日本語の参照表現と接近可能性理論：接近可能性理論の言語間適用性の考察

全文

(1)Title. 日本語の参照表現と接近可能性理論：接近可能性理論の言語間適用性の考察. Author(s). 泉, 朋子; 佐藤, 吉文. Citation. 北海道教育大学紀要, 人文科学・社会科学編, 58(2): 101-114. Issue Date. 2008-02. URL. http://s-ir.sap.hokkyodai.ac.jp/dspace/handle/123456789/71. Rights. Hokkaido University of Education.

(2) 北海道教育大学紀要（人文科学・社会科学編）第58巻第2号 JournalofHokkaidoUniversityofEducation（HumanitiesandSocialSciences）Vol．58，No．2. 平成20年2月 February，2008. JapaneseReferringExpressionsandAccessibilityTheory： TheExaminationofaCross−LingulSticApplicabilityofAccessibilityTheory. IZUMITomoko andSATO Yoshifumi. l）epartmentofEnglishLanguageEducation，SapporoCampus，HokkaidoUniversityofEducation. 日本語の参照表現と接近可能性理論：接近可能性理論の言語間適用性の考察. 泉朋子・佐藤古文北海道教育大学札幌枚英語学研究室. ABSTRACT ThisstudyinvestigatedtheapplicabilityofAriel’sAccessibilityTheory（2002）toJapanesereferring. expressions．AccessibilityTheoryclaimsthatthereisacorrelationbetweentheinformationstatusofa. referentanditsreferringexpression．Themoreretrievabletheinformationis，thelessinformative，rigi. andmoreattenuatedformofreferringexpressionisused．Weexamined335nounphrasesfroma Japanesenovelbycalculatingeachreferent’sdegreeofaccessibilityandtheformofreferringexpression. usedforthereferent．TheresultofthestudyindicatedthechoiceofreferringexpressioninJapanese. mostlycorrespondedtoAriel’sAccessibilityTheoryasshorterformswereusedtorefertoreferents Withhigherinformationstatus，Whilelongerformswereusedtorefertoreferentswithlowerinformation. StatuS．However，theresultalsoindicatedreferringexpressionssuchaspersons’names，kinshipterms，. andtitlesdidnotfollowAccessibilityTheory．Weclaimthatthechoiceofthesereferringexpressionsis affectedbyasociolinguisticfactorandnotbyAccessibilityTheory．Thisisbecausehowoneaddressesa personplaysanimportantroleinsignalingthesocialrelationshipbetweenthespeakerandthereferent inJapanesesociety，SOtheselectionofaddressingnamesismadebasedonthesocialrelationship．. ringexpressionsinJapanese．InSectionl，We Introduction. Inthispaper，WeWillexamineifAccessibil− ityTheory（Ariel，2002）canbeappliedtorefer−. WillsummarizeAccessibilityTheoryproposed byAriel（2002）．InSection2，WeWillcompare. thedifferencesbetweenJapanesereferringex−. 101.

(3) IZUMITomoko・SATO Yoshifumi. pressionsandEnglishreferringexpressionsand. informationaboutthereferent．Accordingto. introducethreedifferentreferringexpressions. Ariel，theaccessibilityofreferringexpressions. WhicharepeculiartoJapanese．Forthepurpose. isdeterminedbythreefactors：informativity，. Ofcomparison，eXCerptSfromaJapanesefiction−. rigidity，andattenuation（Ariel，2002，p．32）．. alnovelanditslミnglishtranslation willbeused．. Arielclaimsthat“themoreinformative，rigid. Section3willbetheresultsofthestudyin. andunattenuatedanexpressionis，thelower. Whichweinvestigatedtheaccessibilitiesof12. thedegreeofaccessibilityitcodesandvicever−. differentkindsofreferringexpressions（335. sa（p．32）．’’Basedonthesethreefactors，Ariel. tokensintotal）fromtheJapanesefictional. proposesthefollowingaccessibilityhierarchyof. novelDiipuRibaa“DeepRiver’’（Endo，1996. EnglishreferrlngeXpreSSions．. ／1994）．InSection4，WeWilldiscussifthere−. SultsofthestudycorrespondtoAccessibility TheoryadvocatedbyAriel．Thelastsectionwill. （1）AccessibilityhierarchyofEnglishre− ferringexpressions（Ariel：2002，p．31）. be theconclusion．. Low Accessibility Sectionl：AccessibilityTheoryandSelectionof ReferringExpressions Ariel（2002）claimsthatreferringexpress−. Fullname＋modifier＞fullname＞longdefi− nitedescription＞shortdefinitedescription＞ 1astname＞firstname＞distaldemonstrative ＋modifier＞proximatedemonstrative＋mod−. ions，Whichareusedtorefertoaparticularob−. ifier＞distaldemonstrative＋NP＞proximate. ject，areSelecteddependingonhowaccessible. demonstrative＋NP＞distaldemonstrative（−. the referentisin the hearer’s mind．The term. NP）＞proximate demonstrative（−NP）＞. accessible”isusedtorefertothedegreeofdif−. StreSSedpronoun＋gesture＞stressedpro−. ficultyofretrievinginformationaboutthere−. noun＞unstressedpronoun＞cliticizedpro−. ferentinone’smind．Thehighertheaccessibil−. noun＞verbalpersoninflections＞zero. High Accessibility. ityofthereferent，theeasieritistoretrievethe informationaboutthereferent，andthelower theaccessibilityofthereferent，theharderitis. Accordingtothishierarchy，fullNPssuchas. toretrievetheinformation．. mynephewJohnSmith，arelowaccessibility. AccessibilityTheory（Ariel，2002，p．29）. attemptstodescribetherelationshipbetweena referringexpressionandthedegreeofre−. markers，Whilepronounssuchasheanditare hなhaccessibilitymarkers．. Thedegreeofaccessibilityofareferentit−. ferent’saccessibilityinthehearer’smind．Ariel. selfisdeterminedbycomplexfactors（p．34）．. Claimsthatthedegreeofaccessibilitythatare−. Oneofthesefactorsisthetopicalityofthere−. ferrlngeXpreSSionmarkscorrelateswiththein−. ferent．Ifthereferentisthetopicofthedis−. formationstatusofthe referentwithinthe dis−. COurSe，theaccessibilityofthereferentishigh．. COurSe．Referringexpressions，SuChasthepro−. Anotherfactoristhenumberoftimesthe re−. nounhe，andthefullnounphrase（NP）John. ferentismentioned withinthe discourse．The. Smith，marktheirowndegreeofaccessibility，. morefrequentlythereferentismentioned，the. andtheyservetocuethehearertoretrievethe. highertheaccessibilityis．ThenumberofNPs. 102.

