Social Inequality among Sama-Bajau Migrants in Urban Settlements : A Case from Davao City (<Special Theme>Continuity and Reorganization of Identities among the Transnational Communities in East and Southeast Asia)

Social Inequality among Sama-Bajau Migrants in

Urban Settlements : A Case from Davao City

(<Special Theme>Continuity and Reorganization

of Identities among the Transnational

Communities in East and Southeast Asia)

著者名(英)

Waka Aoyama

journal or

publication title

Hakusan jinruigaku

number

15

page range

7-44

year

2012-03

URL

http://id.nii.ac.jp/1060/00002420/

Creative Commons : 表示 - 非営利 - 改変禁止

http://creativecommons.org/licenses/by-nc-nd/3.0/deed.ja

Hakusan　Review　ofAnthropology　Vol．15．　March，2012

Social　Inequality　among　Sama－Bajau　Migrants　in

　　　Urban　Settlements：A Case　from　Davao　City

AOYAMA　Waka★

Keywords：Social　inequality，　Sama－Bajau，　migrants，　urban　settlements，　Davao　City

I　INTRODUCTION

As　the　literature　on　the　Sama－Bajau　in　the　maritime　Southeast　Asia　has　revealed，　the Sama　Dilaut（the　Sama　of　the　sea）abandoned　their　boat－dwelling　practices　by　the　mid　20th

century．　They　were　said　to　be　amalgamated　into　other　Islamized　groups　within　the

encompassing　society，　and　it　was　indeed　what　happened　to　many　former　boat－dwellers　in

Sulu　Archipelago［Nimmo　2001；Sather　1997］．　Armed　conflicts　in　the　region　and　the

fbllowing　deterioration　of　peace　and　security　order　and　economic　hardship　caused　many Filipino　Sinama－speaking　groups　of　people　to　migrate，　legally　or　not，　down　south　across　the border　to　Sabah，　Malaysia　［Kassim　2009］．There　were　also　other　groups　who　fled　Jolo，　Siasi and　Zamboanga　to　other　places　up　north　in　Mindanao，　the　Visayas　and　Luzon．　These　people have　cast　an　image　of　the“Bajau”as“beggars”in　some　cities，　unlike　their　counterparts　in Malaysia［Nagatsu　2001］．　Little　has　been　known　about　their　life　in　non－maritime　urban situations　on　the　side　of　the　Philippines．　Although　there　is　at　least　one　documentary　film　on the　Sama・Dilaut　in　the　urban　settlements　directed　by　Nannette　Matial　in　2008，　no sociological　or　anthropological　studies　have　been　made　to　describe　their　social　and　ethnic identities　and　everyday　life　in　the　urban　market　society，　out　of　their　sea－based　histories． 　　　Given　the　current　situation　of　the　Sama－Bajau　studies，　the　ultimate　goal　of　my　research

is　to　write　an　ethnography　of　the　Sinama－speaking　urban　migrants　in　Davao　City，

Mindanao，　the　Philippines，　in　order　to　explore　how　they　have（re－）organized　their　life　in “poverty”and　become　our　neighbors，　both　literally　and　metaphorically　in　the　urban　market society　in　the　globalized　world．　The　present　paper　will　focus　the　social　inequality　among

★　青山和佳，北海道大学大学院・メディアコミュニケーション研究院；Faculty　of　Media　and

　　　Communication，　Hokkaido　University，　North－17，　West・8，　Kita－ku，　Sapporo　O60・0817，　Japan／ 　　　e－mail：waka＠imc．hokudai．ac．jp 　　　　　　7

Hakusan　Review　ofAnthropology　Vol．15，　March　2012 them：Iwill　speci丘cally　study　the　community　social　status　of　households　and　the　criteria that　have　been　identified　through　their　own　eyes　as　members　of　the　community．　The analytical　framework　and　method，　which　are　to　be　detailed　in　the　fbllowing　chapters，　are based　on　those　originally　suggested　by　Fr．　Lynch［1959】to　examine　social　class　in　a community　in　the　PhilipPines． 　　　The　bulk　of　the　data　fbr　the　present　study　had　been　gathered　through　a　long－term 丘eldwork　starting　from　1998．　This　study　aims　to　provide　a　small　but　a　panoramic　view　of the　Sama・Bajau　community　in　question．　It　is　also　meant　to　search　f（）r　units　of　observation and　analysis　more　appropriate　fbr　further　research，　rather　than　the“village”or　the“ethnic group”as　a　whole　which　had　been　often　used　as　a　unit　of　analysis　when　the　Sama－Bajau still　shared　a　limited　number　of　sea－related　livelihood（mostly　fishing　with　boats）．　Now　that the　livelihood　underlying　their　social　and　ethnic　identities　has　clearly　disapPeared　in　the context　of　Davao　City　at　the　turn　of　the　20th　century，　we　need　to　find　the　units　more　relevant to　understanding　their　way　of　life　and　its　change． 　　　It　should　be　noted　here　that　it　was　problematic　enough　fbr　me　to　actually　identify　who the“Bajau”or　the　Sama・Dila・ut・are　in　Davao　City1）．　As　you　will　find　out　later，　the“Bajau”in the　present　study　is　actually　a　mixture　of　the　Sama　Dilaut（Sama　of　the　sea），　as　I　have mentioned　earlier，　and　the　Sama　Deyaq（Sama　of　the　shore），　those　who　have　always　lived on　the　land　instead　of　being“Sea　Nomads”and　who　used　to　clearly　segregate　themselves from　the　fbrmer．　Both　of　them，　however，　are　identified　as　the“Bajau”in　the　urban　milieu． Both　call　themselves　the“Bajau”under　many　circumstances．　So，　in　this　paper，　I　will　use　the Sama，　the　Bajau　and　the　Sama－Bajau　interchangeably　according　to　the　context，　while　the identifications　they　used　originally　to　refer　to　themselves　in　any　specific　context　will　be retained　in　case　of　quoting　their　own　narratives． 　　　This　research　makes　three　contributions．　As　a　community　study　of　the　urban　poor，　it offers　an　alternative　research　method　which　incorporates　the　norms　and　values　of　the people，　rather　than　purely　objective　indicators　such　as　daily　income　or　consumption　per

head，　into　the　framework　to　analyze“diversity”as　well　as“commonality”within　the

communityl　In　Sama　Studies，　it　adds　a　unique　comparison　between　the　Sama－Bajau　as non－“唐?＝| 垂?盾垂撃?h，　who　struggle　to　create　a　unique　subculture　in　given　situations　and　the Islamized　Sama　in　maritime　Southeast　Asia［Nimmo　2001；Sather　1997］．　As　a　Philippine ）

18

It　is　spelled“Badjao”in　Cebuano，　a　linga　franca　in　the　Visayas　and　Mindanao．

AoYAMA：Social　lnequality　among　Sama－Bajau　Migrants　in　Urban　Settlements study，　it　presents　stories　of　marginalized　minorities　who　are　rarely　seen　in　official　parades of　indigenous　peoples　in　the　town’s　main　streets，　whose　cultural　traits　are　ignored　because of　having　little　social　value　to　protect　or　market，　but　who　are　undeniably　part　of　the Philippines　as‘‘a　singular　but　a　plural　place”［Steinberg　2000］． 　　　Following　this　introduction，　this　paper　comprises　the　five　chapters　namely　Question， Setting，　Method，　Findings　and　Conclusions．

