【作者】 范淑玲；

【导师】 张树铮；

【作者基本信息】 山东大学 ， 汉语言文字学， 2009， 博士

【摘要】 域外译音的研究向来为汉语音韵界所重视。日语中有许多汉字音是从中国汉语中古时期传入的,而且吸收自汉语不同的时代和地区。鉴于汉字在日语中的特殊地位,分析日译汉音对研究汉语中古时代的标准音和方言有重要意义。本文分析了日语上代、中古时期中日语汉字音的历史,并和同时期的汉语中古音进行了比较研究。日语语音史上的上代时期和中古时期正值中国语音史上的中古时期。这时期的日本典籍颇多,本文主要依据《古事记》、《万叶集》、《日本书纪》和字典、佛书及汉籍资料,他们均用万叶假名编纂而成。在分析万叶假名并和中古声母进行比较时,主要依据《古事记》、《万叶集》和《日本书纪》。他们所用万叶假名有近千个。《古事记》、《万叶集》采用吴音音系,《日本书纪》采用汉音音系。吴音以魏晋南北朝时期的南方方言为母胎音,反映的是当时吴语区的音韵情况;汉音以隋唐五代的西北方言为母胎音,反映的是当时西北方音。通过共时和历时比较,我们发现万叶假名反映了中古汉语声母的一些特点。唇音方面,中古重唇音和轻唇音没有区别,而且中古唇音一定是读重唇音。日语汉音介音(i|¨)的唇音字中的介音脱落,反映的是唐后期西北方音中产生了轻唇音。舌音方面,中古舌头音和舌上音没有区别。在唐音中的对应,说明舌上音和照组齿音合流。齿音方面,中古齿头音和正齿音没有区别,而且正齿音照二和照三无别。止摄开口三等、蟹摄开口三等和四等字对应辅音s/z,其余对应ts/dz,我们推测是细音的影响,使中古齿头和正齿音产生了腭化。另外,船母和禅母、从母和邪母不分。全浊声母在唐西北方音中已经有清化现象,不过在魏晋南北朝时代清化现象不普遍,唐五代时已经基本清化。鼻声母在魏晋南北朝时代除泥母有非鼻音化苗头外,其余均保持鼻声母。但在唐五代西北方音中已呈现非鼻音化趋势。泥、娘、日三母,在魏晋南北朝时代没有区别,证明“娘日归泥”的可信。但在唐西北方音中,娘、日有别。喉音方面,影母和喻三相混,喻三和喻四基本区分,但三母都对应单元音,说明三者声母相通,也就是说喻三和喻四是分而有混。喻三和合口匣母都对应w,说明喻三和匣母相混。在和中古汉语韵母进行比较时,我们仍然依据《古事记》、《万叶集》和《日本书纪》的万叶假名,特别是“上代特殊假名遣”,还采用了字典、佛书及汉籍资料。通过分析“上代特殊假名遣”中甲类、乙类的不同对应和合流,可以看出它们反映了中古汉语的重纽现象和韵的合流。“上代特殊假名遣”中甲类、乙类的不同,反映的是日语实际语音的差异,它分别对应的中古汉语的重四重三,说明中古的重纽三四等有语音的差异。“上代特殊假名遣”中イ列(i、(i|¨)),在《古事记》、《万叶集》和《日本书纪》中,甲类所对应韵部只有支、脂,前两书中キ/ki(キ/gi)甲类中,只有一个脂韵字,说明在魏晋南北朝时代,支、脂并未合流,但在唐西北方音中支、脂两韵合流,但区别于之、微两韵。而乙类则覆盖了整个止摄支、脂、之、微四韵,说明唐代北方方言中这四韵已经合流。“上代特殊假名遣”中的ェ列(e、(e|¨)),在《日本书纪》α群中,甲类只对应中古汉语的齐韵和祭韵,而且齐韵全是开口四等,说明唐代北方方音中齐祭相混;乙类只对应哈韵和合口灰韵,而且只是唇音的へ/F(e|¨)和め/m(e|¨)。说明在唐时的中国北方,咍灰合一。“上代特殊假名遣”中的才列(o、(o|¨)),《古事记》、《万叶集》甲类主要对应模韵、侯韵,还有少量的钟韵、宵韵、冬韵、虞韵(只有一字“虞”);乙类对应鱼韵和登韵、之韵、虞韵、蒸韵、咍韵、魂韵、职韵。《日本书纪》的对应也大致如此,这个复杂性也应是中古汉语实际语音的反映。从这个复杂对应中,我们证明了鱼虞模同用、蒸登相押、部分鼻韵尾合流。分析字典、佛书和纪传等汉文资料中日语汉字音标记的不同和变化,可以看出中古汉语韵尾的一些真实情况。入声韵尾,中古汉语的入声韵尾是闭音节音。在魏晋南北朝吴方言中-p、-t、-k三种韵尾三足鼎立,不相混。在唐西北方音中,-p、-t、-k向喉塞音韵尾变化,出现相混的情况。宋代南方方言中入声-p、-t、-k已经合一。鼻音韵尾,中古汉语的鼻音韵尾是闭音节音,依然是-m、-n、-(?)。但在魏晋南北朝吴方言中,-m、-n有相混现象。在唐西北方音中,-n韵尾和-(?)韵尾有消失现象,而且-m、-n、-(?)开始合并。宋代南方方言中-(?)韵尾合并到-n韵尾中。更多还原

【Abstract】 Sounds and transliteration are always an important research area in phonologie.The pronunciation of many Chinese characters in Japanese originated from Mid-century Chinese.The different pronunciations of the Chinese characters in Japanese resulted from its absorption from Mid-century Chinese in various regions and at different times. Due to their special position,analyzing Chinese characters in Japanese is crucial for the study of standard pronunciations and dialects in Mid-century Chinese language.In this dissertation,the author reviews the history of Chinese character pronunciations in both Zyoudai and Mid-century Japanese,and compares these pronunciations with their Mid-century Chinese counterparts.The Japanese Zyoudai and Mid-century periods correspond with the Chinese Mid-century period.During this period,there appeared many Japanese works,among which the authors mainly refers to Kojiki,Manyoshu,Nohonshoki,and dictionaries, Buddhist sutra and translated Chinese works,all of which were compiled in Manyo Kana.The analysis of Manyo Kana and its comparison with Mid-century initiales are based on Kojiki,Manyoshu and Nihonshoki,which consist of about a thousand Manyo Kana. Kojiki,and Manyoshu adopt Goon Phonology while Nihonshoki uses Kanon