(4) JapaneseReferringExpressionsandAccessibilityTheory. thatcouldbeapossibleantecedentalsoaffects. holdscross−1inguistically．Therefore，WeWillex−. thedegreeofaccessibility．Forexample，ifthere. amineifAccessibilityTheorycanbesaidto. are three NPs that could be possible. holdcross−1inguisticallybyinvestigatingthecor−. antecedentsbetweenananaphor（i．e．，arefer−. relationbetweenJapanesereferringexpressions. ringexpression）anditsrealantecedent（i．e．，a. andthedegreeofaccessibilityofthereferents．. referent），thenthedegreeofaccessibilityis lower thanifthereis noNP betweenthem．. Furthermore，factorssuchashowrecentlythe referenthasbeenmentionedandwhetherthe. Section2：RefbrringexpressionsinJapanese Inthissection，WeWilldemonstratethedif−. referentisthesubjectornotcouldaffectthe. ferencesbetweenreferringexpressionsin. degreeofaccessibility．AsArielclaims，OneCan−. JapaneseandthoseinEnglish．Then，WeWillin−. notdecidetheaccessibilityofareferentbased. troducethreedifferentkindsofreferringex−. solelyononefactor（Ariel，2002，p．37）．. pressioninJapanese，Whicharenotincludedin. AccessibilityTheoryclaimsthataspeaker. Ariel’sproposal．. Selectsareferringexpressionaccordingtothe degreeoftheaccessibilityofthereferent．For example，ifamaleboyisreferredtobythepro−. 2．J．J乃deβ乃ffeα乃ddeβ乃ffeαrffeJeβ. UnlikeEnglish，Japanesedoesnotlexically. nounhe，Whichislessinformative，rigidand. expressthedifferencebetweena＋Nandthe. moreattenuated（i．e．，ahighaccessibility. ＋N（e．g．，Gundeletal．，1993，p．284）．Thefol−. marker），theinformationabouttheboyinthe. lowingisanexcerptfromtheJapanesefictional. hearer’smindishighlyaccessible．Thisisbe−. novelDiipuRibaa（Endo，1996）anditsEnglish. cause the hearer can retrieve theinformation. translation（Endo，1994）．1. abouttheboyregardlessoftheverylimitedin− formationconveyedsolelybythereferringex−. （2）Isobe−Wa Shinsatsushitsu−nO. pressionhe．Ontheotherhand，ifthesameboy. Isobe−Nom from theexamination−rOOm. isreferredtobythefullNPmynephewJohn. mado−nO Shita−kara kare−nO rOubai−O. Smith，Whichismoreinformative，rigidandless. Windows−Of−belowfrom hisdismay. attenuated，thentheinformationabouttheboy. WarauyOunikikoeta yakiimoya−nO. inthehearer’smindislessaccessiblebecause. laughing. thehearerneedsto obtainmoreinformation. koe−ga（p．9）．．．. fromthereferringexpressionitselfinorderto. VOICe. retrievetheinformationabouttheboy．Ariel. Mata soto−de yakiimoya−nOkoega. Claimsthatthisrelationbetweenreferringex−. Again，OutSide potatovendorofvoice. pressionsandreferentsholdscross−1inguistical1y．. kikoeru（p．10）．. Thatis，inanylanguages，themoreaccessible. can be heared. potatovendorof. thereferentisinthediscourse，thelessinforma− tive，rigid，andmoreattenuatedformisusedto. “The voice ofastreetuendorpeddling. referto thereferentand viceversa．Because. roastedsweetpotatoesbelowthewindow. ArielmakesherclaimbasedsolelyonEnglish. ofthe examination−rOOm Came back to his. data，thereisnoevidencethatthetheorytruly. earslikeasneeringmockeryofhisdismay．. 103.