II　THE　QuESTION

1．　Literature　review：social　inequality　as　a　subjective　ranking　of　social　status Social　class　and　social　stratification　have　been　one　of　the　primary　concerns　in　sociology． Empirical　studies　in　industrialized　countries　such　as　the　United　States　and　Japan　abound． In　cases　of　developing　countries　including　the　Philippines，　a　wide　disparity　between　haves and　have－nots　has　been　one　of　the　central　issues　in　social　studies．　Such　researches，　whether conducted　in　developed　countries　or　in　developing　countries，　normally　utilize　the　results　of large・scale　national　survey　conducted　by　public　and　private　institutions．　In　either　case， however，　investigations　into　social　status　among　people　in　specific　sectors　of　a　nation　are scarce．　Furthermore，　when　it　comes　to“disadvantaged”groups　such　as　indigenous　peoples and　ethnic　minorities，　the　image　of　an“egalitarian　and　homogenous”society　can　often　gloss OVer　UnCOmfOrtable　realitieS． 　　　Fr．　Frank　Lynch　was　the　first　anthropologist　to　conduct　a　community－based　study　on social　class　in　the　Philippines　in　1950s．　He　chose　a　lowlander　Christian　society　in　a　town called　Canaman　located　in　the　rice・growing　Bicol　Province　of　the　southern　Luzon．　He　stayed 20months　to　gather　empirical　data　in　the　area． 　　　Acomm皿ity　can　be　divided　into：1）Political，2）Economic　occupational　and　3）Social classes． 　　　Social　classes　are　the　most　popular　and　most　ambiguous　among　the　three　because　the value　scale　is　not　simple　Iike　those　of　political　and　economic　occupational　classes． 　　　Lynch’s　major　contribution　lies　in　the丘rst　apPlication　of　a　method　that　produces　a composite　scale　to　empirically　analyze　a　social　class　in　the　Philippines．　In　the　case　of“social classes，”Lynch　stresses，“the　individual　is　being　rated　on　many　scales　simultaneously；he　is being　considered　according　to　his　position　in　the　economic　and　political　power　systems，　the occupational　structure，　the　community　scale　of　moral　behavior，　and　other　related　scales” 　　　　　　9

Hakusan　Review　ofAnthropology　Vol．15，　March　2012 ［Lynch　1959：8］． 　　　Based　on　the　results，　he　wrote　Socia／Class　1［刀aBiko11わ㎜［Lynch　1959］，　which revealed　that　there　were　two　social　classes　that　complement　each　other　in　social，　ecollomic and　political　relationship　in　the　community．　It　also　proved　that　such　ranking　of　individual households　was　frequently　related　to　the　level　of　their　wealth　and　economic　security． 　　　The　next　important　research　on　this　type　of　social　inequality　in　the　Philippines　was conducted　by　James　F．　Eder，　who　studied　Cuyo　migrants　in　a　settlement　in　San　Jose，　a pioneer　farming　village　located　in　Palawan　lsland，　in　1970s．　His　research　resulted　in　a　book entitled　閲o　Sha」η　Sueceθdタ　：∠48打ヒrultui「al　develoρment　a刀ゴ　soeial　inequality　on　a 肪吻ine　fi刀］ロtier［Eder　2009（1982）］． 　　　In　his　book　he　extended　the　method　of　analysis　developed　by　Lynch，　and　recorded　the emergence　and　institutionahzation　of　social　inequality　as　he　ranked　social　status　in　the lifetime　of　the　single　community．　He　provided　a　description　of　how　the　residents　in　the　Cuyo settlement，　which　was　initially　homogenous　both　in　economic　and　social　aspects，　grew unequal　as　they　experienced　development　on　the　agricultural　frontier　over　the　years．　His conclusion　stressed　the　diversity　of　circumstances　and　the　diversity　of　human　motives　and capabilities　to　explain　the　emergence　of　social　inequalities． 　　　Both　Lynch　and　Eder　applied　a　research　method　that　integrated　quantitative　data　and qualitative　data，　utilizing　sociological　survey　and　anthropological　observation』b　measure the　social　status　among　the　people　in　the　communities，　each　of　them　carefully　selected

members　of　the　community　as　a　rating　panel　and　then　took　interviews　with　them

indiVidually．　The　raters　were　asked　to　place　the　families　of　the　community　in　order according　to　their　own　criteria　fbr　higher　and　lower　standing．　In　other　words，　such placement　also　used　the　subjective　perceptions　of　the　individual　raters　as“insiders”of　the communities　rather　than　used　only　objective　indicators　such　as　the　level　of　income．　Most importantly，　the　concept　of　social　status　in　the　two　studies　was　presented　as　a　composite scale　that　integrated　various　subjective　infbrmation　each“insider”rater　provided　about　the social　status　of　the　rated　members；it　was　not　considered　as　something　measurable　by　a simple　and　single　index　such　as　wealth，　income　and　prestige．　In　this　sense，　the measurement　of　social　status　suggested　in　their　studies　can　be　called“total　status judgment，”or“overall　social　standing．” 　　　This　method　has　two　advantages　in　the　measurement　of　social　inequality：（1）Since　it emphasizes　the　subjective　judgment　of　the　people　in　specific　milieus，　it　helps　reduce　the

AOYAMA：Social　Inequality　among　Sama・Bajau　Migrants　in　Urban　Settlements possibility　fbr　the　researcher，　who　comes　from　a　different　culture　as　well　as　has　been trained　in　a　particular　academic　discipline，　to　naively　choose　a　priori　indices　inappropriate fbr　the　measurement　of　well－being　among　the　people　in　the　research　site；and（2）generating asubjective　and　composite　scale　of　social　status，　it　enables　the　researcher　to　comprehend the　sense　of　social　inequality　among　the　people　from　multiple　perspectives　including　both economic　and　non－economic　ones． 　　　This　method　is　of　course　not　problem・free．　And　its　weakness　lies　in　its　strength：the subjective　and　composite　scale　obtained　from　data　analysis　cannot　be　immediately　broken down　into　single　indices　in　an　attempt　to“reconstruct”the　studied　society　in　a　more “objective”description．　As　Lynch　and　Eder　carried　it　out，　however，　this　problem　could　be solved　through　a　careful　collection　of　the　criteria　the　raters　used　for　placement．　The　diverse criteria　suggested　by　the　rater　could　be　aptly　reduced　to　a　certain　number　of　categories， each　of　which　is　given　weight　according　to　the　number　of　the　raters　who　employed　the　same criterion　or　those　similar　in　essence．　These　procedures　will　reveal　the　categories　behind　the ranking　of　social　status　among　the　members　in　the　community．　With　these　categories，　the researcher　is　ready　to　examine　the　values　and　norms　embedded　in　the　community　and therefore　regulate　the　behavior　of　its　members． 　　　Although　this　method　may　appear　primarily　quantitative　in　data　collection，　it　does require　a　certain　degree　of　rapport　to　be　established　between　the　researcher　and　the researched．　As　the　method　is　meant　to　explore　the　subjective　evaluations　of　their　well－being in　everyday　life，　integration　into　the　community　is　essential　fbr　the　researcher　throughout the　course　of　the　research，　which　includes　design，　implementation，　data　analysis　and　its interpretation．　Both　Lynch　and　Eder　demonstrated　in　their　studies　that　the　method　will　be most　applicable　when　properly　combined　with　other　data　collected　about　the　community from　related　aspects．　Such　data　will　be　also　helpful　to　understand　the　context　in　which　the subjective　scale　on　social　inequality　is　analyzed　and　interpreted．　Thus，　it　is　fundamental　to collect　background　inf（）rmation　of　the　community，　including　socioeconomic　characteristics　of households，　prior　to　the　implementation　of　the　social　inequality　survey． 2．Analytical　framework：definition　of　key　words　and　key　hypotheses　of　this　study To　make　it　clear，　the　study　reported　here　is　f（）cused　on　the　social　inequality　among　the Sama－Bajau　in　urban　settlements　in　Davao　City，　the　Philippines．　As　to　the　definition　of　the keywords，　this　study丘）llows　those　suggested　by　Lynch［1959］： 11