Phonology. With southern dialects of the Wei,Jin,and Southern-and-Northern-Dynasties as its mother son,Goon reflects the phonetic features of WU region at that time;Kanon,with northwestern dialects of Sui,Tang and Five-Dynasties as its mother son,reflects the dialects of northwestern regions.Through synchronic and diachronic comparison,we can find that Manyo Kana bears some features of Mid-century Chinese initiales.In regards with labiales,there is no difference between heavy labiales and light labiales, and the Mid-century labiales must be heavy ones.The disappearance of Kanon mediale (i|¨) in labiales reflects that light labiales originated from northwestern dialects in the late Tang Dynasty.As for linguals,there is no difference between Mid-century apicales and dorsales. Through comparison with corresponding linguals in Tang dynasty,we know that dorsales and ZHAO(照) rubrique dentales merged.As to dentale,there is no difference between Mid-century CHI TOU(齿头) sounds and dentales;so do ZHAO(照) 2 and ZHAO(照) 3 in dentales.ZHI(止) groupe KAIKOU (开) 3 division,XIE(蟹) groupe KAIKOU(开) 3 division and KAIKOU(开) 4 division correspond with consonnes s/z,others with ts/dz.So we hold that it is the influence of mediales that palatalized Mid-century CHI TOU(齿头) sounds and dentales.Besides,CHUAN(船) initiale and SHAN(禅) initiale are the same and so do CONG(从) initiale and XIE(邪) initiale.Some complete vocalized initiales had been devocalized in Wei,Jin,and Southern-and-Northern-Dynasties periods,although this phenomenon was not very common then.Until Tang Dynasty and Five-Dynasties periods,most complete vocalized initiales had been devocalized.In Wei,Jin,and Southern-and-Northern-Dynasties periods,nasal initiales remained unchanged,except that "NI(泥) initiale" tended to be denasalized.But in Tang Dynasty and Five-Dynasties periods,northwestern nasal initiales had gradually been denasalized.The three initiales NI(泥),NIANG(娘) and RI(日) were the same in Wei,Jin,and Southern-and-Northern-Dynasties periods,which proves that NIANG(娘) and RI(日) originated from NI(泥).On the other hand,NIANG(娘) and RI(日) were different in northwestern dialects of Tang Dynasty.When it comes to laryngales,YING(影) initiale and YU(喻) 3 are hard to tell apart, and YU(喻) 3 and YU(喻) 4 can usually be distinguished.All of the three initiales correspond to voyelles,which indicates that they are correlated with each other,that is, YU(喻) 3 and YU(喻) 4 are sometimes distinct and sometimes exchangeable.YU (喻) 3 and HEKOU(合) XIA(匣) both correspond to w,which shows that they are exchangeable.As is mentioned above,the comparison between Manyo Kana and Mid-century Chinese is based on Manyo Kana in Kojiki,Manyoshu,Nihonshoki,and "Zyoudai Special Kanadzukai" as well.Moreover,the author also refers to dictionaries,Buddhist sutras and translated Chinese works.Within "Zyoudai Special Kanadzukai",type A and type B correspond to different finales but they merge as well,which reflects the CHONG NIU(齿头) phenomenon and the merging of rimes in Mid-century Chinese.The difference between type A and type B in "Zyoudai Special Kanadzukai" reflects different