(5) IZUMITomoko・SATO Yoshifumi. ．．（p．7）Hecouldstillhearthevoiceofthe. bothareexpressedbythesamekinshipterm. potatouendoroutside（p．8）．’’. tsuma“wife．’’Therearenopossessivepronouns. noranydeterminersattachedtoisuma．Because Both the first and the second mention of. theusageofbarekinshiptermsisverycom−. yakiimoya”potatovender’’areexpressedbya. moninJapanese，WeWillinvestigateifthese. bareNPinJapanesewhileEnglishdifferentiates. barekinshiptermsexpressadifEerentdegreeof. thesetwobyusingindefiniteanddefiniteartic−. accessibilitycomparedtootherreferringex−. 1es．Itwouldbe difficultto examinethe differ−. pressions．. enceinaccessibilitybetweena＋Nandthe＋ NinJapanesebecausethedifferenceisnotlex−. Referringtoapersonbythenameofhis OCCupationisalsocommoninJapanese．. icallyexpressed．Forthisreason，WeWillex−. Cludethedifferenceinaccessibilitybetweena. （4）Wife：Oishasama−niatta？. ＋Nandthe＋NfromJapanesereferringex− pressions．However，aSWillbediscussedinthe. Doctor（polite）（you）met？ Isobe：Aa．. nextsection，Japaneseusesbarekinshipterms andtitlesinawaywhichisnotusuallyusedin. Yes． Wife：Oishasama nanto osshattemashi−. English．Wewillinvestigateifthesetworefer−. ta？（p．10）. ringexpressionsmarktheirowndegreesof. Doctor what said？. accessibility． “Wife：Didyoutalktothedoctor？ Isobe：Um−hm．. 2．2．点前血叫pねrmβα〝d耶fJeg. Inthenovelusedforthedatacollection，. Wife：What．．．didhesay？（p．8）′′. barekinshiptermssuchas“Wife”and“Mother”. arefrequentlyusedasreferringexpressions．. Again，WhiletheEnglishtranslationusesthe. ThefollowinglSaneXample．. twodifferentreferringexpressionsthedoctor andhe，thesamereferringexpressionOishasa−. 「3）Tkuma−nidouus（｝O tSukouka，atama−nO mfi！−tO howlie tell. head. ma“Doctor（polite）’’isusedtorefertothe same referentwithin the discourse．For this. naka−de sagashihajimeta．Byounin−nO. reason，WeWillalsoinvestigateifthosetitlesas. inside search started patient. areferringexpressionhaveacertaindegreeof. eibinsa−de tsuma−WaIsobe−nO…．（p．10）. accessibilityinJapanesereferringexpressions．. SenSitivity W7h寸寸om ofIsobe 2．3．CJα朗f帯βrざ. “He scanned his brainfor the mostconvinc−. BasedonthestudybyDowning（1986），. inglietotellhisu？ifb．Shewouldsurelysee. Ariel（2OO2，p．48）claimsthatJapanesenumeral. throughtheworkingsofhismindwiththe. Classifiershaveanaccessibilitydegreewhichis. keensensitivityoftheafnicted．’’. betweenpronounsandlexicalNPs．Thefollow−. ingisanexampleoftheuseoftheclassifiernin， Whilethedifferentreferringexpressionsh W室斥andsheareusedintheEnglishtranslation，. 104. whichis usedfor ahuman referent．.

(6) JapaneseReferringExpressionsandAccessibilityTheory. 5）Sorekara3−nin−u？a Sukoratetsugaku−nO. −Titles. Then 3＿（二且＿Nbm scholasticism. amarinimeikainariron−O tSukatte．．（p．191）. Inthenextsection，WeWilldemonstratethere−. Very Clear logic used. Sultsofthestudythatexaminedtheaccessibil− ityoftheJapanesereferringexpressions．. “Then，uSingtheall−tOO−tranSlucentlogicof SCholasticism，they…．（p．118）’’. Asshown，thenumberplusclassifier3−ninis. usedwhilethepronountheyisusedintheEn−. Section3：Materials，method，andresultsofthe. Study Inthissection，WeWillexplainthematerial，. glishtranslation．Downing（1986，pp．364−365）. themethodandtheresultsofthestudythatin−. ClaimsthatJapaneseclassifiers，Whichcontain. VeStigatedtheaccessibilityofJapaneserefer−. informationaboutthenumberofreferents，are. ringexpressions．. used toreintroducereferents whichhavebeen textuallyevokedbuthavebeenunused（Prince， 1981）．Becausetheuseofclassifiersispeculiar. 3．J．肋ferfαJ. ThefirstchapteroftheJapanesefictional. toJapanese，WeWillcategorizeclassifiersasone. novelDiipuRibaa“DeepRiver’’wasusedfor. kindofreferringexpressionsandexaminethe. datacollection．2Thechapterconsistsof32. degreeoftheiraccessibility．. pages，andittellsastoryaboutamiddle−aged manliobeand hiswifeKeiko．Thenoveliswrit−. 2．4．g〃mmα叩0′励cfわ〝2. tenfromapartialomniscientperspective．. In this section，We have shown the dif−. ferencesinreferringexpressionsbetweenEn− glishandJapanese．Inordertoinvestigateif. 3．2．肋αg〃r（ヲme〝由. ThemethodproposedbyToole（1996）was. Accessibility Theory can be applicable to. adoptedtomeasuretheaccessibilityofeachre−. Japanesereferringexpressions，theaccessibility. ferringexpression．AsmentionedinSectionl，. Ofthefollowing12differentNPswillbeex−. thedegreeoftheaccessibilityofareferentina. amined．. discourseisaffectedbyseveralfactors．Toole focusesonthreefactors：distance，tOpicality，and. （6）−NullNPs. COmpetition．Thedistancebetweenareferring. −Reflexives. expressionandthelastmentionofthesamere−. −Pronouns. ferentwasmeasuredfrom ascaleofzeroto. −Demonstratives. fourdependingonthedistance．Ifareferring. −Demonstratives ＋N. expressioniswithinthesamepropositionofthe. −FirstNames. lastmentionofthereferent，thentheaccessibil−. −LastNames. ityofthereferrlngeXpreSSionisfour．Ifthelast. −FullNames. mentionofthereferentistwoormoreepisodes. −LongNPs. beforethereferringexpression，thentheacces−. −Classifiers. Sibilityiszero．Topicalitywasmeasuredbythe. −KinshipTerms. frequencywithwhichthereferentwasmen−. 105.