Hakusan　ReView　ofAnthropology　Vol，15，　March　2012 1．“Social　status”of　individuals　means　the　subjective　rating　given　by　the　rater．　Then individuals　may　be　arranged　according　to　social　status　in　a　kind　of　continuum，　or　in ‘‘_{窒≠獅汲奄獅〟D”} 2．Systems　of“social　stratification”should　be　distinguished　from　systems　of　ranking： the　former　are　present　only　where　there　are　at　least　two　separate　strata　with　a qualitative　difference　between　them． 3．Systems　of“social　stratification”here　are　called　systems　of“social　inequality” featuring　strata　culturally　recognized　as　qualitatively　distinct． 4．One　kind　of　stratification　is“class，”which“may　be　de丘ned　as　a　system　of　social inequality　featuring　two　or　more　qualitatively　distinct　orders　of　people　ranked　as superior　and　inferior，　with　relative　ease　of　mobility　between　these　two．” 5．The“inequality”have　three　dimensions；namely，　the　political，　economic　occupationa1， and　the　social． 　　　Now，　to　explore　the　social　inequality　in　the　Sama－Bajau　community　in　the　research　site， 1　wil1　state　the　matter　more　concretely，　and　in　the　fbrm　of　hypotheses： 1．The　community　of　the　Sama－Bajau　in　the　research　site　is　composed　of　individual households　that　can　be　arranged　according　to　social　status　in　a　continuum，　or　ranking． 2．There　is　an　implicit　agreement　among　the　raters　that　the　community　has　at　least　a system　of　status　ranking． 3．There　are　shared　criteria　among　the　raters丘）r　the　placement　of　the　households　such as　social，　economic　and　political　categories．

III　THE　SETTING

This　chapter　summarizes　the丘ndings　from　the　fieldwork　I　carried　out　using　separate　but

related　approaches　over　thirteen　months　from　February　1998　to　March　1999．　This

summary　is　aimed　at　providing　the　reader　with　background　infbrmation　about　the　research site　aro皿d　the　time　when　I　conducted　a　research　on　social　inequality　among　them　in　1999． Specifically，　1　gathered　a　baseline　data　through　a　household　survey　in　the　research　site　from February　2　to　March　6，1999，0f　which　results　will　appear　in　the　last　part　of　this　chapter．　I have　used　ethnographic　present　to　describe　the　Iocal　contexts　the　Sama－Bajau　in　Davao

AOYAMA：Social　Inequality　among　Sama・Bajau　Migrants　in　Urban　Settlements City　were　in　during　those　days． 1．Impoverishlnent　of　the　Sama－Bajau　in　T（）day’s　PhilipPines As　we　mentioned　in　the　introduction，　the“Bajau”in　the　urban　contexts　in　the　PhilipPines have　been　known　fbr　its　impoverishment　and　marginalization．　The　name　of　the“Bajau” may　be　even　used　as　a　synonym　of“beggar，”aderogatory　term［Nagatsu　2001：226］．Unlike the　Sama　Dilaut　in　Malaysia，　the　Sama　in　the　Philippines　lacks　an　official　vehicle　fbr upward　social　mobility2）． 　　　Government　policies　fbr　ethnic　minorities　in　the　Philippines　are　considered　rather advanced　in　Southeast　Asia．　As　a　normative　framework，　the　1987　Constitution　stipulates that　the　rights　of　indigenous　communities　and　cultural　minorities　and　their　diversities should　be　respected　3）．　As　an　institutional　framework，　the　nation　is　known　f（）r　its “Indigenous　Peoples　Rights　Act　of　1997”（Republic　Act　No．8371），　which　acknowledges　the rights　of　such　communities　to　their　ancestral　domains．　Moreover，　an　autonomous　region　has

been　officially　created　fbr　Muslims　in　the　Southern　Philippines，　aside　from　an

administrative　region　in　Cordillera　inhabited　by　other　ethnic　minorities． 　　　Despite　the　establishlnent　of　these　f｛）rmal　frameworks，　most　of　the　ethnic　minorities are　still　tantalizingly　out　of　the　reach　of　the　benefits　of　social　integration［UNDP　2004：68］． This　is　partly　because　these　frameworks　implicitly　set　the　prerequisite　of　community organization　befbre　they　can　exercise　their　rights　as　individual　members　of　a　minority group　moving　towards　social　integration．　In　addition，　it　requires　a　considerable　amount　of documentation　to　complete　the　procedures，　which　people　with　disadvantage　in　educational attainment　would丘nd　it　virtually　impossible　to　process　by　themselves　without　support from　concerned　parties　such　NGOs．　In　reality，　there　are　quite　a　few　groups　of　ethnic minorities　who　find　it　hard　to　organize　themselves　and　stand　up　f（）r　their　rights．　One　of them　is　the　Sama・Bajau　in　urban　settings．

　　　It　has　been　often　noted　that　social　organizations　among　the　Sama・Bajau　are

comparatively　loose　and　transient　with　leaders　who　individually　lead　only　a　small　number of　families．　However，　recent　studies　by　Nagatsu　on　the仇晦α（or　Bajo）in　Indonesia　have 2）　Conversion　to　Islam，　however，　has　led　to　reconfigure　their　identities　as“legitimate　Malay”and 　　　concomitantly　proVided　them　a　strong　vehicle　to　move　upward　in　the　local　society［Nagatsu　2001］． 3）　For　reViews　on　normative　and　institutional　frameworks　of　the　state　policies　f（）r　indigenous　peoples 　　　in　the　Philippines，　see　ILO［1993］，Eder　and　McKenna（eds．）［2004］． 　　　　　　13

Hakusan　ReView　ofAnthropology　Vol．15，　March　2012 reported　the　emergence　of　alliances　that　bring　scattered　Sama　communities　in　different islands　together　and　form　a　social　movement，　in　response　to　changing　incentives　in　local politics　and　interventions　by　international　NGOs　such　as　WWF4）．　Such　movement　cannot　be seen　among　the　people　called　the“Bajau”in　the　urban　areas　of　the　Philippines．

2．The　Sama　in　Davao　City

Furthermore，　it　is　difficult丘）r　the　government　to　provide　protection　fbr　the“Bajau”in　the urban　contexts，　including　the　present　case　in　Davao　City．　Most　of　them　are　originally refugees，　who且ed　their　home　waters　or　lands　due　to　the　deterioration　of　peace　and　order and　the　subsequent　economic　predicaments　in　Sulu　and　Zamboanga　regions． 　　　Davao　City　is　we11－known　fbr　its　ethnic　diversity，　which　is　considered　as　a　major resource　f（）r　its　regional　development　in　tourism．　For　instance，　the　city　annually　celebrates “Kadayawan，”which　is　a　fもstival　featuring　the　diversity　of　local　ethnic　and　cultural communities　with　various　kinds　of　presentations　such　as　parades　in　the　streets　and handicrafts　in　trade　fairs、　However，　it　does　not　seem　that　the　Sama－Bajau　in　Davao　City　is fully　recognized　as　such　an　ethnic　minority　with　such“marketable”cultural　resources． 　　　People　tend　to　think　that　the“Bajau”poverty　is　due　to　their‘‘laziness”or‘‘apathジ’In fact，　people　commonly　adopt　a　negative　image　of　the“Bajau”way　of　life　and　attitudes　and attribute　their　poverty　to‘‘poor”or“bad”culture．　In　this　social　milieu，　even　if　the Sama－Bajau　were　selected　as　a　target　group　fbr　any　poverty　eradication　program　of　the government，　it　might　be　simply　treated　as　part　of　the　urban　poor　with　no　specific　cultural tralts　to　apPreclate． 　　　Besides，　resources　f（）r　the　urban　poor　are　not　only　limited　but　also　unevenly　allocated． Because　most　of　the　Sama・Bajau　still　do　not　vote，　they　lack　connections　with　politicians　and this　deprives　them　of　access　to　various　resources．　Small　in　number　and　politically unimportant，　the　Sama－Bajau　are　disadvantaged　in　competing　with　more　dominant　ethnic groups　over　the　distribution　of　the　government　budget　and　resourees　that　aid　agencies　from the　outside　bring　in． 4）　Apresentation　of　Dr．　Nagatsu　Kazufumi（T（）yo　University）titled“An　Incubation　Study　on　Social 　　　Dynamics　of　the　Maritime　Southeast　Asia：Focusing　on　the　Population　Flow　and　the　Making　of　Sea 　　　Folks，”at　the　Annual　Report　Meeting　of　FY　2010，　lnternational　Program　of　Collaborative　Research， 　　　Center　for　Southeast　Asian　Studies，　Kyoto　University，　March　24，2012．

