQV &KDQJLQJRXUVHOYHV WREHWWHUVKDSHWKHIXWXUHMusicology Australia 27 7KLHEHUJHU1LFKRODV 6WHSVWRZDUGDJUDPPDUHPEHGGHGLQGDWD,Q(SSV3DWULFLD $OH[DQGUH$UNKLSRY HGV New challenges in typology: Transcending the borders DQGUH¿QLQJWKHGLVWLQFWLRQV %HUOLQ 1HZ Nicholas Thieberger WKLHQ#XQLPHOEHGXDX Linda Barwick [email protected] MELANESIAN LANGUAGES ON THE EDGE OF ASIA: CHALLENGES FOR THE 21ST CENTURY Nasal m