(7) IZUMITomoko・SATO Yoshifumi tioned．Ifthe referentis mentioned more than. Classi重ers. 28. 2．8. twicewithinfourpropositions，thentheaccessi−. KinshipTerms. 43. 1．16. bilityistwo．Ifthereisnomentionofthere−. Titles. 27. 0．7. ferentwithinthefourpropositions，thenthe. Total. 335. accessibilityiszero．Competitionismeasuredby thenumberofmatchingentitiesbetweenan. Asshown，therewereonlythreetokensfor. anaphoranditsrealantecedent．Matchingen−. firstnamesandfourtokensforfullnames．Be−. titiesare those whichhave the samefeaturefor. CauSeOfthelimitednumberoftokens，Wedo. person，numberandgenderasthereferentand，. notassumethattheaveragedegreeofaccessi−. asaresult，COuldbepossibleantecedentsofthe. bilityforthesereferringexpressionsisreliable．. anaphor（Toole，1996，p．274）．Iftherearemore. However，WeWilldiscussthereasonwhythere. thantwomatchingentities，theaccessibilityis. WereVeryfewtokensforthesetwocategories. reducedby2．Ifthereisnomatchingentity，the. laterin thissection． Basedontheaveragedegreeofaccessibil−. accessibilitydoesnotchange． Byfollowingthesecriteriaproposedby. ity，theaccessibilityhierarchyofreferringex−. Toole，therewouldbeninedifferentdegreesof. pressionsinJapaneseisasfollows：. accessibility．Themaximumissix，Whichmeans. themostaccessible，andtheminimumisminus. （7）AccessibilityhierarchyofJapanesere−. ferring expressions. two，Whichistheleastaccessible．. Low Accessibility. 3．3．点錯血魅. examinedinthisstudy．Foreachtoken，We. LongNPs＞Titles＞Kinshipterms＞＊First names＞Lastnames＞Demonstrative＋N＞. Calculatedthedegreeofaccessibilitybyusing. Pronouns，Classifiers＞Demonstratives＞Null. Toole’scriteria．The numberoftokensand the. NPs＞＊Fullnames＞Reflexives. Atotalof335referringexpressionswere. HighAccessibility. averagescoreoftheaccessibilityofeachrefer− ringexpressionareasfollows．. Thedistributionofthereferringexpress− Table3．1：Thenumberoftokensforeachreferring. expression. graphbelowrepresentsthedistributionofrefle− ＃oftokens MeanAccessibility. NullNPs. 45. 3．67. Reflexives. 12. 4．55. Pronouns. 45. 2．8. Demonstratives. 29. 3．24. Demonstratives＋N. 41. 2．61. ＊FirstNames LastNames. ＊FullNames LongNPs. 106. ionsofeachcategoryisalsoexamined．The. 3. 42 4. 16. 1．3 1．38 4. −0．81. Ⅹ1VeS．.

(8) JapaneseReferringExpressionsandAccessibilityTheory. Table3．2：DistributionoftheAccessibilityofReflex− ives（M＝4．55）. Willdiscusstheaccessibilitiesoftworeferring expressionswhicharepeculiartoJapanese， namelykinshiptermsandtitles． 4．J．Aα闇扇動服砂丘壷r（ⅠⅣ鳥yorJαpα〝eぷere危rrf〝g 叩reββわ朋. Thefollowinggraphsindicatetheaccessi− bilitydistribudonofeachreferringexpression．「「. Table4．1：NullNPs（M＝3．67）. 4 4. ■b. Null NPs. 3. O. 3 2. 5. 2. Thexaxisrepresentsthedegreeofaccessibil−. 1. 0. 1. 辞. ity．Theyaxisrepresentsthepercentageof NPswhichbelongtoeachdegreeofaccessibil−. ■b. ity．Forexample，50％ofthereflexiveshavean. O. accessibilityof5and18％ofthemhavean. 5 0. accessibilityofeither30r4．Graphsofallthe. −2 −1 0 1 2 3 4 5 6 Accessibility. ■b. CategOrieswillbeshowninthenextsectionfor. O. dataanalysis．However，Wewi11notprovidethe. Table4．2：Demonstratives（M＝3．24）. graphsoffullnamesandfirstnamesduetothe Demonstratives 0. limitednumberoftokens．Inthenextsection，. 8 0 5 0. WeWilldiscussiftheresultsofthestudyindi−. 4 0 3. Cate that Accessibility Theory applies to. 0. ㌔. 2 0. Japanesereferringexpressions．. 1 0. Section4：ApplicabilityofAccessibilityTheory. 2 −1 0. ityTheorycanbeappliedtoJapanesereferring ■hU O 3 3 ■．■リ 20. expressions．Forthispurpose，WeWillfirstin−. VeStigateiftheorderingoftheaccessibilitiesof referringexpressionsproposedbyAriel（i．e．，. 一宇. Pronouns. 2. J O ・Ol ■．・■l. termsofthedegreeofaccessibility．Lastly，We. 5. ■．hJ. EnglishandJapanesereferringexpressionsin. 4. ㌔. Willexaminewhythereisadi＃erencebetween. 3. Table4．3：Pronouns（M＝2．8）. Inthissection，WeWilldiscussifAccessibil−. Japanesereferringexpressions．Ifitisnot，We. 2. Accessibility. to Japanese referring expressions. （1）inSectionl）isthesameasthatofthe. 1. 2 −1 0 1 2 3 4 5 6 Accessibility. 107.

(9) IZUMITomoko・SATO Yoshifumi Table4．4：Classi五ers（M＝2．8）. Table4．5：Demonstrative＋N（M＝2．61）. Table4．7：KinshipTerms（M＝1．16）. Table4．8：Titles（M＝0．7）. Demonstrative＋N ■. 2. 1 0. 1. 2. 3. 4. 5. 6. A（）CeSSibility. Table4．6：Lastnames（M＝1．38）. Table4．9：LongNPs（M＝−0．81）. Last names 3. ■．︻︶. ■. 3 0 ︻■. 2 ㌔. 0. 2. ■h︶. ・t−. 0. ・t−. ■．■■. 2 −1 0 1 2. 3. 4. 5. 6. A（）CeSSibility. 0. Asshown，theuseofeachreferringexpression. graphofthedistributionofnullNPs（i．e．，Table. inJapaneseisnotrandom．Somereferringeエー. 4．1）withthatoflongNPs（i．e．，Table4．9），itis. pressionsareusedforhighlyaccessiblere−. ClearthatnullNPsareusedforhighlyaccessi−. ferents whileothers are usedforless accessible. blereferentswhilelongNPsareusedforless. referents．Forexample，ifonecomparesthe. accessiblereferents．Statisticalanalysisalsoin−. 108.