AOYAMA：Social　lnequality　among　Sama・Bajau　Migrants　in　Urban　Settlements 3．　Research　Site Davao　City　is　located　on　Mindanao　lsland　in　the　Southern　Philippines（Figure　1）．　Blessed with　natural　resources（e．g．　minerals，　and　farm　and　marine　products）and　other　natural advantages　in　Mindananao5），　the　city　is　known　as　a　center　of　the　development　in　the　region， serving　as　one　of　the　leading　Production　bases　of　agri－industry　and　also　as　a　major trade－oriented　port　city　in　the　country．　As　of　May　2000，　it　was　home　to　a　population　of approximately　1．15　million　people　and　ranked　fburth　in　size　as　a　single　chartered　city nationwide　and　the丘rst　in　Mindanao． 　　　With　relatively　stable　peace　and　public　order，　Davao　City　absorbs　migrants　and refugees　from　trouble－plagued　areas　such　as　the　southwest　regions　of　Mindanao　and　other inland　regiolls，　where　political　stability　is　yet　to　be　seen．　During　the　1970s，　the　armed con且ict　between　the　Philippine　government　and　the　Muslim　separatists　reached　a　peak；it ravaged　local　communities，　resulting　in　the　evacuation　of　the　people　from　Zamboanga　and Sulu　to　other　cities　like　Davao．　Some　of　the　pioneer　migrants　to　the　research　site　in　our study　were　refUgees　at　that　time．　Due　to　the　destabilized　political　situation　and　the accordingly　deteriorating　economics　in　those　areas，　people　continued　to　flock　to　Davao　City． The“Bajau”and　their　Muslim　neighbors　in　the　present　paper　can　be　equally　considered　as refugees　in　this　sense．　They　both　f（）rm，　more　or　less，　the　urban　poor　problem　in　the　local context　of　the　city．

Egure　1　Map

5）　Including　the　Sulu　Archipelago．

15

Hakusan　Review　ofAnthropology　Vol．15，　March　2012 　　N

w‘L

〈’v・’”SibutY 　　　　L二＿＿」

㌢

　　　The　research　site　is　situated　in　an　islet　called‘‘Isla　Bella6），”which　is　a　squatter　area　on the　eastern　shore　of　the　city．　The　government　was　planning　to　release　the　ownership　of　the land　to　the　residents　at　the　time　of　our　research．　As　of　1997，　the　islet　had　an　area　of　about nine　hectares．　The　estimated　number　of　households　was　1，300　with　a　population　of　about 10，000people．　By　ethno－linguistic　grouping，　approximately　60％of　the　total　population belonged　to　major　Muslim　groups（e．g．　Maranao　and　Tausug）．　In・other・words，　the　Muslims were　the　majority　in　the“Barangay”（the　smallest　administrative　unit　in　the　Philippines）， although　in　the　larger　context　of　Davao　City　they　were　extremely　small　in　number　against Christian　populations（mostly　the　Cebuano）7）．　Indeed，　the　Captain　and　seven　of　the Barangay　Councilors　were　the　Maranao，　and　the　area　was　known　as　a　Muslim－dominated area　in　the　city．

　　　We　attempted　complete　enumeration　of　the　184　households　in　two　Sama－Bajau

））

ハ07

Some　of　the　names　of　indiViduals　and　places　are　changed　for　the　protection　of　privacy． According　to“Ethnicity”（self－claimed）in　the　2000　population　census，　Cebuano　accounted　for 33．32％of　the　total　population　of　1，145，033　in　Davao　City．　Meanwhile，　of　the　three　major　Muslim groups　in　the　Philippines，　the　Tausug　only　accounted　fc）r　O．79％and　the　Maranao　O．66％，　and　the Maguindanao　accounted　f（）r　O．17％．　By　the　way，　although　the　accuracy　of　the　survey　is　doubtfu1　in comparison　to　the　population　of　the　study　site，　the　Bajau（Sama．Di7aut）accounted　for　O．08％（928 people）and　Sama　Dilaya　O．0024％（28　people）［NSO　2000】．　In　the　context　of　Davao　City， non－Christian　populations　are　considered　as　minorities．　In　terms　of　indigenicity　in　Davao， Christians　originally　from　North（the　ethnic　Cebuanos　mostly），　lowland　Muslims（the　Maranao　and the　Maguindanao），other　ethnic　groups　from　Zamboanga　and　Sulu（the　Sama，　the　Laminusa　and the　Tausug）are　should　be　categorized　all　as　migrants［Hayase　2007］．

AoYAMA：Social　Inequality　among　Sama－Bajau　Migrants　in　Urban　Settlements setlements（of　which　172　households　are　from　the　area　called　Macao8），　and　12　households from　China　Pikas　respectively）．　T（）make　it　a　comparative　study，　we　also　collected　data　of

the　entire　180　non－Bajau　households　in　non－Bajau　dominated　communities　g）．　The

Sama－Bajau　settelements　embrace　a　population　of　981（of　which　51．7％were　male，47．5％ female　and　O．8％unknown），　while　the　non－Bajau　settlements，1，044（48．1％male，51．5％ female　and　O．8％unknown）．　The　non－Bajau　consist　of　f（）ur　groups，　namely，　the　Cebuano， Tausug，　Maranao　and　Laminusalo）．　The　Cebuano　are　Christians，　whereas　the　latter　three are　Muslims．　The　Sama－Bajau　consist　both　land－based　and　originally－sea－based　Sinama speaking　groups　from　various　places，　mostly　from　the　region　of　Zamboanga　del　Sur　with some　from　that　of　Sulu． 　　　While　Table　l　and　Table　2　summarize　the　key　socioeconomic　characteristics　of　the residents　in　the　two　settlements，　we　will　f（）cus　on　their　livelihood　in　the　next　chapter　f（）r　the purpose　of　this　study．　The　shift　of　livelihood　is　one　of　the　major　changes　they　have experienced　as　they　migrated　to　the　city．　It　also　helps　us　understand　their　confinement　to the　marginalized　segment　of　the　urban　labor　market． 4．The　Sama－Bajau　as　the　Urban　Poor（at　the　margin　of　the　urban　labor　market） It　is　fbund　in　our　1998／99　household　survey　that　one　of　the　significant　differences　in socioeconomic　characteristics　between　the　Sama－Bajau　and　non－Sama－Bajau　is　educational attainment．　The　data　f（）r　this　item　was　collected　only　from　the　respondents　of　age　7　and　over． Among　the　Sama－Bajau，　the　average　length　of　schooling　is　1．2　years　f（）r　male（with　the mode　of　O　year）and　O．9　year　for　female（with　the　mode　of　O　year）．　In　other　words，　most　of the　Sama－Bajau　respondents　never　attended　school，　while　a　number　of　the　non－Sama－Bajau

respondents　at　least丘nished　elementary　school（6　years）．　On　top　this　disparity，

improvement　in　educational　attainment　over　generations，　which　is　common　among　the

non－Sama・Bajau，　could　not　be　seen　among　the　Sama－Bajau． 8）　The　name　of　the　area　inhabited　by　the　Sam　in　Isla　Bella． 9）　The　data　on　non－Bajau　were　collected　from　all　the　households　in　the　two　areas　named“purok”（this 　　　Barangay　is　divided　into　seven　puroks）：one　is　adjacent　to　Macau；and　the　other　includes　China 　　　Pikas． 10）The　Laminusa　also　belong　to　the　Sama　ethno・1inguistically．　They　are　a　subgroup　of　the　land・based 　　　Sama　from　Siasi．　In　the　research　site，　though，　they　often　tried　to　explain　to　us　that　they　were 　　　different　from　the　“Bajau”by　emphasizing　that　they　practice　Islam．　Compared　to　the　selfidentified 　　　“Bajau”indeed，　fewer　houses　owned　by　Laminusa　were　built　along　the　coast．　They　also　had 　　　relatively　higher　educational　attainment． 　　　　　　　17

Hakusan　ReView　ofAnthropology　Vol．15，　March　2012 Table　1 Basic　Attributes　of　Sama－Bajau　Residents　in　Isla　Bella（Macao　and　China　Pikas） Number　ofhouseholds　and　population Total　number　ofhouseholds 184 （Macao） （172） （China　Pikas） （12） Average　number　ofhousehold　members 5．3 Average　number　ofchildren　per　household