pronunciations in Japanese.The two types correspond respectively to CHONG (重) 3 and CHONG(重) 4 in Mid-century Chinese,which proves CHONG NIU(重纽) 3 division and CHONG NIU(重纽) 4 division are different in pronunciation.In Kojiki,Manyoshu and Nihonshoki,type A ofイcolumn(i、(i|¨)) in "Zyoudai Special Kanadzukai" correspond only to ZHI(支) rime and ZHI(脂) rime.And type A ofキ/ki (ギ/gi)in Kojiki and Manyoshu only correspond to the ZHI(脂) rime.This proves that by Wei,Jin,and Southern-and-Northern-Dynasties periods,ZHI(支) rime and ZHI(脂) rime had not merged together.While in northwestern dialects of Tang Dynasty,ZHI(支) rime and ZHI(脂) rime merged but they were different from ZHI(之) rime and WEI (微) rime then.On the other hand,type B in Nihonshoki include ZHI(支),ZHI (脂),ZHI(之) and WEI(微) four rimes of the ZHI(止) groupe;this indicates that in northwestern dialects of Tang Dynasty,the four rimes had already merged together. Inαgroup of Nihonshoki,type A ofエ(e、(e|¨)) in "Zyoudai Special Kanadzukai" correspond to QI(齐) rime JI(祭) rime in Mid-century Chinese,with QI(齐) rime being KAIKOU(开) 4 division.This shows that in northwestern dialects of Tang Dynasty,QI(齐) rime and JI(祭) rime merged.Type B correspond with HAI(咍) rime and HEKOU(开口) HUI(灰) rime,which only applies in labialesへ/F(e|¨) andめ/m(e|¨).This proves that in the north part of China in Tang Dynasty,HAI(咍) rime and HUI(灰) rime merged.In Kojiki and Manyoshu,type A of Columnオ(o、(o|¨)) in "Zyoudai Special Kanadzukai" correspond primarily to MO(模) rime and HOU(侯) rime,in some cases ZHONG(钟) rime,XIAO(宵) rime,DONG(冬) rime and YU(虞) rime;type B correspond to YU (鱼) rime,DENG(登) rime,ZHI(之) rime,YU(虞) rime ZHENG(蒸) rime,HAI(咍) rime,HUN(魂) rime and ZHI(职) rime.There are similar kinds of corresponding in Nihonshoki.The complexity of rime corresponding here reflects actual pronunciations in Mid-century Chinese.And through the complex corresponding,we have verified that YU(鱼) and YU(虞) are exchangeable,ZHENG(蒸) and DENG(登) rhyme,and some nasal consonne finales merge.Through the analysis of such Chinese resources as dictionaries,Buddhist sutras and biographies,and different marks in the pronunciation of Chinese characters in Japanese and the changes of these marks,the real situation of some Mid-century consonne finales are reflected.Occlusive consonne finales:occlusive consonne finales in Mid-century Chinese were syllabe overte.In WU dialect of Wei,Jin,and Southern-and-Northern-Dynasties periods, consonne finales -p,-t and -k were distinct.While in northwestern dialects in Tang Dynasty,-p,-t and -k developed into glottal stop consonne finales,merging with each other.In southern dialects of Song Dynasty period,-p,-t and -k had merged into one occlusive consonne finale.Nasal consonne finales:nasal consonne finales in Mid-century Chinese were syllabe overte,including -m,-n and -(?).In WU dialect of Wei,Jin,and Southern-and-Northern-Dynasties periods,-m and -n began to merge.In northwestern dialects of Tang Dynasty,-n and -(?) started to disappear,and -m,-n and -(?) began to merge.In southern dialects of Song Dynasty period,-(?) was incorporated into -n.更多还原