(10) JapaneseReferringExpressionsandAccessibilityTheory. dicatesasignificantdifferencebetweennull NPsandlongNPs（p＜．001）．Furthermore，the. （8）IsotxO3，4kagetsu−Wanyuuinshinakucha 30r4months 一光inthehospital. Orderingoftheaverageaccessibilityratesof. naranai．DemoO 4kagetsu−gOniwa. thosereferringexpressionsmostlycorresponds. should． But. totheorderingproposedin Ariel（i．e．，（1）in. kanariyokunarutoOitteorareru. Sectionl）．Forexample，thereisasignificant. Very Wellbewill. 4monthslater thatsaid. differencebetweentheaccessibilityofnullNPs，. （Endo，1996，p．11）. Whicharelessinformative，rigid，andmore attenuate，andthatofpronouns，Whicharemore. “Isobe：（You）aregoingtostayinthehos−. informative，rigid，andlessattenuate（p＜．005）．. pitalforthreeorfourmonths．But「ゐq）said. TheaverageaccessibilityrateoflongNPs，. 付ou）would be alot better after four. Whicharethemostinformative，rigid，andthe. months（Endo，1994，p．8）．’’. leastattenuate，isthelowestamongtherefer− ringexpressions．Forthisreason，WeaSSume. ThereisnoexplicitNPusedin（8）．whileEn−. AccessibilityTheoryisapplicabletoJapanese. glishusespronouns，JapaneseusesnullNPs．. referringexpressions．Exceptforthereverse. Thefollowingisanexampleoftheuseofa. Orderofpronounsanddemonstrativesandthe. Japanesepronoun．. Samedegreeofaccessibilitybetweenpronouns andclassifiers，theorderingoftheaverage. （9）“Kondo−Wa田kitono. kaiwaka．’’. accessibilitiesofreferringexpressionsisthe. “Thistime withatree conversation. SameaSthatinEnglish．Thelessinformative，. to kare−Wafuan−00kakusutame. rigid，andmoreattenuatedformsareusedas. that he. highaccessibilitymarkerswhilethemorein−. SeSeraWaratte miseta．. formative，rigid，andlessattenuatedformsare. 1aughed. anxiety tohide. tried（p．13）. usedaslowaccessibilitymarkers． “Sonowyouaretalkingtotrees？Helaughed 4．2．乃eαCCeggfふfJf抄0′Jqpα〝egepr（〉〝0〃〝g. at仇河toshieldhisownuneasiness．. WhiletheaccessibilityofEnglishpronouns ishigherthandemonstratives，thisisnottrue. Inthesentenceabove，boththepronounkare. forJapanesepronouns．Statisticalanalysisin−. “he’’andanullNPareused．Theaccessibilityof. dicatesthatthereisnosignificantdifferencebe−. thenullNP，indicatinghiswife，is5whilethe. tween pronouns and demonstratives in. accessibilityofkare“he’’isonlyl．Thatis，. Japanese（p＝．128）althoughthemeanaccessi−. JapaneseusesanullNPtorefertothehigher. bilityofdemonstrativesishigherthanthatof. accessibilityreferentandapronountoreferto. pronouns．Weassumethisrelatestotwofactors. theloweraccessibilityreferent．ForbothNPs，. that are peculiar toJapanese．Oneis that. however，Englishusespronouns（i．e．，heand. Japaneseallowsnullsubjectsandobjects（e．g．，. her）．weassumethisisthereasonwhythe. Guasti，2002，p．154）．Thefollowingisanexam−. averagedegreeofaccessibilityofJapanesepro−. ple．. nounsseemstobelowerthanthatofEnglish pronouns．Thatis，becauseJapanesehasanop−. 109.