35

Total　population 981 Gender（ratio　to　the　total　population） Male _5L7％ Female 47．5％

Age

Male Female

ve「age　age 22．3years　old 22．8years　old

Age　distribution（ratio　to　the　total　population） 0－6years　old 2LO％ 22．1％ 7－15years　old 20．4％ 16．6％ 16－29years　old 29．4％ _325％ 30－49years　o［d 20．6％ _2L7％ 50years　old　or　up 8．6％ 7．1％ Educational　attainment（question　intended　only　fbr

respondents　seven　years　old　and　up） Male Female

Average L2　years 0．9years

Mean

Oyear Oyear

Self二identification（ratio　to　the　total　population）

Sama

97．8％ （Land－based／5α〃1αDのノαg） （65．1％） （Sea－based／∫α〃20　Dj1αμり （32．7％） Laminusa 1．5％ Cebuano 0．5％ Ata 0．1％ Tausug 一 Marano 一 Others 一 Religions　and　behefS（rat▲o　to　the　total　population） Ancestor　worship（mbog） 80．3％ Musli 10．8％ Christia 8．9％ （Pentecosta1 （4．8％） （Roman　Catholic　and　others （4．1％） Place　of　birth（ratio　to　the　total　population） Other　than　Davao　City 68．7％ Davao　City 31．3％ Note：total　may　not　be　100％due　to　the　exclusion　ofunknown　answers． Source：1998　household　survey　conducted　by　the　author．

　　　　　　A°YAMA：S・・i・I　I・・q・・lity・m°ng　Sama’Bajau　Mig「ants　i”　U’ban　Sgttlements Table　2　Basic　Attributes　and　Migration　History　of　Sama－Bajau　Household　Heads Tbtal　number　of households 184

Gender

Male

89．1％

Female

10．9％

Age

Average

40．Oyears　old

Median

35．5years　old

Mode

25．Oyears　old Place　of　b輌rth Davao　City 5．4％ Places　other　than　Davao　City 94．6％

Migration history （N＝170migrants） Number　of　trans企rs　with　change　of“home”address

Average

　　　．_P．4tlmes Place　of　birth（Province） Zamboanga　del　Sur 78．7％

Sulu

12．6％ Bohol／Cebu／Surigao　City 3．9％ Other　provinces　in　Mindanao 3．5％

Others

_12％

Place　of　theprevious　residence befbremoving　to　the　present　address（Province）

Zamboanga　del　Sur 74．7％

Sulu

9．2％

Davao　City　and　Davao　del　Sur 8．0％

Other　provinces　in　Mindanao 2．3％

Cebu／Surigao　City 2．3％

Metro Manila

and　its　neighboring　areas 2．3％

Major　reason丘）rtransfbr　f士om　the　place　of　birth

Deterioration　of　security　situations 66．7％ Hardship　of　lifb 14．4％

Invited　by relatives　who　had already　migrated　to　other　places 8．6％ Looking丘）r　livelihood 3．3％

Marriage

2．9％

Looking　fbr　housing 12％

雄ected　by　tidal　waves 0．6％

Main　channel　to　gather　infbrmation about　the　present　address befbre　transfbr

Through　family　and　relatives 82．8％ Directcontacts（pioneer　migrants） 14．9％ Through　friends　and　neighbors 2．3％ Year　of　transfbr to　the　present　address 1965・1969 1．1％ 1970・1974 9．7％ 1975－1979 8．0％ 1980－1984 10．2％ 1985’1989 14．8％ 1990－1994 30．7％ 1995・1999 25．6％ Note：total　may　not　be　100％due　to　the　exclusion　of　unknown　answers． Source：1998　household　survey　conducted　by　the　author．

19

Hakusan　Review　ofAnthropology　Vol．15，　March　2012 　　　　This　disadvantage　of　the　Sama－Bajau　against　the　non－Sama－Bajau　in　educational background　may　be　directly　reflected　in　its　inability　to　gain　employment　in　the　local

government，　given　its　comparatively　poor　human　capital　and　low　level　of　potential

productivity．　While　the　non・Sama－Bajau　experience　the　diversi丘cation　of　livelihood　after he／she　migrates　to　Davao　City　and　are　gradually　absorbed　in　the　urban　Iabor　market，　the Sama－Bajau　goes　through　only　a　limited　and　slow　shift　of　livelihood．　Most　of　the　household heads　were　fishermen　in　their　places　of　origin（71．4％of　the　respondents），　and　even　after they　came　to　the　city，　the　ratio　decreased　rather　slowly：51．2％of　the　Sama・Bajau　household heads　claim　to　be　still　engaged　in　fishing　in　our　survey．　In　addition，　intergenerational

occupational　mobility　is　extremely　low　among　the　Sama・Bajau　compared　with　the

non－Sama－Bajau． 　　　Nevertheless，　some　of　the　Sama－Bajau　household　heads　shifted　their　livelihood　from fishing　to　non－fishing　activities、　The　Sama－Bajau，　working　in　non－agricultural　sectors concentrates　on　the　self－employment　sector：33％of　the　households　reportedly　run　some kind　of　business　of　their　own（“nigosyo”）、　This　relatively　high　ratio　of　self・employment　can be　f（）und　in　the　non－Sama－Bajau　households　as　well．　It　should　be　noted，　however，　that　the variety　of　business　is　extremely　limited　in　the　case　of　the　Salna・Bajau．　There　are　practically only　two　types　of　business：（1）shell　and　pearl　vending（18．8％）and（2）used－clothing vending　or‘‘ukay－ukay”in　the　local　Cebuano　language（9．4％）． 　　　Another　typical　livelihood　the　Sama－Bajau　engages　in　is　mendicancy，　or　begging（5．9％）． There　are　many　public　markets　of　various　sizes　in　major　population　areas　of　Davao　City where　various　kinds　of　fresh　f（）od　are　sold．　This　affluence　in　the　local　markets　makes　it

possible　fbr　the　Sama－Bajau　not　only　to　beg　f（）r　spare　change　but　also　fbr

about－to・be・discarded　items：vegetables，　fruits　or　fish　no　longer　fresh　enough　to　be marketable．　Cassava　and　rice，　their　staple　fbod，　though，　have　longer　shelf　lives　and　cannot be　procured　from　begging．　In　reality，　there　are　obviously　more　Sama－Bajau　beggars　than our　questionnaire　survey　grasped，　and　most　of　them　roam　the　markets　and　streets　begging f（）rtheir　survivaL　One　of　the　factors　behind　this　tendency　lies　in　the　decline　of　fishing among　the　Sama－Bajau．　Unless　he／she　switches　to　another　livelihood　quickly，　the　line　of least　resistance　would　be　fbr　the　individual　members　of　the　household　to　beg　to　make　ends

meet　somehow．

　　　The　disadvantage　in　the　limited　type　of　livelihood　among　the　Sama－Bajau　naturally con丘ne　them　to　a　lower　socioeconomic　status　than　the　non－Sama・Bajau　as　shown　in　some

20

AoYAMA：Social　lnequality　among　Sama－Bajau　Migrants　in　Urban　Settlements indicators　of　economic　standards　of　living．　For　instance，　the　average　income　per　day　of　the household　head　from　his／her　main　livelihood　was　177．7　pesos　among　the　non－Sama－Bajau， whereas　it　was　only　153．9　pesos　among　the　Sama－Bajau11）．　Other　indicators　including　the level　of　expenditures，　the　access　to　the　credit　market，　housing　conditions　and　the　possession of　durable　goods　cのfirm　that　the　Sama’Bajau　are　generally　disadvantaged　compared　to　the other　ethnic　groups．

IV　THE　METHOD

1．Characteristics　of　the　Sample

This　study　follows　the　method　of　Lynch　I1959］and　Eder［2009（1982）］which　we　have

adapted　to丘t　the　situation　in　our　research　site． 　　　First，　we　selected　twenty　individuals　from　twenty　of　the　184　households　in　the　research site　to　fbrm　a　panel　of　raters．　The　selection　rate　we　set　was　over　ten　percent　of　the　total number　of　the　households．　Candidates　fbr　the　raters　were　initially　selected　through　three stratifications，　namely：place　of　birth，　livelihood（state　of　labor）and　level　of　income，　based on　the　results　of　the　basic　household　survey　conducted　prior　to　this　research（see　the

previous　chapter）．　Recent　migrants（within　six　months）and　temporary　migrants　were

exempted　from　the　sample　because　of　their　limited　knowledge　about　their　neighbors． 　　　Then　we　talked　to　each　of　the　candidates　to　make　sure　that　he　or　she　understood　the nature　of　our　interview　and　was　willing　to　accept　us　as　interviewers．　We　tried　to　secure raters　from　different　age　and　sex　categories．　Consequently，　we　narrowed　the　nulnber　of raters　down　to　twenty　f（）r　our　interviews．　The　raters’characteristics　are　presented　in　Table　3． 11）As　of　November　in　1999，　the　exchange　rate　was　US＄1＝about　P40（one　peso　is　about　three　yen）． 　　　The　legitimate　minimum　wage　in　the　non・agricultural　sector　was　158　pesos　a　day．　The　poverty　line 　　　was　defined　as　the　level　of　income　required　to　meet　one’s　nutrient　requirements　and　other　basic 　　　needs．　As　of　1997，　the　poverty　line　in　the　Southern　Mindanao　Region（including　Davao　City）was 　　　per　capita　annual　income　of　10，440　pesos．　The　date　on　the　household　income　collected　in　our 　　　questionnaire　survey　should　be　considered　to　be　higher　than　the　actual　level　of　income　that　the 　　　respondents　were　earning．　This　is　partly　because　the　data　was　gathered　by　self－assessment　of　each 　　　respondent，　and　also　because　we　assumed　in　the　quantitative　analysis　that　the　respondent　could 　　　regularly　earn　the　amount　of　the　income　they　claimed．　Through　a　long－term　participatory 　　　observation　later，　we　came　to　learn　that　the　level　of　monthly　household　income（with　more　than 　　　one　gainfUl　earner）could　widely　vary　from　the　maximum　of　20，000　pesos（as　an　exceptional　case） 　　　to　the　minimum　of　2，600　pesos， 　　　　　　　21

Hakusan　Review　of　Anthropology　Vol．15，　March　2012 　　　　The　interviews　were　conducted　from　January　22　to　April　27　in　1999．　It　had　been　one year　and　a　half　since　I　entered　the　research　site　with　my　research　assistant，　a　native speaker　of　Cebuano　who　lived　in　an　area　adjacent　to　the　research　site　in　the　same　barangay． We　already　had　a　baseline　data　on　demographic　and　socio・economic　characteristics　of　the raters　and　the　households　to　be　rated．　We　had　also　been　integrated　into　the　community　to the　extent　that　we　could　exchange　everyday　conversation　with　the　residents． 　　　　Once　the　panel　of　raters　was　determined，　we　began　to　visit　the　twenty　raters individually　at　their　homes　and　ask　them　to　rank　their“Bajau”neighbors　according　to　their 　　　　　　　Table　3　Socioeconomic　Characteristics　of　the　Raters

Frequency

Ratio 皿）tal 20 100．0％

Gender

Male

11 55．0％

Female

9 45．0％

Age

20－29years　old 5 25．0％ 30－39years　old 6 30．0％ 40years　old　and　up 9 45．0％ Civil　status

Married

20 100．0％ Place　of　birth（stratification　fbr　random selection　1） Rio　Hond 5 25．0％ Busbus！Jolo 5 25．0％ Sangali 4 20．0％ Margos　sa　Tubig 2 _10．0％

Taluksangay

1 _5．0％

Others

3 _15．0％ Status　of　laborllivelihood（stratification　fbr　random　selection　2）

Homemaker

5 25．0％

Fishermen

4 20．0％ “Ukay－ukaプ（secondhand　clothes）vendors 4 20．0％ Pearl　and　shell　vendors 3 15．0％

Mendicants

2 10．0％ Mat　weavers（fbr　domestic　use） 1 _5．0％

Unemployed

1 _5．0％ Level　of　income　in　the　community・（stratification 長）rrandome　selection 3） 50％（below　100　pesos） 8 40．0％ 75％（100°200pesos） 5 25．0％ 100％　（above　200　pesos） 7 35．0％ Note　a：the　quatile　category　the　rater’s　household　belongs　to　in　the　distribution　of　household income　per　day　in　the　studied　community． Source：1998　household　survey　conducted　by　the　author．

22

AOYAMA：Social　Inequality　among　Sama－Bajau　Migrants　in　Urban　Settlements social　status．　As　the　interviews　were　meant　to　be　personally　conducted，　we　tried　to　secure as　much　privacy　as　possible　fbr　each　of　our　raters　at　the　beginning．　We　were　later　to　adjust this　policy　mainly　because　the　one－room　structure　of　most　of　their　houses　made　it　almost impossible　to　stop　other　members　of　the　household，　and　sometimes　their　neighbors，　who were　curious　to　see　what　was　going　on，　from　sharing　their　own　opinions，　if　not　advices，　on the　ranking　of　their　neighbors　with　our　designated　rater．

　　　Consequently，　some　of　the“persona1”judgments　made　by　our　individual　raters

eventually　turned　out　to　be　highly“collective”judgments．　But　as　we　will　see　later　in　this paper，　the　analysis　of　the　data　reveals　that　there　is　an　implicit　agreement　among　the　raters， beyond　the　boundaries　of　households，　that　the　community　has　at　least　a　system　of　status ranking（our　second　hypothesis）．　Then　we　may　presume　the　influence　of　the　unexpected participation　of　non・raters　in　our　interviews　as　minimum．　In　addition，　we　also　learned through　the　interviews　that　the　community　could　be　divided　into　a　certain　number　of groups　according　to　their　social　status　and　that　within　the　same　group　its　members　share certain　criteria　fbr　social　standing．　In　such　a　case，　involvement　of　members　from　the　rater’s household　or　from　other　households　around　it　should　not　be　taken　as　a　major　disturbance　in our　research． 　　　We　had　also　to　consider　that　most　of　the　raters　in　our　interview　were　illiterate．　To　make it　easier　f（）r　them　to　identify　each　household　in　consideration，　we　prepared　index　cards，　each representing　one　of　the　184　households　in　the　research　site　and　bearing　the　names　of　the members　of　the　household、　Then　we　read　out　the　cards　to　the　raters　so　that　they　could　place them　in　ranking　according　to　their　perceived　social　status．　The　fbllowing　are　the introductory　questions　that　we　asked　each　rater　in　our　interview12）： 1 Do　you　think　that　people　are　different　from　each　other　in　any　given　place？ 　　　This　question　is　to　confirm　that　each　rater　perceives　the　existence　of　differences　among individuals　in　a　particular　location．　Once　the　rater　says　yes，　we　proceed　to　the　next questlon； 2． Do　you　think　it　is　possible　to　rank　some　people　in　higher　positions　and　others　in　lower positions？ 12）The　three　questions　below　were　originally　asked　in　Cebuano，　a　lingUa　franca　in　the　Visayas　and 　　　Mindanao，　including　Davao　City．