(11) IZUMITomoko・SATO Yoshifumi. tion（andmostofthetimesitispreferred）to. kersthanEnglishpronouns．Asmentionedin. usenullsubjectsandobjects，referentswhich. Sectionl，themoreinformativeareferringex−. arehighlyaccessiblearereferredtobynull. pressionis，thelessaccessiblethereferentis．. NPs．Ifareferentisnotaccessibleenoughtobe. Forthesereasons，WeaSSumethatthe. referredtobyanullNP，thepronoun tendsto. accessibilityofJapanesepronounsislowerthan. beusedasthereferringexpression．InEnglish，. that of English pronouns．Thisis because. Ontheotherhand，theuseofnullsubjectsand. Japanesehasanoption，Whichisoftenprefer−. nullobjectsisgrammatica11yrestricted，SOeVen. red，tOuSenullNPsandJapanesepronouns. referentswithhighaccessibilitytendtobere−. haveextrainformationtoexpresstheemotional. ferredtobypronouns．. associationbetweenthespeakerandthere−. Thesecondfactorthatexplainsthere−. ferent．Asshown，althoughtheorderingofthe. 1ativelyloweraccessibilityofJapanesepronouns. accessibilityofreferringexpressionsinEnglish. relatestotheuseofpronounsinJapanesesocie−. isdifferentfromthatofJapanese，thedifference. ty．AccordingtoDowning（1986，p．365），the. CanbeexplainedbasedonAccessibilityTheory．. useofthethirdpersonpronounssuchaskare “he’’carriessomeobligatoryemotionalasso−. Ciationsbetweenthespeakerandthereferent．. 4．3．乃eαeee朋fふfJf抄0′eJα朋fβe柑. AsopposedtoAriel’sprediction，theacces−. Forexample，ifonereferstohisbossbykare. Sibilityofclassifiersinthisstudyisashighas. “he，’’itwouldbeconsideredthatthespeakeris. thatofpronounsandthereisnosignificantdif−. notrespectingtheboss．Althoughtheimplica−. ferencebetweenthem（p＝．950）．Thisresult. tionscarriedbytheuseofthethirdpersonpro−. alsocontradictswiththestudyofDowning. nounsvary，Downingclaimsthattheuseofthe. （1986）．Downingshowsthatthemeandistance. thirdpersonpronounsisimproperifthere−. betweenaJapaneseclassifieranditsantecedent. ferentissomeonewhomthespeakerrespects．. islonger than that of a pronoun andits. Theuseofthethirdpersonpronounsispossible. antecedent（p．360）．Thatis，theaccessibilityof. ifthereferentisafriendclosetoone’sownage，. the referentofaclassifierislowerthan thatof. asibling，aStranger，apublicorhistoricalper−. apronoun．Then，Whyistheaccessibilityofclas−. SOnWithwhomonehasnopersonalrelation−. Sifiersinthisstudythesameasthatofpro−. Ship．Becausethenovelusedinthisstudyis. nouns？Weassumethisdifferenceiscausedby. Writtenfromtheperspectiveofathirdparty，it. thedifferentdefinitionsoftheanaphorofaclas−. WOuldbepossibletousethethirdpersonpro−. sifier．In Toole（1996），the antecedentofa. nounsasreferringexpressions．However，be−. groupentityisdefinedaseitherthementionof. CauSethepronounsthemselveshaveextrain−. thegroupitselforoneoftheindividualsofthe. formationsuchas“expressingtheemotional. group，Whicheveriscloser．Forexample，inthe. associationbetweenthespeakerandthere−. SentenCebelow，Md7γWOuldbedefinedasthe. ferent（Downing，1986，p．365），’’itwouldbe. antecedentofthepronounthey（p．271）．. possibletosaythatJapanesepronounsare moreinformativethanEnglishpronouns．Ifthis istrue，thenonecouldexplainwhyJapanese pronounsareusedasloweraccessibilitymar−. 110. qO）JohnsawMary．T7ieywavedtoeach other．.

(12) JapaneseReferringExpressionsandAccessibilityTheory. However，inDowning（1986），theantecedentof. assume this is because the main character of. aclassifieris definedasfollows．. thischapterisreferredtobyhislastname lsobe．400fthe42forlastnamesareactually. “Whentheanaphoricmentionwaspre−. lsobe．Itwouldbeplausibletoclaimthatthe. Cededbymentionsofonlyasubsetofits. averageaccessibilityofthelastnameinthe. referents，thesementionswerenotcounted. dataisrelativelyhighbecausemostofthe. asantecedents，butratherasmentionsof. tokensrefertothemaincharacter，Whichis. interveningreferents（pp．372−373）．’’. Salient，frequentlymentionedandoftenatopic．. Itistherefbrentofoneparticularlastname Ifonefollowsthisdefinition，thenMaryin（10），. Whichishighlyaccessibleinthereader’smind．. Whichiscountedastheantecedentoftheyin. Similarly，theaverageaccessibilityofthekin−. Toole（1996），WOuldbecountedasaninterven−. Shiptermsishigherthanthatoftitlesbecause. ingreferent．Whenwereexaminedtheaccessi−. Ofthereferentsofthesereferringexpressions．. bilityofclassifiersbyfollowingDowning’s. Thesecondmaincharacterofthechapteris. Criteria，theaverageaccessibilitydroppedto. Isobe’swife．Becausesheisreferredtobythe. l．6．Therefore，itisclearthatthehighdegree. kinshiptermtsuma“wife’’mostofthetime，the. Ofaccessibilityofclassifiersinthisstudyisdue. averageaccessibilityofkinshiptermsishigher. tothefactthatToole’smeasurementclassifies. thanthatofthetitles．Althoughthereareonly. anindividualmember of a group as an. fourtokens，theaverageaccessibilityoffull. antecedentwhileDowningcountsitasanin−. namesisreallyhigh（i．e．，4．5）whilethatofthe. terveningreferent．Inordertoinvestigatethe. firstnameisbetweentheaccessibilityofthe. accessibilityofJapaneseclassifiers，afurther. kinshiptermsandthatofthetitles．. Studywillbeneededtoseeifthementionofan. Then，howcouldoneaccountforthisfact？. individualofagroupactivatestheaccessibility. IfonefollowsAriel’sproposal，fu11namesshould. Ofthegroupreferentorlowerstheaccessibility. havealoweraccessibilitythanfirstandlast. Ofthegroupreferent．. names．Kinshipterms，Whichcouldbecatego−. rizedasshortdefinitedescriptions，Shouldalso 4．4．乃eαCCeざざfふfJf抄0′丘f朋んわfer〃ばα〝dfffJeざ. AsshowninSection3，thereareonlyfour. haveloweraccessibilitythanlastandfirst names．However，although the number of. tokensforfullnamesandthreeforfirstnames. tokensissmal1，fu11namesmarktheexceptional−. inthe textusedforthe datacollection．Onthe. 1yhighaccessibility（i．e．，4．5）inthedata．Simi−. Otherhand，therearemorethan40tokensfor. 1arly，theaccessibilityofkinshiptermsissome−. kinshiptermsand27tokensfortitlesobtained. Whathigherthanthatoffirstnames．Infact，. fromthetext．Furthermore，StatisticalanalysIS. thereisnostatisticallysignificantdifferencein. indicatesthatthereisnosignificantdifference. accessibilitiesamongthesereferringexpress−. intheaverageaccessibilitiesamonglastnames，. ions．. kinshipterms，andtitles（1astnamesandkin−. Weassumethesereferringexpressions. Shipterms，p＝．556；kinshiptermsandtitles，p. actuallyhavethesameor，atleast，aSimilarde−. ＝．31；titlesandlastnames，p＝．127）．Theaver−. greeofaccessibility．InJapanesesociety，how. ageaccessibilityofthelastnameishigh，butwe. Onereferstoapersonisanimportantfactorto. 111.