23

Hakusan　Review　ofAnthropology　Vol．15，　March　2012 　　　This　question　is　to　confirm　that　each　rater　thinks　he！she　is　able　to　express　the difference　among　people　in　a　system　of　status　ranking．　We　also　used　simple　gestures　of moving　our　hands　up　and　down　to　visualize　the　concept　clearer　to　the　raters．　While　all　of the　twenty　raters　agreed　to　the　procedure，　some　asked　questions　about　the　researcher’s　own criteria　in　status　ranking：“What　do　you　mean？Are　you　asking　if　some　are　rich　and　others are　poor？Or　any　difference　other　than　that？”In　response　to　such　questions，　we　replied　that we　wanted　them　to　use　their　own　criteria　for　ranking　rather　than　ours　and　that　we　would ask　them　to　explain　their　own　criteria　once　their　placement　of　the　households　was　finished． 3．Who　do　you　think　is　in　the　highest　position？ 　　　This　question　asks　each　rater　to　name　the　individual　whom　he　or　she　thinks　ranks highest　in　social　status．　After　the　rater　answered，　we　picked　the　card　with　the　name　of　the chosen　person　representing　the　household　and　handed　it　to　the　rater　who　placed　it　on　the floor．　This　household　represented　the　top　starting　point　for　the　placement　of　the　other households． 4．　Is　there　anybody　else　who　belongs　to　the　same　position　as　this　person？ 　　　If　any　other　individual　was　chosen，　the　card　of　household　that　he　or　she　belonged　to　was 丘led　with　those　in　the　highest　position，　and　they　were　put　together　as　the　highest　group． Then　we　repeated　the　same　procedures　until　no　more　names　would　be　called　out　fbr　this category．　Thus　the　households　that　fbrmed　the　highest　group　were　determined．　After　this， we　proceeded　with　the　selection　of　households　next　to　the　top　using　the　same　procedures repeating　the　same　process　to　go　on　to　the　next　group　until　all　the　households　were　divided into　a　certain皿mber　of　groups　according　to　their　positions　in　ranking、 　　　Some　of　the　raters　fbund　the　standardized　procedures　described　above　rather　difficult． So　alternatively，　we　simply　read　out　the　names　of　husband　and　spouse，　or　if　both　or　either　of them　were　absent，　the　name　of　the　person　who　appeared　to　be　the　head　of　the　household， from　the　index　cards、　The　cards　were　numbered　in　sequence　from　number　one　to　number 184according　to　the　location　of　their　dwelling　in　the　research　site．　Thus，　neighboring households，　most　of　which　were　related　to　each　other　by　blood　or　marriage，　had　numbers close　to　each　other．　Consequently，　the　raters　who　prefbrred　the　alternative　procedure　tended to　divide　the　households　into　a　certain　number　of　groups　first　and　then　ranked　the

24

AOYAMA：Social　lnequality　among　Sama－Bajau　Migrants　in　Urban　Settlements households　in　the　groups　according　to　their　social　status13）． 　　　Whether　the　rater　preferred　the　first　procedure　or　the　latter，　or　mixed　both　in　the course　of　our　interview，　it　was　clear　that　no　rater　opted　to　arrange　all　the　184　individual households　from　the　first　to　the　last　places　in　a　clear　sequence．　Instead，　all　of　them　finished the　placement　of　the　households　in　a　number　of　small　groups．　The　number　of　groups　varied from　two（high　and　low　categories）to　five（highest，　high，　middle，　low　and　lowest　categories）． We　shall　call　this　number　of　groups　in　assessment　the“number　of　placements”or“number of　classi丘cations”in　the　following　analysis． 　　　The　raters　were　also　encouraged　to　reconsider　the　grouping　and　ranking　of　the households　by　rearranging　the　households，　sub　grouping，　or　integrating　two　groups　into　one， etc．　until　they　became　satisfied　with　the　results　of　their　assessment．　Households　that　the rater　had　too　little　knowledge　about　were　carefully　removed　from　the　assessment．

　　　Once　the　placement　was丘nished，　we　confirmed　the　number　of　the　assessment

categories　that　the　rater　used，　and　if　the　rater　mentioned　specific　terms　to　refer　to　any　of the　categories，　we　collected　those　terms　as　well．　Although　we　were　constantly　taking　notes on　the　criteria　f（）r　ranking　the　raters　referred　to　during　the　placement　process，　we　still f（）rmally　asked　the　rater　to　explain　his　or　her　criteria　of　higher　and　lower　standings　befbre closing　our　interview．　We　shall　call　this　survey　the‘‘social　inequality　survey”hereafter．

V　THE　FINDINGS

1．　Status　classifications （1）Number　of　status　classifications Through　our　interviews，　it　was　clear　that　the　community　of　the　Sama－Bajau　in　the　research site　was　composed　of　individual　households　that　could　be　arranged　according　to　social status　in　a　continuum，　or　ranking． 　　　The　total　number　of　ratings　possible　was　20　raters　x　184　households，　or　3，680．　The　total number　of　ratings　actually　recorded　was　3，365（9工4　percent）．　Missing　ratings　arose　mostly because　the　rater　did　not　know　the　household　or　had　too　little　knowledge　to　make　a placement（Table　4）． 13）For　more　details，　see　Chapter　V　3．（1）of　this　paper．

25

Hakusan　Review　ofAnthropology　Vbl．15，　March　2012

Table　4　Twenty　Status　Classif］cations Number　ofhouseholds　placed in Tbtal

Number　of Rate　of

Rater

category households

　　　　　　， モ≠狽?№盾窒撃? 1 2 3 4 5 classi丘ed placementa 1 3 36 43 103 ． ． 182 98．9％ 2 3 20

45

117 ． ． 182 98．9％ 3 3 12 61 104 参 ■ 177 96．2％ 4 4 14

66

49

55 ■ 184 100．0％ 5 3 77 23

84

一 一 184 100．0％ 6 3 8 73 101 一 一 182 98．9％ 7 3 9 82 63 ・ 一 154 83．7％ 8 3 8 27 126 ・ 一 161 87．5％ 9 3

47

59

48

■ ． 154 83．7％ 10 4 16 56 75 33 ． 180 97．8％ 11 3 26 30 107 一 ■ 163 88．6％ 12 4 34 64

44

22 ◆ 164 89．1％ 13 3

20

57 83 ． 一 160 87．0％ 14 2 147 37 ． 一 ■ 184 100．0％ 15 3 20 28 123 ． 一 171 92．9％ 16 4 11 14 97 95 一 177 96．2％ 17 3 52 24 32 ． ． 108 58．7％ 18 5 12 8

86

60 5 171 92．9％ 19 3

96

30 37 ■ 一 163 88．6％　．

20

4

68

26

40

30 一 164 89．1％ Ibtalb

_9L4％

Notes：a．　Number　of　total　households　classified／total皿mber　of　households（184）tlOO 　　　　　b．Total　number　of　scores　actually　recorded／total　number　of　ratings　possible　de　100． Source：1999　social　inequality　survey　conducted　by　the　author． 　　　　The　rating　procedure　generated　different　numbers　of　status　placement　fbr　each household，　reflecting　the　difference　in　perception　among　the　raters．　However，　thirteen　out of　the　20　raters（65．0％of　the　total　raters）settled　on　3　placements，　while　5（25．0％）used　fbur． The　remaining　two　raters（5％each）reported　2　and　5　placements　respectively．　Table　4 compares　their丘ve　classiflcations． 　　　　It　is　statistically　proven　that　the　rater’s　income　level　was　not　correlated　to　the　number of　classifications　he　or　she　produced　in　the　rating　procedure．　Prior　to　the　selection　of　the raters，　we　had　stratified　the　184　households　of　the　community　into　three　categories　by　level of　daily　income（below　100　pesos，100　to　200　pesos　and　above　200　pesos，　see　Table　3）， considering　that　one’s　relative　economic　position　may　affect　his　or　her　perception　of　social status　order　in　the　community．　Nonetheless，　the　result　shows　that　regardless　of　their earning　position，　most　of　the　raters，　or　eighteen　out　of　the　20　in　our　sample，　agreed　to