(13) IZUMITomoko・SATO Yoshifumi. Signaltherelationshipbetweenthespeakerand. Thechoiceofthesereferringexpressionsisde−. thereferent（e．g．，Holmes，2000，ChlO）．Refer−Cidedbysocialfactorsandnotbytheaccessibil−. ring to a person by his first name is only. ityofthereferent．Thisiswhythereisnoclear. allowedifthespeakerisolderorthesameage. differenceintheaccessibilitiesamonglast. andveryfamiliarwiththereferent．Referringto. names，kinshiptermsandtitles．Weassumethat. apersonbyhisfullname，incontrast，implies. thesetermsactuallybelongtoonecategory，. thatthespeakerisnotfamiliarwiththere−. Whichissomethinglike“addressingnames．. ferent．Bytheterm“familiar’’wemeanthatthe. Thereasonwhytherearefewtokensforfirst. referenthasasociallycloserelationshipwith. namesin the dataisbecause referentsthatare. thespeaker．Thisisnotthesame“familiarity’’. accessibleenoughtobereferredtobytheir. definedinGundeletal．（1993，p．278），Whichin−. firstnamesareactuallyreferredtobykinship. dicatesthecognitiveretrievabilityofthere−. terms and titles．The use of first namesis. ferent’sinformation．ThefollowinglSaneXam−. avoidedbecauseitindicatesasociallyclosere−. pleoftheuseofthefu1lnamefromthetext．. 1ationship between the referent and the speaker．AlthoughMulken（1996）claimsthat. 仕t）“Zonjiteorimasu．Demo．．．’’To “（I）knew．. But’’. that. theaccessibilityofpropernamesishigherthan thatoffullnames，theresultsofthisstudydo. 胸TYLSeMitsuko−Wa Shizukanakoede．．．（p．22）. notshowthistendency．Rather，thesedifferent. Naruse Mitsuko quietly. kindsofreferringexpressionsareeffectively usedtorefertothesocialrelationshipbetween. ‘“Iunderstand．But．．．’Mitsukosaidsoftly. thespeakerandthereferent．. （p．15）．’’ 4．5．g〃mmα叩OrgeC￡わ〝4. Surprisingly，althoughthefullnamejVdruseMit−. Inthissection，WehaveexaminedifAcces−. SukoisusedintheoriginalJapanesetext，the. SibilityTheorycanapplytoJapanesereferring. SameperSOnisreferredtobyonlyherfirst. expressions．Althoughtheaccessibilityof. nameMitsukointheEnglishtranslation．We. Japanesepronounsdoesnotseemtobeashigh. assume thatthefullnameis used toindicate. asthatofEnglishpronouns，WeClaimthatthis. thefactthatthemaincharacterliobe doesnot. isduetothehighavailabilityofnullNPsand. likethewomanNaruseMitsuko．Referringto. extrainformationconveyedbythethirdperson. Miisukobyherfullnameindicatesthesocialdis−. pronounsinJapanese．. tancebetweenthespeakerandthereferent，. Exceptfortheorderbetweendemonstra−. anditdoesnotmeanthattheaccessibilityof. tivesandpronouns，theorderoftheaccessibil−. NaruseA4itsukoislow．The referenthas been. ityofreferringexpressionsinJapanesefollows. mentionedseveraltimesintheprecedingprop−. AccessibilityTheoryadvocatedbyAriel（2OO2）．. OSitions．Therefore，itcanbesaidthatreferring. Themoreinformative，rigidandlessattenuated. expressionssuchasfirstnames，1astnames，full. areferringexpression，thelessaccessiblethe. names，titles，andkinshiptermsareveryimpor−. referentisandviceversa．Althoughtheaccessi−. tanttoindicatetherelationshipbetweenthe. bilityofclassifiersinthisstudydoesnotfollow. SpeakerandthereferentinJapanesesociety．. Ariel’sclaim，itisduetothedifferentcriteria. 112.