26

AOYAMA：Social　Inequality　among　Sama－Bajau　Migrants　in　Urban　Settlements classify　the　households　in　the　community　into　3　or　4　groups　according　to　their　social　status． 　　　It　was　also　clear　that　the　raters　generally　had　knowledge　about　lifestyles　of　their neighbors．　There　were　no　correlations　between　the　raters’attributes　such　as　place　of　birth， 1ivelihood　and　level　of　income，　and　the　number　of　the　households　they　were　able　to　evaluate in　our　interviews．　As　Table　4　indicates，　nineteen　of　the　20　raters　were　able　to　evaluate　the social　standing　of　more　than　80　per　cent　of　the　households　in　community．　The　only　rater who　had　managed　to　place　only　58．7　per　cent　of　the　households　was　a　married　male relatively　old，　estimated　between　late　50s　and　early　60s，　among　the　raters．　Even　though　he had　resided　in　the　research　site　fbr　a　decade　by　the　time　of　our　interview，　the　space　and scope　of　his　daily　life　was　so　limited　that　he　was　unconcerned　with　others　beyond　his relatives　and　neighbors　close　to　his　dwelling． （2）Naming　of　status　classifications The　descriptions　that　the　raters　gave　to　the　social　strata　they　identi丘ed　imply　that　social ranking　was　perceived　to　be　related　to　difference　in　economic　status　within　the　community． Sixteen　out　of　the　20　raters　used　Sinama　words　such　as“daアahan”（rich）and“miskin” （poor）to　name　the　status　classifications．　Some　of　them　prefbrred　to　apply　more　modified expressions　such　as‘‘sida　daアahan”（the　richest）to　detail　the　di脆rences　in　strata　they provided，　especially　if　the　number　of　strata　they　identified　was　three　or　more．　Others employed　slightly　different　vocabulary　reflecting　their　places　of　origin． 　　　In　addition，　two　of　the　remaining　4　respondents　applied　livelihood－related　descriptions to　name　the　social　strata　they　recognized．　One　named　the　classifications　from　top　to　bottom according　to　livelihood　group，（1）‘‘palangre”（line　fishing）and‘‘bubu”（丘sh　trap）fishing，（2） any　kind　of“negosyo”（small　business），（3）“ukay・ukay”（secondhand　clothes）vending，　and （4）‘‘panaq”（spear　gun）fishing．　The　other　called　each　classification　by　the　name　of　a　person whom　the　rater　identified　as　a　leader　of　the　group　which　normally　composed　of　relatives and　close　neighbors，　and　engaged　in　the　same　type　of　livelihood． 　　　The　rest　of　two　referred　to　characters　or　personalities　of　people　who　belonged　to　each classification　they　set，　using　words　meaning“kind”and“not　kind．” 2．The　composite　status　distribution The　analysis　of　the　data　revealed　the　existence　of　an　implicit　agreement　among　the　raters that　the　community　had　at　Ieast　a　system　of　status　ranking．　This　was　statistically　proven by　the　fbllowing　procedures． 　　　　　　　27

Hakusan　Review　ofAnthropology　Vol．15，　March　2012 　　　As　we　have　seen　in　the　previous　chapter，　different　raters　employed　different　numbers　of strata　to　assess　the　social　status　of　the　184　households　in　the　research　site．　The　resulting twenty　sets　of　status　ratings　were　not　immediately　comparable　to　each　other、　The　first analytical　task　was　to　integrate　these　ratings　on　a“single　objective　scale”［Eder　2009 （1982）：213］． 　　　In　preparation　for　the　analysis，　I　needed　to　test　inter－rater　agreement　on　the　20　sets　of ratings．　To　do　this　1　applied　Kendall’s　coefficient　of　concordance（Kendall’s　W）14）to　the　data． The　result　showed　that　the　test　statistic　W　was　O．82　with　p＝0．000＜significant　level　O．05：it was　proven　that　there　is　generally　high　agreement　on　the　placement　of　the　184　households among　the　raters　despite　the　difference　in　number　of　status　strata　identified　by　each　rater and　the　difference　in　placement　of　each　household　he　or　she　decided　in　the　interview． 　　　Next　I　converted　all　raw　scores　to　percentile　scores　and　then　averaged　the　twenty percentile　scores　for　each　household［Eder　2009（1982）：213；Lynch　1959：95］．This　was　done through　the　procedure　described　below． 　　　First，　the　percentile　score，　or　the　cumulative　percentile　rating，　fbr　each　household　by individual　rater　was　calculated　through　the丘）llowing　steps： （1）Counting　the　total　number　of　households　that　a　particular　rater　placed　in　each　level on　his　or　her　own　scale；（2）dividing　the　numbers　obtained　from（1）by　the　total　number of　levels　that　the　particular　rater　employed　in　order　to　generate　the　percentile　score　f（）r each　household15）；and（3）apPlying　these　steps　to　all　the　sets　of　the　original　ratings． 　　　Then，　the　mean　status　score， out　through　completing　the　steps： or　the　composite　rating，　f（）r　each　household　was　figured 14）If　the　test　statistic　W　is　1，　then　all　the　respondents　have　been　unanimous，　and　each　respondent 　　　had　assigned　the　same　order　to　the　list　of　objects　or　concerns．　If　W　is　O，　then　there　is　no　overall 　　　trend　of　agreement　among　the　respondents，　and　their　responses　may　be　regarded　as　essentially 　　　random．　Intermediate　values　of　W　indicate　a　greater　or　lesser　degree　of　unanimity　among　the 　　　various　respondents． 15）For　example，　let　us　consider　a　case　in　which　a　rater　employs　a　f（）ur’1evel　classification　to　make　a 　　　placement　of　10　households．　Suppose　the　rater　placed　two　in　the　first　group，　two　in　the　second，　one 　　　in　the　third　and　five　in　the　last．　Cumulative　number　of　households　at　the　bottom　is　five．　Likewise 　　　　　　　， 　　　the　figures　f（）r　the　rest　of　the　three　levels　from　the　bottom　are　6（5＋1），8（6＋2）and　10（8＋2） 　　　respectively．　Therefore，　the　cumulative　percentiles｛for　the　f（）ur　levels　are　50　percent（at　the　bottom）， 　　　60percent，80　percent　and　100　percent（at　the　top），　which　are　given　to　the　indiVidual　households　in 　　　each　group　as　its　percentile　score．　This　procedure　assumes　that　all　intervals　on　the　individual 　　　scales　had　equal　significance　for　the　rater　who　made　them［Eder　1982：213］．

28

AOyAMA：　Social　lnequality　among　Sama－Bajau　Migrants　in　Urban　Settlements （4）Adding　up　the　percentile　scores　of　a　particular　household　and　divide　the　result　by the　number　of　raters　who　were　able　to　make　a　placement　on　the　household；and（5） apPlying　the　procedure　to　all　the　households　in　the　data　set． 　　　Then　I　plotted　these　scores　as　a　fどequency　distribution　table　with　O．025　intervals．　At　a glance，　there　appears　to　be　two　prominent　gaps　at　O．725　and　at　O．875．　Given　these　two　gaps as　breaking　points　in　the　composite　status　distribution，　the　upper　group　would　comprise 12．5percent　of　the　184　households　in　the　research　site，　while　the　middle　and　the　lower groups　contain　29．9　percent　and　57．6　percent　respectively．　As　we　have　seen　in　the　previous chapter，　more　than　half　of　the　raters　also　employed　three　status　classifications　in　the placements　they　made． 　　　However，　the　establishment　of　breaking　points　in　the　composite　distribution　should　be consistent　with　the　other　data　collected　through　interviews　and　observations．　T（）choose　the break，　our　next　task　is　to　examine　the　behaviors　the　raters　showed　in　the　course　of　rating procedures　as　well　as　the　criteria　of　higher　and　lower　social　standings　they　employed．　The breaking　points　should　not　only　be　statistically　rational　but　also　agreeable　with　their　views and　values　expressed　in　their　own　words． 3．Placement　behaviors　and　criteria （1）Placement　behaviors　among　the　raters In　contrast　to　our　expectation　that　the　raters　would　evaluate　individual　households　to　make placements［Lynch　1959；Eder　2009（1982）］，　more　than　half　of　the　raters　in　our　sample initiated　their　rating　procedures　by　sorting　the　households　in　the　community　into　groups which　they　called“」kampong．”16）While　the　word“kampong’re丘rs　to　a　village　or　hamlet　in Malay，　it　was　used　in　the丘）llowing　definition　by　Bottignolo　in　his　book： The　group　of　persons　with　whom　some　kin　relationship　exists，　whether　traceable　or　not． The　same　term　can　be　used　with　reference　only　to　the　local　cluster　of　kin　households ［Bottignolo　1995：266］。 In　other　words，　the　social　status　ranking　was　not　particularly　made　on　the　basis　of　the 16）The　relevance　of　this　term　to　refer　to　the　groups　in　the　given　context　was　conf］rmed　through 　　　interViews　with　some　of　the　key　informants　when　the　author　visited　the　research　site　in　March 　　　2012．The　term　was　explained，　though，　to　basically　mean”relatives”in　the　Sinama　language　used 　　　in　the　area． 　　　　　　　29