(14) JapaneseReferringExpressionsandAccessibilityTheory for the antecedents ofclassifiers usedin this. aspronouns．Furthermore，thedifferencesin. studyandthatofDowning（1986）．. accessibilitiesamongfirstnames，1astnames，. Onethingthatisworthmentioningisthat. fullnames，kinshiptermsandtitlesdidnot. thereisnocleardifferenceinaccessibility. SeemtOfollowAriel’saccessibilityhierarchyin. amongfullnames，1astnames，firstnames，kin−. English referringexpressions．Tn Section4，then，. Shipterms，andtitles．Itseemsthatsocialfac−. Weinvestigatedwhythesedifferencesoccurred．. tors，SuChasthesocialclosenessbetweenthe. ItwasshownthattheaccessibilityofJapanese. referentandthespeaker，ratherthanthede−. pronounswasrelativelylowbecausereferents. greeofaccessibility，affectthechoiceofrefer−. Withhighaccessibilitieswereusuallyreferred. ringexpressionsinJapanese．Thisispeculiarto. tobynullNPs．Furthermore，thethirdperson. Japanesebecausethesamereferentexpressed. pronounsinJapanesehaveextrainformation，. bythefirstnameinEnglishisexpressedbythe. Whichconveystheemotionalassociationbe−. fullnameinJapanese．Callingareferentbyher. tweenthespeakerandthereferent．Weclaimed. fullnameimpliesthatthespeakerdoesnot. thatthesetwofactorsloweredthedegreeof. haveasociallycloserelationshipwiththere−. accessibilityofJapanesepronouns．Thediffer−. ferent．. enceinthedegreeoftheaccessibilityofclassi− fiersbetweenthestudyofDowning（1986）and. Conclusion Inthispaper，WehaveexaminedifAccessi−. thatofourswascausedbyadifferenceinthe measurementofaccessibilitybetweenthetwo Studies．Consideringthefactsmentionedabove，. bilityTheory（Ariel，2002）isapplicablecross−. itseemsthatreferringexpressionsinJapanese. 1inguistically．InSectionl，WeSummarized. COuldbedeterminedbasedonAccessibility. AccessibilityTheoryadvocatedbyAriel．Itwas. Theory．OnethingthatdidnotfollowAccessi−. Claimedthatthechoiceofreferringexpressions. bilityTheorywasreferringexpressionssuchas. dependedontheaccessibilityofthereferent．In. persons’names，kinshipterms，andtitles．We. Section2，WeeXaminedthedifferencesinrefer−. Claimedthatthechoiceofthesereferringex−. ringexpressionsbetweenJapaneseandEnglish．. pressionswasafEectedbythesocialrelationship. In order to apply Accessibility Theory to. betweenthespeakerandthereferentandnot. Japanese，Weintroducedthreedifferentkindsof. bytheaccessibilityofthereferent．Although. referringexpressions：Classifiers，kinshipterms. somesocialfactorshavemoreeffecton the. andtitles．InSection3，Weintroducedthemate−. ChoiceofreferrlngeXpreSSionsovertheaccessi−. rial，methodandresultsofthestudywhichin−. bilityofthereferent，WeCOnCludethatAccessi−. VeStigatedtheaccessibilityofJapaneserefer−. bilityTheoryappliestothereferringexpress−. ringexpressions．Byexaminingtheaccessibility. ionsinJapanese．. degreeofeachreferringexpression，WeprOp−. OSedanaccessibilityhierarchyofJapanesere− ferringexpressions．Itwasshownthatthe accessibilityofJapanesepronounswasnothigh− erthanthatofdemonstratives，andJapanese Classifiershadthesamedegreeofaccessibility. References Ariel，M．（2002）．Accessibilitytheory：Anoverview．In T．Sanders，J．Schilperoord，＆W．Spooren（Eds．），Text R＠resenhdion（pp．29−86）．JohnBenjamins．. 113.

(15) IZUMITomoko・SATO Yoshifumi Clancy，P．M．（1980）．ReferentialchoiceinEnglishand. itiswrittenfromtheperspectiveofthemaincharac−. Japanesenarrativediscourse．InW．L．Chafe（Ed．），. ter．Inthisnovel，themaincharacterchangesevery. r／∼（・♪√〟／・∫／「げ′l・∫．・r〃g〃′／′J・（・．r〟／／J…J／．（川（J／擁印南晶、∫. Chapterandtheuseofreferringexpressions，SuChas. aspectsofnarrativeproduction（pp．127−202）．Nor−. namesofpeople，Changedependingonthemainchar−. WOOd，NJ：Ablex．. acter．Forexample，aperSOnWhoisreferredtoby. Downing，P．（1986）．Theanaphoricuseofclassifiersin. herlastnameinChapterlisreferredtobyherfirst. Japanese．InC．Craig（Ed．），Nounclassesandcategor−. nameinChapter2duetothefactsheisthemain. ization（pp．345−375）．Amsterdam：JohnBenjamins．. Characterofthischapter．AsshowninSection4，. Endo，S．（1994）．DeゆRiver（Ⅴ．C．Gessel，Trans．）．New. thesereferringexpressionsaresensitivetothere− 1ationshipbetweenthemaincharacterandthere−. York：NewDirections．. Endo，S．（1996）．Diipuribaa［Deepriver］．Tokyo： Koudansha．. ferent．Therefore，Wedecidedtoonlyusethefirst Chapterfordatacollection．. Guasti，M．T．（2002）．Langu‘聯αCquisition：Thegrou，th q′Bnmmar．Cambridge，MA：theMITPress． Gundel，J．K．，Hedberg，N．，＆Zacharski，R．（1993）．Cogni−. tivestatusandtheformofreferringexpressionsin discourse．Langu（雛，69，274−307． Holmes，J．（2000）．Anintroductiontosociolinguistics （2nded．）．London：Longman． Mulken，A．E．（1996）．Thegameofthename．InT． Fretheim，＆J．K．Gundel（Eds），Rejbrenceandre− jbrentaccessibility（pp．235−250）．Amsterdam：John. Benjamins． Prince，E．F．（1981）．Towardataxonomyofgiven−neW information．InC．Peter（Ed．），RadicalPragmatics （pp．223−255）．NewYork：AcademicPress． Toole，J．（1996）．Theeffectofgenreonreferential choice．InT．Fretheim，＆J．K．Gundel（Eds．），Rqfbr− enceand7マ々柁nLaccessibiliLy（pp．263−290）．Amster− dam：JohnBenjamins．. Footnotes I DiipuRibaahasbeenpublishedbyseveralpub− 1ishers．Theeditionweusedforthedataanalysiswas firstpublishedin1996．However，thesamenovelhad alreadybeenpublishedbyotherpublishers．Thisis thereasonwhytheEnglishtranslationofDiiPu WaSpublishedin1994，Whichistwoyearsearlierthan theJapaneseversionweusedforthisstudy． 2 Becausethereareonlytwotokensforclassifiers， WeeXaminedallthechaptersofthenoveltoobtaina sufficientnumber oftokensasclassifiers．There are 28examplesin whichclassifiersareusedasreferring expressions．However，forobtainingthetokensof Otherreferringexpressions，WeOnlyexaminedthe firstchapterofthebook．Althoughthenoveliswrit− tcnfromthcpcrspcctivcofathirdparty，SOmCtimcs. 114. （泉朋子札幌校大学院生）（佐藤吉文札幌校教授）.

(16)

日本語の参照表現と接近可能性理論 ： 接近可能性理論の言語間適用性の考察

日本語の参照表現と接近可能性理論：接近可能性理論の言語間適用性の考察